The analysis of geological materials, volume 2

The analysis of geological materials, volume 2

T

HESE

T

ERMS

G

OVERN

Y

OUR

U

SE OF

T

HIS

D

OCUMENT

Your use of this Ontario Geological Survey document (the “Content”) is governed by the terms set out on this page (“Terms of Use”). By downloading this Content, you (the

“User”) have accepted, and have agreed to be bound by, the Terms of Use.

Content: This Content is offered by the Province of Ontario’s Ministry of Northern Development and

Mines (MNDM) as a public service, on an “as-is” basis. Recommendations and statements of opinion expressed in the Content are those of the author or authors and are not to be construed as statement of government policy. You are solely responsible for your use of the Content. You should not rely on the

Content for legal advice nor as authoritative in your particular circumstances. Users should verify the accuracy and applicability of any Content before acting on it. MNDM does not guarantee, or make any warranty express or implied, that the Content is current, accurate, complete or reliable. MNDM is not responsible for any damage however caused, which results, directly or indirectly, from your use of the

Content. MNDM assumes no legal liability or responsibility for the Content whatsoever.

Links to Other Web Sites: This Content may contain links, to Web sites that are not operated by MNDM.

Linked Web sites may not be available in French. MNDM neither endorses nor assumes any responsibility for the safety, accuracy or availability of linked Web sites or the information contained on them. The linked Web sites, their operation and content are the responsibility of the person or entity for which they were created or maintained (the “Owner”). Both your use of a linked Web site, and your right to use or reproduce information or materials from a linked Web site, are subject to the terms of use governing that particular Web site. Any comments or inquiries regarding a linked Web site must be directed to its Owner.

Copyright: Canadian and international intellectual property laws protect the Content. Unless otherwise indicated, copyright is held by the Queen’s Printer for Ontario.

It is recommended that reference to the Content be made in the following form: <Author’s last name>,

<Initials> <year of publication>. <Content title>; Ontario Geological Survey, <Content publication series and number>, <total number of pages>p.

Use and Reproduction of Content: The Content may be used and reproduced only in accordance with applicable intellectual property laws. Non-commercial use of unsubstantial excerpts of the Content is permitted provided that appropriate credit is given and Crown copyright is acknowledged. Any substantial reproduction of the Content or any commercial use of all or part of the Content is prohibited without the prior written permission of MNDM. Substantial reproduction includes the reproduction of any illustration or figure, such as, but not limited to graphs, charts and maps. Commercial use includes commercial distribution of the Content, the reproduction of multiple copies of the Content for any purpose whether or not commercial, use of the Content in commercial publications, and the creation of value-added products using the Content.

Contact:

F

OR FURTHER

INFORMATION ON

The Reproduction of

Content

P

LEASE CONTACT

:

MNDM Publication

Services

B

Y E

-

MAIL

:

[email protected]

The Purchase of

MNDM Publications

Crown Copyright

MNDM Publication

Sales

Queen’s Printer

B

Y TELEPHONE

:

Local: (705) 670-5691

Toll Free: 1-888-415-9845, ext.

5691 (inside Canada,

United States)

Local: (705) 670-5691

Toll Free: 1-888-415-9845, ext.

5691 (inside Canada,

United States)

Local: (416) 326-2678

Toll Free: 1-800-668-9938

(inside Canada,

United States)

[email protected]

[email protected]

L

ES CONDITIONS CI

-

DESSOUS RÉGISSENT L

'

UTILISATION DU PRÉSENT DOCUMENT

.

Votre utilisation de ce document de la Commission géologique de l'Ontario (le « contenu ») est régie par les conditions décrites sur cette page (« conditions d'utilisation »). En téléchargeant ce contenu, vous (l'« utilisateur ») signifiez que vous avez accepté d'être lié par les présentes conditions d'utilisation.

Contenu : Ce contenu est offert en l'état comme service public par le ministère du Développement du Nord

et des Mines (MDNM) de la province de l'Ontario. Les recommandations et les opinions exprimées dans le contenu sont celles de l'auteur ou des auteurs et ne doivent pas être interprétées comme des énoncés officiels de politique gouvernementale. Vous êtes entièrement responsable de l'utilisation que vous en faites.

Le contenu ne constitue pas une source fiable de conseils juridiques et ne peut en aucun cas faire autorité dans votre situation particulière. Les utilisateurs sont tenus de vérifier l'exactitude et l'applicabilité de tout contenu avant de l'utiliser. Le MDNM n'offre aucune garantie expresse ou implicite relativement à la mise à jour, à l'exactitude, à l'intégralité ou à la fiabilité du contenu. Le MDNM ne peut être tenu responsable de tout dommage, quelle qu'en soit la cause, résultant directement ou indirectement de l'utilisation du contenu. Le

MDNM n'assume aucune responsabilité légale de quelque nature que ce soit en ce qui a trait au contenu.

Liens vers d'autres sites Web : Ce contenu peut comporter des liens vers des sites Web qui ne sont pas exploités par le MDNM. Certains de ces sites pourraient ne pas être offerts en français. Le MDNM se dégage de toute responsabilité quant à la sûreté, à l'exactitude ou à la disponibilité des sites Web ainsi reliés ou à l'information qu'ils contiennent. La responsabilité des sites Web ainsi reliés, de leur exploitation et de leur contenu incombe à la personne ou à l'entité pour lesquelles ils ont été créés ou sont entretenus (le

« propriétaire »). Votre utilisation de ces sites Web ainsi que votre droit d'utiliser ou de reproduire leur contenu sont assujettis aux conditions d'utilisation propres à chacun de ces sites. Tout commentaire ou toute question concernant l'un de ces sites doivent être adressés au propriétaire du site.

Droits d'auteur : Le contenu est protégé par les lois canadiennes et internationales sur la propriété intellectuelle. Sauf indication contraire, les droits d'auteurs appartiennent à l'Imprimeur de la Reine pour l'Ontario.

Nous recommandons de faire paraître ainsi toute référence au contenu : nom de famille de l'auteur, initiales, année de publication, titre du document, Commission géologique de l'Ontario, série et numéro de publication, nombre de pages.

Utilisation et reproduction du contenu : Le contenu ne peut être utilisé et reproduit qu'en conformité avec les lois sur la propriété intellectuelle applicables. L'utilisation de courts extraits du contenu à des fins non

commerciales est autorisé, à condition de faire une mention de source appropriée reconnaissant les droits d'auteurs de la Couronne. Toute reproduction importante du contenu ou toute utilisation, en tout ou en partie, du contenu à des fins commerciales est interdite sans l'autorisation écrite préalable du MDNM. Une reproduction jugée importante comprend la reproduction de toute illustration ou figure comme les graphiques, les diagrammes, les cartes, etc. L'utilisation commerciale comprend la distribution du contenu à des fins commerciales, la reproduction de copies multiples du contenu à des fins commerciales ou non, l'utilisation du contenu dans des publications commerciales et la création de produits à valeur ajoutée à l'aide du contenu.

Renseignements :

P

OUR PLUS DE

RENSEIGNEMENTS SUR la reproduction du contenu l'achat des publications du MDNM les droits d'auteurs de la Couronne

VEUILLEZ VOUS

ADRESSER À

:

Services de publication du MDNM

Vente de publications du MDNM

Imprimeur de la

Reine

P

AR TÉLÉPHONE

:

Local : (705) 670-5691

Numéro sans frais : 1 888 415-9845, poste 5691 (au Canada et aux

États-Unis)

Local : (705) 670-5691

Numéro sans frais : 1 888 415-9845, poste 5691 (au Canada et aux

États-Unis)

Local : 416 326-2678

Numéro sans frais : 1 800 668-9938

(au Canada et aux

États-Unis)

P

AR COURRIEL

:

[email protected]

[email protected]

[email protected]

l

Ontario

Ministry of

Northern Development and Mines

The Analysis of Geological Materials

Volume II: A Manual of Methods

Ontario Geological Survey

Miscellaneous Paper 149

Staff of the Geoscience Laboratories

Ontario Geological Survey

1990

CO 1990 Queen's Printer for Ontario 155IN

ISBN 0-7729-7036-X (v.II)

ISBN 0-7729-7034-3 (set)

Publications of the Ontario Geological Survey and the Ministry of Northern Develop ment and Mines are available from the following sources. Orders for publications should be accompanied by cheque or money order payable to the Treasurer of Ontario.

Reports, maps and price lists (personal shopping or mail order):

Public Information Centre, Ministry of Natural Resources

Room 1640, Whitney Block, Queen's Park

Toronto, Ontario M7A 1W3

(416) 965-1348 Collect calls accepted

Reports and accompanying maps only (personal shopping):

Publications Ontario

Main Roor, 880 Bay Street

Toronto, Ontario M7A 1N8

Canadian Cataloguing in Publication Data

Main entry under title:

The Analysis of geological materials

(Ontario Geological Survey, miscellaneous paper, ISSN 0704-2572 ; 149)

Contents: Vol. L A practical guide l A- Vander Voet and C. Riddle v. II. A manual of methods l staff of the Geoscience Laboratories.

ISBN 0-7729-7034-3 (2 v. set). - ISBN 0-7729-7035-1 (v. I). - ISBN 0-7729-7036-X (v. II).

1. Rocks-Analysis-Laboratory manuals. I. Vandervoet, A. H. M., 1945-. II. Riddle, C.

III. Ontario. Mines and Minerals Division. IV/Ohtario Geological Survey. V. Series."

QE433.A52 1990 552'06 C90-092526-4

Every possible effort is made to ensure the accuracy of the information contained in this report, but the Ministry of Northern Development and Mines does not assume any liability for errors that may occur. Source references are included in the report and us ers may wish to verify critical information.

If you wish to reproduce any of the texj, tables or illustrations in this report, please write for permission to the Director, Ontario Geological Survey, Ministry of Northern

Development and Mines, 11th Floor, 77 Grenville Street, Toronto, Ontario M7A l W4.

Parts of this publication may be quoted if credit is given. It is recommended that refer ence be made in the following form:

Ontario Geological Survey 1990. The analysis of geological materials, volume II: a manual of methods; Ontario Geological Survey, Miscellaneous Paper 149.

Commerce-250-90

TABLE OF CONTENTS

MINERAL SCIENCE SECTION

Method/Technique Page Header Section

The Technique of Sample Preparation

Assay Sample Preparation

Whole-rock Sample Preparation

Carbonate and Trace Element Preparation for Soils

Conodont Separation and Preparation

Grain-Size Analysis

Introduction Grain-size Analysis ................. .MS4-1

Conventional ASTM D422-72 Grain-Size Analysis Grain-size Analysis-1 ............... .MS4-2

Grain-Size Analysis by Particle Sizing Grain-size Analysis-2 ............... .MS4-10

Atterberg Limits (ASTM D423-66) Atterberg ......................... .MS5-1

Separation of Minerals Using Heavy Liquids

Mineral Separation Using the Frantz

Magnetic Separator

Rock and Mineral Identification for

Prospectors and Geologists

Optical Mineralogy

Sample Preparation ................ .MSI-l

Assay Preparation ................. .MS l-2

Whole-rock Preparation .....-........ .MS l-4

Carbonate Preparation ............. .MS2-1

Conodont Preparation .......,...... .MSS-1

Heavy Liquid Separation ............ .MS6-1

Magnetic Separation ............... .MS7-1

Mineral Identification ............... .MSS-1

X-ray Powder Diffraction

X-ray Diffraction Identification

Determination of Specific Gravity

Optical Mineralogy ................. .MS9-1

XRD ............................. .M10-1

Clay XRD ........................ .MSI 1-1

Specific Gravity ................... .MS 12-1

Determination of Moisture Content of Soils

(ASTMD2216-71)

Soil Moisture ...................... .MS 13-1

Fire Assay

Introduction

Conventional Fire Assay Method for Au and Ag

Fire Assay ........................ .MS14-1

Fire Assay-1 ...................... .MS 14-2

Fire Assay Preconcentration for Determination of Fire Assay-2 ...................... .MS 14-3

Au,Pt,PdbyGFAAS

Nickel Sulphide Fire Assay Fire Assay-3 ...................... .MS 14-4

Normative Programs Norms ........................... .MS15-1

Alknorm for Feldspathoidal and Melilitic Igneous Rocks Alknorm .......................... .MS16-1

ELEMENTAL ANALYSIS SECTION

Method/Technique

Protocols for Sample Digestion and

Analytical Determination

Sample Dissolution

Graphite Furnace Operating System

Operation of the Atomic Absorption Spectrometer

Liquid Handling System

Determination of Major Elements by XRF

The Philips PW14OO Fluorescence Spectrometer

System

Determination of the Major Elements

(M1,M2,M3)

X-ray Fluorescence Spectroscopy

Appendix A: LOI Corrections

Appendix B: Method Development

Appendix C: Setting Up an Analytical Package

Determination of Major Elements

(SiO2, A12O3, Fe203 , MgO, CaO)

Flame Atomic Absorption Spectroscopy

Determination of Major Elements

(SiO2, A12O3, CaO, MgO)

Classical (Gravimetric) Method

Determination of Titanium as (TiO2)

Colorimetric Method

Determination of Phosphorus (P2O5)

Colorimetric Method

Determination of Sodium and Potassium

Flame Photometric Method

Determination of Water

Combustion/Infrared Absorption Method

Determination of Carbon and Sulphur

Combustion/Infrared Absorption Method

Determination of Total Fe

Volumetric (Titrimetric) Method

Determination of Ferrous Iron

Titrimetric Method

Page Header Section

Protocols ................... ...... .EA1-1

Sample Dissolution ................. .EA2-1

Graphite Furnace .................. .EA3-1

VarianAA775 ................... .. .EA4-1

Liquid Handling System ............. .EA5-1

Major - XRF ...................... .EA6-1

Majors-XRF ..................... .EA6-6

Majors-XRF ........... .......... .EA6-14

Majors-XRF . .... .... .. .. ........ .EA6-16

Majors - XRF ..... ........ .. .. .. .. .EA6-21

Majors - AA .,... ... ... . y .. .. .. .. . .EA7-1

Majors - Classical ... .. ............ .EA8-1

Titanium ......................... .EA9-1

Phosphorus . .. ...... .... .. .. .. .. .. .EA10-1

Sodium/Potassium ....... .... ...... .EA11-1

Water. .... .. .. .. .. ...... ...... ... .EA12-1

C'orboni'Sulphur ... .. ...... ........ .EA13-1

Ferrous ..... .. ................... .EA15-1

IV

Method/Technique

Page Header Section

Determination of Trace Elements (T3)

X-ray Fluorescence Spectroscopy

Appendix A: Trace Element Determinations -

Measurement Programs

Measurement Program for Gallium

Measurement Program for Tantalum

Measurement Program for Ce, La, and Nd

Measurement Program for Cesium

Appendix B: Mass Absorption Calculation Program for Apple Ile

Appendix C: Mass Absorption Values for Selected

Silicates SRM's

Appendix D: MA Factors (Sr Ka Wavelength)

Appendix E: XRF Detection Limit Calculation

Appendix F: Rate Correction Parameters

Traces-XRF ...................... .EA16-1

Traces -XRF ...................... .EA16-4

Traces -XRF ...................... .EA 16-7

Traces-XRF ...................... .EA16-9

Traces - XRF ...................... .EA16-11

Traces - XRF ... ................... .EA16-13

Traces-XRF ...................... .EA16-14

Traces-XRF ... .. ................. JEA16-18

Traces - XRF ............. ......... J5A16-19

Traces - XRF ............. ......... .EA16-20

Traces - XRF ...................... .EA16-21

Traces -AA ....................... .EA17-1 Determination of Trace Elements

(Co, Cu, Cr, Ni, Pb,Zn, Ba, Li, Ag, Cd, Mn, Fe)

Atomic Absorption Spectroscometry

Determination of Trace Elements by ICP-OES

Overview of Trace Element Determination by

ICP Optical Emission Spectroscopy

The Trace 2 (T2) Package

Appendix A: Generation of T2 Final Reports

Aqueous Samples Analysis Package

Tentative Spectroscope Analysis Package (TSPA)

Appendix A: Generation of the TSPA Certificates

Determinatio of Trace Elements by ICP-MS

Sample Preparation for the T4 and T5

ICP-MS Package

The Elan-250

Trace 4 (T4) Elements

Trace 5 (T5) Elements

Determination of the Rare Earth Elements

(La, Ce, Nd, Sm, Eu, Yb, Lu) Thorium, Tantalum,

Hafnium, Uranium, and Scandium

Instrumental Neutron Activation Analysis

Determination of Gold, Platinum and Palladium

Graphite Furnace Atomic Absorption Spectroscopy

Determination of Gold in Natural Waters

Solvent Extraction and Electrochemical

Atomization Method

Determination of Fluoride

Automated Colorimetric Method

Traces - ICP-OES .................. .EA18-1

Traces - ICP-OES .................. .EA18-14

Traces - ICP-OES ......... ...... ... .EA18-21

Traces - ICP-OES . ................. .EA18-24

Traces - ICP-OES . ................. .EA18-25

Traces - ICP-OES ..... ........ .. .. . .EA18-26

Traces ICP-MS ....... ............. JBA19-1

Traces - ICP-MS ................... .EA 19-5

Traces - ICP-MS ..... ........ .. .. .. .EA19-13

Traces - ICP-MS ................... .EA19-18

Traces - NAA ........... .......... .EA20-1

Au,Pt,Pd ........................ .EA21-1

Gold .......................... .. .EA22-1

Flouride ......................... .EA23-1

Method/Technique Page Header Section

Determination of Chloride

Automated Colorimetric/Flow Injection Analysis Method

Chloride ..................... .... .EA24-1

Determination of Arsenic, Antimony and Bismuth

Hybrid Generation - Atomic Absorption Spectrometry

Determination of Selenium

Hybrid AAS - Automated Flow Injection Analysis

Method

Determination of Mercury Cold Vapor

Flameless Atomic Absorption Method

As,Sb,Bi......................... .EA25-1

Selenium ........... ........ ...... .EA26-1

Mercury ............. ............. .EA27-1

Carbonate Carbon ................. .EA28-1 Determination of Carbonate-Carbon

Coulometric Method

Determination of Moisture

Determination of Acid Insoluable Residue

Determination of the Calcite to Dolomite Ratio

Chittick Apparatus Method

Moisture ......................... .EA29-1

Acid ............................. .EA30-1

Chittick ...........................EA31-1

Sample Preparation

THE TECHNIQUE OF SAMPLE PREPARATION_____________

The preparation of samples represents the single most important step in the analysis of geological materials.

Undetected errors in the handling of rocks such as: sample mixup, contamination with dust or rock frag ments, and introduction of components from the grind ing equipment can lead to erroneous conclusions in the interpretation of the data.

The Laboratories uses two sample preparation proce dures on a routine basis:

1. Preparation of samples for assay work, and

2. Preparation of samples for whole-rock geochemi cal analysis.

MS1-1

assay r reparation

ASSAY SAMPLE PREPARATION

Introduction:

2. Crushing the samples sequentially using a jaw crusher

Assay preparation is employed whenever a geologist or prospector requires an indication of base or precious metal potential. Assay samples are given high priority and a rapid turnaround. The assay preparation should not be used when geologists require detailed geochemical work.

3. Riffling to split the samples - retaining one split for grinding and another split as a back-up

4. Grinding the split in the mill on the vibratory ring-pulverizer for the appropriate length of time.

Safety advisory:

When working in the grinding area:

6. Be aware of samples requiring special handling.

These samples will be flagged, e.g. As/Pb contain ing. If you identify a potential hazard then it is your responsibility to flag the samples.

5. Removing pulp, after grinding, onto glazed paper and transferring to pulp bags.

General comments

1. Wear a full face-piece respirator unit to eliminate the possibility of silicosis; THIS LABORATORY

AREA IS MONITORED UNDER A SILICA

CONTROL PROGRAM.

Ensure that the mill is cleaned before and after use with compressed air in the blow-out area. When com pressed air will not remove the dust, use quartzite as a cleaner. The grinding area should be kept clean by using the central vacuum unit.

2. Wear a face shield or safety goggles.

3. Wear ear muffs. Sound levels from the jaw crushers can exceed 100 dB.

Apparatus:

4. Wear a laboratory coat at all times in the area, and take it off before leaving.

5. Do not eat, smoke, or drink in the grinding area.

- Containers for laying out the samples on the bench-top in a systematic fashion

- Large jaw-crusher with steel plates

- Riffle to split samples

- Vibratory ring-pulverizers equipped with two chrome-steel mills (a tungsten carbide and ceramic mill can be used if specifically re quested). When large quantities of pulp are re quired, two large chrome-steel mills are available.

7. Ensure that the main exhaust system is switched on during grinding operations.

Reagents:

- Quartzite

8. Observe safety procedures specific to each piece of equipment. Ensure that the lids on the jaw crushers are closed immediately after the sample is introduced. The mills on the vibratory ring-pul verizer (also called a Swing-Miller) should be firmly secured before the equipment is switched on.

Procedures:

1. Samples are collected from the sample reception area along with a series of labels and a list of samples. Any special instructions accompanying samples should be clarified with the sample recep tion technician.

9. Uncontrolled cleaning of equipment and clothes using compressed air is not acceptable.

Technique:

2. Labels are affixed to bags and/or containers used to store crushed rock and final pulp.

Assay sample preparation involves the following pro cedures:

3. Samples and bags are placed in previously cleaned containers in ascending order of sample number.

Any materials affixed to the samples, except mark ings, are removed.

l. Sorting of samples in preparation for grinding

MS 1-2

Assay Preparation

4. Wet samples are dried in a hot-air drier.

5. The exhaust fan is switched on, and the gate to the jaw crusher is opened. The jaw crusher is cleaned using compressed air and a wire brush.

Quality Control:

When the mill cannot be cleaned with compressed air, quartzite is used.

6. Observing necessary safety precautions, the rock sample is crushed.

Assay grinding is used in the preparation of samples for geologists and prospectors who do not require whole-rock geochemical analysis.

7. The crushed sample is split in the vented riffler.

8. Reject material is retained in a pulp bag.

9. About 100 g of crushed rock is placed in a cleaned chrome-steel mill. The mill is secured to the vibratory ring-pulverizer, and ground for an ap propriate length of time (depending on hardness).

Typically, 90 seconds is sufficient.

A record is kept of the sequence in which the samples are ground, and of the identity number of the container used for each sample. This record is of use if a high grade sample is identified in the analytical stage; it allows contamination of subsequent samples to be checked.

Productivity:

10. The mill is transferred to the blow-out area, and the rock powder is collected on glazed paper, and transfered to a previously labelled pulp bag.

Typically, a technician can complete forty-five assay grinds per day. One hundred assay grinds are possible under ideal sample conditions.

General comment. Use two chrome-steel mills so that one is grinding while the other is being cleaned with compressed air.

MS 1-3

whole-rock Preparation

WHOLE-ROCK SAMPLE PREPARATION

Introduction:

This preparation technique is applied whenever the geologist requires detailed whole-rock geochemical analysis. Samples submitted in batches for whole-rock analysis are usually given a job number.

The Braun pulverizer equipped with ceramic plates is sometimes used for coarse-grained jaw crushed rock to improve mineral liberation for mineral separation procedures.

Reagents:

Safety advisory:

Refer to the safety advisory as detailed in the Assay

Preparation method (Page Ml-1).

- Acetone

- Quartzite

Procedures:

Technique:

Whole-rock sample preparation involves the following procedures:

1. Samples are collected from the sample reception area along with a series of labels and a copy of the job sheet. Any special instructions accompanying samples should be clarified with the sample recep tion technician.

1. Sorting of samples in preparation for grinding

2. Determining specific gravity for each sample

3. Crushing the samples sequentially using a jaw crusher (retaining a hand sample)

2. Labels are affixed to containers used to store crushed rock and final pulp. The sample prepara tion technician assigns a laboratory number in addition to the sample number, which determines the sequence in which the samples are ground.

4. Riffling to split sample - retaining one split for grinding and another split as a back-up

3. Hand samples of the rocks are retained in the original sample bags.

5. Grinding the split in a mill on a planetary ball mill for an appropriate length of time

6. Sieving the pulp and re-grinding the material (^

170 mesh)

7. Removing pulp, after grinding, onto glazed paper and transferring to labelled bottle

General comment. Ensure that equipment is cleaned before and after use with quartzite as a cleaner, and compressed air in the blow-out area. The grinding area should be kept clean by using the central vacuum unit.

4. The remainder of the samples are placed in pre viously cleaned containers in ascending order of sample number. Any materials affixed to the samples including tape, enamel, lichen and paint are removed. NOTE: If the samples are weathered or altered, the technician should inform the sample reception technician, who will contact the geologist to ascertain whether the weathered material should be removed. It is up to the

geologist to remove weathered surfaces and other markings.

5. Wet samples are dried in a hot-air drier.

Apparatus:

6. Specific gravity is determined (refer to the method of Determining Specific Gravity on page M12-1).

- Containers for laying out the samples on the

bench-top in a systematic fashion

- Balances for the determination of specific gravity

- Small jaw-crusher with steel plates

- Riffle to split samples

- Planetary ball mill (99.89fc pure alumina)

- 170 mesh nylon sieve

- Central vacuum unit

7. The exhaust fan is switched on, and the gate to the jaw crusher is opened. The jaw crusher is cleaned using compressed air and a wire brush.

8. Observing necessary safety precautions, the rock sample is crushed.

9. Crushed sample is split in the vented riffler.

MS 1-4

10. Reject material is retained in a labelled pulp bag.

11. 100 g of crushed rock are placed in a cleaned ceramic mill along with an appropriate number of grinding balls (normally about five 20 mm, or two

30 mm and three 20 mm balls). Two or four mills are bolted onto the turntable of the planetary ball mill, the lid is closed, and the machine is switched on for a period of upto 15 minutes (depending on sample hardness).

12. The powder and ceramic balls are removed from the mill and screened through a 170 mesh nylon sieve onto a piece of glazed paper. Any material which is -i- 1 70 is re-ground, until the entire sample passes a 1 70 mesh. The agate mortar and pestle are used when the sample is too small for re-grinding

(i.e.,

13. The rock powder is rolled on the glazed paper and transferred to a labelled bottle.

Quality Control:

The mills are cleaned with quartzite and compressed air.

Under an in-house quality control program, three samples in selected jobs are ground in duplicate.

A record is kept of the sequence in which the samples are ground, and of the identity number of the container used for each sample. If a high grade sample is iden tified during analysis, subsequent samples prepared in the same apparatus are be checked for contamination.

Productivity:

A technician should be able to complete thirty grinds

(samples) per day.

Whole-rock Preparation

MS 1-5

Carbonate Preparation

CARBONATE AND TRACE ELEMENT PREPARATION FOR SOILS

Introduction:

This sample preparation procedure is used to prepare soil and regolith samples for Chittick analysis, major analysis and/or trace element analysis as described in

The Elemental Analysis Section of this manual.

Procedure:

1.1. The dry sample is hand broken to pass No. l O mesh sieve. Special care is taken not to crush any larger particles in the soil sample. The amount of sample dried depends upon the grain-size.

1.2 . Soil passing the No. 10 sieve is thoroughly mixed until homogeneous. Portions of this mixture are sieved through No. 80 and No.

200 sieves. Material passing No. 200 is kept for carbonate analysis (12-15 g).

1.3. Another portion i s sieved through number 120 and 400 sieves. Material passing No. 400 is kept for trace element analysis (12-15 g).

Sieves No. 80 and 120 are used to protect the fine mesh of the No. 200 and 400 sieves.

MS2-1

Conodont Preparation

CONODONT SEPARATION AND PREPARATION

Introduction:

Conodonts are microfossils. They are used to establish the age of the strata in which they are found, and to correlate rocks in different areas.

Procedure:

1.1.

Crumble the sample gently between two layers of clean paper.

1.2.

1.3.

1.6.

Transfer the sample to a 3 or 4 li ter beaker and cover with water for 2-3 days. Decant.

Add commercial H2O2 to just cover the sample.

1.7.

1.4.

1.5.

Add hot water to half-fill the beaker if the sample is still coherent, otherwise proceed to 1.8. step 1.5.

Add 6 g of "Calgon" and 6 g of bicarbonate of l .9. soda to the beaker containing the sample and water. This causes an exothermic reaction in which clay is dissolved and dispersed. Cold water may be added to lessen the reaction.

After 30 minutes, dispersal is complete and the residue may be washed and sieved. If pyrite is present, the production of sulphuric acid is avoided by adding 6 g of sodium bicarbonate and then boiling on a hotplate until disintegra tion occurs.

Wash the residue through 3 sieves. An 18 mesh sieve is used to collect larger microfos sils; a 85 and 170 mesh sieve is used to collect smaller microfossils.

If any undisintegrated rock fragments remain on the 18 mesh sieve, repeat steps 1.5-1.6.

Wash the residue from each sieve into filter papers in funnels, and dry the residue.

Transfer the residue to dry bottles for sub sequent study.

MS3-1

Grain-size Analysis

GRAIN-SIZE ANALYSIS

Introduction:

Detailed knowledge of the physical characteristics and structural properties of soils is essential in helping engineers determine soil behavior and performance in a variety of construction conditions. Grain-size analysis is widely used in the engineering classifica tion of soils. It is useful in assessing soil permeability and capillarity which are important considerations in predicting soil behavior in cold climates.

An important use, particularily with coarse soils, is in designing inverted fillers for dams and levees. Grain- size analysis is also used in the study of sedimentary rocks as grain-size distributions are a function of the processes forming the sediment.

Grain-size analysis may be performed by either:

1. A conventional sieve-hydrometry (ASTM D422-

72) technique.

2. The use of a limited number of sieves, a particle sizing system, and a clay fraction determination by hydrometry.

The latter procedure is not well suited to samples with high clay content (greater than 25^o less than 2mu in size).

Table GSA1 shows the relationship between sieve number and particle size.

TABLE GSA1. PARTICLE SIZE TABLE

AND SIEVE INFORMATION

U.S. Mesh

4

8

5

6

7

10

12

14

16

18

20

25

30

35

40

45

80

100

120

140

170

200

230

270

325

400

500

Inches

0.187

0.157

0.132

0.111

0.0937

0.0787

0.0661

0.0555

0.0469

0.0394

0.0331

0.0280

0.0232

0.0197

0.0165

0.0138

0.0070

0.0059

0.0049

0.0041

0.0035

0.0029

0.0024

0.0021

0.0017

0.0015

0.0014

Microns

62

53

44

37

31

590

500

420

350

177

149

125

105

88

74

4760

4000

3360

2830

2380

2000

1680

1410

1190

1000

840

710

Millimeters

0.59

0.50

0.42

0.35

0.177

0.149

0.125

0.105

0.088

0.074

0.062

0.053

0.044

0.037

0.031

4.76

4.00

3.36

2.83

2.38

2.00

1.68

1.41

1.19

1.00

0.84

0.71

Safety advisory:

1. When sieving samples wear ear muffs, a mask and eye protection.

2. The Hiac Royco system works on the basis of laser sizing. When removing the laser sizer for cleaning, ensure the power is turned off to avoid accidental exposure to laser radiation.

MS4-1

Lrrain-size s\naiysis-i

METHOD l

CONVENTIONAL ASTM D422-72 GRAIN-SIZE ANALYSIS

Method:

A conventional grain-size analysis involves the fol lowing techniques: weighing 50 g (for sandy material), or 100 g

(for silty or clay soils) is retained in a 250 ml capped flint glass jar. The remaining -10 mesh material is retained, and may be required for other tests.

1. Sample preparation - drying sample and dispers ing clumps

2. Separating the various size fractions - sieving and hydrometry

1.4. 125 ml of the deflocculant solution (see

Reagents section for preparation) is added to the sample, and is allowed to soak into the sample overnight. The sample is then trans ferred to a mixing container, and mixed with the jet mixer for l minute.

3. Calculating results based on measured weights

Apparatus:

- Sieves (2000,850,425,250,125 and 63 microns)

- Soil breaker

- Sedimentation cylinders (46 cm high, 6.4 cm wide)

- Thermometers (accurate to 0.50C)

- Hydrometer (ASTM standard hydrometer, graduated to read in either specific gravity of the suspension or grams per liter of suspension)

- Jet mixer (with electric motor producing

10,OOORPM)

1.5. The sample is washed into a graduated cylinder which is then filled up to the level

Reagents:

- Sodium hexametaphosphate (deflocculating agent used to break the charge attraction of clay particles). A solution of this chemical is made by dissolving 40 g of sodium hexametaphosphate in l liter of distilled water

1.7.

line with distilled water. Before taking any readings, each cylinder is covered with a rub ber stopper and systematically inverted for a period of l minute to completely mix the soil suspension with the water.

1.6. At the end of l minute, the cylinders are set down, and hydrometry readings are taken at the following time intervals (measured from the beginning of sedimentation): 2,5, 15,30,

60 and 240 (4 hrs), and 1440 (24 hrs) minutes.

The hydrometer scale would normally be read at the base of the meniscus, but the soil suspen sion makes this impossible, so the reading is taken at the top of the meniscus.

The temperature of the suspension is checked with a thermometer.

Procedure:

1. Sample preparation

1.1. The soil sample is collected from the sample reception technician along with a job sheet.

The sample bag (normally of paper construc tion) is opened and placed into the drying cupboards for a period of at least one week.

1.8.

After the final reading has been taken, the sample is transferred to a No. 230 (62 micron) sieve and washed with tap water until the wash water is clear. Material left on the sieve is then dried in an oven at 1100C. This portion is then sieved. For clay samples, the following sieves are used: No. 10, 20, 40, 120, and 230. For sands, the following sieves are used: No. 10,

14, 18, 25, 35,45, 60, 80, 120, 170 and 230.

l .2. Dry samples are removed from the driers and systematically broken using a mechanical soil breaker, and a mortar with a rubber pestle to avoid breaking individual particles.

1.9. The portion retained on each of the sieves is weighed using a balance, and the percentage passing each sieve is recorded.

1.3. The sample is then passed through a number

10 (2000 micron) sieve, and a sub-sample

1.10. Calculations are made using the results from the above procedure.

MS4-2

Grain-size Analysis-1

2. Calculation of results

For an ASTM hydrometer (152H), the calculation reads in grams of soil per liter of suspension.

The results are presented graphically by plotting the diameters of the particles on a horizontal logarithmic scale and the cumulative percentage smaller than a given size on the vertical axis (Figure GSA1).

2.1.

Percentage retained on any sieve =

Weight of soil retained

~, , . , r

. i

Total weight of soil

A *-W-' ' C

The final report should include: l . The maximum particle size

2. The percentage passing each sieve

2.2.

Percentage finer than any sieve size =

100*7c - Cumulative percentage retained

3. A description of the sand and gravel shape and hardness

2.3.

Percentage of soil remaining in suspension at the level at which the hydrometer measures the density of the suspension =

RxA

W

4. Specific gravity

5 . Any problems in dispersing the 1 000 micron frac tion

Quality control:

where, R = corrected hydrometer reading

A = correction factor applied to the readingof hydrometer 152H

(Refer to Table GSA4)

W = Oven-dry weight of soil dispersed and used for hydrometry readings (g)

Analysis of the in-house Geoscience Laboratories soil standard should accompany each job. Hydrometry should be performed on a 50 g sample of the standard which can be reused.

2.4. Diameter of soil particles (mm) =

The gradation curve for this material is shown in

Figure GSA l based on data presented in Table GSA5; and should be reproduced with a precision of 5^c or better.

where, K =

L =

T =

A constant depending on the temper ature of the suspension and the specific gravity of the soil particles

(Refer to Table GSA2)

Distance from the surface of the sus pension to the level at which the density of the suspension is meas ured (cm) (Refer to Table GSA3)

Interval of time from beginning of sedimentation to the taking of the reading (min.)

MS4-3

TABLE GSA2.

VALUES OF K FOR USE IN FORMULA FOR COMPUTING DIAMETER

OF PARTICLE IN HYDROMETER ANALYSIS

24

25

26

27

28

29

30

19

20

21

22

23

16

17

18

Temp.

0C

2.45

0.01531

0.01511

0.01492

0.01474

0.01456

0.01438

0.01421

0.01404

0.01388

0.01372

0.01357

0.01342

0.01327

0.01312

0.01298

2.50

0.01505

0.01486

0.01467

0.01449

0.01431

0.01414

0.01397

0.01381

0.01365

0.01349

0.01334

0.01319

0.01304

0.01290

0.01276

2.55

0.01481

0.01462

0.01443

0.01425

0.01408

0.01391

0.01374

0.01358

0.01342

0.01327

0.01312

0.01297

0.01283

0.01269

0.01256

Specific Gravity of Soil Particles

2.60

2.65

2.70

0.01457

0.01439

0.01421

0.01403

0.01386

0.01369

0.01353

0.01337

0.01321

0.01306

0.01291

0.01227

0.01264

0.01249

0.01236

0.01435

0.01417

0.01399

0.01382

0.01365

0.01348

0.01332

0.01317

0.01301

0.01286

0.01272

0.01258

0.01244

0.01230

0.01217

0.01414

0.01396

0.01378

0.01361

0.01344

0.01328

0.01312

0.01297

0.01282

0.01267

0.01253

0.01239

0.01225

0.01212

0.01199

2.75

0.01394

0.01376

0.01359

0.01342

0.01325

0.01309

0.01294

0.01279

0.01264

0.01249

0.01235

0.01221

0.01208

0.01195

0.01182

2.80

0.01374

0.01356

0.01339

0.01323

0.01307

0.01391

0.01276

0.01261

0.01246

0.01232

0.01218

0.01204

0.01191

0.01178

0.01165

2.85

0.01356

0.01338

0.01321

0.01305

0.01239

0.01273

0.01258

0.01243

0.01229

0.01215

0.01201

0.01188

0.01175

0.01162

0.01149

MS4-4

19

20

21

22

23

24

25

26

27

28

29

13

14

15

16

17

18

30

31

32

33

0

1

2

3

4

9

10

11

12

8

7

5

6

TABLE GSA3. EFFECTIVE DEPTH BASED

ON HYDROMETER AND ON

CYLINDER OF SPECIFIED

SIZES.

HYDROMETR 152H

Hydrometer

Reading

Effective

L (cm)

Hydrometer

Reading

Effective

L (cm)

14.2

14.0

13.8

13.7

13.5

13.3

13.2

13.0

12.9

12.7

12.5

12.4

12.2

12.0

16.3

16.1

16.0

15.8

15.6

15.5

15.3

15.2

15.0

14.8

14.7

14.5

14.3

11.9

11.7

11.5

11.4

11.2

11.1

10.9

7.3

7.1

7.0

6.8

6.6

6.5

6.4

8.3

8.1

7.9

7.8

7.6

7.4

10.7

10.6

10.4

10.2

10.1

9.9

9.7

9.6

9.4

9.2

9.1

8.9

8.8

8.6

8.4

6.3

6.1

5.9

5.8

5.6

5.5

47

48

49

50

51

52

53

54

34

35

40

41

42

43

36

37

38

39

44

45

46

63

64

65

66

67

55

56

57

58

59

60

61

62

Grain-size Analysis-1

MS4-5

Lrram-size anaiysis-i

01

1*1

00

50

CD

S* o so

ScS

f^U

(A

-D

^ co tp,

-o

CO

(u

S

U)

v\

•C:

r1

o

: o

?

-O

1

^8 v/s -C.

by.

P en o

OQ u

In

t?

53

O

^

K)

S

O)

od rs o y

H

oQ

oQ

oO

CO

OQ

-w

Ui

(T*

(T*

X)

KC

8

rO

^

hM o

X]

JO

UJ

^ o a

Ci

^

8

u

M

S 3

* 0

Figure GSA5.

MS4-6

Grain-size Analysis-1

10

100

C*SE

Grain size

1000

10000

Figure GSA1. Grain-size analysis.: comparison of CLOGS (case 2) and BARRINGER (case 1).

MS4-7

Lrram-size Analysis-1

S A P L E

Hydrometry rnksrons

50-1000

i

B.

Calculate

K

+14,

+18 and +2S

Dry and black sieve to -10

—— i — a—

Sieve:

+H+18.+25 and +230 and weigh.

Collect -230

c.

Calculate *fc cont ribution from each channel using weight of

+230, -25

I

Compile data into one grain-size distribution

D.

Calculate % ributlon from each channel using dry weights and microns

l

Figure GSA2. Flow chart for grain-size distribution calculation.

MS4-8

-size /inaiysis-z

METHOD 2

GRAIN-SIZE ANALYSIS BY PARTICLE SIZING

This method involves the use of both modified ASTM methods and the Hiac Ryco particle sizing system. It is best applied to the determination of particle sizes for samples which are sandy, or where specific informa tion is required about particle sizes between 2 and 700 microns. Automated grain-size analysis is not suited for routine application as a substitute for ASTM tech niques for clay-rich samples.

than 2.25 microns is determined using the particle size analysis instrument equipped with a wet sensor. The particle size distribution in the fraction finer than 2.25 microns is determined by conventional hydrometry.

1.1. Upon receipt, samples are place in drying cupboards for a week.

Method:

l .2. The sample is broken in a soil breaker, rolled, and about 50 g of the -10 sieve fraction is retained for analysis.

The method consists of the following techniques:

1. Sample preparation - samples are dried and clumps of particles are broken down using a soil breaker

2. Separation of soil particles - the sample is sieved to -10 mesh, and wet sieved in the presence of a deflocculant (See previous ASTM D422-72 method) using a 14, 18, 25, and 230 sieve.

1.3. A 2 g sub-sample consisting of a repre sentative fraction of the -10 mesh fraction is weighed and placed in a 250 ml jar, and 20 ml of a dispersant is added (a solution of sodium hexamethaphosphate in distilled water - 40 g per liter).

3. Calculation of results - a program is used to com pile the data into an ASCII MS-DOS file, which is then reported to the geologist.

Apparatus:

- Sieves

- Soil breaker

- Mortar and rubber pestle

- Grain-size analysis system

- Hydrometer cylinders and hydrometer

- Thermometer

1.4. The -10 mesh fraction in the dispersant is wet-sieved through -14, -18, -25, and -230 sieves. Care is taken to wash any fine material through the sieve using a small amount of distilled water. No more than 200 ml of dis tilled water should be used. Particles resting on the four sieves (+14, + 18, -f-25, and +230) are washed onto filter papers, dried in the oven at 1100C, and then weighed to 0.001 g.

The -230 mesh particulates in the dispersant are poured into the original 250 ml jar, and any residual particles are washed out. The suspen sion is made up to 250 ml.

Reagents:

- Sodium hexametaphosphate (See previous

ASTM D422-72 method for preparation instruc tions - Page MS4-3)

1.5. As outlined below the -1-230 fraction is analyzed using the dry sensor on the particle size analysis system, while the -230 fraction is analyzed using the wet sensor.

2. Calibration of the system

Procedure:

1. Preparation of sample

2.1. The main control unit is switched on at the back and allowed to warm up for half an hour before calibration.

The distribution of panicle sizes coarser than

710 microns (retained on No. 25 sieve) is determined by wet sieving. The distribution of particles coarser than 62 microns, but finer than 710 microns is deter mined using a particle size analysis system equipped with dry sensor. The particle size distribution in the fraction finer in grain size than 62 microns but coarser

2.2. A calibration must be established for the 32 channels of this system for both the wet and dry sensor. Once this calibration is completed, it can be used repeatedly, the only precaution routinely needed is to check that the channel thresholds have not drifted away from those values listed in the original calibration. When

MS4-10

Grain-size Analysis-2

switching from the wet to the dry sensing modes, it is necessary to completely change these settings. To view the thresholds of each channel, the special function key is used, and the number of the channel is entered on the keypad (e.g. 01 for channel number l, etc.).

2.3. The system is configured to use calibration

No. 3 for the dry sensor and No. 2 for the wet sensor. A linear relationship exists between the threshold voltage of the sensor and the size of the particle, from which the millivoltage corresponding to a particular size can be es tablished. These thresholds are kept constant for every analysis performed on the particular sensor.

3. Operation of the computer and processor

5. Analysis using the wet sensor

The wet sensor is used to determine the distribution of

-230 mesh particles in a solution containing disper- sant. As this instrument is equipped with a threshold sensor, the smallest particles which it can count are

2.25 microns in size. Specific attention is drawn to the following requirements:

5.1. The sensor must be cleaned with alcohol,

5.2. micro solution and water before operation.

Between samples, the sensor should be cleaned with distilled water.

The concentration limit of the sensor is de pendent on the grain-size of the sample. Very dilute solutions must be used. This can be achieved by taking l ml of the sample solution

(250 ml containing up to 2 g of sample), and making up to one liter.

3.1. The instrument runs under the control of an

IBM-PC AT clone equipped with an EGA card, monitor, math co-processor, and an extra serial port. The system runs a program called

PDAS directly from the DOS COMMAND shell. Operations are explained directly by means of help menus for each function selected. Run mode parameters are selected as each run is set-up.

4. Analysis using the dry sensor

A Hiac Royco (Pacific Instruments) particle size analysis system is used to determine the grain size distribution in the solid fraction (+230, -25). Detailed instructions regarding the calibration of the instrument are presented earlier in this section. Specific attention is drawn to the following points.

5.3. Analyses should be conducted in duplicate.

6. Particle size distribution data from the sensors

Data are stored at the end of each run in a job file; two files are maintained for each job - one for each sensor.

Once both wet and dry runs are completed, the ASCII translation option is invoked to create ASCII files from the PDAS data sets; this is done sequentially for each sample and standard. The ASCII files are saved on floppy disk, and the original PDAS data file is deleted from the system.

7. Analysis of the clay fraction by hydrometry

4.1.

4.2.

A sample consisting of about 200,000 par ticles is introduced into the feeder chute, such that no more than 500 particles pass the sensor in each second of operation. The distribution of particle sizes is registered in 32 channels with preset size intervals.

Care must be taken not to saturate or block the sensor. Saturation may result in chains of par ticles being counted as single large particles.

Too rapid throughput may cause the sensor to become blocked. Under no circumstances must particles larger than 25 mesh be intro duced into the sensor.

A sample of soil passing No. 10 sieve weighing 50 g

(for clay-rich soils) or 100 g (for sandy soils) is allowed to soak overnight in a solution of the dispersant

(125 ml). The soil and water are then transferred into a mixer for l minute, and washed into a graduated cylinder. The cylinder is filled to the 100 ml mark with distilled water. The covered cylinder is inverted repeatedly for l minute before commencing the read ings, and readings are then taken after l hr, 2 hrs, 3 hrs., and 4 hrs. The hydrometer scale should be read at the top of the meniscus. A working temperature of 230C is assumed, but if very different, then a correction is applied in the calculation (Table GSA2).

8. Grain Size Calculations

4.3. Each sample should be run in duplicate. The average of the data from each analysis is used.

The calculation of grain-size distribution follows the flow chart shown in Figure GSA2. The percentage of particles of size less than 2.25 microns is determined using the following calculation:

MS4-11

uram-size Analysis-Z

8. l . Percentage of soil in suspension

SS ~

W

where, SS =

R = a =

W = percentage of soil remaining in sus pension at the level at which the hydrometer measures the density of the suspension.

corrected hydrometer reading correction factor to be applied for the reading of the hydrometer

(See Table GSA4) oven-dry weight in grams of soil dis persed and used for the hydrometer readings

TABLE GSA4.

Specific Gravity

8.2.

(g/cc.)

2.95

2.90

2.85

2.80

2.75

2.70

2.65

2.60

2.55

2.50

2.45

CORRECTION FACTOR (a)

Correction Factor

Diameter of soil particles (D)

(a)

0.94

0.95

0.96

0.97

0.98

0.99

1.00

1.01

1.02

1.03

1.05

of percentage of soil in suspension against the diameter of the soil particles.

9. The percentage of particles with sizes between 7 1 0 and 2000 microns.

The weight percent of sample from the -i- 14, +18, and H-25 sieves can be used to calculate the weight percent contribution in each of these grain- size intervals (i.e A = wt. retained on No. 14/100, etc.).

10. The percentage of particles (PL^c) with sizes be tween 2.25 and 65 microns is determined by

100 where, PL =

W =

A,B,C,

D and percentage of particles being ana lysed by the wet sensor original weight of sample in grams prepared for grain-size analysis

(not hydrometry).

SL = percentage of sample retained on

No. 14, 18, 25, and 230 sieves, re spectively (i.e. weight/100), percentage of particles with grain sizes less than 2.25 microns (deter mined from hydrometry).

The contribution of particles in each grain-size interval determined by the wet sensor is then cal culated by multiplying the percentage in that chan nel (for each of the 32 channels) by the overall percentage of particles analyzed by this sensor

(PL).

1 1 . The percentage of particles with sizes between 65 and 710 microns.

where, D = diameter of particle (mm)

K = constant depending on the temper ature of the suspension and the specific gravity of the soil particles

(Table GSA2)

L = distance from the surface of the sus pension to the level at which the density of the suspension is known (cm).

T = interval of time from beginning of sedimentation to the taking of the reading (min.).

8.3. The percentage of particles of size 2.25 microns or less (SL) is calculated from a plot

This is determined by multiplying the count in each of the 32 channels by the percentage con tribution (by weight) made by the particles falling in this grain-size interval (e.g. D).

12. Recalculation of data

Hydrometry and sieve data for each sample are entered into an ASCII file using a GWBASIC program

(GRAIN.ANY). This information is then combined with the data from the ASCII files for the wet and dry sensors using a program which generates an ASCII file and report.

MS4-12

Quality Control:

Analysis of the in-house GLOGS soil standard should accompany each batch of samples. Hydrometry deter minations on the GLOGS standard are made without using stock material, as the amount of this standard used in a hydrometry determination would be 50 g.

The method produces a slightly different gradation curve compared to the conventional ASTM method.

Productivity:

A technician should be able to complete 40 samples per week with the automatic system.

Grain-size Analysis-2

MS4-13

Atterberg

ATTERBERG LIMITS (ASTM D423-66)

Introduction:

TABLE ATT1.

CONSTANTS USED IN THE

The Atterberg Limits represent the liquid and plastic limits of a soil, which in turn are used in the classifica tion.

No. Blows

The Atterberg limits are frequently used in specifica tions for controlling the type of soil used in a variety of engineering projects.

Procedure:

1.1. The soil sample is placed on a glass plate and

1.2. thoroughly mixed with distilled water. Part of the soil mixture is then placed in the Liquid

Limit device, which consists of a brass cup and carriage. With a spatula, soil is leveled and trimmed to a depth of l cm at the point of maximum thickness. With the use of a groov ing tool, the soil is divided along the diameter through the center line of the cup.

By turning the crank, the cup is lifted and dropped at a rate of two revolutions per second until the two halves of the soil cake come into contact at the bottom of the groove along a distance of 1.3 cm. The number of drops needed to close the groove for a distance of l .3 cm is recorded.

1.3. Part of the soil that flowed together is taken and placed in a container; the wet weight is recorded, and then the soil is oven dried to

1100C. The sample is re-weighed, and the dry weight is recorded. This procedure is dupli cated. The procedure is repeated until the sample acquires a consistency that will require

25(-i-7-10) drops to close the groove.

17

18

19

20

21

14

15

16

10

11

12

13

28

29

30

31

32

33

34

22

23

24

25

26

27

35

36

37

Quality Control:

DETERMINATION OF

ATTERBERG LIMITS

k

0.979

0.985

0.990

0.995

1.000

1.005

1.009

1.014

1.018

1.022

1.026

0.906

0.913

0.920

0.927

0.934

0.941

0.948

0.955

0.961

0.968

0.974

1.030

1.034

1.038

1.041

1.044

1.047

2. Calculations

LI = Wt. water x 100 x

Wt. dry soil

The GLOGS soil standard is analyzed on a regular basis and the results are compared with in-house ac cepted values for this standard (Table ATT2).

Productivity:

where, k =

LI =

PI s constant depending on the number of blows (See Table ATT1) liquid limit.

Wt. water x

100

Wt. dry soil

A technician should be able to complete 40 samples per day for non-clay material and 10 samples per day for clay-rich soils.

where, PI = plastic limit.

MS5-1

Atterberg

E w a

Q -n m o n n H n

cm 11 a x o OJ

w ji

O m m a o CD >

1

O

Limit

•a P

H-

a

H

(D

0

• ^

^^|

H

Oi

9 ft h

(D

5

H- tt ft

D

S

CO

0

H

**

O

Bf

3

S!

ft s

CO

0

H

ft

O

Q|

3 o

O

V ol

S n

(A H

2. -n

^ P"^

M

Ul

* ro

0

M

H

Ul

H

O\

'

-J

U)

U)

M

Ul

O\

"

M

U)

*!

10

M

U)

*

O

00

00 ro

Ul

M ro

M

* o

U) ro

U)

U) ro

*

M vo

Ul

0^ fVi

•D

^ C

12

0

H

H Z!

ro

CO l^S

(A i

II i 1

O1 a o

3 i ?

O

H*

Ul

Ul

M

00

M

U)

M to

0) ro

-j •j

00

U)

CO

O

S z

(A i i

0 (A

I I

St

O

Ul

H*

-j

00 ro

LJ vo

Ul

M en

U)

U)

Ul

Ul

00

M

H* ro

-

H H

X I cr

(D

T)

H

P tt

U) ro ui

V

O ft

Ho

•a

H-

CL r~

5^

H

^

^

S

H

•o

O

3D

•o m

Tl

O

CO

g

O

^

30

O

o

Heavy Liquid Separation

SEPARATION OF MINERALS USING HEAVY LIQUIDS

Introduction: Procedure:

Heavy liquids are used to fractionate mineral phases from each other, by differentiating the minerals based on their inherent density. Separated minerals may then be weighed and/or analyzed by conventional geochemical techniques.

1. Preparing liquids for use

Lowering or increasing (for sodium polytungstate only) the densities of heavy liquids:

Mineral separations using heavy liquids require that the density of the liquid is such that the mineral to be separated either sinks or floats in the liquid while the rest of the sample does the opposite.

The densities of the first three liquids (Table HLS1) can be lowered by diluting with a number of miscible organic solvents, including: benzene, carbon tetrachloride, acetone, and alcohols.

Heavy liquids most widely used in the laboratory are shown in Table HLS1.

Sodium polytungstate is soluble in water at room temperature. Solutions should be stored in air-tight containers to prevent evaporation of water.

2. Recovery or cleaning of organic heavy liquids

TABLE HLS1. HEAVY LIQUIDS.

Liquid Density g/cc at 200C

Depending on the nature of the chemical used to dilute the heavy liquid, the organic heavy liquid may be recovered from diluted mixtures by fractional distilla tion at reduced pressure (for methyl iodide), or simply by evaporation in a vented hood.

Bromoform

Tetrabromoethane

Methyl iodide

Sodium polytungstate

2.89

2.964

3.325

Adjustable up to 2.9

NOTE:The first three chemicals are light sensitive, and should be stored in dark-coloured con tainers.

When acetone is used, it may be completely removed by streaming water through the diluted heavy liquid until a suitable density marker (specific gravity block) or mineral fragment placed in the liquid floats. The purified liquid is then separated from the excess water and finally dehydrated by shaking with calcium chloride followed by filtration.

Sodium polytungstate will tend to crystallize if the solution is too concentrated.

3. The fumes from these liquids are heavier than air.

Use down draft hood if possible.

3. Separation procedure

Safety advisory:

Separations may be made at any particle size down to a limit of about 10 microns. Below 150 microns, a centrifuge becomes necessary to accelerate the settling rates of the fine particles.

1. The organic heavy liquids are toxic and should be handled with great caution. Do not open bottles, or use outside of a well vented hood. These liquids are known carcinogens. Staff should read the per tinent MSDS sheets.

The weight of material processed may vary from a few milligrams to 200-300 g.

2. Ingestion and contact with skin must be avoided.

The sample must be thoroughly cleaned and free of all clay particles. This may be done with the aid of a dispersing agent such as sodium hexametaphosphate or sodium silicate, after which the sample is dried and dry-sieved to obtain the size of grains required for separation (usually -60 * 1 20 mesh).

3.1. Take a portion of the sample (-60 ± 1 20 mesh) and weigh a minimum of l g to within 0.01 g.

MS6-1

Heavy Liquid Separation

3.2. Transfer the sample into a glass separating funnel (fitted with stopper) filled with 200 ml of heavy liquid.

3.3. Gently stir the grains into the liquid to ensure complete wetting. Grains with densities greater than the density of the liquid will settle to the bottom of the funnel (heavy minerals); light minerals will float on the surface of the liquid. During the separation, stir the grains

3-4 times, depending on the quantity of material used.

where, PH = percentage of heavy minerals in the sample,

WH

= weight of heavy minerals in

W grams, and s weight of material used in grams.

PM s

(WH-WN)x-~

3.4. Drain the heavy minerals into a filter paper

(Whatman -4 qualitative).

4. Calculations

PH = W H x

100

W

3.5. Wash the collected minerals 3 or more times with a suitable solvent (acetone) to remove all trace of heavy liquid.

where, PM = percentage of magnetics in the sample, and

WN = weight of non-magnetics.

Productivity:

3.6. Dry the sample in an 1100C oven for 3-5 days.

A technician should be able to complete 25 samples over a period of 2 days.

3.7. Weigh and dry the heavy minerals collected

(to 0.001 g).

3.8. Remove all magnetic minerals from the sieve.

3.9. Weigh the non-magnetic fraction to 0.001 g.

MS6-2

Magnetic Separation

MINERAL SEPARATION USING THE FRANTZ MAGNETIC

SEPARATOR

Introduction:

Many different minerals may be separated from each other on the basis of their magnetic susceptibility. For practical purposes minerals can classed into three groups depending on their magnetic properties:

1. Ferromagnetic: the most magnetic minerals eg magnetite. These are actually strongly paramag netic minerals.

greater than the barrier field are deflected by it and stay on the upper side of the chute. As the strength of the barrier field is increased it is able to deflect minerals with weaker and weaker susceptibilities and thus separate minerals on the basis of their magnetic proper ties. The range of magnetic susceptibility of various minerals is given by Rosenblum (1958), Hinter (1959) and Hutchison (1974). Note that the currents given by these authors are for the isodynamic separator, the newer barrier separator uses much lower currents.

2. Paramagnetic: most minerals are paramagnetic and are attracted by a magnetic field.

3. Diamagnetic: weakly magnetic minerals repelled by the magnetic field eg quartz, feldspar.

The Frantz magnetic barrier separator consists of a vibrating inclined chute, partly divided along its length, along which minerals are fed; specially designed pole pieces exert a strong magnetic barrier field along the centre of this chute. Mineral separation takes place in the middle of the chute by balancing magnetic forces on each mineral against the gravita tional forces. The divider separates the two mineral streams directing them into bins at the end of the chute.

For ferromagnetic minerals the residual magnetic field is still too strong for successful separation of these minerals. The Frantz low field control can be used to separate these minerals as it reverses the direction of the current to establish a hysteresis loop that reduces the magnetic field to zero. When separating paramag netic minerals, material is fed into the inner channel of the chute which has a side inclination such that gravitational forces push the feed into the other outer channel. For diamagnetic separations this is reversed: the feed is in the outer channel and the side inclination is toward the pole pieces.

Safety advisory:

The vibrating chute has a forward inclination which controls the speed at which minerals move down the chute. This is normally set at about 20 to 30 degrees.

The chute also has a side inclination the setting of which controls the gravitational force on the mineral grains. This is usually set at about 15 degrees for the first few passes but may be increased to around 60 degrees in the later stages of a mineral separation. The strength of the barrier field is controlled by a rheostat; magnetic separations separate the strongly magnetic minerals first from the sample and proceed in order of decreasing magnetic susceptibility. Therefore, the bar rier field is initially very weak so that the ferromag netic and strongly paramagnetic minerals can be removed.

As the mineral stream enters the middle of the chute the gravitational forces on the mineral mixture will act so as to impel the mineral grains toward the lower edge of the chute. As they travel through the centre of the chute they interact with the magnetic barrier field.

Those minerals less magnetic than the field will carry on through this region towards the lower edge of the chute; those minerals with a magnetic force equal to or

1. The high magnetic fields generated by the pole pieces may have a harmful effect on pacemakers and hearing aids. These fields will also cause damage to watches, calculators, magnetic disks, bank and security cards, etc.

Method:

1. Sample preparation - the sample should be washed to remove fine powders which could alter the magnetic properties of the mineral to which it is adhering and also clog the separator. Sizing is also important; while the best grain size to work with will depend on the rock, the most common size range used is -100 +140. Finer sizes may be used if the proportion of composite mineral grains is unacceptably high. Finally a hand magnet is passed over the sample to remove the ferromag netic minerals which can clog the flow of material in the chute unless removed.

2. Mineral separation - the material is placed in the feed hopper and using an appropriate feed rate is fed onto the vibrating chute. The fastest way to proceed is to divide the material by half during

MS7-1

Magnetic separation

each pass making sure that the mineral of interest is entirely in one of the halves. The magnetic range necessary for the separation of that mineral will soon be arrived at. At any stage if the volume of material is sufficiently small 0:50 ml) heavy li quids may be used to further concentrate the mineral required.

The final stages of separation will take place at high side inclinations and changes in the magnetic field between passes will be very small. The most important point to remember is that each rock is different and that there are no set procedures to follow; trial passes are always necessary to see how a mineral mixture reacts to different settings of the magnetic field, feed rate and side slope.

Apparatus:

- Magnetic separator

- Binocular microscope

- Clean paper to cover the working surfaces to prevent contamination and recover mineral mix tures when they are split

- Hand magnet to remove ferromagnetic minerals

- Plastic vials to store the various mineral fractions

- Compressed air, brushes etc. for cleaning the instrument

- Titanium-tipped tweezers

1.3. The sample is placed in a deep container and washed in water to remove the rock flour.

Swirl the water through the sample and let stand still for 15-20 s before decanting the water. Be careful not to lose micas when decanting; they often float on the water sur face as a result of surface tension effects.

Washing should continue until the decanted water is clear. The final rinse should be with acetone or alcohol to speed drying. Dry the sample under a heat lamp or in a warm oven overnight.

l .4. When dry, sieve the sample using clean nylon cloth sieves. The usual size is -100 * 140 mesh but the grain size of the rock will be the major factor in deciding what size should be used.

The sieve clothes can be placed over the mouth of a 500 ml jar attached with rubber bands, containing approx 100 ml sample.

Shake the jar until no more grains pass through. When about 50 ml of the desired size range is available start the magnetic separation and complete the sieving of the rest of the sample later. Before putting the sample in the feed hopper pass the hand magnet over it once more to remove any leftover magnetic grains.

Reagents:

None required.

Procedures:

1.1. The rock sample is reduced to its component mineral grains by crushing it in a jaw crusher followed by pulverization in a mill. In the

Braun pulverizer, several passes, reducing the gap between the plates each time, are neces sary to avoid too much fine material. A roll mill, if available, produces a better result, as the crushing action here as opposed to the ripping action of the former results in more liberated grains.

l .5. The magnetic separator should be thoroughly cleaned prior to starting the separation. Brush the pole pieces carefully, as well as the trough behind them which often contains minerals from the previous separation. Stand the pole pieces vertically and brush them out. Sample cups, the feed hopper,and other surfaces in contact with the sample should be wiped with tissue paper soaked in alcohol. The chute must not be wiped or touched with the fingers; if it is greasy, wash it in detergent and hot water and let it dry. Otherwise, blow it clean and reassemble the pieces.

1.2. A hand magnet should be passed over the sample at this stage to remove iron filings and magnetite which could stain the other minerals. These should be saved in case they are needed at a later time. The hand magnet should be in a plastic bag so that the poles are not coated by the magnetic particles; alterna tively, an enclosed plunger-type magnet may be used.

1.6. The initial settings are 100 side slope, 200 forward inclination. The feed is into the inner channel for the separation of paramagnetic minerals and the side slope is forward, away from the pole pieces. The chute is angled so that the top end is under the pole pieces and the bottom end outside of them. The gate at the bottom end of the feed hopper and the vibra tion of the hopper are adjusted so that flow rates are about 5 ml per minute; in the final stages a feed rate of less than l ml/min is appropriate. The vibration of the chute is also adjusted so that the travel time of grains from one end to the other is about 30 s.

MS7-2

Magnetic Separation

1.7. The initial magnetic setting is dependant on the sample but 0.1 A is normally used. The current is increased on successive passes.

Once the more magnetic minerals which could clog up the chute have been removed the sample may be run at a current high enough to divide it into two equal fractions with the mineral of interest entirely in one. This will speed up processing particularly if the sample is large. Also relatively high flow rates can be used as the deterioration in the separating ability of the Frantz is not of great concern at this stage. The sample volume is halved on each successive pass until the mineral of inter est is concentrated. Note that in the separation of diamagnetic minerals the feed is into the outer channel of the chute which is sloped in toward the pole pieces.

1.10. Often the range of magnetic properties in the

1.11. mineral being separated are such that it does come clean at a relatively narrow magnetic range and therefore the final product will con tain some impurities. Most of these may be removed by heavy liquid separation.

The final product will still contain a few im purities which will have to be removed by hand-picking under a binocular microscope.

Quality Control:

A grain mount should be made of the final product to demonstrate its purity (refer to the section on optical mineralogy, page M9-1).

Productivity: l .8. The final stages of the separation take place at much slower feed rates so that the separation is precise and entrainment of grains does not occur. The side slope is also increased to about

600 so that the gravitational component is in creased with a corresponding increase in the magnetic field. This allows for a more precise separation as small differences in magnetic susceptibility can now be exploited. Special feed troughs are available so that the grains do not spill and the feed rate is narrowly control led.

Variable. Dependant on the volume and the difficulty of the separation.

Bibliography:

Flinter, B.H. 1959. The magnetic separation of some alluvial minerals in Malaya. Am. Mineralogist, 44, pp

738-51

Hess, H.H. 1966, Notes on the operation of Frantz

Isodynamic magnetic separator. Unpubl. notes in

Frantz manual.

l .9. The settings in these final stages are a matter of trial and error. Frequent checking of the separated products using the binocular micro scope, grain mounts and the X-ray diffrac tometer will help control the quality of the separation.

Hutchison, C.S., Laboratory Handbook of petrographic techniques. Wiley, New York, 1974,527 pages.

Rosenblum, S., 1958, Magnetic susceptibilities in the

Frantz isodynamic magnetic separator, Am.

Mineralogist, 43, pp 170-73.

MS7-3

Mineral Identification

ROCK AND MINERAL IDENTIFICATION

FOR PROSPECTORS AND GEOLOGISTS

Introduction:

The identification of minerals is an essential first step in the classification of rock samples (see Vol. I, Chap ter 15). The initial examination of a rock may be carried out with the naked eye, with a hand lens, or with a stereo microscope. Sufficient information may be obtained at this stage for a preliminary identification which can be confirmed using grain mounts or thin sections in the optical microscope, XRD, or other more advanced techniques.

lies. Pyrite and galena have a well-developed luster.

Non-metallic minerals have luster which can be described in a number of ways (e.g: vitreous - resem bling glass; greasy - as if coated with a thin layer of oil; silky - resulting from fibrous aggregations of parallel crystals; adamantine - having a hard brilliant luster, pearly - resembling mother of pearl; resinous - resembling resin).

Colour:

Some of the tests and observations which may be used by the mineralogist in a visual identification are listed here together with other information. This informa tion may be incorporated into a report for the geologist or prospector.

For many minerals, colour is a reliable guide to iden tification. In the metallics, fresh surfaces frequently reveal quite diagnostic colours (e.g. yellow in chal copyrite). Nonmetallics display a greater variety of colour.

Streak:

Visual Examination

Hardness:

One of the most useful tests in mineralogy is the simple scratch test, comparing the ease of leaving a scratch mark on the mineral. A series of l O minerals were used by Mohs to develop a hardness scale in a range of l to

10. Each mineral in the scale can be scratched by the one above it. The scale intervals are approximately equal except that diamond is about 30 times harder than corundum.

Most minerals have a hardness of less than 7. Minerals with a hardness of up to 2.5 can be scratched with a fingernail; hardness of up to 3 can be tested with a copper coin; a knife can scratch minerals with hardness of less than 5.5; a porcelain streak plate can scratch minerals with hardness of less than 6.5; a steel file will scratch minerals with hardness of less than 7.

Streak is the colour of the powdered mineral. It is usually obtained by rubbing the mineral over an un- glazed porcelain streak plate. Metallic minerals produce quite distinctive streaks.

Crystal Habit:

Crystals vary considerably in shape depending on rates of growth, impurities present during growth, and the nature of the host. Nevertheless, some mineral phases are often characterised by particular shapes called crystal habits. Well known terms are fibrous, acicular

(needle-like), columnar, tabular, scaly, and micaceous. In addition, form names are often used such as cubic, prismatic, pyramidal, etc.

Cleavage, fracture and parting;

Hardness is a measure of the strength of the bonding of the atoms in a crystal, and consequently can vary depending on the direction in which it is scratched.

For example, kyanite forms narrow bladed crystals which have a hardness of 4-5 along the axis, and 6-7 at right-angles to the axis.

Many crystals break along smooth planes which are parallel to possible crystal faces. Such planes are called cleavage planes. Cleavages are repeated by the symmetry of the crystal in exactly the same way on all faces. A cleavage may often be described as perfect, good, distinct, imperfect or poor depending on its development.

Luster:

The appearance of the surface of a mineral in reflected light is a quality termed luster. On the basis of luster, minerals can be divided into metallics and non-metal-

Fracture refers to the shape of surfaces formed by breaking a crystal in a direction other than the cleavage. Fracture may be conchoidal, even, uneven, or hackly.

MS8-1

mineral laemijicanon

Twinned crystals:

A twinned crystal is formed of two or more individuals of the same mineral joined together according to a defined law. They may be joined as contact twins on a plane, or as penetration twins.

MS8-2

Optical Mineralogy

OPTICAL MINEROLOGY

Introduction:

Rock and mineral samples are examined under the polarizing optical microscope as thin sections or grains, to allow examination of rocks in transmitted light.

Fine grinding for the production of sections suitable for microprobe analysis is performed on a rotating lap in two stages; a cover slip is not applied. In the first stage, 6 micron diamond paste is used, and in the second step, l micron diamond paste is used.

2. Determination of refractive index

Procedures:

1. Preparation of thin sections

1.1. l .2.

1.3.

A small chip of rock or mineral is sampled, or a thin slice ^5 mm thick) is cut from the specimen using a diamond saw.

The chip or slice is ground flat and smooth on the surface using progressively finer abrasive, e.g. Carborundum powder, starting with 80 grade and finishing with 600 grade.

The smooth surface is cemented to a glass slide (usually 2"xl") using Canada balsam or

Lakeside 70 cement.

In a normal thin section, it is rare for the contacts between mineral phases to be vertical and for the surfaces to be smooth. If there is a difference in refractive index between the mineral and cement, the irregularities (stepping or curving of the grain margin) concentrate orscatterlight by reflection and refraction.

The effect is to give an impression of three-dimen sional relief (high relief). If the difference in refractive index is small, the irregularities will hardly be visible, and the mineral has low relief. Relief may be observed in plane polarised light with a low or medium power objective, with the diaphragm nearly closed and with the condenser assembly lowered.

1.4.

1.5.

The other surface is ground down until the rock section is 300 microns thick. Progres sively finer abrasives are used as the section becomes thinner, and finishing is done with

600-grade Carborundum powder. The thick ness is gauged by observing the interference colours of common minerals such as quartz or olivine.

After cleaning excess cement from the surface of the section, it may be covered with a glass slip, cemented in place with Canada balsam.

Description of relief is subjective, but can be reference to the refractive index (RI) of Canada balsam

(RI - 1.537). For example, rutiles have a high relief, whereas quartz has a low relief. The Becke line test may be applied to determine whether low relief minerals have higher or lower refractive indices than

Canada balsam. Under similar light conditions to those described above, low or moderate relief minerals concentrate light as a thin bright white line along their margins. This line, termed the Becke line, will move inwards or outwards if the mineral grain is brought slowly into or out of focus. The Becke line moves into the substance of higher RI, when the dis tance between the objective and section is increased.

Canada balsam may be thinned with xylol to ease spreading. The mounting is done on a hot-plate which serves to drive off xylol and harden the cement.

A more accurate means of determining RI involves the use of loose or crushed grains of a mineral in a series of liquids of known RI. The relief of the mineral will disappear when its RI, matches that of the liquid.

Some poorly consolidated samples require impregna tion with a cement before a section can be made.

Heating the sample in Canada balsam is often suffi cient.

Loose and crushed grain mounts may be made from unconsolidated sediments and loose minerals. The grains are spread on a glass slide, and immersed in

Canada balsam. The mount is covered with a glass slip. For crushed grain mounts, an immersion liquid capable of supporting the cover-slip by capillary ac tion is sufficient

A small amount of the mineral is crushed to a powder, and placed on a glass slide. The powder must be neither to fine nor too coarse as a fine powder is too hard to observe, yet a coarse powder will tend to float.

A glass slide is placed on the dry powder, and the RI liquid is introduced at the margin of the coverslip using a dropper. The procedure is repeated with dif ferent RI liquids until the RI of the mineral is matched to that of the liquid. When the RI of the mineral is within 0.002 of the RI of the oil, colored Becke lines are seen.

MS9-1

upticui mineralogy

RI liquids are obtained commercially, or can be produced in-house (Shelly, 1975).

This is performed by putting the mineral section in extinction (Figure OM2) thus producing two vibration directions E-W and N-S.

3. Determination of birefringence

The birefringence or partial birefringence of a mineral section is determined in thin section by observing interference colours in crossed-polarised light with the diaphragm below the stage open, and using low-power objectives. The Michel-Levy colour chart presents the relationship of birefringence to the thickness of the section (horizontal lines) and interference colour (ver tical lines). Figure OM1 illustrates how the chart is read. For example, a mineral with colour A and thick ness B, has a birefringence, or partial birefringence of

C. To perform this determination, several grains of a mineral in a thin section are cut parallel to different crystallographic directions, and hence intersect the indicatrix in different orientations. The interference colors will vary from black in sections cut parallel to the circular section to a maximum for those cut parallel to e and CD, or X and Z. The grains with the highest interference colors will therefore be diagnostic of the birefringence. Preferred orientation in layered or banded rocks often introduces a bias in the section, and caution is required in the interpretation of the inter ference color.

Interference colours are normally split into orders cor responding to the amount by which two rays are out of phase. The simplest method of determining the order of the colour is to look at the edge of the mineral grain. Edges of grains which are wedge-shaped show a range of interference colours from 1st order grey at the thin edge, inwards towards progressively higher order colours. By counting the colour rings, the order of the colour in the main part of the mineral section can be determined.

In thin sections which are 30 microns thick, quartz may be used as a control on the birefringence of other phases, as it invariably shows first-order pale yellow.

The stage is then rotated through 450 noting whether clockwise or anticlockwise to determine the colour produced by the mineral. Insertion of the plate is accompanied by a change in the colour. If retardation occurs (i.e. the interference colour increases), then the slow direction of the plate is parallel to the slow direction of the mineral. If compensation occurs (i.e. the interference colour decreases), then the slow direction of the plate is parallel to the fast direction of the mineral.

The fast and slow directions are partially diagnostic in the identification of mineral phases.

3.2.

Interference figures are obtained as follows: focus on a grain with a high-power objective, and center the grain in the field of view, center the microscope, and raise the sub-stage condenser, opening the sub-stage diaphragm. Using cross-polarised light, insert the

Bertrand lens. The stage should be rotated when ob serving interference figures so that the pattern of the isogyres can be noted.

Interference figures indicate whether the the mineral is uniaxial or biaxial, and with the aid of the sensitive tint plate, the optic sign can be determined. 2V can be obtained on biaxial minerals. These properties can all be determined from grains cut parallel to circular sections of the uniaxial or biaxial indicatrix. Therefore, the best grains are those with the lowest interference colours (black or grey). Such grains give a uniaxial cross, or a biaxial optic-axis figure. Isotropic minerals will not produce an interference figure.

3.3.

Interference figures, determination of optic sign, and measurement of 2V

Use of accessory plates to determine optic sign

3.1. Fast and slow direction: the use of acces sory sensitive tint plates

The l-wavelength plate (made of gypsum or quartz) has the thickness and orientation to produce Ist-order pink. This plate has two vibration directions in the 450 position which are marked on the plate as fast and/or slow. The accessory plates are used to increase the equivalent thickness of the thin section and to compen sate for the wavelength difference produced by a mineral.

To determine the optic sign, interference figures and sensitive tint plate are used together. In uniaxial figures, the e-vibration directions are radial to the CD direction. Insertion of the plate results in retardation or compensation in alternate quadrants. Minerals with positive e are slow, whereas minerals with negative e are fast. Thus, as shown in Figure OM2, the change in pattern is diagnostic of positive or negative uniaxial minerals.

In biaxial figures, the pattern of retardation and com pensation in relation to the isogyres enables the posi-

MS9-2

Optical Mineralogy

tions of X,Y, and Z to be determined, and the optic sign of the mineral.

have characteristic shapes and habits that can aid in the identification of minerals.

3.4. Interpretation of figures

3.4.1. Uniaxial cross interference figures (Figure

OM2).

Cleavage is commonly observed as a set of cracks or bands in a mineral section. Geavage vertical to the section appears as fine cracks, but cleavage oblique to the section will produce broad dark bands. The num ber of cleavages, and their relative orientations and relations to the crystal faces are diagnostic features of certain mineral phases.

The mineral is uniaxial or biaxial with 2V close to O0.

The optic axis is approximately vertical in the section.

The positive or negative character can be determined.

Off-centred figures can be used provided the centers can be seen. In these circumstances, the optic axis may be 20-300 from vertical. When the crossed isogyres move out of the field of view, there is no guarantee that the figure is not biaxial, and therefore these figures should not be used.

The extinction angle may be used to document aspects of shape and cleavage. The angle between a vibration direction and a cleavage or prominent crystal face is termed the extinction angle. If the angle is zero, the mineral has straight extinction. Extinction angles are particularly useful in the determination of plagioclase determinations.

3.4.2. Biaxial optic-axis figures (Figure OM2) 5. Colour and pleochroism

If the middle of an isogyre remains in the centre of the field of view during rotation of the stage, the mineral is biaxial (Figure OM2). The degree of curvature of the isogyres enables 2V to be estimated. For a 2V of 900, the isogyres remain straight during rotation. The cur vature becomes large at 2V less than 200. At O0, the figure resembles a uniaxial cross. The positive or negative character can be determined with the tint plate (Figure OM2). If 2V is 900, the mineral is neither positive nor negative.

Most mineral phases are transparent in thin section, but a few mineral phases are opaque. Colour is judged with the analyzer removed. Colour can be a useful property, but is rarely diagnostic.

Opaque minerals are best studied using a microscope fitted for reflected light work.

Anisotropic minerals frequently display variations in colour or intensity of colour when the stage is rotated in plane-polarized light, a phenomenon termed pleochroism.

If the isogyres remain in the field of view when the stage is rotated (Figure OM2), the mineral is biaxial, and the acute bisectrix is near vertical in the section.

Positive or negative character can be determined, and

2V estimated when the isogyres are in the 450 position.

If the isogyres barely separate, 2V is very small (less than l O0), whereas if 2V is 500' the isogyres move to the edge of the field of view.

6. Point counting to determine the modal mineralogy

If the isogyres move slowly out of the field of view when the stage is rotated, the mineral is biaxial, and the acute or obtuse bisectrix is near vertical. The 2V is moderately large. The optic sign cannot be deter mined if the bisectrix is close to vertical.

If the isogyres move quickly out of the field of view when the stage is rotated (Figure OM2), then the optic sign can be determined, but the distinction between uniaxial and biaxial is not clear-cut.

In order to assess the relative proportions of mineral phases, it is often necessary to utilize a polarizing microscope equipped with a point counting stage.

Such a stage is mounted on the rotating stage, and is capable of being moved by a finite and constant dis tance in either direction parallel to the cross-hairs of the microscope. At each point on the thin section or polished section (reflected light microscope), the phase is identified visually, and the stage is advanced simultaneously as the counter increments the total number of points counted for that particular phase.

After counting at least 2000 points (for a fine-grained sample), the modal composition of the sample is deter mined.

4. Crystal shape and cleavage

Shape is not a diagnostic feature in thin-section, but is a useful guide. Euhedral crystals of many minerals

7. The reflected light microscope

A variety of properties can be studied in reflected light, but these are generally qualitative in character, and

MS9-3

i** in trier

rarely provide mineral phase.

an unequivocal identification of a

11. Internal reflections

Qualitative optical properties are made with either the polarizer inserted (and the analyzer out) or with both polarizer and analyzer inserted. Observations are made in air or with oil immersion objectives. Colour, reflectance, birefringence, and reflection pleochroism are observed in plane polarized light; anisotropism and internal reflections are observed under crossed polars.

Some minerals examined in polished section are transparent and others opaque. Internal reflections are observable in the more transparent of the opaque minerals.

12. Hardness

8. Reflectance

The amount of light incident on a polished surface of a particular mineral depends on its reflectance. Both colour and degree of polishing can produce an anomalous amount of reflection. Although the eye cannot measure reflectance directly, it is possible to arrange minerals in order of reflectance.

Polishing and scratch hardness are useful tools in the identification of opaque minerals. To observe the polishing hardness, the KALB light line is used (Shel ley, 1975). To observe this line, focus on the boundary line between two mineral grains, lowerthe stage so that the sample begins to go out of focus, and observe a

'line' of light which will move towards the softer mineral provided there is significant relief.

13. Structural and morphological properties

9. Bireflectance and Reflection Pleochroism

Cubic minerals remain unchanged in reflectance and colour on rotation of the stage whatever the orientation of the grains. Basal sections of hexagonal and tetragonal crystals will also appear the same as cubic minerals.

Crystal form and habit are recognizable in reflected light. Some minerals form well-developed crystals

(euhedra), whereas others are anhedral. All of the standard terms used for transmitted light can also be used.

Cleavage and parting are sometimes seen in polished surfaces as rows of triangular pits.

10. Anisotropism

Twinning of crystals is often observed in reflected light.

When the polished surface of a cubic mineral is ex amined under crossed polars, it is found to remain in extinction in all positions of the stage, whatever the crystallographic orientation - it is said to be isotrophic.

Variation in the colour of anisotropic minerals, when the stage is rotated, can cautiously be used as an aid in identification.

Bibliography:

Shelley, D., Manual of Optical Mineralogy, Elsevier,

Amsterdam, 1975.

MS9-4

Optical Mineralogy

BIREFRINGENCE c

RCTAROATION

Figure OM1. Diagram illustrating how the Michel-Levy chart is used.

MS9-5

All +VC Schemes reversed for -ve.

Optic-axis figure of a biaxial +V6 mineral using a sensitive tint plate or quartz wedge, b= blue, y - yellow, arrows indicate movement of colour rings when wedge inserted.

ve

- ve

Distinction of biaxial+VC from biaxial -ve acute bisectrix interference figures using a sensitive-tint plate or quartz wedge, b = blue, y = yellow; arrows indicate movement of colour rings when wedge inserted.

Distinction of biaxial -f ve from biaxial -ve acute bisectrix interference figures using a sensitive-tint plate or quartz wedge, b = blue, y = yellow; arrows indicate movement of colour rings when wedge inserted.

Figure OM2.

MS9-6

XRD

X-RAY POWDER DIFFRACTION

Introduction:

X-ray powder diffractometry (XRD) is a widely used diffraction technique for the precise and rapid iden tification of cry staUine materials. However, complica tions occur when mineral mixtures are used and it is preferable to use monominerallic powder mounts when an unequivocal result is desired.

The following paragraphs summarise the principles of the technique. In the X-ray diffractometer, a collimated beam of Cu X-rays (other radiations such as Mo, Cr,

Fe, (fe Co, can be used but Cu X-rays are the most commonly used) falls on a finely ground sample in which the crystallites have a random orientation. Dif fraction of the X-rays occurs from planes of atoms making up the crystalline lattice. The direction in which the diffracted beam travels is dependant on the spacing between successive parallel planes of atoms in the lattice and this is given by the Bragg equation:

X = 2d sin0 where, d = the interplanar spacing

K ^ wavelength of the X-ray, e.g. l.542A for Cu K

6 = angle of incident radiation to the plane of atoms is an X-ray sealed tube system which consists of a heated tungsten filament firing electrons at a copper anode (for Cu X-rays) from which X-rays are emitted.

The electrons are accelerated towards the anode by means of a high potential which is supplied by the high voltage generator. Large amounts of heat are generated by this process and water cooling is essential. The intensity of the X-rays is directly dependant on the current and voltage used, which must never exceed the rating for that tube. Tube life will be prolonged by use of a voltage and current below the maximum.

The diffractometer has a specimen holder which rotates at an angle (0) relative to the parallel beam. The detection system is on an arm rotating at an angle of

20. Both the incident and diffracted beams are col limated by a series of parallel slits. The divergence of the beam is further limited by a divergence slit which is normally l 0; a narrower slit of 1/40 is used if low 20 angles are to be used. The size of the receiving slit on the detector arm governs the width of the line profile; narrow slits will therefore allow better resolution. The use of narrow divergence or receiving slits reduces peak intensities and minor peaks will be lost in the background. The detector normally used is a propor tional counter. The diffracted beam is also passed through a crystal monochromator which allows only

Cu Ka radiation to pass thus ensuring a monochromatic beam.

Different planes within the crystal will each give a reflection (hkl reflection) at a particular angle, thus defining a unique diffraction pattern characteristic of that particular crystal somewhat analogous to a fingerprint. In a randomly oriented sample all planes giving reflections will be suitably oriented to give a reflection. Some minerals tend to have a preferred orientation because of their habit or tendency to break along certain cleavage directions. These minerals, the most notable of which being the clay minerals and micas, will only give a few of the total refections possible. Recognition of these minerals can be difficult but usually they have one intense line (normally the

001) that can be used to identify them; the difficulties arise when two minerals having 001 lines close together are present in the same sample. This is often the case in mixtures of clay minerals, for which XRD is the most reliable method of identification; therefore, a variety of special techniques have been developed to reliably identify clays.

Instrumentatioa The X-ray diffractometer consists of a source of radiation, a diffractometer, and a detection and counting system. The most common type of source

Safety advisory:

l. As X-rays are involved, film badges must be worn and the working area is restricted to authorized personnel only. Although the instrument has been constructed in a manner to ensure there is no X-ray leakage, a geiger counter should be regularly used to check for this possiblity. Note that the shutter should be closed when the operator is changing the removable slits to avoid leakage of X-rays from the slot.

Method:

X-ray powder diffractometry involves the following steps: l. Sample preparation - the sample is ground to a fine powder in alcohol and the slurry is smeared onto a glass slide. To ensure randomly oriented mounts, a back-packing technique should be used. Prepara tion of clay mineral samples is given in the follow ing section.

MS10-1

2. The sample is scanned in the diffractometer generally from 50 to 700 26.

Check that the powder is sufficiently fine before packing it; further grinding in the mor tar and pestle is usually necessary.

3. The peaks are located on the chart and the d-spac ing calculated.

4. The pattern is indexed and the constituent phases are identified.

1.4. The powder mount is placed in the diffrac tometer using the spring clip to hold it. The sample should be right up against the aluminum block above the spring clip for it to be evenly illuminated by the X-ray beam.

5. Special treatments for clay minerals and metamict

(see below) minerals may be used.

l .5.

IMPORTANT! - turn the water cooler on.

Apparatus:

- Agate mortar and pestle

- Petrographic glass slides or sample mounts

- Drying lamp

- Furnace (to 12000C)

- Desiccators

- Ultrasonic bath

- 600C oven

1.6. The diffractometer will have been set up by the Philips service engineer. The settings for the detector and measuring circuits are stand ard and are given in the manual. They do not need to be changed in ordinary operation.

l .7. Turn the mains power on. The diffractometer arm should zero itself; if not, turn it off and on again until it does so.

Reagents:

- Alcohol

- Ethylene glycol

- Hydrochloric acid, HC1,20?c

- Hydrogen peroxide, H202

- Sodium hexametaphosphate

Procedures:

1.8. Turn on the power to the X-ray tube. The voltage and current dials should be at zero or else the safety device will not allow the tube to turned on. Increase the voltage gradually to

40 kV and then the current to 20 mA. These are the standard settings and cover most types of analysis.

1.9. Check that the divergence and receiving slits are the l 0 slits.

LI. Using a mortar and pestle, crush a small piece

^0.5g) of the rock or if possible the individual minerals to a fine powder, in alcohol, (which helps prevent loss of material due to fragments flying off). The optimal size is about 10 microns and the powder should not feel gritty in the mortar.

1.10. Set the 26 ranges on the dials: the usual range is 50 to 700 26. Set the minimum position at the lower angle and advance the diffractometer arm to it before opening the shutter or else the beam will shine directly into the detector and may damage it.

l .2. Spread the slurry evenly over two-thirds of the glass slide and let it dry, under a heat lamp if necessary. Ideally, the sample should be one layer thick. If only a small amount of sample is available a low background holder such as a specially cut quartz crystal or an iron plate should be used.

1.11. Turn the chart recorder on and advance the paper until the pen is at a major division. The chart drive is set at 2 cm per minute in normal operation. Open the shutter and press the auto matic start button to begin the scan. The scan rate is set at T 26 per minute.

1.12. When the scan is over close the shutter and change the sample. The next scan can be started as above.

1.3. Some samples are received already crushed for chemical analysis. If sufficient powder is available fill one of the special aluminum well mounts. A glass slide is held against the front and the powder is poured in through the open ing at the side, tapping the mount so that the powder is packed down. This method is useful when one wants to avoid preferred orientation.

1.13. Peaks on the previous chart can now be iden tified and their 26 angle marked. A rule of thumb for distinguishing peaks is that they be at least two times background. The d-spacing can then be read of a chart which gives the

MS10-2

XRD

d-spacing for a given 29 angle for Cu radia tion. The peaks are then ordered in terms of intensity; for most purposes only the three or four most intense peaks need be ordered but the process for mixtures is more complicated.

The indexing of the pattern can now take place.

2. Indexing

For mixtures of several minerals the procedure is much more complex as the ordering of intensities is difficult.

In most rock samples, however, there are recognizable minerals and their peaks e.g. quartz at 26.60 29. The full pattern of each mineral is subtracted and the search is then done on the residual pattern. Identification of trace phases in such mixtures can be almost impossible particularly if these minerals are strong absorbers of the particular radiation used e.g., iron minerals with Cu radiation.

The Hanawalt system is used for the indexing of pat terns. This system takes the three strongest intensities on the pattern and uses the JCPDS search manual which lists the intensities for almost 48,000 patterns.

This is a huge volume of information which would make searches long and difficult as well as giving many different possibilities for each pattern. For tunately, there are a number of ways of dealing with this information overload. As this laboratory deals almost exclusively with minerals we need only to use the JCPDS Mineral sub-file which lists every known naturally occurring substance - some 3,500 in all.

Secondly, some preliminary information about the un known mineral is usually available; this would include knowledge of the mineral properties (colour, hardness, etc), crystallography and mineral association. All of this information is helpful in matching the pattern. The most useful data would be the chemistry of the mineral as the minerals are listed chemically.

Special problems may occur with mineral solid solu tions where the mineral composition lies between two end-members e.g. olivine where the composition may lie between Mg and Fe end members. The diffraction lines will shift accordingly between the two extremes and allowances will have to be made for this shift when the search is being conducted.

3. Identification of metamict minerals

Certain minerals which contain radioactive elements may not give a pattern or one that is impossible to match. This is because the radioactive disintegrations have disrupted the crystal lattice forming what are termed metamict minerals. In most cases the original pattern may be recovered simply by heating the sample at 10000C for one hour and re-running it. One possible complication is minerals that oxidize e.g. UO2 to U3O8.

Indexing a pattern using the Hanawalt system, the following steps may be used:

Quality Control:

2. l.

2.2.

Identify the three strongest lines.

Turn to the section in the search manual cover ing the strongest line.

The above technique is a qualitative means of iden tification, although abundances can be estimated using the relative intensities of each phase present. The absorption effect is variable for each mineral present, so these relative intensities can be misleading - par- ticularily if iron minerals are present.

2.3. Taking the second strongest line, scan the second column in the section and match it.

2.4. Looking at the first and third lines in this area attempt to match the unknown, using the five other lines forconfirmation and any additional information that is known about the sample.

The d-spacings may not agree exactly because of instrumental problems or solid solution but the overall pattern should be the same. Note too, that variations in the intensity of some lines may occur because of a coarse sample or preferred orientation.

A quartz sample should be run regularly to check that the pattern obtained is the same as the standard JCPDS one. If not there may be an alignment problem. The detection limits of minor phases depends entirely on their chemistry and on whether the mineral is likely to have a preferred orientation which may enhance one of its lines. A detection limit of l ^ is assumed by most workers (Klug and Alexander, 1971)

Productivity:

Eight to ten samples can be run and indexed in a day.

2.5. For a proper match the JCPDS card should be examined and all the peaks indexed with ref erence to the card.

Bibliography:

Jenkins, R. &L de Vries, J.L., 1972, An introduction to

X-ray powder diffractometry, Philips, 41 pages.

MS 10-3

AKU

Klug, H.P. (fe Alexander, L.E., X-ray diffraction pro cedures for percrystalline and amorphous materials.

Wiley, New York, 1974, 716 pages.

Starkey, H.C., Blackmon, P.O., & Hauff P.L., 1984,

The routine mineralogical analysis of clay-bearing samples, USGS Bulletin, 1563.

MS 10-4

ClayXRD

X-RAY DIFFRACTION IDENTIFICATION

Introduction:

Clay minerals by definition occur as particle less than

2(4. in size. Their identification poses special problems because of their small particle size. Over the last four decades clay mineralogists have developed a variety of techniques for clay mineral identification of which

XRD has been the most important by far. A large number of pre-treatments have become available for positively identifying clays but only two of these are routinely used: glycolation and heating. The prepara tion of clay samples is also different as the clay size fraction has to be separated out. In the Geoscience

Laboratories the preparation forms part of the grain size analysis procedure which is usually requested at the same time (Refer to page MS4-1).

Method:

The method consists of the following techniques:

1. Clay preparation.

1.2. HC1; NaOAc buffer, or CH3COOH removal for carbonates.

1.2.1. If there is a large amount of carbonate material, add 50 ml of 5N NaOAc.

l .2.2. If there is a small amount of carbonate, add 25 ml IN HC1 and dilute solution to 250 ml.

l .2.3. Repeat treatments as necessary.

1.2.4. The pH of the solution should be no more acidic than 3, for prolonged periods.

1.2.5. This procedure should leave a clear solution within an hour or two, with a slurry at the bottom; decant this solution.

l .3. Treat with sodium hexametaphosphate to dis perse the suspension.

2. Clay separation.

3. Clay tile preparation.

1.3.1. To the 25 ml soil-water suspension, add 25 ml of 50 g/1 stock solution of sodium hexametaphosphate.

l .4. Wet-sieve through "230 mesh.

4. X-ray diffraction.

Procedures:

1. Clay preparation for X-ray diffractometry

Clay preparation is performed from the -230 mesh

(-63 micron) fraction. The three pretreatments given below are only used where excessive organic or car bonate content will interfere with the analysis. Disper sal of the clay minerals should be carried ultrasonically and the use of a deflocculant such as sodium hexametaphosphate should be a last resort.

2. Clay separation

2.1. Start with 1000 ml suspension of dispersed solution

2.2. Shake cylinder vigorously for one minute with a stopper over the top.

2.3. Let suspension stand for either 6 hours and 3 minutes or 3 hours and 2 minutes and draw off circa. 75-100 ml of suspension from depths of 8 cm or 4 cm respectively, using a rubber bulb pipette.

1.1. Add H2O2 to remove organic (colloidal) mat ter.

3. Clay tile preparation

1.1.1. Add 5 ml amounts of 307c H202 to a soil-water mixture of l :1 (ea. 25 g of soil); heat lightly and stir occasionally.

1.1.2. Add H202 until the reaction stops or use light heat (ea. 800C for several hours) to complete the reaction.

3.1. Let the 100 ml suspension settle for a day or

3.2. When ready to prepare the tiles, carefully decant 90 ml of the clear liquid and stir the remainder to a slurry.

3.3. so in a covered jar.

Fill an eye-dropper with the slurry and spread it evenly over two glass slides and a porous ceramic tile.

MS11-1

3.4. Allow to dry. The timing of this preparation should be such that the samples are just dry when the XRD scan is made. Late afternoon is preferable as the tiles can air-dry overnight on the bench.

or a flow sheet for clay mineral identification as in Starkey et al. (1984).

Quality Control:

3.5. The preparation above will give a sample that has a preferred orientation and with an intense

001 line which is desirable. If the full pattern is required a randomly oriented mount will have to be prepared by drying the sample and back-packing the sample in an aluminum holden

4. X-ray diffraction

4.1.

4.2.

Run one of the glass slides from 20 to 38C using a 1/40 divergence slit.

29

Place the other glass slide in a desiccator with a bowl of ethylene glycol in the bottom and leave in an oven at 600C for at least 4 hours.

The sample must be run immediately on removal from the desiccator.

The above technique is a qualitative means of iden tification, although abundances can be estimated using the relative intensities of each phase present. The absorption effect is variable for each mineral present, so these relative intensities can be misleading - par- ticularily if iron minerals are present.

A quartz sample should be run regularly to check that the pattern obtained is the same as the standard JCPDS one. If not there may be an alignment problem. The detection limits of minor phases depends entirely on their chemistry and on whether the mineral is likely to have a preferred orientation which may enhance one of its lines. A detection limit of l Ve is assumed by most workers (Klug and Alexander, 1971).

Productivity:

Eight to ten samples can be run and indexed in a day.

4.3.

4.4.

The sample on the porcelain tile should be heated to 4000C for at least a half-hour and then run on the XRD.

Bibliography:

Jenkins, R. SL de Vries, J.L., 1972, An introduction to

X-ray powder diffractometry, Philips, 41 pages.

This sample is then reheated to 5500C for at least a half-hour and re-run. The heated samples should be run immediately on removal from the furnace.

Klug, H.P., (fe Alexander, L.E., X-ray diffraction pro cedures for percrystalline and amorphous materials.

Wiley, New York, 1974, 716 pages.

4.5.

Interpretation of the patterns may carried out using the indexing procedure described above

Starkey, H.C., Blackmon, P.O., & Hauff P.L., 1984,

The routine mineralogical analysis of clay-bearing samples, USGS Bulletin, 1563.

MS11-2

Specific Gravity

DETERMINATION OF SPECIFIC GRAVITY

Introduction: Procedures:

Specific gravity is determined on a routine basis for job samples that have been submitted for whole-rock geochemical analysis. Specific gravity can also be determined for other rocks and minerals where the geologist or prospector has clearly identified this test.

1. Weighing of samples

1.1. Weigh dry sample on top-loading balance to at least one decimal place.

Specific gravity is often used to confirm whether a rock is felsic, intermediate or mafic. Mafic rocks will have a higher specific gravity because of the presence of iron and magnesium bearing minerals.

l .2. Place the sample in the water bucket so that it is covered; remove and suspend on mesh bas ket beneath the balance and record the weight of the sample suspended in water to at least one decimal place.

Safety advisory:

1. There is no specific safety advisory for this method.

Method:

2. Calculation of results

S.G.

=

DW

DS - SW

where, S.G.

^ specific gravity

DW s dry weight, and

SW s suspended weight

1. Weighing of the dry sample, followed by weighing of the sample when wetted.

Quality Control:

2. Calculation based on the difference of the two weights.

Apparatus:

The Geoscience Laboratories has a rock 'standard' which should be checked at the beginning and end of each job.

Productivity:

- Top-loading balance

- Bucket of water large enough to handle most samples, with suspended mesh basket

A technician should be able to complete 100 deter mination per day.

Reagents:

Additional notes:

No specific reagents are required for this method.

1. The balance should be calibrated often.

2. Note any sample peculiarities, e.g., vesicles.

3. Do not use this method on fragile samples such as clays.

MS12-1

DETERMINATION OF THE MOISTURE CONTENT OF SOIL

(ASTM D2216-71) ______________

Introduction:

Moisture (or water content) of soil is the ratio ex pressed as a percentage, of the weight of water in a given mass of soil to the weight of the solid particles.

Procedure:

1.1. Record the weight of a clean, dry container and lid (Wl). A tight-fitting lid is essential.

1.2. Place a crumbled sample loosely in the con tainer and replace the lid. Weigh the container, lid, and contents (W2).

1.3. Remove the lid and place container, lid, and contents in an oven at 1050C until a constant weight is reached (generally overnight).

1.4. After drying, remove container, lid, and con tents from oven, replace the lid, and allow to cool.

l .5. Weigh the container, lid and contents (W3).

Calculation :

W2-W3

96 moisture in sample = TTTT* 100 9fc

Soil Moisture

MS13-1

Fire Assay

FIRE ASSAY

Introduction:

Fire assaying is used to determine the precious metal content of rocks and ores. Determinations of gold and silver present in abundances exceeding 0.01 oz/ton for

Au and 0.10 oz/ton for Ag are made by fire assay and gravimetric techniques. Gold, platinium and palladium present in lower abundances are determined by Atomic

Absorption Spectroscopy (see Section EA21) follow ing fire assay preconcentration.

8. Read MSDS sheets for all chemicals used in this method, i.e. litharge, sodium carbonate, silica, and borax glass.

Method:

There are three fire assay methods:

1. A regular gold and silver fire assay utilizing gravimetry.

Special caution is required when dealing with unusual rock matrices, and it is critical that the flux composi tion be modified according to the composition of the sample. If an estimate of the gold or silver content is available, then the weight of the sample used can be adjusted.

2. A fire assay/graphite furnace technique for Au, Pt, andPd.

3. A fire assay/nickel sulphide concentration techni que, with analysis by instrumental neutron activa tion analysis.

Precious metals are normally present in nature at con centrations of less than l ppm. Furthermore, their distribution within a sample may be very in- homogeneous. The fire assay procedure enables a large sample to be analyzed in order to optimize precision and accuracy.

Safety advisory:

Method #1 consists of the following techniques:

1. Sample preparation - the weighing and mixing of the rock powder and flux

2. Preconcentration - the melting and pouring to form a button, removing and cubing of the button, and cupelling to produce a prill

A lead flux is used to prepare the button. Lead is a known carcinogen, and has been linked to various chronic forms of cancer. The working area is a lead control area. Specific attention should be paid to the following points:

3. Determination - weighing and parting

4. Calculation of the results - weighing of the gold and converting this weight to ozAon

1. Ensure that fluxes are mixed ONLY in the vented hood.

2. Do not eat, drink or smoke in the room.

3. Wash your hands after handling flux.

If concentrations of less than 0.01 oz.Aon Au, Pt, or Pd, or 0.10 oz/ton Ag are to be determined; the prill is passed to the Chemistry subsection for dissolution and analysis by graphite furnace AA (Method 2). The nickel sulphide fire assay (Method 3) is detailed later in this section.

Apparatus:

4. Wear a fire-resistant lab coat in the fire assay area.

Keep this coat in the area and do not remove it from the assay suite except to be laundered.

5. Wear a full face shield and insulated gloves when using the furnace.

6. Perform all parting processes in a fume hood.

7. Wear safety glasses when not wearing the full face shield.

- Two regular fire assay furnaces used to prepare beads. One is used for ore-grade material, and the other for low abundance samples

- One cupellation furnace for the preparation of the prill

- Pouring plates, crucibles, cupels

- Balances capable of weighing to 0.01 mg

- Parting dishes and acid dispenser

- Fume hood

MS 14-1

Reagents:

- The stock flux contains the following:

1.12. Place the lead cube on a preheated cupel

(9500C) and heat in the furnace for ap proximately half an hour, until lead is ab sorbed in cupel.

TABLE FA1.

1.13. Remove the silver prill and note any peculiarities.

Component Weight (g)

9fc litharge sodium carbonate silica borax glass flour

66

33.6

9

8.4

3

Nickel sulphide buttons require the use of:

55

28

7.5

7

2.5

1.14. Brush and accurately weigh the silver prill using a fine balance (to 0. l mg). Place the prill in a parting dish.

1.15. Digest silver prill in hot nitric acid and wash the residue with distilled water (fume hood operation).

1.16. Anneal and then accurately weigh the gold using a fine balance (to 0.01 mg).

- nickel carbonyl

- sulphur

- silica gel

- sodium carbonate

- sodium metaborate

1.17. Record all observations and results in a daily log.

Method 1: Conventional Fire Assay Method

1.18. Calculate silver weight and record gold and silver results in report. Assuming 1/2 assay ton, multiply the weight Ag (in g) by a factor of 2 and convert to oz/ton by division by

29.166.

Procedures:

1.1. Weigh out rock pulp (14.583 g, 1/2 assay ton) into a 20 g crucible.

1.2. Add pre-mixed stock flux (approximately

120 g).

Quality Control:

The determination limits for this method are 0.01 oz/ton for Au and 0.10 oz/ton for Ag.

1.3. Mix rock pulp and flux in crucible.

The precision (2) at 10 times determination limit for

Au is 0.01 ozAon, and 0.10 oz/ton for Ag l .4. The position of each sample in the furnace is noted (it is not possible to write on the crucible unless a GRAPHITE pencil is used).

1.5. Place in assay furnace (preheated to 10250C) and heat for 35 minutes.

Precision and accuracy are monitored by measuring the concentrations of gold and silver in at least one duplicate sample from each batch, and by analyzing the MRB gold/silver standards (MRB25-28) on a monthly basis. Proposed in-house certification data for these standards are: l .6. Pour molten charge into cast iron mold.

1.7. Inspect crucible for lead loss (Figure PAI).

1.8. Note slag colour for possible interferences, l .9. Note size and appearance of lead button.

1.10. Break slag and free the 20-25 g lead button.

1.11. Hammer the lead button into a cube.

TABLE FA2.

Standard Gold oz/ton ppm

MRB 25

MRB26

MRB27

MRB28

1.44

0.37

0.016

1.45

49

13

0.5

50 oz/ton

2.2

10.1

45.1

94.5

Silver ppm

75

346

1546

3240

MS 14-2

Fire Assay-2

Productivity:

Method 2: Fire Assay Preconcentration with

Determination of Au, Pt, Pd by GFAAS

A technician should be able to complete up to 100 samples per day.

Additional Notes:

Procedure:

2. l. Weigh 10 g of rock pulp.

1. Bead colour - the presence of high concentrations of the platinum group elements may give the silver bead a dark grey colour.

2.2. Mix the pulp with 120 g of the standard flux

(refer to conventional fire assay method, page MS 14-2).

2. Bead shape - flat silver beads usually contain some base metals, i.e. zinc.

2.3. Add one drop of silver nitrate (AgNO3) solu tion to make a bead of about 15 mg.

3. Special assays (assays of complex or unusual ores): changes to the stock flux or standard assay method are often necessary with complex or un usual ores.

2.4. Mix the pulp and reagents in a crucible.

2.5. Place the crucible in preheated furnace at

10250C and heat for 35 minutes.

4. In the standard assay, the dilution of the pulp is approximately 1:5. However, an unusual chemistry of the pulp can still result in anomalous results. For example, excess silica may cause precious metals to be lost by slagging or matte formation. A knowledge of the pulp, fluxes, and reagents available will allow for the adjustment of the assay process to give accurate and reproducible results.

2.6. The position of each sample in the furnace is noted (it is not possible to write on the crucible unless a GRAPHITE pencil is used).

2.7. Pour and inspect the crucible for lead loss.

2.8. Note the slag colour for possible interferences.

2.9. Note the size and appearance of the lead but ton.

5. Other assays (bullion, umpire, and fraud case as says): these categories of assay, although identical to the regular gold and silver assays or special assays, require replicate analysis.

2.10. Break the slag and free the lead button (20-

25 g).

2.11. Cube the button with a hammer.

6. The various reagents (fluxes, reducing, oxidizing, and desulphurizing agents) and processes used during regular fire assay work are:

A) Acid fluxes such as silica (SiO^) and borax glass (Na2B4O7) for use with basic ores

2.12. Place the lead cube on a preheated cupel

(9500C) and heat in the furnace for ap proximately 30 minutes, until the lead is ab sorbed by the cupel, in the vented atmosphere.

B) Basic fluxes such as sodium carbonate

(NaCO3) and litharge (PbO) for use with silicic ores.

2.13. Remove the silver prill and note any peculiarities.

2.14. Pass the bead on to the Chemistry subsection for graphite furnace AA analysis.

C) Oxidizing agents such as potassium nitrate

(KNO3) used with strongly reducing ores such as sulphides, by roasting at 6500C.

Quality Control:

D) Reducing agents such as carbon-flour and

Argol used with strongly oxidizing ores.

The determination limit for Au is 2 ppb, and l ppb for

Pt and Pd.

E) Desulphurizing agents such as Na2C03, PbO, and KN03 help to remove excess sulfur from the charge.

The precision (2o) at 20 ppb is 10 ppb for Au, and 5 ppb for Pt and Pd.

MS 14-3

r ire Assay-

Blanks and the MRB-27 standard reference material are processed with every batch of samples for quality control purposes.

Additional Notes:

STAGE 1:

If the pulp contains ^0 Wt. 9fc S, then the sample size is selected so that the button contains only 10 g sulphur, i.e. if the sample contains 309fc S, then the amount of pulp used is 1000/30 = 33.3g.

1. For samples with low concentrations OcO.O l ozAon

Au, Pt or Pd and -cO.lO oz/ton Ag) of precious metals, it is preferable to use the dedicated low blank furnace.

If l Wt. 9fc < S < 20 Wt. 7c, 50 g of pulp is used, and sufficient S is added so that the button will contain 10 g sulphur,

2. Other assays (bullion, umpire, and fraud case as says): these categories of assay, although identical to the regular gold and silver assays or special assays, require replicate analysis.

i.e.,

Wt. S added (g) = content ofpulp(Wt.

Method 3: Nickel sulphide fire assay

Procedure:

If the S content is less than l Wt.%, only 40 g of sample is used, and 10 g of S is added.

The noble metals are inhomogeneously distributed in rocks and ores, forming discrete noble metal minerals and possibly occurring in solid solution in rock-form ing minerals, chromites, and sulphides. The nickel sulphide fire-assay technique quantitatively collects all the platinum group metals (PGE) and gold from a large sized sample. The bead is dissolved in 12 N HC1; the residue is collected on filter paper and irradiated.

STAGE 2:

The nickel content of the button should be ea. 16 g. To achieve this, the amount of Ni added is given by:

Wt. Ni added(g) =

Wt. Ni in sample(g) x Wt. sample(g)

100

The noble metals on the filter paper can be determined by instrumental neutron activation analysis. Full details of apparatus, reagents, and sample preparation are given in Hoffman et al. (1978).

STAGE 3:

Silica is added to the charge:

3.1. Preparation of charge

3. l .2. Samples are prepared in duplicate such that a

24-34 g button is prepared which contains ea.

16 g Ni, 10 g S, and all noble elements. The charge composition is calculated in four stages:

If S^ Wt. 9fc, then no silica is added.

3.1.1. Samples are ground following standard pro cedures (pages MS l-2 to MS l-5), and sub mitted for geochemical analysis for the elements Ni, Cu, Co, Cr, and S. Nickel and sulphur values are critical in the determination of the charge composition. Elevated copper values may indicate low concentrations of some of the PGE. High chrome samples may indicate the presence of chromite which will interfere with the precious metal recovery.

If 8 Wt. 9SKS*cl5 Wt. 9fc, then 5 g of silica is added.

If 15 Wt. 9SxS^O Wt. 9fc, then 10 g of silica is added.

If 30 Wt. 9fc^40 Wt. 9fc, then 13 g of silica is added.

STAGE 4:

For each sample, 60 g of Na2B04 and 30 g of Na2CO3 is added to the charge.

3.1.3. A standard reference material (SARM-7) together with one of the MRB PGE standards

(MRB19-22) is included with each batch of samples. The SARM-7 standard may be prepared by mixing 9 g S with 15 g Ni, and adding 35 g of pulp. 60 g Na2BO4 and 30g

NaCO3 are also included, but no silica is added.

MS 14-4

Fire Assay-3

3.1.4. A 30 g assay crucible is used for the charge, and the components are thoroughly mixed.

The position of the charge in the furnace is noted for each sample to avoid confusing the samples (it is not possible to write on the crucible unless a GRAPHITE pencil is used).

3J. Determination by INAA

The PGE can be determined using a procedure involv ing two counts following Hoffman et al. (1978).

Rhodium and palladium can be determined during the first count, and the remainder of the PGE during a second count.

3.1.5. The charge is fused for 90 minutes at 10000C in the Globar furnace. When the reaction is complete, the charge is removed from the furnace and cooled.

3.1.6. The nickel sulphide button is removed from the crucible with a hammer, and weighed.

Rhodium and palladium are determined by irradiation of samples serially for 5 minutes and then allowed to decay for 60 seconds so that the very short-lived radioisotopes can decay. Samples and standards were counted for 200 s for the 109mPd and 104mRh gamma-rays. Peak areas, corrected for background, are compared for samples and standards.

3.1.7. The nickel sulphide bead is broken with the

Rocklabs 'little smasher' (a hand held steel pulverizer), and the fragments are milled in zirconia or alumina mills to produce a fine powder of -70 mesh.

3.1.8. The sample is weighed to determine the loss during grinding, and samples are passed on to the Chemistry subsection for the concentra tion of the PGE, before determination by

INAA.

Palladium, platinum, osmium, ruthenium, iridium, and gold are determined up to 40 days after irradiation.

Samples, ore standards, and elemental standards are sealed in plastic irradiation canisters and irradiated for

16 hours. Integrated peak areas, corrected for back ground, are used together with yields from crushing procedure, decay constant, irradiation time, counting time, and duration of count, to determine the con centrations in the unknowns.

3.2. Concentration of the PGE

Full details of the counting procedure are provided at the SLOWPOKE reactor facility.

The procedure for the concentration of the PGE from a crushed nickel sulphide button is outlined below:

3.4. Data Reduction

3.2. l. The crushed button is heated with 400 ml of

12 N HCL in a covered beaker on a hot plate in a fume hood until the sample completely dissolves.

Two IBM-PC BASIC programs are used to reduce data; they prompt the user for information and input data. Copies of the programs are available from the

Geoscience Laboratories.

Quality Control:

3.2.2. The solution is cooled and filtered under vacuum and washed with about 400 ml of distilled water. It is imperative that all the hydrochloric acid is removed; that includes any acid caught at the base of the vacuum filter apparatus. The compositional character of the filter paper is very important; Millipore (SWP

047-00) SM-type 5 micron 47 mm filter papers are best suited for this technique.

All samples and standards are analyzed in duplicate.

SARM 7 and at least one of the MRB standards (MRB-

19 through 23) are included with each batch of samples.

Bibliography:

3.2.3. The filter papers are folded twice. The quad rant is then sealed inside special plastic using a bag-sealing unit. The bag is then itself sealed within another bag ready forthe irradia tion and counting procedure. NOTE that it is important that all bag seals are good and stur dy; this will avoid the escape of radioactive material from the bag during rough handling subsequent to the irradiation.

Hoffman, E.L.,Naldrett, A.J.,Van Loon, J.C.,Han- cock, R.G.V., and Manson, A., 1978, The Determina tion of all the Platinum Group Elements and Gold in

Rocks and Ore by Neutron Activation Analysis after

Preconcentration by a Nickel Sulphide Fi re-Assay

Technique on Large Samples, ANALYTICA

CHIMICA ACTA, Vol. 102, pp 157-166.

MS 14-5

r tre f\ssuy-j

TABLE FA3. DETERMINATION LIMITS .

Element

Determination Limit (ppb)

TABLE FA4. PRECISION (2a).

Element Concentration (ppb)

Precision f'%)

Rh

Pd

Pt

Ir

Ru

Os

Au

1

50

5

0.1

3

2

0.1

( 200 s count)

( 200 s count)

(2500 s count)

(5000 s count)

(5000 s count)

(5000 s count)

(5000 s count)

Rh

Pd

Pt

Ir

Os

Ru

Au

135

320

290

75

30

150

42

5

12.5

10

6.7

10

13

14

Precision (2a) values quoted by Hoffman et al (1978), based on multiple analysis of an internal standard (L. S.

4) are:

Accuracy may be judged with reference to the ex pected values of SARM 7, and with reference to the results for the MRB-series standards. These values, expressed in ppb, are shown in Table FA5.

TABLE FAS.

Standard

SARM 7

MRB-19

MRB-20

MRB-21

MRB-22

MRB-23

Au

(ppb)

310

450

269

5616

1344

2810

Pt

(ppb)

3740

728

1395

69872

23551

31643

Pd

(ppb)

1530

1784

1154

243929

71219

90126

Os

(ppb)

63

10

24

271

101

145

Ir

(ppb)

74

37

48

287

84

126

Ru

(ppb)

43

65

108

730

25

339

Rh

(ppb)

24

228

149

2362

5876

915

MS 14-6

Fire Assay-3

i 50 mm

(approx.)

(approx.)

5O mm

(approx.)

Figure FA l.

MS 14-7

Norms

NORMATIVE PROGRAMS

Introduction:

centages are then assigned to the normative minerals according to the steps in the norm calculation.

The CIPW system of rock classification was formu lated about the turn of the century by Whitman Cross,

Joseph Iddings, Louis Pirsson and Henry Washington.

TABLE NORM1.

The system recalculates the bulk chemical composi tion of a rock into a hypothetical assemblage of stand ard (normative) minerals. The calculation of the norms for different types of rocks, provides a method of classification by quantitatively studying and com paring the norms of unknown rocks.

The elements used in the calculations are usually ex pressed in the oxide form, and are given in Table

NORM l. The CIPW norms (weight norms) are calcu lated by assigning amounts of the molecular properties to a standard set of mineral molecules as outline in

Table NORM2. The molecular properties are the weight percent of the elemental analysis divided by the molecular weight.

Elemental

Oxide

SiO2

A12O3

FC2O3

FeO

MgO

CaO

NaaO

K2O

TiO2

P205

S

MnO

Approx.

Molecular

Weight

60

102

160

72

40

56

62

92

80

142

32

71

Elemental

Oxide with

Cation

Si02

AlOi.5

FeOi.5

FeO

MgO

CaO

NaOo.5

KOo.5

TiO2

P02.5

S

MnO

Approx.

Molecular

Weight

The percentage of each mineral molecule is then cal culated by multiplying the molecular amount of each constituent of each mineral by its molecular weight.

Normative Program Algorithm

The Swiss mineralogist and petrographer, Paul Niggli, introduced an important change to the norm calcula tion by discarding the weight units and introducing the

"equivalent molecular unit". These are based upon the number of cations in the mineral or oxide, e.g.

Alteration of Raw Chemical Data

1. The Fe2O37FeO ratio is adjusted according to the equation:

7cFe2O3

= %TiO2 + 1.5

CaO.Al2O3.2SiO2 ~ 5 An

(formula has 5 cations) l An is equal to a quantity of anorthite containing one cation. Cation equations can thus be written easily, e.g.

If Fe2O3 is less than this, no change is made; if it is greater, the excess is converted to FeO.

2. The total percentage is recalculated to 100 without the volitiles H20*, H2O' and C02.

4 En = 3 Fo * l Q

3. The cation 7o for MnO is added to that of FeO.

2(MgO.SiO2) = 2MgO.SiO2 + SiO2

Algorithm

In the norm calculations with Niggli's method, the molecular proportions are abandoned in favor of cation proportions. With this method all the elemental oxides constituents are reduced to one cation (See Table

NORMS).

Once the chemical data have been adjusted, as indi cated above, the program calculates the Norms based on the following steps:

1. Apatite is formed from P and 1.67 times this amount of Ca.

To calculate the cation proportions, the weight percent is divided by the equivalent molecular weight and multiplied by 1000. These figures are added and per centages of each cation iscalculated. The cation per

2. Pyrite is formed from S and half this amount of Fe.

80

71

32

71

40

56

31

47

60

51

80

72

MS15-1

Norms

TABLE NORM2. THE NORM MINERALS

USED

Salic

Group

Quartz

Corundum

Orthoclase

Albite

Anorthosite

Leucite

Nepheline

Kaliophilite

Femic Group

Wollastonite

Enstatite

Ferrosilite

Acmite

Magnetite

Hematite

Ilmenite

Apatite

Pyrite

Hypersthene

Diopside

Olivine

Hedenbergite

Q

C

OR

AB

AN

LC

NE

KP

WO

EN

FS

AC

MT

HM

IL

AP

PY

HY

DI

OL

HE

SiO2

A12O3

K20.Al2O3.6SiO2

Na20.Al2O3.6Si02

CaO.Al2Os.SiO2

K20.Al2O3.4SiO2

Na2O.Al2O3.2SiO2

K2O.Al203.SiO2

CaO.SiO2

MgO.SiO2

FeO.SiO2

Na2O.Fe2O3.4SiO2

FeO.Fe2O3

FC2O3

FeO.TiO2

3 (CaO.P2O5).CaF2

FeS2

(Mg,Fe)O.SiO2

CaMgSi2O6

2 (Mg,Fe)O.SiO2

CaFeSi2O6

Table NORM3.

Elemental

Oxide

Weight Equiv.

*7o Molecular

Weight

Cation Cation

Proportions 9fc x 1000

SiO2

AlOi.5

FeOi.s

FeO

MgO

CaO

NaOo.5

KOo.5

TiO2

P02.5

S

MnO

49.10

16.21

2.87

6.84

5.04

8.90

3.53

2.76

3.59

0.54

0.00

0.05

56

31

47

80

71

32

7

60

51

80

72

40

818

317

36

95

125

159

114

59

45

8

0

11

46.0

17.9

2.0

5.3

7.0

9.0

6.4

3.3

2.5

0.5

0

0.1

4. If CO2 is included in analysis, calcite is formed from CO2 and an equal amount of Ca (if required

CO2 can be included in the program).

5. Orthoclase and albite are formed provisionally from K and Na respectively combined in the right proportions with Al and Si.

6. If there is an excess of Al over K and Na, it is combined with the right proportions of Ca and Si to form Anorthite.

7. If there is an excess of Al over the Ca it is calcu lated as Corundum.

8. If in (5) there is an excess of Na over Al it is to be combined with an equal amount of Fe3* and twice as much Si to form Acmite.

9. The remaining Fe3"1" is assigned to magnetite with half this amount of Fe .

.3-1- -.

10. If there is still an excess of Fe , it is calculated as

Hematite.

11. Wollastonite is formed from the amount of Ca left and an equal amount of Si.

12. Enstatite and ferrosilite are formed provisionally from all the Mg and Fe2*.

13. If there is an excess of Si, it is calculated as Quartz.

If step (13) is true, the norm calculation is finished: otherwise, Si has been assigned beyond the original amount and is a negative quantity. The following steps bring Si back to O by the minerals of a lower degree of silification substituting in part or wholly for those minerals that were formed provisionally.

14. The amount of Mg and Fe2* used to form Enstatite and Ferrosilite are added and the relative amounts of Mg and Fe2* calculated using the ratios noted here.

Fe'

Mg

Mg + Fe

2+

Mg + Fez

15. If the amount of Ca in Wollastonite is greater than the total Mg+Fe all the Mg and Fe will be convened with Ca to Diopside and Hedenbergite respectively. Diopside (MgCa2Si02) is equal to 4 times the amount of Mg.

3. Ilmentite is formed from Ti and an equal amount ofFe.

MS 15-2

Norms

Hedenbergite (FeCa2SiO3) is equal to 4 times the amount of Fe2*.

18. If these still is not enough Si in the analysis, Albite is turned into Nepheline according to the equation:

5 AB = 3 Ne * 2 Q

Wollastonite is now equal to its original value minus 2 (Mg+Fe2*).

Enstatite and ferrosilite now equal 0.

19. If the analysis is very low in Si, Orthoclase is in part or wholly converted into Leucite.

16. If the amount of Ca in Wollastonite is less than the total Mg+Fe2* all of the Wollastonite will be con verted to Diopside and Hedenbergite with the right proportions of Mg and Fe2"1".

5 OR = 4 LC + l Q

20. In rare cases there is not even enough Si to form

Leucite.

Diopside =

Mg + Fe

u . , .

f-——

Then Kaliophilite is formed.

4 LC = 3 KP + l Q

Bibliography:

The remainder of Mg+Fe2"1" ^ to the original minus

Ca of Wollastonite) is calculated as Hypersthene.

There is no change in total Si in the above 3 steps because the minerals have been converted to other minerals with the same silicon content.

Earth, T.F.W., Calculations and Classification in

Theorectical Petrology, John Wiley and Sons, New

York, 1952, pp 76-82.

17. The necessary amounts of Hypersthene (if any) are converted to Olivine according to the equation:

Irvine, T.N., and Baragar, W.R.A., A Guide to the

Chemical Classification of the Common Volcanic

Rocks, Canadian Journal of Earth Sciences, Vol. 8,

1971, pp. 525-526.

4 HY = 3 Ol * l Q

Johannsen, A., Calculation of the Norm in a Descrip tive Petrography of the Igneous Rocks, Vol. l, 1950, pp. 88-92.

MS 15-3

Alknorm

ALKNORM FOR FELDSPATHOIDAL AND

MELILITIC IGNEOUS ROCKS

Introduction:

Calcium orthosilicate is not calculated since it is quite unrealistic (Washington, 1915).

Chayes and Yoder (1971) have underlined some of the inadequacies of the CIPW normative system when dealing with feldspathoidal and melilite bearing lavas.

King (1965) experienced the same inadequacies for strongly alkaline undersaturated plutonic rocks. As a result, a calculation similar to mat for the standard

CIPW norm (Washington 1917; modified by Barth

1931) has been devised for these rarer igneous rocks.

3. The significance of the roles of Ca and Fe, as well as the alkalis is demonstrated. In particular, in stead of using all the ferric iron early in the cal culation to form acmite and magnetite, as in the

CIPW system, here some is retained till later to enable the possible formation of andradite.

The new calculations have been specifically created for ijolitic, nephelinitic, and melilite-bearing suites of igneous rocks but is also applicable to other under saturated and strongly alkaline rocks which lack Ca- plagioclase. This would include the peralkaline syenites, alkali pyroxenites, peridotites, and nepheline syenites; ultra-alkaline rocks and volcanic equivalents of the classificatory system described by Nockolds

(1954); and many of the feldspathoidal peralkaline and metaluminous rock types defined by Shand (1949).

The criterion governing the use of this norm is that it is appropriate for undersaturated rocks when either or both acmite or calcium ortho-silicate (CS) appear in the CIPW norm.

The main features of this norm are:

4. It allows more realistic comparison with the mode and facilitates geochemical and mineralogical in terpretation. However, amphibole and micas are not calculated, nor is ilmenite because it is not normally a common constituent in strongly alkaline and oxidized magmatic rocks:

Algorithm:

1. Determine the molecular proportions of the con stituents in the chemical analysis by dividing the weight percentage of each by the molecular weight. Add MnO to FeO, and SrO and BaO to

CaO.

2. If molecular SiO2 > FeO * MgO + CaO ± 6(Na2O

+ K2O), then omit stages 3ei, 3eii, 3eiii; 4A; 4B;

4C; 4E; and 4K.

1. It allows the calculation of nepheline as

Na3KAl4Si4O16, the formula preferred by Hamil ton and Mackenzie (1960) and Earth (1963). In the CIPW norm, nepheline is calculated free of the kalsilite molecule (KAlSiO4), which frequently gives rise to leucite in the norm and unnecessary silica deficiency elsewhere.

3. Allot molecular amounts in the order and in the ratios listed below.

a. P2O5 : CaO ~ 3:10 for apatite b. CO2 : CaO = 1:1 for calcite

C02 : MgO = 1:1 for magnesite

2. It favors the formation of the minerals pervoskite, sphene andradite, kalsilite, and melilite (both aker manite and gehlenite end members). A more ac- tualistic melilite (i.e. showing the replacement of

Ca by Na, and Al by Fe3*) is not calculated because the geochemistry of melilite is not clear. The formation of the iron-bearing akermanite by desilicating diopside in IV, J of the calculation is governed by the relation: c. SO3 : Na2O - 1:1 for themardite d. F:Ca(O) = 2:1 for fluorite ei. K2O : A12O^1:1 ii.A!203 :Na2Osl:l

4 Ca(Mg,Fe)Si2O6 - 3SiO2 2Ca2(Mg,Fe)SiO7 +

(Mg,Fe)2SiO4 iii.combine i and ii to give

(K2O.Al2O3):(Al2O3.Na2O^ 1:3 for neph eline

MS16-1

AiKnorm

iv. residual K2O.A12O3 for kalsilite, leucite or orthoclase v. residual Na2O.Al2O3 for carnegieite or al bite f. residual Na2O:Fe2O3 - 1:1 for acmite g.residual Al203 :CaO = 1:2 for gehlenite h. residual A12O3 for corundum i. residual Na2O for Na2SiO3 j.TiO2:CaO= 1:1 for sphene or pervoskite (or rutile i f no CaO)

J. In 3, CaO.(MgO + FeO).2SiO2 for diopside

(wo, en and fs); if SiO2 is insufficient, dis tribute between x molecules diopside, y molecules akermanite (2CaO.(Mg,Fe)

0.2810^, and y/2 molecules olivine, given by

2x 4- 2.5y = available SiO2; and olivine formed to I. (N.B. Akermanite is recalculated in the norm as both the ordinary akermanite molecule Ca2MgSi2O7, and the Fe-akermanite molecule, Ca2FeSi207 in a manner similar to that for olivine.

K. If silica is insufficient for stages A-I; create the normative molecules in A-I and akemanite and olivine in J; and declare the silica deficien cy (Si-def); if, however, silica remains, then continue allocation of SiO2.

L. ZrO2.SiO2 for zircon; if no silica or insuffi cient, then remainder of ZrO2 for baddeleyite.

M. In 10, CaO.TiO2.SiO2 for sphene, or as far as possible (remainder left as pervoskite k.If FeO remaining CaO, then allot excess

FeO:Fe2O3 = 1:1 for magnetite; if not then continue to 12.

1. (FeO + MgO) :CaO = 1:1 for diopside (the ratio of MgO and still available FeO is calcu lated, and is maintained in 12/13, and any akermanite subsequently formed) m. residual (FeO -H MgO) for olivine n. Fe2O3 : residual CaO = 1:3 for andradite o. residual Fe2O3 for hematite, or residual CaO for wollastonite

4. Allot silica to make the following normative minerals:

N. Convert B to K2O.Al2O3.4SiO2 for leucite or if possible K20. Al2O3.6SiO2 for orthoclase

(if the available SiO2 is not sufficient for the conversions, then distribute between x molecules leucite and y molecules kalsilite, given by x * y = available K20 and 4x + 2y s available SiO2 or between x molecules or thoclase and y molecules leucite, given by x+y

= available K2O and 6x + 4y = available SiO2, as appropriate)

A. In 3, 3Na2O.K2O.4Al2O3 .8SiO2 for nepheline

B. In 3, K2O.Al2O3 .2SiO2 for kalsilite

C. In 3, Na2O.Al2O3.2SiO2 for carnegieite

D. In 3, Na2O.Fe2O3.4SiO2 for acmite

E. In 3,2CaO.Al2O3.SiO2 for gehlenite

F. In 3, Na2O.SiO2 for Na2SiO3

G. In 3, 3CaO.Fe2O3.3SiO2 for andradite

H. In 3, CaO.Si02 for wollastonite (wol)

O. Convert C to Na20. Al2O3.6SiO2 for albite; if insufficient Si02, then distribute between x molecules albite and y molecules carnegieite, where x * y = available Na2O, and 6x -i- 2y = available SiO2.

P. If silica remains, convert nepheline A to albite -H orthoclase; if the available SiO2 is not sufficient to convert all the nepheline, then distribute between x molecules albite -f or thoclase, and y molecules nepheline, given by x 4- 4y = available Na2O * K2O in A, and 6x

+ 8y = available SiO2 (but, maintain

Na2O:K2O = 3:1 in nepheline); add any or thoclase formed to N and albite to O

Q. Any residual SiO2 is calculated as quartz.

I. In 3,2(MgO + FeO).SiO2 for olivine (fo and fa) 5.

Determine the percentage weights of the norma tive minerals by multiplying the molecular propor-

MS16-2

tions of the minerals in 3 and 4 by the appropriate molecular weight.

Bibliography:

Le Bas, M.J., A Norm for Feldspathoidal and Melilitic

Igneous Rocks, Journal of Geology, Vol 81,1973, pp.

89-96.

Alknorm

MS 16-3

Protocols

PROTOCOLS FOR SAMPLE DIGESTION AND ANALYTICAL

DETERMINATION IN THE ELEMENTAL ANALYSIS SUBSECTION

Introduction:

Selection of Determination Technique

In general, the instrumental techniques used for each package are:

Selection of Samples to be Monitored with the

Screen Program

Samples received for analysis in the Geoscience

Laboratories are classified: (1) under assigned job numbers or (2) as "assay" samples. Jobs comprise up to 50 samples and can require extensive analytical work including up to 70 elemental determinations for each sample. Priorities for these jobs are assigned by the Chief in consultation with the client geologist.

"Assay" samples require a faster turnaround time and can require as little as one elemental determination; clients include the private sector and priorities for completion are set immediately and updated weekly.

1. AU samples requesting the complete T2 and T4 package or any part of such packages require a

Screen evaluation.

2. All samples requesting barium and chromium

(AAS) or tin (ICP-MS) require a Screen evalua tion.

Elemental analysis is divided into major element analysis, and trace element analysis. Each has a num ber of packages (M l, T l, etc.) which for convenience, group elements according to client requirements or analytical technique. A complete description of these analytical packages can be found in the Geoscience

Laboratories "Analytical Capabilities and Services" booklet.

3. All samples requesting major analysis which have a sulphur content 0. 3 ?o - 3.07c, should be checked for possible high arsenic using the Screen pro gram. Although arsenic is not determined with the

Screen program, its presence will be indicated as a high lead value.

Major Element Analysis

1. A carbon/sulphur determination is made on all samples for which any of the major element pack ages are requested.

In addition to the elemental packages, an XRF Screen is implemented as an unreported monitor package to check for acid-resistant minerals with known acid digestion problems and unusual matrices.

2. If the sulphur content is ^.39fc, proceed with fusion and determination by XRF.

3. If sulphur content is 0.3 9fc - 3.09fc, roast the sample prior to fusion. Request a Screen evaluation on the sample for arsenic content. Proceed with fusion and determination by XRF if the sample contains no appreciable arsenic.

M l, M2, M3 XRF - alternate techniques include

AAS and classical analysis involving gravimetric, volumetric and colori metric techniques.

4. If sulphur content is ^.09fc, or high arsenic is suggested with the Screen evaluation, pass the sample to AAS for preparation and determination.

Tl

Major element analysis will be performed by

AAS when

AAS - including flame, graphite fur nace and hydride generation techni ques.

l .

Less than 3 elements per sample are requested.

T2

T3

T4,T5

Screen

ICP-OES

XRF

ICP-MS

XRF

2.

Sulphur content is ^.

and/or arsenic content is suspected to be high. DO NOT attempt to fuse these samples in platinum crucibles. Damage to the crucible can occur with some sulphide bearing samples. A few hundred ppm As will scour a crucible; a few percent will destroy it.

3. Silica content is between 10 and 30 percent.

EA1-1

Protocols

Programs for XRF determination in this range are not reliable.

Selection of Sample Weight/Volume

4. Iron content is high O507c as Fe2O3) and a glass bead for presentation to an XRF spectrometer is difficult to prepare.

1. A 0.500 g sample, digested, and made to a final volume of 25 ml is the norm for requests for T l and T2 elements. All subsequent dilutions are the responsibility of the personnel performing the final determination. Dilution factor 50.

Selection of Digestion Procedure

Acid attack:

2. Assay samples whose analytes are determined by flame AAS, use a 0.500 g portion of sample made to a final volume of 50 ml. Dilution factor 100.

1. Requests for Tl options and/or the T2 package, options or additions, require a HNO3-HC104-HF attack. Any element determined by ICP-OES re quires this acid attack.

3. The T4 package requires a 0.200 g of sample made to a final volume of 100 ml. Dilution factor 500.

2. Any "assay" sample requiring elements of the Tl options, which will be determined by Flame AAS only, may use a HC1-HF-HNO3 acid attack.

4. Silver and cadmium digestion requires a 1.000 g sample made to a final volume of 25 ml. Dilution f actor 25.

3. Requests for silver and cadmium require an acid digestion without the presence of HQ.

4. Requests for a T4 or T5 package require a detailed

HNO3-HC104-HF attack.

5. Less sample may be used and/or larger volumes chosen when elements normally considered as trace components are present at ore grade levels.

These new dilution factors must be flagged for the attention of the instrument operator performing the final determination.

5. Vegetation samples require dry ashing followed by mineral acid attack.

Quality Control:

Fusions:

1. Instrumental control and stability are monitored by the instrument operator to assure meaningful data acquisition. Appropriate solutions are chosen for this purpose.

1. Samples requiring the determination of tin are fused with lithium metaborate by the sample preparation group and passed to ICP-MS for deter mination.

2. Sample residues, which indicate incomplete acid attack, are fused with an appropriate flux. The indication may be visual or after Screen evalua tion.

2. In-house or certified reference materials and reagent blanks are used with each batch of samples decomposed. These are chosen in consultation with the instrument operator and should account for approximately 107e of the samples decom posed. Quality control data relating to accuracy and precision determined with these reference materials are kept by the instrument operator and should be available upon request.

3. For accurate work, use an acid attack followed by a fusion of any residue using a minimum of flux.

The two solutions are combined and analyzed by the appropriate technique.

Data Evaluation

NOTE: Solutions presented to ICP-OES and ICP-MS must indicate the approximate total salt con tent.

4. Samples known to contain high barium (barite) and chromium (chromite) are fused with lithium metaborate without prior acid attack. If the chromium content is 5^c, sodium peroxide is used as the flux. Only Flame AAS is used for measure ment when the salt content of the solution is high.

1. Duplicate determinations are checked for precision. A discrepancy may be due to in- homogenity of the sample, incomplete decomposi- tion caused by acid-resistant minerals, contamination, loss of instrument control, or dilu tion errors. Each source of potential problem is investigated. If the problem is identified as one that may have affected all samples in a batch, the entire batch must be rerun. If the problem is clearly identified as affecting only the one sample

EA1-2

Protocols

it is repeated and the whole batch need not be rerun.

2. Check the standard reference materials run with each Screen program to determine the accuracy obtained.

2. Determined elemental values are checked with values obtained by the Screen program to indicate the possible presence of acid-resistant minerals.

Pay particular attention to those samples which have been flagged by the sample preparation group as containing undecomposed material after acid digestion. If an inconsistency is indicated the sample is redecomposed using an appropriate fusion method.

3. Adjust analytical values obtained for each sample using the reference material values. For example, if a SRM containing 800 ppm B a yields a result of

500 ppm with the Screen program, all Ba screen values should be adjusted by a factor of 800/500 = l .6 to give a better approximation of the Ba content in the samples near this concentration. DO NOT change any values obtained by other techniques.

3. Evaluation of quality control data is monitored by the instrument operator and kept for the

Supervisor's attention.

Screen Evaluation

4. A significant discrepancy exists when the adjusted

Screen value is higher than the AAS or ICP result by 2096. Significantly lower Screen values, com pared with AAS or ICP results, are brought to the attention of the XRF personnel.

l. Values obtained with the XRF Screen program are approximate.

The decision sequence for major element determina tion in the elemental analysis subsection is sum marized in Figure PROT1.

EA1-3

Protocols

^,

33*

Determine C/S

Are the number of elements requested

3 or less?

-NO-

•YES-

If S content is

Q.3-3%

Roast sample

^.OX

Determination to be made by appropriate methods including sample decomposition with acids or flux and measurement by

* AAS

* Flame photometry

* Colorimetry

* Titration

REPORT

Check for presence of substances which will damage Pt-ware, i.e.,

As, Sb, etc. by obtaining infor mation from Screen and client

Are any present?

-NO-

YES-

Proceed with fusion

Is a fused bead obtained?

YES

Proceed with XRF measurement

Is Si02 content?

DO NOT fuse in platinum

Fe content too high? >5Q%

Prepare pressed powder pellet and measure with XRF carbonate program

XRF calibration in this range is not accurate

Figure PROT1. Decision sequence for major element determination.

EA1-4

Sample Dissolution

SAMPLE DISSOLUTION

Introduction:

The purpose of any sample dissolution technique is to provide a homogeneous solution of the elements of analytical interest. The two most common techniques are acid digestion and flux fusion.

The advantages of acid digestion over fusion include easier operations, less total dissolved solids, less pos sible introduction of contaminants, and adaptability to the use of robotic systems to perform the digestion. Its main disadvantage is that some minerals are resistant to acid attack, and some elements can be lost in the process.

Fusion techniques are favored when acid resistant minerals are present and total dissolution is vital. The use of fluxes, however, is likely to introduce con taminants as well as providing a solution with high salt content. High salt contents are to be avoided where the sample delivery system can be clogged (ICP nebulizer systems).

A combination of the two techniques can be used for accurate work, but the time required makes its less amenable to batch operations. In this case acid diges tion is performed on a sample, followed by fusion of any residue. The solutions are combined to produce a total digestion with minimal salt content.

A list of the more commonly used acids and fluxes is given below.

Acids:

Hydrofluoric Acid:

Nitric Acid:

697c HNO3 (w/w), 16 M oxidizes organic matter prior to perchloric acid attack.

Sulphuric Acid:

H2SO4 (w/w), 18 M is used when a high boiling acid is required. Most effective in removing residual amounts of fluoride, but formation of sulfates can create problems with dissolution and with depressant effects during the determination by AAS.

Fluxes:

Lithium Metaborate/Lithium tetraborate:

Used individually or in combination when total dis solution is required. May be used with graphite or platinum crucibles.

Sodium Carbonate:

Na2CO3 fused in platinum crucibles is used primarily with classical rock analysis methods.

Sodium Hydroxide:

NaOH or KOH (potassium hydroxide) are used (usual ly with nickel crucibles) when a strongly alkaline flux is required.

Sodium Peroxide:

Na2O2 is a powerful oxidant used to attack spinels, zircon, and sulfides.

HF (w/w), 29 M is used to attack most silicates.

HF will attack glass. See Safety advisory.

Perchloric Acid:

607c HC104 (w/w), 12 M used with HF to decompose silicates and is effective in removing residual amounts of fluoride. See Safety advisory.

Hydrochloric Acid:

37 '9fc HG (w/w), 12 M dissolves most oxides, sulfides and carbonates. In combination with nitric acid (3:1) it forms aqua regia.

Safety advisory:

CAUTION: Exercise extreme care when using any acids and fluxes. Their use should only be attempted after the appropriate MSDS sheets have been read and the safe handling and first aid procedures understood.

Acids should only be handled in a fume hood with proper ventilation and with proper protective equip ment worn. Proper ventilation will be required when handling fluxes which create a dust control problem.

1. Hydrofluoric and perchloric acids require spe cial care in handling. Hydrofluoric acid (HF) may only be used after training in the proper proce dures. Its use is restricted to designated rumehoods. Its hazard lies in the nature of the bum

EA2-1

Sample Dissolution

which can result. Very close attention should be paid to suspected skin contact and treatment ap plied immediately. NEVER handle HF and

HC1O4 without gloves, proper eye protection, and a lab coat.

2. Non-glass labware must be used when handling

HF because the acid attacks silica in laboratory glassware. Perchloric acid may only be used in designated perchloric acid fumehoods with operable scrubbing and/or wash down facilities.

Washdown facilities are required in designated fumehoods to prevent the buildup of anhydrous perchloric salts which can become spontaneously flammable. Explosions can result when organic perchlorates are formed. Ensure that organic material, such as alcohols, organic solvents, and paper products are kept away from the digestion area. Organic material in the sample should be oxidized at a lower temperature with nitric acid prior to digestion with perchloric acid.

Decomposition by Acid Attack

Apparatus:

- Top-loading electronic analytical balance, reproducibility +I- 0.001 g

- Polytetrafluoroethylene (PTFE) beakers, 30 ml,

50 ml, 100 ml capacity

- Polypropylene culture centrifuge tubes, 50 ml, graduated every 5 ml

- Disposable polypropylene, culture tubes with caps, 17 x 100 mm, 16 ml capacity

- Hot plates, Thermolyne 12" x 12" or equivalent

- Teflon (PTFE) stirring rods

- Test tube racks

- Polypropylene acid dispenser, 500 ml

- Nalgene graduated cylinders, 250 ml and 25 ml

Reagents:

- Hydrofluoric acid, HF, 489fc (w/w)

- Perchloric acid, HC1O4, 609fc (w/w)

- Nitric acid, HN03, 697c (w/w)

- Hydrochloric acid, HC1, 37*7o (w/w)

Procedures:

1. Sample Preparation for T l Options and T2

Analytical Package (HNO3 - HC1O4 - HF)

This is a mixed acid attack, generally considered to provide a "total decomposition", and produce a solu tion for determination of trace elements by AAS or

ICP-OES. It is used to break down silicates in rocks, soils, and sediments.

1.1 Prepare an acid mixture to contain 100 ml of

HC1O4, 50 ml of HNO3, and 400 ml of HF using a Nalgene graduated cylinder for measuring the HF volume. Store in a clean polypropylene dispenser used to deliver HF acid, (see Note 1).

l .2. Weigh 0.500 g of sample into a 50 ml PTFE beaker.

1.3. Add 15 ml of acid mixture (see step 1.1) to the beaker using Nalgene labware for delivery.

1.4. Ensure complete wetting of the sample by swirling the beaker.

1.5. Place beakers on hotplates at 1500C and evaporate to dryness overnight.

1.6. Gently tap beakers to cause any droplets of acid condensed on the upper portion of the beaker walls to fall to the bottom.

1.7. Leave beakers on the hot plate until fumes cease.

1.8. Check for more droplets and repeat steps 1.6 and l .7 until acid droplets have been removed.

1.9. To the dry residue, add 2 ml of concentrated

HN03 and heat on hot plate for l minute.

1.10. Remove beaker and add 0.5 ml of con centrated HC1, allowing any vigourous reac tion to subside before placing beaker on hot plate.

1.11. Use a teflon rod to break up the residue while adding 15 ml of distilled water.

1.12. Heat on the hot plate for 15 minutes or until dissolution appears complete.

1.13. Examine the beaker for any incompletely decomposed residue and record its presence for that sample, (see Note 2)

1.14. Transfer the cooled solution to a 50 ml polypropylene centrifuge tube and bring to 25 ml mark using distilled water.

1.15. Cover the test tube with parafilm or a cap and mix well.

EA2-2

Sample Dissolution

1.16. Split the solution into two fractions using 17 x

NOTES:

100 mm polypropylene culture tubes for dis tribution to AAS and/or ICP-OES. (see

Note 3)

1. A clean, empty, supplier HF dispensing bottle may be used for this purpose.

2. The presence of minerals resistant to acid attack may be observed at this time. Check determined values against values obtained with XRF Screen and ascertain whether fusion is required for these samples.

3. Dilution factor for these fractions is 50. ICP-OES will dilute their fraction l: l using the liquid han dling apparatus.

2. Adaptation of Procedure l to a Robotic System

2.8.

The following is a description of the steps taken by a robotic system to provide a final solution identical to that produced by the manual approach given in l above.

2.5. After sample l, rack l has been in the low temperature hot block for 3 hours 30 minutes, the robot will transfer it to position l, of the high temperature block l, set at 1800C. (see

Note 2) The samples will remain at this posi tion for 17 hours.

2.6. If sample l, rack 2 (sample #31) is available, the robot will process this sample as in steps

2.1 and 2.2. This sample is then placed in the vacated position l of the low temperature hot block. Time required for this operation is approximately 3 minutes, (see Note 3)

2.7.

Operations 2.5 and 2.6 are repeated until all samples originally in the low temperature hot block are now in the first high temperature block, and all samples from rack 2 are in the low temperature block.

The robot will remove samples from the low temperature block to the second high tempera ture block after 3 hours 30 minutes has elapsed for each sample. It will remain in a rest posi tion for the full 6 1/2 minutes between samples.

The system requires a weighed sample to be placed at a home position (rack number and position). The

PTFE vessels have a maximum volume of 30 ml, can be capped, and have dimensions 65 mm (height) x 29 mm (diameter).

The system can handle 60 samples in two groups of 30.

Current programming allows only 48 samples to be processed as one group of 30 and one group of up to 18.

2. l.

2.2.

2.3.

Sample l, from rack l is moved to the dispens ing station and 8 ml of HF, l ml of HC1O4 and

2 ml of HNO3 is automatically added, under robotic control, to the PTFE vessel.

Sample is placed in position l on the low temperature hot block, set for 1500C, and al lowed to remain there for 3 hours 30 minutes,

(see Note 1)

Sample 2, from rack l is processed in the same manner as steps 2. l and 2.2.

2.4. All samples from rack l are treated until the maximum of 30 samples have been placed in the low temperature hot block.

2.9. After sample #1 has been in the high tempera ture hot block for 17 hours, it is transferred to the acid dispensing station where 2 ml of

HNO3 are added, a l minute pause introduced,

0.5 ml of HC1 is added, and after a l 1/2 minutes pause, and 0.5 ml of distilled water is added.

2.10. Sample #1 is placed in the low temperature hot block for 8 1/2 minutes, (see Note 4)

2.11. Time will allow sample #2 to be processed as insteps 2.9 and2.10.

2.12. Robot initiates next step for sample #1. (see

Note 4)

2.13. Sample #1 is transferred to dispensing station where 22.5 ml of distilled water are added to the vessel.

2.14. Sample #1 is transferred to the capping station, vessel is capped and placed in its home posi tion.

2.15. Robot has time to process sample #3 as in steps

2.9 and 2.10.

EA2-3

sample uissoiution

2.16. Robot must now process sample #2 as in steps

2.13 and 2.14.

2.17. Samples are processed in the above sequence until all vessels have been capped and placed in their home positions, (see Note 5) This completes the decomposition steps handled by the robotic system.

2.18. Capped vessels are mixed thoroughly, and split into two components for distribution to

AAS and/or ICP-OES measurement, (see

Note 6)

NOTES:

1. The rate limiting time for this initial step is 6 1/2 minutes. Sample 2 will not be processed until this time has elapsed.

2. The robot will note the time remaining of the 6 1/2 minutes before sample 2, rack l needs to be removed to the high temperature block. If suffi cient time is available to initiate another step in the program before this time has elapsed, it will do so.

(see step 2.6)

3. Sample #32 cannot be processed before #2 be cause of insufficient time (see Note 2) and because only one low temperature hot block is available and position 2 is still being occupied by sample #2.

4. This 8 1/2 minutes is another rate limiting time.

Another sample will be processed if sufficient time is available before the next step for sample #1 needs to be initiated.

5. Time elapsed for first sample to complete the process is 20.5 hours; for a batch of 48 samples the time is 24 hours.

6. Dilution factor for these solutions is 50. IGP-OES will dilute their fraction l: l using the liquid han dling apparatus.

3. Fast Sample Dissolution for Tl Options by

Flame AAS (HC1 - HF - HNO3)

This mixed acid attack requires less time to complete than procedure l which uses HQO4. It is used for material received and designated as an "assay" sample which require a faster turn-around time and usually fewer elemental determinations with less stringent ac curacy requirements. The solution resulting from this procedure is suitable for measurement with atomic absorption only.

3. l. Weigh 0.500 g of sample into a 50 ml PTFE beaker, (see Note 1)

3.2. Add 5 ml of HC1 and 5 ml of HF to the beaker, swirling the beaker to ensure complete wetting of the sample, (see Note 2)

3.3. Place the beaker on a hotplate at 1500C and evaporate to dryness (approximately 3 hours).

3.4. Add 5 ml of HNO3 to the dry sample and evaporate to dryness.

3.5. To the dry residue, add 5 ml of HNO3 and heat on the hot plate for l minute.

3.6. Add 20 ml of distilled water, and using a teflon stirring rod, break up the residue to aid in dissolution.

3.7. Continue to heat on the hot plate for 15 minutes or until dissolution appears complete.

3.8. Examine the beaker for any incompletely decomposed material and note its presence,

(see Note 3)

3.9. Cool and transfer the solution to a 50 ml polypropylene centrifuge tube and bring to the

50 ml mark with distilled water.

3.10. Cover the tube with parafilm or a cap and mix well.

3.11. Distribute samples for AAS determination.

(see Note 4)

NOTES:

1. If the sample is an ore material or i s dark in colour it is advisable to use a 100 ml PTFE beaker.

2. At this stage, if the sample contains sulfides, it will be necessary to add 3 ml of HNO3 prior to the addition of HF and cover the beaker with a watch glass. Allow the sample to reflux for 15 minutes on a hot plate, wash the watch glass with distilled water and add the HF.

3. The presence of minerals resistant to acid attack may be observed at this time. Check values deter mined by AAS with those obtained by the XRF

Screen Program. The XRF Screen will be avail able only if chromium or barium have been re quested.

EA2-4

Sample Dissolution

4.

4.

Dilution factor for these solutions is 100. Acid 4.7. concentration is 109fc HNO3.

Sample Dissolution for the T4 and T5 Analyti cal Packages

This preparation is a modification of Procedure l , and requires fastidious attention to detail. It was developed to provide a solution for the determination of rare earth elements and yttrium by ICP-MS. It is the procedure required when ICP-MS is used for analytical measure ment. Other elements determined with this solution are uranium, thorium, thallium, hafnium and tantal- lum.

4.1 Preparation of 10 ppm Ruthenium/ Rhen-ium

Standard

4.1.1. Pipet 20 ml of 1000 ppm Ru solution and 20 ml of 1000 ppm Rh solution into a 200 ml volumetric flask, and make to volume with

HNO3. (see Note 1)

4. l .2. Pipet 20 ml of the 100 ppm Ru and Re solution prepared in 4.1.1 into a 200 ml volumetric flask, and make to volume with lO^o nitric acid. Store solution in a polypropylene bottle.

4 J. Preparation of mixed acid solutions

4.2.1. Solution A: Mix 400 ml of HF, 40 ml of

HC1O4, and 40 ml of HC1 and store in a polypropylene acid dispensing bottle, (see

Note 2)

4.8.

4.9.

Place beaker on hot plate until acid fuming ceases.

Repeat steps 4.6 and 4.7 until all acid has been removed.

Add 15 ml of solution B to the dry residue, place beaker on the hotplate, and evaporate to dryness overnight.

4. 10. Repeat steps 4.6 to 4.8. (see Note 3)

4.11. Add l ml of HNO3 to the dry residue, cool, add

6 drops of HC1 (see Note 4), and 15 ml of distilled water.

4.12. Place the beaker on the hotplate, heat, and evaporate to a volume of 10 ml.

4.13. Add l ml of 10 ppm Rh/Ru standard solution to a 100 ml volumetric flask using a dispenser calibrated to accurately deliver l ml.

4.14. Transfer the sample solution from the beaker into the 1 00 ml volumetric flask containing the

Ru/Re standard, and make to volume with nitric acid.

4. 15. Mix well and transfer a portion of the solution to a 17 x 100 mm polypropylene test tube and cap.

4. 16. Solution is allocated to ICP-MS for measure ment. (see Note 5)

CAUTION: Be aware of the hazards noted at the beginning of this section dealing with hydrofluoric and perchloric acids.

4.2.2. Solution B: Add 30 ml of HC1O4 and 70 ml of HC1 to 380 ml of distilled water and store in a polypropylene acid dispensing bottle.

4.3. Weigh 0.200 g of sample into a 50 ml PTFE beaker.

4.4. Add 12-15 ml of solution A to the samples, and swirl the beaker to ensure complete wet ting of the sample.

4.5. Place the beaker on a hotplate set at 1 800C and evaporate to dryness overnight.

4.6. Gently tap the beaker to allow any droplets of acid condensed on the upper portion of the beaker walls to fall to the bottom.

NOTES:

l . 1 000 ppm Ru and Rh solutions are purchased from manufacturers supplying plasma standards.

2. A clean, empty, supplier HF dispensing bottle may be used for this purpose.

3. This second evaporation to dryness is necessary to remove the last traces of HF. Without this step, low values for the rare earth elements and thorium will result.

4. If the sample contains high iron, additional HC1 may be added to dissolve the iron oxides.

5. Dilution factor for this solution is 500, and con tains 100 ppb Ru and Re used in the internal standardization scheme.

EA2-5

Sample Dissolution

6. For accurate work, use this acid attack followed by a fusion of any residue using a miniumum efflux, combine the two solutions, indicate total salt con tent of solution, and present solution to A AS,

ICP-OES, or ICP-MS for measurement.

5.8. Cover the tube with parafilm or a cap and mix well.

5.9. Distribute the samples for AAS determina tion, (see Note 3)

5. Sample Digestion Requiring the Absence of

Hydrochloric Acid (HNO3 - HF)

This procedure is used when hydrochloric acid will prevent the quantitative dissolution of a required analyte. It can be completed within 6 hours and is used when silver, cadmium, and/or lead (high content) are requested. The resulting solution is suitable for AAS determination only.

NOTE: Preparation l with its longer decomposition

time may also be used if step 1.10 is deleted, i.e., HC1 addition is eliminated.

The presence of hydrochloric acid fumes in the vicinity of the decomposing samples will result in low silver results, caused by the formation of insoluble silver chloride. When cadmium in hydrochloric acid solu tion is taken to dryness, losses in cadmium have been observed. Lead chloride can form and precipitate if the lead content of the sample is ^000 ppm.

5.1. Weigh l.000 g of sample into a 50 ml PTFE beaker, (see Note 1)

NOTES:

The weight of sample may be reduced for samples known to contain high lead; an appropriate final volume of solution is chosen. In these cases the final dilution factor for the solution must be clearly indicated.

2.

3.

The addition of nitric acid may cause a vigorous reaction, with the release of large amounts of nitric oxide fumes. If the sample is suspected to contain sulphide material, moisten the sample with water before adding the nitric acid and cover the beaker with a watch glass after the nitric acid has been added. Free sulfur will collect on the surface of the liquid and may remain in the final solution present to the AA spectrometer. Care must be taken to avoid nebulizer blockage.

Dilution factor for these solutions is 25. acid concentration is lG-12%.

Nitric

6.

Ashing and Solution Preparation for Vegeta tion Samples

5.2. Add 5 ml of HNO3 and 5 ml of HF to the beaker, (see Note 2)

CAUTION: Be aware of the hazards associated with the use of hydrofluoric acid noted at the begin ning of this section.

5.3.

5.4.

5.5.

5.6. Heat for an additional 15 minutes or until dissolution is complete.

5.7.

After any vigorous reaction has subsided, place beaker on a hotplate at 1500C and evaporate to dryness (approximately 3 hours).

Add 5 ml of HNO3 to the dry sample and evaporate to dryness on the hotplate.

Add 3 ml of HN03 to the dry residue, heat on the hotplate for l minute, and add 15 ml of distilled water.

Cool and transfer the solution to a 50 ml polypropylene centrifuge tube and bring to 25 ml with distilled water.

This is a two-stage technique: (1) Dry ashing the vegetation material, and (2) Treating the ashed material with mineral acid to obtain a solution for

ICP-OES determination of the analytes.

Ashing in a muffle furnace at 5000 to 5500C is the technique most frequently used for the decomposition of organic matter. The loss of certain metals (lead, cadmium, mercury, arsenic, selenium) is possible through volatilization, the formation of insoluble sili cates, or retention on the ashing vessel. If the tempera ture for ashing is held at 4500C, volatile metals such as lead, zinc, and cadmium are normally retained. Reten tion on the ashing vessel is reduced if new vessels are

"conditioned" in the muffle furnace at 5500C before use, and kept specifically for ashing purposes.

6.1. Weigh 20.00 g of dry material into a pyrex beaker conditioned for ashing purposes, (see

Notes 1,2)

6.2. Position the beaker on a silica tray and place in the cool furnace.

EA2-6

Sample Dissolution

CAUTION: The furnace must be in a vented fume hood. Ensure that no smoke enters the room.

6.3. Set the furnace temperature control at 2000C and allow the temperature to rise. The furnace door needs to be held open a few centimeters to allow air to enter the ashing chamber and the smoke to be vented.

6.4. Raise the temperature of the furnace in 500 increments, once it has reached 2000C, hold ing each temperature until the formation of smoke has subsided. When an increase in temperature produces no additional smoke, charring is essentially complete and the fur nace doors may be closed.

6.5. Set the furnace to 4500C, and complete the ashing for 5 hours, (see Note 3)

6.6. When ashing is complete, allow the beakers to cool.

6.7. Add 3 ml of nitric acid, and using a fine jet of distilled water, wash the sides of the beaker

(about 5 ml).

6.8. Heat the beakers on the hotplate, and quantita tively transfer the solution to 50 PTFE beakers. Wash the pyrex beaker with small portions of distilled water and add the wash ings to the PTFE beaker.

6.9. Add l ml of perchloric acid to the PTFE beaker and evaporate the contents to dryness using a hotplate set at 1800C.

6.11. Replace the beaker on the hotplate until acid fumes cease.

6.12. Check for more droplets and repeat steps 6.10 and 6.11 until all acid has been removed.

6.13. Add 3 ml of HNO3 to the dry residue, and heat on the hotplate for l minute.

6.14. Remove the beaker and add a few drops of

HC1, allow the reaction to subside, and add 15 ml of distilled water.

6.15. Heat on the hotplate for 15 minutes or until dissolution is complete.

6.16. Transfer the cooled solution to a 50 ml polypropylene centrifuge tube and bring to the

25 ml mark with distilled water.

6.17. Cover the test tube with parafilm or a cap and mix well.

6.18. Add a portion of the solution to a 17 x 100mm polypropylene culture tube and issue to ICP-

OES for measurement of analyte concentra tions, (see Note 5)

CAUTION: Note the hazards associated with the use of perchloric acid at the beginning of this sec

tion, (see Note 4)

ATTENTION: At this time it is necessary to prepare

a reagent blank and solutions of certified reference materials SO-1, SO-2, SO-3, and SO-4 using proce dure 1. HF is required with these samples. When appreciable siliceous material is present, e.g., in leaves and stems of plant matter, hydrofluoric acid may need to be added. However, these materials also contain a high concentration of calcium and the formation of

CaF2 upon addition of HF will deter dissolution and removal of fluoride. If HF is to be used add 2 ml to the

PTFE beaker at this step.

6.10. Remove any acid droplets condensed on the walls of the beakers by gentle tapping of the beaker.

NOTES:

1. The sample weight and beaker size depends on the density of the vegetation material. Usually 20 g of sample and 100 ml beakers are used, however, 10 g or 250 ml beakers are alternate choices.

2. New beakers need to be conditioned at least three times before they are used for ashing. Condition ing involves heating the beakers in a muffle fur nace at 5500C for three hours and removing the beakers from the furnace to cool. Conditioned beakers are cleaned by adding l: l nitric acid to the beakers covered with a watch glass, and allowing the acid to reflux by heating on a hotplate. They are then thoroughly rinsed with distilled water and dried at 105-1100C.

3. The charred material in the beaker may be stirred and left overnight to thoroughly ash the sample.

4. The steps which follow are essentially the same as those used for the T l options and T2 analytical package (procedure A) but without HF addition.

5. Dilution factor for a 10 g sample is 2.5; for a 20 g sample, 1.25. Nitric acid concentration is

EA2-7

Sample Dissolution

Decomposition by Fusion

Apparatus:

- Porcelain crucibles (Coors High-Form, 30 ml)

- Graphite crucibles (32x29 mm, 30 ml)

- Zirconium crucibles with covers (Low-Form,

25ml)

- Magnetic stirrers with teflon stirring bars

- Thermolyne Muffle Furnace

- Fire and heat retardant mittens

- Crucible tongs

- Silica tray

- Nalgene beakers, 250 ml

- Borosilicate glass beakers, 250 ml

- Filtering funnels

- Whatman #41 filtering paper, 12.5 cm

- Top-loading electronic analytical balance, reproducibility +I- 0.001 g

- Volumetric flasks, 100 ml, 200 ml

Reagents:

- Lithium metaborate, anhydrous, LiBO2

- Lithium tetraborate, anhydrous, Li2B4O7

- Sodium peroxide, granular, Na2O2

- Nitric acid, HNO3, 699fc (w/w)

- Graphite powder

NOTE: Use only Na2O2 that is protected by plastic

from the metal storage container.

Procedures:

1. Fusion with Lithium Metaborate/Lithium

Tetraborate

Lithium metaborate is a suitable flux for rapidly decomposing silicates and oxygen-containing minerals such as spinel, chromite, ilmenite, cas siterite, rutile, zircon, gahnite (Zn-spinel), and other resistant minerals.

The lithium borate fluxes do not damage platinum ware if oxidizing conditions are maintained, and allow the melt to be poured from the crucible, either into a platinum mold or into a mineral acid solution. The tetraborate salt is utilized to prepare glass buttons in a platinum mold for presentation to an XRF spectrophotometer. The metaborate flux can be sub stituted for tetraborate, offers a lower fusion tempera ture (8490C versus 9170C), and provides a more fluid melt for pouring and consequently less retention of material in the crucible.

Graphite crucibles are normally used when the melt is to be poured quantitatively into a mineral acid solution, typically 49fc HNO3, for presentation to solution-based instrumentation.

The tetraborate salt, with a relatively higher acidity than the metaborate salt, is more suitable for fusions of basic rocks and minerals (e.g., dolomite). For acidic rocks (high silica) the metaborate salt is preferable.

Flux to sample ratios are usually 5:1.

1.1.Fusion for determination of Barium and

Chromium

This fusion is suitable for barium concentrations of

0.05-40.0 percent and chromium concentrations of 100 ppm to 5.0 percent when determined by AAS. This procedure may also be used on any residue following an acid digestion; the weight of flux being dependent on the amount of residue present.

1.1.1. Weigh 0.200 g of sample into a porcelain crucible, (see Note 1)

1.1.2. Add 1.0 g of lithium metaborate and mix sample and flux with a small teflon coated spatula, (see Note 2)

1.1.3. Transfer quantitatively to a graphite crucible and place on a silica tray, (see Note 3)

1.1.4. Fuse in a muffle furnace for 30 minutes at

10500C. (see Note 4)

1.1.5. Transfer the molten fusion mixture to a 250 ml

Nalgene beaker containing 100 ml of 12 9fc nitric acid.

1.1.6. Dissolve the sample using a magnetic stirrer and teflon coated stirring bar. Complete dis solution requires about 20 minutes, (see

Note 5)

1.1.7. Filter the solution using #41 Whatman filter paper (12.5 cm) into a 200 ml volumetric flask.

Wash the filter paper several times with dis tilled water and make to volume with same.

1.1.8. Solution is sent to AAS for determination of required element. Further preparation for determination will be found in the section

"Trace Elements Determined by Flame

Atomic Absorption".

EA2-8

Sample Dissolution

NOTES:

1. If the samples contain appreciable amounts of sulphide, weighed samples in porcelain crucibles should be roasted in a muffle furnace at 6500C for three to four hours before mixing with the borate flux. Organic matter should be ashed.

1.2.8. Transfer a portion of this solution to a 17 x 100 mm polypropylene test tube, cap, and circulate to ICP-MS for the determination of Sn. (see

Note 3)

NOTES:

l . It is important that samples should be stirred and made up to volume immediately to prevent the formation of polymeric hydrated silicic acid.

2. For samples known to conta^ 3C^ iron oxide, add about 15 mg of pure graphite powder to the porcelain crucible containing the flux and sample.

This will prevent the fused melt from sticking to the bottom of the graphite crucible during pouring.

3. New graphite crucibles need to be conditioned in a muffle furnace at 10000C for 15 minutes.

4. It is important to maintain the fusion for 30 minutes for barium sulphate to be completely at tacked.

5. If the dissolution is not allowed to take place at room temperature, the silica may form polymeric hydrated silicic acid which does not dissolve.

1.2. Fusion for the determination of tin by ICP-

MS

2. For preparation of 10 PPM Ru/Re see 'Sample dissolution for the T4 and T5 Analytical

Packages', Step 4.1. (Page 13)

3. Dilution factor for solution is 1000. Nitric acid concentration is

2. Fusion with Sodium Peroxide

Sodium peroxide (Na2O2) is used to decompose sili cate rocks containing spinels, zircons, arsenides, and sulfides, and tungsten, niobium, and tantalum minerals, and rare earth phosphates. This procedure is used primarily when chromium (as chromite) is present at concentrations ^ percent.

Read notes for procedure l . l above before beginning.

Each batch of samples (15 maximum) must include two blanks and one reference material.

1.2.1. Weigh 0.100 g of sample into a porcelain crucible.

1.2.2. Add 0.300 g of lithium metaborate and mix well with a teflon-coated spatula.

1.2.3. Transfer quantitatively to a small graphite crucible (35mm high x 25mm diameter) and place on a silica tray.

1.2.4. Fuse in a muffle furnace for 15 minutes at

10000C.

2.1. Weigh 0.100 g of sample into a low-form zirconium crucible.

2.2. Add 0.600 g of sodium peroxide and mix by shaking or rotating the crucible carefully.

2.3. Place the covered crucible on a silica tray and

2.4.

2.5. fuse in a muffle furnace at 6500C for 10 minutes.

Remove crucible from furnace and cool.

Hold the crucible with tongs while cleaning the crucible bottom with distilled water.

2.6. Place crucible and fused contents into a 250 ml beaker and cover with a watch glass.

l .2.5. Transfer the molten mixture (red hot) into a 50 ml PTFE beaker containing 25 ml of lO^o

HNO3.

1.2.6. Stir for 20-30 minutes using a teflon-coated stirring bar and a magnetic stirrer. (see Note

1)

1.2.7. After dissolution transfer immediately to a

100 ml volumetric flask containing l ml of 10 ppm Ru/Re standard, (see Note 2)

2.7. Add 50 ml of distilled water and 10 ml of concentrated HNO3 very carefully.

2.8. Using a hotplate, heat for about 15-20 minutes until dissolution is complete.

2.9. Remove the crucible using teflon-coated crucible tongs and wash the crucible with a fine jet of distilled water, collecting the wash ings in the beaker.

EA2-9

Sample Dissolution

2.10. Cool and transfer to an appropriate size 2.11. Submit solution for determination of volumetric flask (minimum volume = 100ml). chromium by AAS. See section "Trace Ele ments Determined by Flame Atomic Absorp tion".

EA2-10

Graphite Furnace

GRAPHITE FURNANCE OPERATING PROCEDURE

Thermo Jarrell Ash Video 22E - CTF 188

Introduction:

The following procedure is provided as a guide for graphite furnace analysis (electrothermal atomization analysis). For more specific details refer to the section in the manuafacturer's operation manual concerning each step in the process.

It is also assumed that the following steps have been taken:

The atomizer cell has been aligned to allow max imum transmission of the light from the hollow cathode lamp. See Section 2.4, page 2-6.

A furnace cuvette has been installed, see Section

3.6.2, page 3-14; and decontaminated, see Section

3.6.4, page 3-17.

WARNING: Ensure "run light" on 188 is off and cell body is cool. This can be monitored by the temperature displayed digitally on the AA CRT.

3. CTF 188

4. Graphics printer

5. AA

6. Autosampler (if used)

Instrument controls:

Front panel of CTF 188

ESC - Stops the furnace cycle from continuing. The

RUN light will go out and the ESC light will stay lit until the furnace cuvette cools below 1500C

NEB AIR -Initiates operation of the vacuum pump for nebulization when aerosol deposition is used.

DOOR OPEN - Opens the furnace cell door for manual pipetting when furnace is cool. Pressing a second time will close door.

RUN - Initiates a furnace cycle. When the FASTAC

ADM (Aerosol Deposition Module) is in use, the fur nace will continue through the number of heating cycles selected under the MODE Key on the AA CRT.

To discontinue the process, press RUN again and the current cycle will be completed and the remaining cycles will not be initiated. Pressing ENTER on the

AA keyboard will cancel the remaining cycles.

Safety advisory:

1. Always ensure that the cell door is functioning correctly before applying power to the atomizer.

2 Never look directly at a hot furnace cuvette during atomization without proper eye protection

(welders goggles).

3. Never touch the cell head (i.e. to change a cuvette assembly) while the "run light" is on, or severe tissue burns may result.

4. Always ensure power to the 188 control module is off before handling the cell head and allow suffi cient time for all graphite such as the cell door, cell body, electrode inserts, cuvette, and as sociated parts (cell windows, washers, and light baffle) to cool before handling.

"POWER ON" Sequence

To avoid communication errors which may lock-up the

AA keyboard, always power up the system in the following order:

1. Water coolant (recirculating pump).

2. Argon gas

Right Panel of Video 22E

Most controls are self-explanatory, functions should be noted.

The following

SIGNAL/BACKGROUND - selects hollow cathode" or "D2 arc" intensity reading respectively for both energy and current indicators. In S/H mode, signal selects monitoring of the low current pulse, while background selects monitoring of the high cur rent pulse on energy and current indicators.

HOLLOW CATHODE -lamp current for channel A is controlled from turret, for channel B by small diameter knob at B. Large diameter knob for A and B controls S/H background current.

RESTART/STANDBY -will reset the micro computer's program counter to 0000 which can correct certain operational errors.

EA3-1

Grahite Furnace

OPERATE -used when entering an analytical pro

gram.

HIGH VOLTAGE A/B -sets the photomultiplier

voltages for A or B.

Keyboard on Video 22E

INT - used in the flame mode of spectrometer.

READ - initiates integration for flame analysis; RUN

key on CTF 188 performs the same function for graphite furnace work.

STD - used with calibration procedures. ber. The I.D. of the sample currently being analyzed is displayed on the bottom of the CRT.

l IBM Compatible

Auto-Graphics

*0 Off - The graphics printer will serve only as an alphanumeric printer. A hard copy of the CRT is made by keying the decimal point on the AA number pad.

1 On - The graphics printer will automatically trigger the CRT graphics after every integration cycle if peak

area integration is selected with graphics.

* l All data - provides a hard copy of the CRT each time an absorbance profile is displayed if auto

graphies was selected.

2 Stats only - lists sample number and statistical sum mary; the graphics printer does not hard copy the CRT.

ENTER - executes the keyed command.

I/O - This key formats the input and output functions of the spectrophotometer.

The CRT displays

I/O: l Recorder

2 Printer

3 Serial Port

4 Time

5 Date

6 OpI.D.

7 Wt. Correction

8 Percent Modulation

9 To exit

The following settings should be made after the I/O key is pressed and the above menu is displayed. For example, to choose Recorder, press l and Enter.

Default choices are indicated by an asterisk "*" and only require the enter key to be pressed to proceed.

Once the settings have been made, press I/O to return to the menu and continue. Exit the menu by pressing

9 and Enter.

NOTE: Do not take the graphics printer off-line while

3 ENTER

*0 Off- deactivates the RS232C interface, l On

O Printer - used to transmit data to an external printer properly configured to accept data.

*1 Protocol - must be entered to use the ADS-200

system. Enter default settings from next CRT dis

plays.

it is printing; doing so would lock-up the AA's microprocessor.

CAUTION: If not using the ADS-200 system, do not select the Protocol option; otherwise, the microprocessor locks-up and the message "waiting for RS232C" appears on CRT

1 ENTER Enables the formatting of the recorder

output. *2 Abs.

Press enter to choose this default setting. Then choose

2 (Element-bkg) from the next menu for corrected absorbance signal. The above process is repeated for

Channel B if in two-channel operation (two elements determined simultaneously).

2 ENTER-—*Enter Sample no.

When first sample no. is entered, all subsequent analyses will be incremented by one. Auto-zero and auto-cal functions will not increment the sample num-

4 ENTER Sets the time of day. Enter the appropriate

two-digit code when prompted.

5 ENTER Sets the date. Enter the appropriate two-

ditit code when prompted.

6 ENTER Enter operator I.D. as a two digit code.

Key ENTER to cancel I.D.

See Section 3.2.10 of the VIDEO 22E Spectrometer

Operator's Manual for further explanation of the I/O key.

EA3-2

Graphite Furnace

AA Spectrometer Keyboard (System Software)

A detailed description of the software menu for each key (Mode, Recall, Store) is provided in the VIDEO

22E Spectrometer Operator's Manual. Default selec tions are indicated with an asterisk * on the CRT. Press

ENTER to select the default condition. To select any of the other options displayed on the CRT, you must press the appropriate key, then key ENTER.

Mode Menu centration, integration, mode, and concentration of calibration standards) for both channels.

5 ENTER To chose statistics. Enter no. of runs as 3 and ENTER. Select *1 to display Mean, SD, RSD.

Note: Selecting O or l as the number of runs indicates no statistics and O is displayed on line 5 of Mode menu.

6 ENTER Chose CTF as atomizer if furnace has not been turned on. AA units will the 188 will default to the mode last used.

The Mode menu is used to set the instrument parameters (element, optics, background, results, statistics, atomizer, autosampler, test and exit) for in strument operation.

The following menu should appear on the CRT after the MODE key is pressed.

Mode:

5

6

3

4 l Element

2

7

9*

Optics

Bkgnd

Results

Statistics

Atomizer

Autosampler

Test

To exit

AA,DB

S/BsOl.0

O

CTF off

1 ENTER Select element or elements if dual channel is used. Enter element number from the resulting table that appears and press ENTER.

2 ENTER Select optics for each channel if AA, DB is not on Mode menu display.

3 ENTER Display the Background menu and select appropriate background for each channel

(D=deuterium arc, S-H =Smith-Hieftje). A two-digit number must be entered; one for channel A, then one for channel B.

NOTE:If D2 background correction has been selected, the CRT will display "D2 Warmup" for a few seconds.

NOTE:To run S-H and D2 simultaneously, enter 4.2 which is S-H in channel A and D2 in channel

B. Entering 2.4 is invalid.

4 ENTER Select l for Cone. (S/E is scale expansion in absorbance mode). Select *1 Normal. Select default condition 1.1 to display all information (con

7 ENTER Turn autosampler on if required. Further description for autosampler (PS-75 Prep Station) fol lows.

9 ENTER To exit the Mode menu.

Autosampler with Dilutor

This accessory provides automatic sample dilution, standard preparation, matrix modification, and stand ard additions preparations fully under computer con trol via the CRT and keyboard of the Video 22E spectrophotometer.

The Prep Station Option is selected through the

MODE key, 7

ENTER

- Select 2 Autosampler with diluter

- Turn on the Sample Changer and Diluter

- Press * l Prime to prime the syringe and fill

- the rinse station.

- The CRT will display (if AA is in the concen tration mode) options for

1 Run samples

2 Calibrate with autosampler

3 Standard addition

9 To exit

Calibration

2 ENTER Change to read 0.0 mis matrix modifier.

Bulk AZ A AC?

BULK VOL^S.O mis times O*

This allows you to prepare large volumes of auto cal solutions. You are given the option of having as many times the previously selected Final Volume (in this case - 5.0 mis) as desired (max. 45 mis) prepared in the

AZ and AC vessels. Enter O to bypass.

EA3-3

Grahite Furnace

Select *1 YES for rinse between samples.

Select 3 Both for dual channel operation. You are now ready to enter the concentration of the standards through the keyboard.

NOTE:STD C is the autocal standard and should be

midrange (0.100 ppm). STD Z must be zero.

NOTE: The highest standard concentration used in

channel A is used by the microcomputer to calculate the concentration of the stock analyte which you must prepare and place in the STOCK vessel.

Key in the concentrations for each standard in channel

A. When the desired number of standards have been entered, key STD. Enter the appropriate concentration for STD C in channel B (0.100 ppm).

NOTE: Enter concentration of STD C to provide three

significant figures. Because the PS-75 prepares standards from the same mixed stock solution, entry of the first standard in channel

B establishes the other standard values.

The CTR displays a diagram of the sample changer to aid in setup, for example:

STKCONC A- B-

AUTO CAL A- 0.100 B- 0.100

X = SAMPLE/STD O = EMPTY TUBE

RNS STK —— row 1 000-

*** A/Z A/C

3

4 ———

LOAD RACK AT THIS TIME

*1 Prepare

2 Prepare and Run

9 To exit

STK CONC's are the concentration of A and B ele ments in the stock analyte solution to be prepared and placed in the second large vessel from the left.

Provide sufficient empty test tube (this example has 3) for standard preparation.

If *1 Prepare is selected, the Prep Station will only prepare the standards.

If 2 Prepare and Run is selected, both preparation and analysis will proceed automatically.

When using the Sample Changer with the CTF188,

do NOT press the escape (ESC) key on the 188

unless an emergency situation requires immediate cooldown of the furnace. If the AA microprocessor locks-up, restart the system by turning off the

Sample Changer power, go to RESTART/STDBY on the AA, back to OPERATE, and re-initiate power to the Sample Changer.

NOTE: In the furnace mode, the Sample changer auto

matically pauses in the "sipper up" position between each sample while nebulizing air for several seconds to clean-out the capillary to the FASTAC.

Once standards have been prepared and analyzed, the

CRT displays the calibration curve for channel A.

Follow the prompts shown. When channel A has been accepted, the calibration curve for channel B will be displayed.

Run Samples

l ENTER Change to read 0.0 mis matrix modifier. To

use the "manual dispense" mode, enter a value of zero for the Sample Uptake Volume. When prompted for

"Dispensed Volume", enter the volume to be placed into each sample test tube. In the "manual dispense" mode, 75 samples can be analyzed. With matrix modification, 37 empty test tubes are required, there fore, 37 samples can be analyzed.

CRT displays options for program to run samples.

CRT displays option for Initial AZ A AC?. NO is normally selected following calibration; YES is nor mally selected following the recall of a stored calibra tion curve.

CRT displays Auto Deposit Adjust if FASTAC and concentration mode is being used. If selected the system will automatically cut the deposition time in half for any sample which gives a CONC HIGH read ing.

NOTE: Make sure initial deposition time is an even

number so the ADA division will not intro duce inaccuracies due to arithmetic founding.

For example 6.5 sec would be cut to 3.2 seconds.

CRT continues to display options similiar to those in

CALIBRATION above.

EA3-4

Graphite Furnace

Recall Menu

The Recall menu is used to examine calibration cur ves, instrument conditions, and to review or change furnace methods. The following menu is displayed when the RECALL key is pressed and if CTF is the method of atomization indicated in the Mode menu:

Recall:

4

5

2

3

Instrument Conditions

Furnace Method by Element

Furnace Method by Matrix

Current Method

6 Furnace Curve

9* To exit purge flow rates, and integration time for the develop ment of new furnace methods.

Once a desired change is requested by typing its ap propriated number from the menu followed by press ing ENTER a new menu will result. The change to be made is indicated by *. The change can be made by typing the desired temperature, etc. followed by press ing ENTER. If no change is to be made at the *, press

ENTER to proceed to the next item in the category to be changed. After all stages have been completed the

CRT redisplays the modifications.

Purge flow rates are indicated by the following:

If you wish to introduce air (or other auxiliary gas) into the cell body during PYR1, type l for YES and press

ENTER.

stop flow low flow medium flow high flow

2 ENTER WiU display the element table. Select

desired element and ENTER to display recommended instrument conditions (HCL current, bandwidth, wavelength) which are the same for flame and CTF measurements.

Press RECALL to return to element table. Press

ENTER to give initialization menu for CTF use, then press RECALL to return to Recall menu. A second element may be entered for dual channel use.

3 ENTER Will display available furnace methods for

the current element of interest. Select the desired furnace method by typing its repective number and pressing ENTER. The furnace method is displayed on the CRT. At this point if * l is chosen (Run as is) the analysis will start after the RUN key on the 188 is pressed. Pressing 2 (Change) will allow you to alter furnace parameters, method title, etc. beginning with the matrix name. Once the matrix name is chosen, other parameters can be changed as indicated on the new menu (see below).

4 ENTER Will display available furnace methods

for a given matrix to use with a different element.

Other options which follow are similar to those when

3 ENTER is pressed.

5 ENTER Assumes you wish to change the furnace

paramenters for the current zero method in RAM. This is the method which is in current use or which was used last.

Integration under furnace operation is the time frame

during which the absorbance signal is measured by the system microprocessor. In most cases, integration is set to start at stage 4 (Atomization). Make sure peak area has been selected for concentration results and statistics to be based on peak area values. Integration time may be changed next, followed by integration delay time (usually 0.0). The delay time is the time during which the microprocessor waits until beginning the stage chosen for integration. At this stage the Plot

vs Time selection menu will appear if the graphics

option is included. Chose option *2 (default) to obtain a plot of total absorbance (solid line), corrected absor bance (dotted line) and a temperature profile.

FASTAC option permits adjusting of the fastac delay

and deposit times once the option has been selected.

FASTAC delay is a time period which the sample is allowed to rinse through the spray chamber and then be drawn to waste by vacuum. The recommended time delay is 6 seconds but may vary according to capillary length. FASTAC deposit is the time period during which the sample is drawn through the nebulizer into the cuvette. Higher concentrations require shorter deposit times and vice versa.

Modifying Furnace Parameters

Modifying Furnace Parameters is reached through op tions 3, 4, or 5 of the Recall menu. It permits you to alter furnace temperatures, ramp times, hold times,

The Furnace Curve portion of the Recall menu acts

as a storage for calibration curves which have been previously stored by the operator and placed in this menu. Storage capacity is 40, numbered O through 39.

After typing 6 and pressing ENTER the first 10 curves are depicted on the CRT. Further pages can be dis-

EA3-5

Grahite Furnace

played by pressing ENTER. The desired calibration curve can be displayed by typing the appropriate num ber (O to 39) and pressing ENTER. Pressing ENTER again will print the curve. The Recall menu can be displayed by pressing the RECALL key. When a calibration curve is recalled, the printer sample number will revert back to the next sample number that existed at the time the curve was stored. The number dis played with the curve selection menu is the program number associated with the curve. The furnace pro gram is recalled from the RECALL menu.

Store Menu

The STORE key is used to accept (store) newly developed furnace methods and furnace curves. Make sure atomizer is set for CTF (selection 6 of Mode menu). Furnace curves will only be displayed when in the concentration mode (selection 4 of Mode menu).

Back up stored methods with a hard copy using the graphics printer.

If you wish to save a newly developed method you

MUST do so through the STORE key while this new

method is recognized as the current method (see

recall menu), or it will be erased when a new method is selected or developed.

As each new method is accepted, it is assigned the next highest number in the user created series (150 to 349).

In addition, you may also overwrite existing methods which you have created. Only the methods created by you can be stored or overwritten. Factor supplied methods can be used to develop new methods but cannot be overwritten.

About Furnace Methods

The 188 microprocessor has an overall storage capacity for 349 furnace methods, divided into two categories of permanent factory supplied methods and storage space for methods developed by the operator.

Space allocations are as follows:

- Factory installed methods for ultra-pure water (l to 52)

- Expansion space for future factory supplied methods for ultra-pure water (53 to 69)

- Factory installed methods for various elements in certain matrices (70 to 129)

- Expansion space for future factory supplied methods (l30 to 149)

- Storage and retrieval space for operator created furnace methods (150 to 349). These methods are maintained only if they are stored through the

Store menu.

The Store menu is self explanatory. It will prompt you to store a new method in its appropriate space or to overwrite a similar method that you have created but modified. If ENTER is pressed without typing O or l the initialization menu will appear without storing the method.

To store furnace curves CONC must be selected for the results option of the Mode menu. It is important to

remember that a furnace curve cannot be stored unless the newly developed furnace method it was

created for has been stored first. Furnace curves

must have method numbers in order to be recalled by the microprocessor.

NOTE: Furnace curves are temporary unless stored through this menu. Because there is room for only one curve in temporary storage, any curve which has not been stored is erased when a new curve is created.

Selection number 3 (Store Menu), Curve (No CTF), is offered as a method of storing curves without affliation with a furnace method number. It functions exactly as selection 2, except that these curves do not have, and do not require, method numbers because they are stored solely by element.

To overwrite a furnace curve, press STORE key, type

2 and press ENTER, and overwrite the desired curve by typing its appropriate number and pressing ENTER.

Press ENTER again to overwrite.

Typical Furnace Operation Steps

FIRST Follow the "Power On" Sequence to begin.

SECOND Set the software parameters for furnace

operation through the Mode menu. See notes under

AA Spectrometer Keyboard (System Software).

NOTE: Set option 3 of the Mode menu first, i.e., single or dual channel, with or without background.

Choose elements (option 3)

Check I/O settings. See notes "Instrument Controls -

Keyboard on Video 22E". Make sure Serial Printer is

OFF if not using the ADS-200 system.

THIRD Set the instrument parameters on the AA

spectrophotometer according to conditions stored in the microprocessor memory (Recall Flame Condi tions). Set wavelength, bandpass, hollow cathode

EA3-6

Graphite Furnace

lamp current, background current. See Section 4.2, page 4-16 of the Video 22E Operator's Manual for insertion and alignment of the hollow cathode lamp.

NOTE: Adjust the PM voltage so that the energy meter

is in the green zone. Optimize alignment of the HCL by adjusting the knob located near each spring and by rotating the HCL.

The following is a list of parameters for elements currently being determined by graphite furnace atomic absorption spectrometry.

NOTE: In the case of low intensity resonance lines that

require upper range high voltage settings, the

D2 arc may exceed the HCL intensity, even at the minimum operating current of the D2 arc.

In this case, increase the lamp current and decrease the PM voltage, or insert the D2 arc

neutral density filter into the beam of the D2

arc. See Figure 4-16, page 4-61. Reset the D2 arc by keying A/Z and READ.

NOTE:To switch off the D2 arc, go through the

MODE key and press 3 ENTER to change to

"A" or "A-bkg (S-H)".

Element

Au

Pd

Pt

Lamp

Current

(ma)

5

5

5

Band width

Background Correction

(nm)

1.0

0.3

0.5

Wave length

(nm)

242.8

247.6

265.9

S-H

Bkg

(ma)

3.0

2.5

2.5

Smith-Hieftje

See Section 4.4.21, page 4-58 of the Video 22E

Operator's Manual. S-H background correction is selected from option 3 of the Mode menu. Set the background current and adjust PM voltage so that the energy meter is in the green zone. Increase Signal

HCL current until its energy matches that of the Back ground.

FOURTH Select the Recall menu and call up the

furnace method for analyzing the required element/s.

Send the furnace method/s to the 188 CTF microprocessor by pressing ENTER. Modify any fur nace parameters at this time. Press the RECALL key to display the list of stored calibration curves.

NOTE: If no curves are stored, refer to notes under

"Autosampler with Dilutor", Calibration with autosampler.

At this time, return to Mode menu, press 7 ENTER, and select 'autosampler with diluter'.

FIFTH Proceed with calibration if required, or ana

lyse samples. See notes under "Autosampler with

Dilutor", 'Calibration' or 'Run Samples'.

Further information can by obtained by referring to section 4.3.3, Calibration Procedure, in the Video 22E operator's manual for creating curves.

Deuterium Arc

The intensity of the D2 is automatically set during the autozero routine to match the intensity of the hollow cathode lamp. Select D2 from option 3 of the Mode menu. Key A/Z and READ to allow the D2 arc current to increase until the D2 reference intensity is balanced with the reference (lo) beam of the HCL.

NOTE: Under furnace operation the RUN key on the

188 is used to initiate the analysis instead of

READ.

EA3-7

VarianAA775

OPERATION OF THE ATOMIC ABSORPTION SPECTROMETER

Varian AA775

Ensure that the proper instructions found in the operator's manual for the atomic absorption unit have been read. The Varian AA775 is installed with a number of safety features which include:

Burner interlock, preventing the lighting of the flame without the proper burner in place.

Flames have different temperatures and analytical characteristics, not only because of the gases used but also because of their ratio.

The following guide is used to describe the flame:

NOTE: This system only works if the socket and cable

attached to the burner is connected.

Automatic flame shut-down with low fuel rates and when requested by operator. Proper sequence for this procedure is performed.

Flame sensor, to shut off fuel when flame is extin guished.

The general procedures to follow with any atomic absorption unit set up are:

1. Preliminary Safety Check

l. l Compressed gas cylinders securely fastened.

1.2

1.3

Exhaust ventilation fan on and operating.

Burner head clean and corrected installed.

1.4

Water trap or loop properly filled.

1.5

Adequate supply of compressed gases to com plete the task.

Coolest: "reducing" - Fuel rich, luminous,

air-acetylene glows bright yellow; nitrous oxide-acetylene a blinding pink-white.

Medium: "stoichiometric" - Balanced fuel

and oxidant, air-acetylene has a slight yellow tinge, nitrous oxide-acetylene has about 15 to

20 mm high red feather.

Hotest: "oxidizing" - Fuel lean, air-acetylene

is blue, nitrous oxide-acetylene has about 5 to

10 mm high red feather.

2.5 Solution uptake rate.

2.5. l Adjust to rate of 6 ml per minute if a variable rate nebulizer is used.

2.5.2 Check for blockage in the nebulizer or solu tion deliver tube.

2.5.3 Check spoiler bead position in bung.

2. Parameters to be Set

Parameters are set according to the information recom mended by the manufacturer or contained in the infor mation sheets for each element. These are found at the end of each section of this manual relating to deter mination by atomic absorption. In addition the follow ing optimization procedures should be performed.

2.1 Lamp alignment.

2.2 Instrument gain.

2.3 Burner alignment.

2.4 Flame stoichiometry.

3. Sensitivity Check

The absorbance of a check standard is analyzed and compared with previous data. If the value is within

109fc of the mean of previously amassed data the operator can proceed. If not, then optimization proce dures should be repeated, or until the problem is found.

4. Calibrate Instrument

Use appropriate standards which encompass the nor mal range of samples or use the technique of bracket ing the sample with high and low standards.

5. Analyze Samples

Reagent blanks, control standards and samples are analyzed. Calibration is checked at intervals. Use a distilled water rinse between samples and standards.

Samples with concentrations outside of the calibration

EA4-1

VarianAA775

limit are diluted to fall within the working range and reanalyzed.

The analytical values are transcribed manually or cap tured by the instrument microcomputer.

EA4-2

Liquid Handling System

LIQUID HANDLING SYSTEM

A liquid handling system to complement the robotic system used for the decomposition of samples is avail able for certain applications. The instrument is a Gil-

son Model 222-401 and is programmable to perform

dilutions, pipetting, dispensing, and partitions or any combination of these. Programs are installed for 1:2, l :5, and 1:10 dilutions for up to 44 samples using the standard culture tubes. A maximum of 110 samples can be diluted using culture tubes when five racks of test tubes are available The procedures are performed automatically and unattended and are initiated through a keypad although the system may be subjugated to a host computer via an optional interface. Programs can be written according to your needs or existing programs modified.

A variety of test tube racks is available to use with the system. The racks are individually coded so that the probe will find test tube positions. To use a rack of your choice will require extensive initialization for each test tube position.

Four glass syringes (0.5, 1,5, and 10 ml) covering a range of 2 to 11000 microliters are available. There is a 5 ml and a 10 ml syringe using a piston with a

PTFE/ekonel seal for use with acids. The following table shows the performance expected for 5 ml and 10 ml syringes respectively.

of diluted solution - 2.5 ml of sample and 2.5 ml of diluent.

PROGRAM FOR 1:1 DILUTIONS

(Stored Under File Number 22)

1 RACK CODE 22/1

2 INPUT CO/0/1/44

3 AO-0

4 FORA^/4

5 FORB~1X11

6 AO=AO+l

7 IFAOCO

8 HOME

9 PRINT AO 11

10 TUBE A/B/1

11 HEIGHT

12 ASPIRO/5/2

13 HEIGHT 70

14 ASPIRO/4000/5

15 RACK CODE 22/2

16 TUBE A/B/2

17 HEIGHT 120

18 DISP 0/8005/7

19 RINSE

20 DISP 0/2000/9

21 NEXT B

22 NEXT A

23 HOME

Set Volumes

(microliters)

5000

2000

1000

500

200

100

50

20

10000

5000

500

200

100

Accuracy 9fe

0.5

0.5

1.0

2.0

3.0

1.0

2.0

3.0

5.0

0.5

0.5

0.5

0.5

Precision 9fc

(30 aliquots)

0.2

0.3

0.5

1.0

0.05

0.05

0.08

0.1

0.05

0.05

0.2

0.3

0.5

With the 5 ml syringe the maximum volume of sample that can be dispensed is 5 ml (sample + diluent). This means that l :2 dilutions will yield a maximum of 5 ml

NOTES: A complete explanation of programming

can be found in the manufacturer's User's Guide for

221 and 222 Sample Changers. The above program uses a 10 ml syringe to give 8 ml of diluted sample.

l. A height of O is taken as the bottom of the test tube..

Any other height designation is taken as the num ber of millimeters above the bottom of the test tube, i.e., HEIGHT X sends the probe X mm above the the test tube bottom.

2 Certain coded racks are programmed to work with a designated vertical arm. The Geoscience

Laboratories' system is equipped with the 123 mm

(standard) length of vertical arm. Code 22 racks are designated to hold test tube of 18 mm o.d.x 150 or 16 x 160 mm. and are programmed for use with the long arm-needle kit (183 mm). At the Geo science Laboratories, culture tubes are only 100 mm high and can be used with the 123 mm arm since the movement of the probe is not obstructed by the test tubes. Therefore when using racks coded 22, the system assumes you are using the

EA5-1

Liquid Handling System

183 mm arm, and will not accept any height value below 60. This value of 60 must be taken as O when using racks coded 22 with the 123mm arm.

In the example program above the true height of the probe from the bottom of the test tube is the value given minus 60 mm.

PROCEDURE FOR PERFORMING DILU

TIONS

The following dilution programs have been installed for use with a 10 ml syringe:

1:2

1:5

1:10 dilutions: dilutions: dilutions:

File Number 22

File Number 15

File Number 110

DILUTOR ABSENT may appear. Messages

MODEL TEST and MODEL 222 V3.O will

appear.

4. Prime the system by pushing PRIME on keypad.

To stop prime push PRIME again.

5. Press EDIT/SAVE/FILE key to search for the required file.

6. Press ENTER when the required file is found and continue to press ENTER until FILE STORED appears on display.

7. Press START to initiate program.

Instructions:

1. Make sure all tubing, keypad, probe, and syringe are installed correctly (see manual). Place samples into test tube racks such that the y-axis (B in program, maximum 11) is loaded first. Make sure sufficient sample and diluent is available for the complete run.

8. A message will appear on display requesting input of number of samples to be run. Note: Above program is currently set for a minimum of l and a maximum of 44.

9. Program will now complete dilutions and return probe to HOME position.

2. Plug in dilutor and sample changer.

10. When finished wash system out with distilled water by replaced the acid bottle with distilled water and operating the PRIME until a complete rinse is accomplished.

3. Switch on dilutor (Model 401) then sampler

(Model 222) in this order, otherwise a message

EA5-2

Majors - XRF

THE PHILIPS PW1400 X-RAY FLUORESCENCE SPECTROMETER

SYSTEM

Introduction:

The Philips PW1400 XRF spectrometer, a wavelength-dispersive X-ray fluorescence spectrometer, is used to nondestructively analyze the matrix-modified samples. The spectrometer consists of a 100 kV generator with current and voltage stabilization. The X-ray tube is a side-window tube; various tubes with different target materials are avail able in the laboratory, e.g., rhodium (general use), chromium, molybdenum and tungsten. The maximum power rating of system is 3 kW. The tube system is cooled via a Haskris water chiller.

The system offers a choice of analyzing crystals (up to six mounted on an "assembly"). Crystal positions are:

1 LiF200

2 LiF220

3 PET

4 Ge

5 T1AP

6 LiF420

(default crystal in parameter set)

The system has two detectors - a flow counter and a scintillation counter. These can be used individually or in tandem.

The spectrometer is equipped with a microprocessor for control and processing of data. There is a four-posi tion sample holder in the unit. A 72-position automatic sampling device is interfaced to the system. The sys tem can be fed both manually and via the sampler.

The instrument and microprocessor are driven by a

Digital Micro POP Minicomputer with dual disk drive,

10 Mbyte Winchester hard disk using the RSX-11M operating system. Video terminal (with printer for screen dumping) and LA 100 Decwriter terminal are both used for either primary system operation and/or secondary result generation. Terminals can be used to access different software routines as long as the same data files are not accessed simultaneously.

Procedures:

1. Turning the system ON

1.1. Confirm that the Haskris Chiller is turned on

(water mains must be on). The On/Off switch is flat in the 'On' position and the lights are on.

l .2. Confirm that the supply of clean dry air to the hydraulic system of the spectrometer is con nected (gauge should read 60 psi).

l .3. Confirm that the current and high voltage set tings on the instrument are set to their mini mum values of 20 kV and 10 mA. Settings are controlled from the computer and do not need to be adjusted at the instrument.

1.4. Confirm that the PI O gas and air are turned on: detector gas flow should be at 0.51/hr and the capacity should be above 1.0 bars.

NOTE: If tanks are changed or any major change in

PI O gas pressure occurs, the gas leak reset button (at the side panel) must be activated before setting-up the instrument.

1.5. Depress the switches for the main and high voltage power located on the upper left corner ofthePW1400.

NOTE: If the X-ray tube has not been in use for some time, the tube should be warmed-up (slowly) according to the instructions accompanying the tube.

l .6. Turn the computer on using the I/O switch on the left side of the VT240. This will "boot" the computer into its operational state.

NOTE: If the instrument is on, booting the computer will activate the vacuum pumps in the spectrometer.

2. The Software - Starting the Computer

The operating files (including the RSX-11M operating system) are stored on the Winchester hard disk. The

Baud rate on the terminal is fixed at 9600.

NOTE: Commands given by the operator are under lined and are followed by a carriage return [R].

When the computer is turned on, the following mes sage appears at the printer:

EA6-1

Majors - XRF

"KDF11B-BE ROM V0.8"

128KW MEMORY

9 STEP MEMORY TEST

STEP 123456789

TOTAL MEMORY ERRORS = O

CLOCK ENABLED

BOOTING FROM DUO

(DUO: = Winchester drive)

>

> ; To run the X14 task, type:

MCR X14 or .X14

> ; To run the Regression task type: MCR REG or .REG

and date (hr:mn dd-mmm-yy and RETURN) [S] :

2.1. Enter the time and date in the specified format and push RETURN. The system then starts up and requires no input until it requests:

ABO — Task not active

*Should you mount DU1 : ? (y/n) :N[R]

(DU1: = disk drive)

*Should you mount DU2 : ? (y/n) : N[R]

2.2. Enter N (as shown above) to both of these queries.

The computer then performs a few more steps, in structs the operator how to use the PW1400 software and logs off, leaving the system in the general monitor system (cursor = ^.

The structure of RSX-11M is too complex to discuss in this manual. It should be understood, however, that operating system programs are generally stored in sub-units called directories and are accessed by log- ging-on. The logging instruction requires both a software name and a password. For the PW1400, the instruction required is: LQGxl47xl4.

2.3. Enter LOG x 147x14 and push return. The in strument will respond with:

; @<EOF>

The system continues to transmit messages for some time but these can be ignored by the operator under routine conditions until it prints:

The new cursor (S) indicates that the system is in the correct directory. Various RSX-1 1M commands are available to initialize disks, locate directories, etc.

2.4.

Activate the system by typing:

MCRX14

.X14 or

The system will respond with a message in dicating ownership of software, etc. and ends with the system cursor (*).

NOTEiThe PW1400 software operates in two modes: automatic mode (cursor = "*") and manual mode (cursor = "C"). In the automatic mode, the operator can communicate with both the microprocessor which controls the instrument and with the computer which stores the information. In manual mode, the operator can communicate only with the microprocessor.

In order to establish the proper communications be tween the terminals, the following steps are required:

2.5. At the printer, type ASP (assemble system parameters). There are only two items of in terest in the system parameters: the terminal type and whether or not the spectrometer is connected.

The first line will read:

RSX-11M BL35B

RSX11M

[1,54] System dd-mmm-yy hrrmn Logged on Terminal

TTO: (TTO:~ LA100)

Good morning

TERMINAL TYPE=PR:=

(i.e. waiting for input)

2.6. Since the terminal is a printer and this is the default device, simply press return. If the required terminal type is a graphics terminal, simply enter GT.

> ;8LOGIN.CMD

EA6-2

Majors - XRF

2.7. Continue depressing the return key until the following message appears:

SPECTROMETER CONNECTED N instrument so information is acquired in the correct manner.

The information stored in the Databank includes:"

2.8. Exit the ASP routine by pressing /[R].

2.9. Check the status by typing PSP. If the status is not correct, repeat steps 2.5. to 2.8.

2.10. Exit from the program by typing END. The S cursor will return.

Before the spectrometer can analyze samples, it must know what information is required and how the data are to be acquired. After acquisition of data, the com puter must know how to process the data to provide corrected results.

Two separate areas, or "memory banks" are used to store this information. These are referred to as the

Databank (DB) and the Parameter Bank (PB). Infor mation relating to a particular set of samples is stored under one name that is common in both banks. The

Databank provides information that is used by the computerto calculate corrected DATA. The Parameter

Bank is used to set the physical PARAMETERS on the

- link programs (LP), which "interface" between the measuring program and the calculation parameters (corrections, calibrations, etc.)

- calculation parameters (CP)

- correction parameters (RC) (count-rate correc tions)

NOTE:These are all concerned with data processing.

2.11. After getting the S cursor type MCR X14 or

.X14. This will initiate the PW1400 software.

The ASP program can be used to change the terminal type (PR or GT) at any time.

The system can be run entirely through the video terminal. The screen-dump printer (L A50) can be used to obtain print-out of results (PON = printer on). The printer can be turned on via a switch on the left side.

3. Running the Samples

Once the Measurement Program (MP), Link Program

(LP), Rate Correction Set (RC), and the Calculation

Parameter (CP) Sets have been defined, it is possible to run samples on a routine basis. The development of these programs is described in the Appendix C follow ing 'Determination of Major Elements - X-ray

Fluorescence Spectrometry' (page EA6-6).

It is important to appreciate the manner in which the computer and microprocessor control the instrument and how the operator can control the system. Operator commands to the system software are via three letter acronyms.

To read what is in the DB type: PDB (Print DATA

Bank)

The contents of the PB include:

- listing of channels and channel conditions

- listing of measurement programs, i.e., channels monitored, times used in counting.

- listing of job programs, i.e., sequence of spectrometer instructions.

NOTE:These listings are all concerned with the physical settings of the instrument. They en sure acquisition of data will be carried out in the correct manner.

To read what is in the PB, type: PPB (Print Parameter

Bank).

In order to use the X14 program for the analysis of samples, a previously developed Databank must be

LOADED into memory. Following this, the associated

Parameter bank must be READ into memory.

All trace element instrument parameters are found in

Parameter/Databank files labeled OTHER.

Channel conditions for trace elements of interest

(stored in OTHER) are listed in the table on the fol lowing page.

3.1.

3.2.

Load the required databank: enter LDB

XXXXXX. (LDB ~ Load Databank)

Read the associated Parameter bank: enter

RPB.

The PB read into the microprocessor cor responds to that stored in conjunction with the

Databank currently in memory.

EA6-3

Majors - XRF

ELEMENT FLT COL DET XTL ORO UPL LWL KV MA ANGLE+OFFS-OFFS

Cu 1

Nb

Ga

Ta

Ce

Nd

La

Cs

Y

Zr

Pb

Rh

Rb

Th

U

Sr

Ka

Ka

La

La

L(3

La

La

La

Ka

Ka

Ka

La

Ka

Ka

Ka

Lp

NO

NO

NO

NO

NO

NO

NO

NO

NO

NO

NO

NO

NO

NO

NO

NO

F

F

F

F

F

F

F

F

F

F

F

F

F

F

C

F

S

S

S

S

S

S

F

F

F

F s s s s s s

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

70

60

70

80

75

80

75

75

75

75

75

75

85

75

80

75

30

25

25

15

25

15

25

25

25

25

25

50

15

25

15

25

65

65

65

65

50

50

50

50

50

50

50

40

25

50

50

50

30

30

30

30

40

40

40

40

40

40

40

40

20

40

40

40

45.005

26.580

27.440

26.125

25JOO

23.755

22.470

33.900

18.300

21.345

38.845

38.465

71.600

72.125

82.865

91.865

1.80

.60

.80

.46

1.80

2.00

.50

.30

.70

1.00

4. Manual Operation

4.1.

To store concentration information on the hard disk enter: WDD XXXXXX (Write Data to

Disk filename XXXXXX)

4.2.

4.3.

4.4.

If you wish to print intensity data to the printer as the analysis proceeds, enter: PIN (Print

Intensities)

Place the samples in sampling cups.

Start the measuring program by entering:

SMX (Start Measurement Program). The pro gram will respond with "POSITIONS".

4.5.

4.6.

4.7.

4.8.

4.9.

Respond with: 12341 (Use all four turrets, wait for sample i. d.) The system will bring the first position (#1) into the loading position, open the port and wait for the sample.

Load the sample into the waiting turret (posi tion

Type the sample name. Press [R].

The system will shut the port and bring turret position #2 to the loading position. In case of error (i.e., sample mix-up, no sample, etc.), the instruction can be cancelled by using the /[R] command.

Repeat steps 4.6 through 4.8. until all samples have been analyzed:

"POS # l ABC-1234

(enter sample numbers)

"POS # 2 : " ABC-1235

. . . . . . . . . .etc.

5. Use of Sample Changer

The 72-position sample changer is used to expand the capabilities of the PW1400 as an "unattended" instru ment. Samples are loaded in the sample trays

(6 samples/tray) and trays are placed in the sample changer. Identification cards are inserted into the ap propriate slots of the trays. These cards indicate the tray number and the program (MP/LP) number. As the card is read by the optical reader on the sample changer, the correct measurement sequence will be performed. Samples are identified in the output by the tray number followed by the position number. Samples can be identified in a separate data file (see manual) or samples can be identified using separate lists.

Care must be taken with the automatic sample changer.

The plastic trays and identification cards can be sources of poor performance and both should be changed regularly. New supplies should be ordered from Philips Electronics.

5.1. Enter RES (RESET) to clear any previous condititions.

5.2. Enter ASC 20 (Assemble Sample Changer - tray #20)

EA6-4

Majors - XRF

In this case, the system is informed that tray #20 is the The MP to be employed is indicated on the i.d. card, last to be analyzed in the sequence. In some therefore it is possible to run a sequence of samples for

Databanks, MP1 involves the measurement of a Cu different analytical "packages".

blank at reduced power for 10 seconds. By having this sample in the first sample position of tray #20, with the 5.3. Enter SMX. (this will initiate the analysis).

card indicating MP1, the Cu blank is read and the sample changer sequence stops and powers the system down at the end of the run.

EA6-5

Majors - XRF

DETERMINATION OF MAJOR ELEMENTS (Ml, M2, M3)

X-RAY FLUORESCENCE SPECTROSCOPY

Introduction:

X-ray fluorescence spectroscopy is used for the deter mination of the ten major elements: Si, Al, Total Fe,

Mg, Ca, Na, K, Ti, P and Mn. The results of the analyses are reported as percent oxides. The Philips

PW1400 sequential (wavelength dispersive) spectrometer performs the determinations using two analytical programs, one calibrated for a wide range of silicate rocks (Ml and M2 packages), in which the samples are presented as fused glass discs containing about 15^o rock pulp, and the other for a range of carbonate rocks (M3 package), in which samples are presented as pressed powder pellets.

In the determination of major elements in silicate rocks

(Ml and M2 packages), the sample is matrix-matched by mixing with a heavy absorber, lanthanium oxide

(La2O3). A fusion with lithium tetraborate flux is carried out. The heavy absorber minimizes the dif ferences in mass absorption (MA) between the samples. For carbonate rocks (M3 package), samples are prepared as rock powder pellets using a boric acid backing. No MA correction is employed in the car bonate rock program.

Geochemical Importance:

Geochemistry deals with two main topics: the com position of the earth and the chemical processes that control the distribution of the elements in space and time. The former is of particular interest to analytical chemists concerned with the analysis of geological materials.

The earth can be divided into zones or shells: the core, mantle, crust, hydrosphere, and atmosphere. The core is approximately 3500 km in radius and is composed of 90.8^o Fe, S.6% Ni, and Q.6% Co which is the approximate composition of iron meteorites. The mantle is about 3000 km thick and is assumed to be composed of mainly peridotite, a rock composed main ly of olivine with some pyroxene, both silicate minerals containing Fe and Mg. The crust is composed of mainly granitic simatic and sialic materials differing in relative proportions in different locations. The term simatic comes from silica and magnesia. Similarly the term sialic comes from silica and alumina. These materials are complex and are not homogeneous.

Table MXRF-1 summarizes the approximate com positions of four common rock-types.

The relative abundance of the 14 most common ele ments in the crust is:

The standard classes of rocks are igneous, sedimen tary, and metamorphic. As a rough approximation, because the last two classes are derived from the first, the average composition of the accessible part of the crust can be represented by that of igneous rocks.

TABLE MXRF-1.

Component

Si02

A12O3

FC2O3

FeO

CaO

Na2O

K2O

MgO

TiO2

P205

MnO

C02

Peridotite

43.9

4.0

2.5

9.9

3.5

0.6

0.2

34.3

0.8

0.1

0.2

-

(Compositions are given in Percent by Weight)

Basalt Granite Sediment

49.9

16.0

5.4

6.5

9.1

3.2

1.5

6.3

1.4

0.4

-

0.3

70.8

14.6

0.2

0.1

-

1.6

1.8

2.0

3.5

4.1

0.9

0.4

44.5

10.9

4.0

0.9

19.7

1.1

1.9

2.6

0.6

0.1

0.3

13.4

Upper Crust

55.2

15.3

2.8

5.8

8.8

2.9

1.9

5.2

1.6

0.3

EA6-6

Majors - XRF

These eight oxides make up 977c of the average ig neous rock and eight elements account for 999fc of the weight. All other elements can be considered to be

'minor' or 'trace' elements in most rock types.

TABLE MXRF-2. AVERAGE COMPOSITION

OF IGNEOUS ROCK

Oxide

SiO2

AlzOs

FeiOs

FeO

CaO

Na2O

K2O

MgO

TiO2

P205

MnO

H2O

Percent by

Weight Element

59.1

15.3

3.1

3.8

5.1

3.8

3.1

3.5

1.0

0.3

0.1

1.2

TOTAL 99.4

O

Si

Al

Fe

Ca

Na

K

Mg

Percent by

Weight

46.6

27.7

8.1

5.0

3.6

2.8

2.6

2.1

98.6

Percent by

Volume

93.8

0.86

0.47

0.43

1.0

1.3

1.8

0.3

100.0

Safety Advisory:

1. X-rays can cause both somatic and teratogenic damage. It is mandatory that workers wear a dosimeter badge (service provided by Health and

Welfare Canada, Health Protection Branch).

Pregnant women should not work in the X-ray laboratory.

Radiation is measured in terms of dose or dose- rate. The unit of biological radiation exposure is the RAD (Radiation Absorption Dose), which is the quantity of radiation dose of any kind (alpha, beta, gamma, or X-radiation), that results in the absorption of 100 ergs or energy per g of biological tissue. The unit of measure of X-ray dose (the radiation given off by a radiation source) is the roentgen (R). The unit of biological radiation damage is the REM (Radiation Equivalent Man).

The maximum permissable whole-body dose rate is 5 REMs per year or 3 REMs per quarter. This is equivalent to: 5 Roentgens per year, or 100 milliroentgens (mR) per 40 hour week, or 2.5 mR per hour. The maximum permissable accumu lated life-time whole-body dose in roentgens is calculated by subtracting 18 from the workers present age and then multiplying by 5. Body parts have associated maximum dose levels. Health and

Welfare Canada should be consulted for additional information.

The spectrometer should be thoroughly monitored with a radiation-survey meter when it is installed, when it is moved and at routine 6-month intervals.

The X-ray tube head and specimen area should be monitored each time a tube is changed or if the compartment is disturbed. All monitoring should be done at a strong spectral line of short wavelength (e.g. Mo). The survey meter should be calibrated and sensitive to X-ray energies up to

100 KeV. Readout should be in mR per hour. The meter should not indicate more than 0.5 mR/h at 5 cm from any accessible part of the instrument.

The instrument is equipped with shielding and interlocks which minimize the possibility of con tact with X-rays. The principal sources of ex posure are at the interface between the tube head and the specimen chamber, the chamber itself when samples are inserted and removed (there could be leakage from a malfunctioning shutter), and the analyzer crystal which may be exposed during servicing. Leakage is more serious when working at high voltage and current than when working at low energy and intensity levels.

The primary beam from the tube is very dangerous, generating megaroentgens per minute.

Due to the presence of interlocks the possibility of exposure is extremely low.

2.

The high-potential cable of the instrument (from the generator to the tube) must be secured at both ends. There is still possible danger from high voltage even when the power is off, due to the presence of capacitors in the high voltage circuit.

Operation and maintenance of the instrument should only be carried out by trained technicians.

It is wise to avoid working in close proximity (1-2 meters) to the instrument for long periods of time.

Eye protection should be worn and appropriate precautions for open-flame use should be taken during sample preparations.

EA6-7

Majors - XRF

Method:

The method consists of the following techniques:

1. Matrix modification by either fusing or pressing samples into pellets.

2. Irradiation and measurement of characteristic X- ray fluorescence using a Philips PW1400 wavelength dispersive, X-ray spectrometer.

3. Calculation of final results using calibration cur ves stored in the instrument's computer.

Samples are either fused with lithium metaborate (sili cate rocks) or pressed into pellets (carbonate rocks).

The prepared samples are analyzed using an X-ray fluorescence spectrometer. Carbon, sulphur and loss on ignition data must be acquired prior to preparing samples for X-ray analysis.

Apparatus:

- Balance capable of weighing l -20 g, 0.001 g

- Platinum crucibles (with rim)

- capacity 27-28 ml

- weight 40 g

- Platinum molds

- internal diameter = 40 mm

- heavy construction, weight = 35 g

- Snap Cap vials (12 dram)

- Triangular support

- Retort Stand

- Meker burner cated in the following section, with ap propriate adjustments made in the weight of flux to counteract the effect of the LOI. For samples showing LOI > 10*8?, the weight of rock powder to be roasted is adjusted prior to roasting such that the weight after ignition is

1.500 g. Calculations for LOI correction of sample material are presented in Appendix A.

Reagents:

- Lithium tetraborate flux

- anhydrous

- JMC 99.997c

- specified LOI -cl9fc @ 8000C

- Lanthanum oxide*

- Baker ^8^c, supplied with lot analysis

- Wetting agent: NH4I

*The lithium tetraborate and lanthanum oxide are pur chased in batches and are identified by the manufacturer's batch number. When a new batch is started, it is checked (in triplicate) for LOI and the new blank beads are monitored carefully for contaminants.

The LOI should be approximately G.40% for the flux and l .309fc for the lanthanum oxide. The weight of reagents used in the sample preparation is adjusted on the basis of the LOI data for these reagents.

Procedures:

A complete introduction and detailed description of how to use the Philips PW1400 X-ray Fluorescence

Spectrometer system is presented on page EA6-1.

Additional Apparatus for Automated Procedure: 1. Matrix Modification

- LECOFX-200Fluxer Model 601-600

- Platinum/Gold alloy crucibles: (grain stabilized)

35 mm high x 37 mm ID top x 20 mm ID bottom

- Platinum/Gold alloy casting dishes: (grain stabi lized) 30 mm diameter x 8 mm deep

Samples are prepared for analysis by either pressing into pellets or by fusing the material into glass beads.

Silicate rock powder samples are fused into glass beads. Carbonate rock powder samples are pressed into pellets.

NOTE: Because fused beads are prepared using Pt crucibles and molds, all samples to be fused must be submitted for carbon and sulphur analysis - whether or not these elements are to bc reported. If samples are found to contain less than G.3% S, the samples can be fused without pretreatment. For samples containing

G.3-5% S, the samples are roasted before fusion. Roasting is carried out by placing l .500 g of the sample in a ceramic crucible and heating in a furnace for two hours, with the temperature increasing from 800-

12000C. The roasted sample is used as indi-

1.1. Silicate Rock Analysis:

Manual Preparation of Fused Beads

1.1.1. Weigh the following into a snap-cap vial and, wearing gloves, homogenize: rock powder: lanthanumoxide: tetraborate flux:

1.500g,±0.001g

0.887 g,± 0.001 g

7.326 g,± 0.001 g

It should be noted that the weights listed above are based on 'Pure Reagents' i.e., reagents with Q.0% Loss On Ignition. The amount of

EA6-8

Majors - XRF

reagent actually added will depend upon the

LOI of the specific reagent batch. For ex ample, typical weights for the lanthanum oxide and flux are 0.918 and 7.355 respective ly. The weights are calculated according to the following formulas:

Weight F lux =

Weight La2Oi =

1326x100

100 -Ve LOI flux

0.887*100

100 -

The final weight of the fused bead should be approximately 9.71 g.

LOI calculations for samples are presented in

Appendix A of this document.

1.1.2. Transfer the mixture to a platinum crucible and add a small amount of wetting agent

(NHJ).

1.1.3. Place the crucible in the triangular support on the retort stand and use the Meker burner to heat, on low heat, for 3-4 minutes.

1.1.4. When the iodine fumes have been driven off and the sample begins to melt, place the mold on the crucible and increase the heat.

1.1.5. Periodically during heating, remove the mold and stir the contents by swirling (using Pt- tipped tongs). Replace the mold and continue heating.

1.1.6. When thoroughly molten and mixed, pour the melt into the hot molds. Swirl the melt gently to cover the bottom of the mold.

1.1.7. Allow the mold and contents to cool.

1.1.8. Remove the bead from the mold.

1.1.9. Label the mold with a gummed tag on the upper surface (that is not in contact with the

Pt) and place it in a clean sample bag to protect it from contamination.

1.1.10. Crack and re-fuse imperfect beads (uneven thickness, rippled, nonhomogeneous colour distribution).

These beads are now ready to be analyzed by the

Philips 1400 XRF spectrometer.

1.2. Silicate Rock Analysis:

Automated Preparation of Fused Beads

A Leco FX-200 Fluxer is used to prepare rock powder samples for XRF analysis. Prepare sample/fluxAan- thanum oxide mixtures as described in the manual procedure above.

1.2.1. Ensure that crucibles and molds are clean.

1.2.2. Inspect top of burners. If necessary, use the wire brush to clean the burner heads.

l .2.3. Ensure that gas supplies are sufficient and are turned on.

1.2.4. Ensure that the spark electrode and ther mocouple have not been disturbed. They should be located 3 mm from the pilot arm.

l .2.5. Turn the fumehood on.

l .2.6. Turn the power switch to the ON position.

l .2.7. Push and hold the PUSH TO IGNITE button until the PILOT LED stays on (20-25 seconds).

1.2.8. Add prepared sample mix to the crucible.

1.2.9. Select the number of burners required by pressing the BURN SEL key the appropriate number of times and finally pressing ENTER.

l .2.10. Determine if the current program is correct by successively pressing the ENTER key to ad vance through the previously entered steps.

l .2.11. Place the crucibles into the holders and secure them by sliding the sleeve over the rims of the crucibles.

1.2.12. Place the molds in position and press the

START button.

l .2.13. When the cycle is complete (3-4 minutes), use the tongs to remove the casting dishes and molds.

1.2.14. Remove the bead from the mold and label it with a gummed tag (on the surface that was not in contact with the Pt).

1.2.15. Crack and re-fuse imperfect beads (rippled, nonhomogeneous colour distribution etc)

EA6-9

Majors - XRF

Shut-Down: l .2.16. Turn power ON/OFF switch to OFF. l .2.17. Turn gas supplies off. l .2.18. Turn the fume hood off.

13.

Preparation of Pressed Powder Pellets for

XRF Determination of Major Elements in

Carbonate Rocks:

Samples of carbonate rock are prepared for XRF analysis as rock powder pellets.

Apparatus:

- Ring press and die (40 mm)

- Aluminum "former sleeve"

- Plexiglass plunger

- Boric acid measuring vial (25 ml)

- Spex pellet press (Model 3624B)

Reagents:

- Boric acid (granular)

- Poly vinyl alcohol (2^c w/v in distilled water)

1.3.8. Leave the pellet f ace-down on a cellulose wipe for 24 hours to allow the binder to dry.

A well-made pellet should have no cracks, an even surface and the rock powder should be centered within the outer ring of boric acid (approximately 2 mm). If a pellet is considered unsuitable, a new one is made from a fresh subsample of rock pulp.

It is important to keep the die clean. Boric acid and rock powder can build up on the surfaces and con taminate subsequent samples.

A thorough cleaning with a cellulose wipe after each pellet is removed will generally suffice. Methanol should be used on the wipe to give a more thorough cleaning. The surface in contact with the rock powder must then be treated with care. It can become pitted and must be polished by gently rubbing the surface with a fine abrasive moistened with methanol.

Application and release of pressure on the die-press should be even and slow. The die must be placed centrally and must be level. Failure to do so can result in uneven loading and mechanical failure.

A Philips PW1400 wavelength dispersive X-ray fluorescence spectrometer is used to nondestructively analyze the matrix-modified samples.

A general description on how to use the Philips

PW1400 X-ray spectrometer is presented on page

EA6-1, and the measurement of trace element analyte concentrations is documented starting on page

EA16-1.

Procedure:

1.3.1. Wearing gloves, weigh 4.0 g ± 0. l g of -200 mesh rock pulp into a 25 ml snap-cap plastic vial.

1.3.2. Add three drops of polyvinyl alcohol solution

(to serve as a binder) and blend into the pow der by stirring with a nickel spatula.

1.3.3. Transfer the sample to the 40 mm die to which an aluminum "former sleeve" has been added.

l .3.4. Pack the contents with a plexiglass plunger to form a compact puck.

1.3.5. Remove the sleeve and plunger and add the boric acid powder, placing it on top of the sample. A measuring vial in the boric acid container indicates the amount to be added.

1.3.6. Complete the assembly of the die and form the pellet by application of 15 tons pressure for 15 seconds using pellet press.

1.3.7. Removed the pellet from the die and label it with a felt-tipped marker on the boric acid side.

2. Calculation of Final Results

All necessary calculations are performed automat ically by the system computer.

Major element analysis results are entered on a "Major l and Major 2 Analysis Worksheet". The Major l

(Ml) package includes the 10 major components, LOI and a TOTAL. C/S data are normally obtained for Ml samples. The S is required to prepare the sample

(roasting etc.). The Major 2 (M2) package includes

C/S, H2O\ H2O, FeO and LOI in addition to the 10 major components.

The LOI is not included in the TOTAL but is useful for comparing the total volatile content. In this case, the LOI is corrected to account for oxidation of ferrous ion.

EA6-10

Majors - XRF

Component

SiO2

A12O3

FC2O3

MgO

Na2O

K2O

CaO

P205

TiO2

MnO

LOI

C02

S

H2O*

H2O'

TABLE MXRF-3 . RANGE AND PRECISION FOR MAJOR ELEMENT PACKAGES

30

0

0

0

Range Ve

0 -

0

0

0 -

0

0

0.4 -

0.1 -

0.03 -

0.2 -

0.1 -

0.05 -

Silicate Rocks

2

10

1

6

8

4

3

1

10

10

15

1

80

20

15

20

Precision Ve

0.8

0.3

0.2

0.3

0.5

0.15

0.15

0.05

0.12

0.015

0.4 ,

0.1

0.02

0.2

0.2

0.1

0

0

0

0

0

0

0 -

0

0

0

0.4 -

0.1 -

0.03 -

-

-

-

50

1

1

1

4

22

5

3

8

4

1

1

2

Carbonate Rocks

Range'% Precision 9e

0.02

0.02

0.4

0.1

-

-

0.02

-

0.2

0.1

0.02

0.2

0.4

0.02

0.4

0.02

During fusion, any FeO in the sample is converted to

Fe2O3 by the reaction:

2FeO

= Fe2O3

As the reaction proceeds, the sample gains weight: the weight gained is equal to 0. 1 1 1 x

The true ignition loss, CORLOI, is therefore:

CORLOI = LOI * 0.111(9fcFeO) and is compared to the total volatile content.

The concentration of Fe2O3 in the sample is equal to the difference between the total iron measured in the sample by XRF and the FeO contribution.

Fe2O3 = Total Iron as Fe^-, - 1 . 1 1 1 x (^cFeO)

Each sample is identified by sample number, batch identifier and batch number. Reports are double checked for agreement in duplicates and totals before being submitted to the Section Supervisor.

Quality Control:

It is important to distinguish between short-term

(batch) precision and long-term (multi-year) precision.

The Ontario Geological Survey carries out many multi-year projects. The analytical data used in the resulting report can come from rocks collected and submitted for analysis over a period of several years.

For the geological interpretations to be significant, they must be based on a realistic evaluation of the long-term precision.

One of the Laboratories blind-duplicate quality control programs involves the insertion of one subsample each, of three in-house reference materials (a granite, a basalt and a syenite), into the routine analytical work on a monthly basis.

It is the precision data from this program that are used to derive our advertised analytical capabilities. The stringency of the test allows us to have confidence that the figures we quote are realistic even if applied to multi-year projects. The rock powder samples are introduced 'blind' into the routine workflow and are subject to no special procedures as often happens when an analyst is aware that QC samples are being run.

Research-grade jobs require precisions to be sig nificantly better than those quoted above. These are achieved by rigorous batch control.

The Laboratories has performed a blind study on in- house basalt and granodiorite standards. The data are produced below and indicate the precision typically to be expected when determined by XRF.

EA6-11

Majors - XRF

TABLE MXRF-4. IN-HOUSE BASALT REFER

ENCE MATERIAL (MRB7)

Component Mean

Precision

(2o)

N

SiO2

A12O3

FC2O3

MgO

CaO

Na2O

Ti02

K2O

P2O5

MnO

CO2

LOI

49.14

12.65

13.61

6.13

9.53

2.45

1.96

0.69

0.25

0.189

0.67

2.95

0.40

0.10

0.20

0.18

0.08

0.14

0.06

0.01

0.01

0.005

0.02

0.13

Total rock analysis is characterized by an analysis total

(including LOI and TOTAL VOLATILES) of lOO^c for most samples. If the total value lies outside the range 98.5-101.5*8?, the sample are double-checked.

Possible reasons for "poor totals" may include:

15

15

15

15

15

15

15

15

15

15

15

15

- disagreement between LOI and total volatiles

- error in FeO determination

- error in C/S determination

- errors in sample preparation

- high trace element concentrations

- highfluoride

- high sulphur

- high non-carbonate carbon content

Bias can result from such causes as:

- contamination of standards

- contamination or degradation of analytical crys tals

- changes in the high voltage power supply or other electronic problems

- detector problems

Certified reference materials (CRM's) or Standard reference materials (SRM's) covering a wide range of element composition are used for the creation of calibration curves. Precision and accuracy are monitored through the use of identified duplicates, blind duplicates, in-house reference materials (MRB

'standards') and SRM's.

TABLE MXRF-5.

IN-HOUSE GRANODIO

RITE REFERENCE

MATERIAL

Component

SiO2

Al2Os

FC2O3

MgO

CaO

Na2O

TiO2

K2O

P205

MnO

CO2

LOI

Mean

60.52

14.83

9.03

2.76

3.53

3.27

0.77

2.07

0.18

N.D.

0.26

1.72

Precision

(2o)

0.23

0.10

0.09

0.16

0.04

0.10

0.03

0.02

0.02

N.A.

0.01

0.12

N

A comprehensive series of reference rock samples, duplicate analyses, blank determinations and calibra tion procedures provide the laboratory with sufficient data to assess both the short-term (batch) and long- term (multi-year) precision and accuracy of the data generated. From compilations of large numbers of QC determinations it is possible to outline the long-term analytical capabilities (as advertised in the

Laboratories' Analytical Capabilities Document).

Determination limits, precision and accuracy capabilities should be checked on an ongoing basis in order to assure quality performance in the laboratory.

10

10

10

10

10

10

10

10

10

10

10

10

Productivity:

A technician should be able to complete 60 samples per day.

Additional Notes:

Some commonly encountered problems and possible solutions are discussed below: l. Problems with analysis of glass fusion beads.

If the analysis total lies outside the 98.5-10^ range, it is important that all parameters are checked.

(a) Check the job sheet and see if the rock-type is unusual. Check the agreement between LOI and total volatiles. Check for high sulphur. See if any trace work has been requested and check trace results for unusually high concentrations.

EA6-12

Majors - XRF

(b) Check the LOI sheet to see if LOI calculations are correct.

(c) Analyze the obverse side of the bead, or if neces sary, crack and refuse. It could be necessary to remake the bead from a new sample of rock pulp.

(d) If LOI and total volatiles disagree, re-do C/S analysis and, if necessary, perform another LOI determination. If C/S and LOI are confirmed, then additional volatiles may need to be determined.

(e) Check the Fe07Fe2O3 results. If appropriate, repeat the FeO determination.

(f) If a series of beads are giving problems, it may be useful to check for errors in sample identification or mix-up in samples.

(g) When all else fails, the 'conflicting' elements can be analyzed by classical methods in the chemistry sub-section.

2. Erratic analytical results associated with the XRF spectrometer. Persistent problems should be brought to the attention of the Supervisor and may require a service call.

3. Fused discs containing low (lQ-30%) silica will produce poor results (below the calibration range of the method). Powder pellets are generally good for silica determinations in the range O-1 M. Thus silica in the range ID-30% is difficult to determine.

Classical analytical methods may be used for samples containing this level of silica.

Bibliography:

Mason, B., Principles of Geochemistry, Wiley and

Sons Ltd., New York, 3rd Edition, 1966, 329 pages.

Nockolds, S.R., Average Chemical Composition of

Some Igneous Rocks, Bull. Geol. Soc. America, V.65.

Read, H.H., and Watson, J., Introduction to Geology,

Vol. l, Principles, MacMillian, 1962.

EA6-13

Majors - XRF

APPENDIX A

LOI CORRECTIONS

When the sample has substantial loss on ignition, the weights of sample or flux must be adjusted to produce a finished bead of the required weight.

CASE.1: LOI = 2-1096 (ADD MORE FLUX)

When the LOI of the rock powder is ^9fc buKlO 96, additional flux is added, equal in weight to the predicted ignition loss of the l .500 g subsample of rock powder. This ensures that a 10096 analysis total still refers to the whole-rock sample, i.e., total bead weight remains the same.

Additional flux weight = 1.500 x LOI(sample)

Analytical results are reported as obtained from the instrument.

CASE.2: LOb* 109fc (ADD MORE SAMPLE)

If rock LOI is greater than 1096, it is unlikely that the sample is a silicate rock and that a good analysis will result.

If the analysis is carried out in spite of this, then the weight of sample is increased such that the ignited sample weight in the bead is equal to l .500 g.

Wt. sample = 1.500 x 1007(100-96LOI(sample))

The final analysis total (ea. 10096) will relate to the ignited sample. Results must be factored to relate to the original sample. All concentrations are multiplied by the correction factor:

Correction factor (100-96LOI(sample)V100

All LOI corrections are noted on the LOI sheet. For CASE, l, the additional flux weight is noted in pencil. For

CASE.2, the total weight of rock powder required is noted in red.

Example of LOI corrections:

Base (finished) weight of bead (assuming 096 sample LOI^ 1.500 4- 7.326 * 0.887 = 9.713 g.

Percentage of rock powder in bead ~ (1.500/9.713) x 100 ^ 15.4496

Example- 1: Sample = 10096 SiO2 LOI = 096

96SiO2 inbead = 1 5. 4496 equivalent to 1 0096 (rock)

Example - 2: Sample = 9696 SiO2 LO^49fc

No Correction:

Wt. Si in 1.500 g samp^ 0.96 x l .500 = l .440 g

Wt. finished bead = 9.713 - 0.060 = 9.653 g

96SiO2 in bead = (1.440/9.653) = 14.929k

Result expected from XRF analysis ^ (14.92/15.44) x 10096 ^ 96.6196

EA6-14

Flux correction:

Wt. Si in 1.500 g samp^ 0.96 x 1.500 ~ 1.440 g

Additional wt. flux used ^ 0.06 g

Wt. finished bead s 9.713 g

7cSiO2 in finished bead ^ M.83%

Result expected from XRF analysis = (14.83/15.44) x 1009fc = 96.027c

Example - 3: Sample ^ 859fc SiO2 LOI = I59c

No correction:

Wt. Si in 1.500 g samples 0.85 x 1.500 = 1.275 g

WtJinished bead = 9.488 g

9fcSiO2 in finished bead = U.44%

Result expected from XRF analysis ^ (13.44/15.44) x lOO^c = Sl.03%

Sample weight correction:

Wt. sample used = (100/85) x l .500 s l .765 g

Wt. Si in 1.765 g sample = 1.500 g

Wt. of finished bead = 9.713 g

^oSiO2 in finished bead = 15.44 7c

Result expected from XRF analysis s 100% before correction factor

After application of factor (85/100), result = 85 Ve

Majors - XRF

EA6-15

Majors - XRF

APPENDIX B

METHOD DEVELOPMENT

Setting Up a Routine Analysis Program - Majors

This section is included to illustrate some of the methodology that has been used to develop the routine methods.

A knowledge of these procedures will enhance the operator's ability to perform accurate routine analysis.

Establishing the Parameters

Creation of the various "banks" and indication of how to use instructions is illustrated through the example of setting up a routine analysis program.

The example shows the process from an empty disk, an empty parameter bank and no data bank.

Turn computer on - MCRX14 as outlined in EA 16.

Select the appropriate communications (i.e., terminal at printer, etc.). The system should be in automatic mode indicated by cursor (*).

A. Creation of the Databank, "MAJORS":

Enter *WDB MAJORS

B. Creation of the Parameter bank (PB):

The first information to be entered in the PB are the measuring conditions for each of channel or analyte element.

This can be done in either manual or automatic mode. The example shows the assembly of the iron channel.

Enter at the cursor * ACH (Assemble CHannel). The system will respond with

NAME: :- FE

Enter the required symbol, e.g. type FE, as above. The system will respond with:

LI NE: KA :- [R] (default = Ka = K alpha)

Enter [R] to select K alpha default value. The system will respond with:

FLT: NO :~ [R] (filter Y/NO: default = NO)

Enter [R] to select no filter. The system will respond with:

COL:F :~ C (collimator fine (F) or coarse (C))

Enter C to select coarse. The system will respond with:

EA6-16

Majors - XRF

DETrFS :- [R] (detector: F = flow, S = scintillation, FS = tandem)

Enter [RI to select tandem. The system will respond with:

XTL:1 :- [R] (LiF200 default)

Enter [RI to select the LiF200 crystal. The system will respond with:

ORO: l :- [R] (order of line)

Enter [R] to select the first order. The system will respond with:

UPL:0 :- 8,0 (upper limit on window of pulse height distribution - PHD)

Enter 80 to select the upper limit on the window of the pulse height distribution. The system will respond with:

LWL: O : - L5 Gower limit on window of pulse height distribution - PHD)

Enter 15. to select the lower limit on the window of the pulse height distribution.

In order to assemble a channel, typical UPL and LWL values can be assumed. Values can be reset after PHD analysis.

The system will respond with:

KV MA: O 0:- 60 50 (kV, Ma desired)

Enter 6050 to specify 60 kV and 50 Ma. The system will respond with:

ANGLE :0 :- 57.470 (enter 2-theta angle)

Enter 57.470 to specify the 2-theta angle. The system will respond with:

* OF F S : O : ~ [R] (off-set on high angle side)

Enter [RI to specify O offset on the high angle side. The system will respond with:

- OF F S : O : - [R] (off-set on low angle side)

Enter [R] to specify O offset on the low angle side.

If you wish to subtract background from the peak at angles on either or both sides of the peak, the offset values

+OFFS or -OFFS are entered as the number of degrees above or below the peak, not the actual 2-theta of the background.

EA6-17

Majors - XRF

NOTE: Any time the operator wishes to interrupt this or any other routine, type a slash followed by a return

(/[R]). For example, when the prompt gives a +OFFS and the operator does not want any offsets, a "/[R]" will exit the operator from the ACH routine.

To change a channel parameter for an element already on file, the operator enters:

ACHFE and values given as defaults will be those entered previously. To change a parameter, the new value is typed in.

Once an element has been entered into the table, it cannot be removed. One way to deal with unwanted channels is to rename them using a nonsensical symbol, e.g., XX, YY, ZZ. As new channels are required, these channels can be renamed.

To see the conditions established for any specific channel, the instruction is:

PCHFE

The system will respond with:

ELEMENT FLT COL DET XTL ORD UPL LWL KV MA ANGLE +OFFS -OFFS

FE Ka NO C FS l l 80 15 60 50 57.470 .00 .00

To see all conditions for all channels in the parameter bank, enter the instruction: PCH. The system will respond with:

ELEMENT FLT COL DET XTL ORD UPL LWL KV MA ANGLE +OFFS -OFFS

Fe

Al

Ti

Mn

K

Ca

Si

Na

Mg

P

Ka

Ka

Ka

Ka

Ka

Ka

Ka

Ka

Ka

Ka

NO

NO

NO

NO

NO

NO

NO

NO

NO

NO

C

C

F

C

F

C

C

C

C

C

F

F

F

F

F

F

F

F

FS

F

3

5

1

1

5

4

1

1

3

1

1

1

1

1

1

1

1

1

1

1

75

75

75

75

80

75

75

80

75

80

25

25

25

25

15

25

25

15

25

15

60

40

60

40

60

40

40

60

40

40

50

75

50

75

50

75

75

50

75

75

57.470

144.540

85.925 .80

62.950

136.730

113.085

108.670

55.080

45.225

140.935 1.00

.80

1.00

1.00

It is useful to include in each parameter bank assembly, the following channels for instrument check:

EA6-18

Majors - XRF

ELEMENT FLT COL DET XTL ORD UPL LWL KV MA ANGLE+OFFS-OFFS

Cu1 Ka

Al 1 Ka

NO

NO

C

F

S

F 3

1

1

1 85

75

15

25

25

25

20

10

45.005

144.620

Optimization of channel parameters

Angle Calibration

Once a channel has been defined, parameters can be optimized. A sample containing a high concentration of the analyte is used for this purpose.

There are four sample positions in the changer, l 2 3 SL 4, of which two are of interest, the LOADING and the

MEASURING. To load a sample into a specific location, the instrument sample holder must first be brought into the loading position:

Enter SLP l (Sample Loading Position 1). This will bring sample location (1) into the loading position.

Place a sample cup containing an appropriate sample into the sample chamber.

Bring sample location #1 into the measuring position by entering SMP 1.

To optimize the ANGLE, the instrument is first set at the appropriate channel (e.g. Fe).

Set the spectrometer to the appropriate channel by entering SCHFE (Set channel)

This will set the correct conditions for Fe measurement.

To calibrate the peak position, the system will perform a short scan around the 2-theta angle listed in the PB.

The instrument will determine the "true" peak position.

Enter CANFE (Calibrate ANgle Fe). The system will respond with:

TIME: 1.0:- [R] (count time per point ^ default l s)

Enter [RI to select the default

RANGE: 0.2:- (±0.1 degree around the 2-theta entered (in 0.0050 steps))

The system performs the scan, finds the new peak and reports both the original and the new peak positions. The message ends with a request to save the new angle (default s yes): [R] saves the new angle; NO rejects the new angle. The new angle is automatically stored in the parameter bank.

EA6-19

Majors - XRF

Optimization of Channel Parameters

PHD Selection

Window selection can be upgraded by PHD analysis.

PHDFE

The system responds:

WINDOW: 5 :- (57c window default - 2^o often used)

TIME: l :^ (time at point s l s)

NOTE: If PHD is done in manual mode, the output is a numerical listing of counts. If PHD is done in automatic

mode, a graphical display is output.

Select or confirm UPL and LWL settings and if necessary, enter new values in the PB using the ACH mode. To obtain total counts in window selected previously, run PHD and after the first line has been printed, type /[R] to stop.

Storing Parameters

After all channels have been entered, they are automatically stored in memory at the microprocessor. To store them on the disk (under the same name as the databank) type

DPB

For initial storage, the system may ask for FILENAME, to which the DB name is given. In this example, the instruction DPB creates (or updates) the file MAJORS.PBK.

EA6-20

Majors - XRF

APPENDIX C

SETTING UP AN ANALYTICAL PACKAGE

Once a series of element channels has been established in the Parameter Bank, they should be combined into an analytical package. This requires the construction of various programs:

MP = measurement program

LP = link program

CP s calculation parameters

RC s rate (count) corrections

The Measurement Program (MP)

The measurement program controls the actual acquisition of counts, i.e., channels to be counted, counting time, counting mode, etc., and contains the following information:

- Mask used (aperture); (l = large: 2 = small)

- Count mode; absolute counts or ratio

- Channels and count times

The Measurement Program is stored in the Parameter Bank and there may be up to 63 such programs in any PB.

It is therefore important to keep track of program numbers (identifiers).

For example, if the majors package were to be assigned to measurement program #1 in DB/PB MAJORS, then the instructions would be:

AMP l

MASK = l : = [R]

ABS s YES: = [R]

CHAN PT PC

(default * 1)

CHAN refers to the name of the channel to be measured. This can be either a true analyte channel, a background or offset, Compton line, etc.

PT refers to the counting time in seconds.

PC refers to the maximum number of counts desired: generally this is left to its default value, i.e., no count limit:

GEO.

For the MAJORS, the program includes counting at all major element channels and the offset positions for Mn,

Ti, P, Mg and Na. Counting times for elements range between 20 and 40 seconds, whereas offset positions are counted for only 10 sec.

MASK

ABS

CHAN

0

0

0

0

0

0

PT

0 OEO

OEO

OEO

OEO

OEO

OEO

OEO

PC

^K40

: = Fe 20

: = Mn- 10

: = .......entering all channels

EA6-21

Majors - XRF

The final MP should look like the one printed below (and accessed from the computer by instruction PMP 1).

PMP1

MP1

MASK : l

ABS : YE

CHAN

K

Fe

Mn-

Mn

Ti

TI+

Ca

Si

Al

P

P+

Mg-

Mg

Na-

Na

40

40

10

20

40

40

40

PT

40

20

10

10

10

40

10

40

PC

GEO

GEO

GEO

GEO

GEO

GEO

GEO

GEO

GEO

GEO

GEO

GEO

GEO

GEO

GEO

When a measurement instruction is given, the system will slew to the K position (first channel on list), count for

40 seconds, then continue on the Fe position. If there is more than one sample in the turret, each sample will be measured at the K channel first, before changing to the next channel.

If the operator wishes to see the raw intensity counts, the instruction PIN ^ Print Intensity) should be entered before the measurement instruction is given.

Rate Correction

The MP for the major elements shown above allow for background correction using the offset position. This is only one of the ways in which the count data can be corrected or adjusted. The three principal rate corrections that can be applied are:

- subtraction of a constant count rate

- subtraction of count rate taken at an offset position or at any other channel position (interference)

- ratioing to count rate at another channel

The RC Table

If the count rate at any given position (channel) contains a "constant" contribution from the spectrometer background, sample cup, additive to sample, etc., this contribution can be subtracted from the observed (raw) count rate before other corrections are made to it. This is done in the RC (Rate Correction) Table under the column BKGR.

The RC table is also used to control contributions to peak count from other channels (analytes) or background.

The contribution from the various channels should be identified independently before filling in the RC table.

The coefficients used in the correction are entered in the table as 'L' factors. Up to four such corrections can be applied to a single channel.

Examples are shown below:

EA6-22

NR

4

5

2

3

1

Majors - XRF

Example (1) Peak - Background:

Both channel offsets and independently defined channels can be used to define a peak background. All or part of the counts at the background position can be subtracted from the peak counts to give a net (P-B) count rate. This is useful for samples where differences in the background can contribute strongly to the peak count rate. If the background is not totally flat, a background correction factor can be applied. For example, to determine background contributions for a given combination of trace elements using a single background position, a quartz pellet is run as a blank, and raw intensities printed out. Background correction factors, defined as

[CPS(analyte)]7[CPS(background)] are calculated and entered as the L values in the RC table.

Example (2) Interference Correction:

The regression package of the PW1400 uses "alpha" factors to account for and correct inter-element effects.

This requires that the interferent channel be both measured and calibrated. The alpha correction uses the concentration of interferent to make the correction. This is effective for small interferences. Gross interferences caused by direct line overlaps should be accounted for by

- selecting an alternate line for the interferent,

- running a sample containing a high concentration of interferent and no analyte; calculation of net counts at the analyte position; calculation of ratio of net analyte CPS to raw CPS at the interferent channel to obtain the correction factor, which is then entered as the 'L' value in the RC set.

If different measurement programs and link programs (see later) are stored in the same PB/DB combination for varying combinations of a group of elements, the same background positions and RC files can be used to build the link programs.

The RC table for the major elements (under DB MAJORS) is constructed as follows: 90*^ of the counts observed at the offset position is chosen (arbitrarily) as the background counts for the five elements concerned.

To assemble a correction set:

ARC l

BKGR:

CHL1:

LI :

CHL2:

= [RI

= Ti-f

= 0.9

= /[El

This must be repeated for each set of correction factors. To see a specific entry, the instruction PRC# is given, where # is the number desired. To see the entire table, the instruction PRC is given to get:

BKGR

0

0

0

0

0

CHL1 LI CHL2 L2

P+

Mg

Na

Mn

0.9000

0.9000

-0.9000

-0.9000

-0.9000

0.0000

0.0000

0.0000

0.0000

0.0000

CHL3 L3 CHL4 L4

0.0000

0.0000

0.0000

0.0000

0.0000

0.0000

0.0000

0.0000

0.0000

0.0000

EA6-23

Majors - XRF

Internal Ratios and Mass Absorption Corrections

To account for mass absorption effects between samples, several methods are used, i.e., ratioing of analyte counts to background and Compton scatter peaks.

For example, if the MA of a particular standard reference material is known from major element contribution calculations, then the MA of an unknown sample can be approximated as follows: if then

MS = MA standard

MU s MA unknown sample

RS = cps at reference peak for standard

RU = cps at reference peak for unknown sample

The net intensity observed at the analyte channel for the unknown sample can be corrected for MA effects, by multiplying by a factor, i.e.,

, ^ lUxMU

MS

where T = MA-corrected net intensity

IU = net observed intensity

This, in turn, is equal to:

r = lUxMSxRS lUxRS

MSxRU RU

Since RS is expected to be constant, the MA-corrected intensities are expected to be proportional to IU/RU, where RU can be the counts at either a background position or at the Compton scatter position. The latter generally gives the better results.

NOTE:There is a program for Apple Ile comparing the background, Compton-scatter and major element computations of MA value. Application of this program to results observed for a series of reference materials indicate that MA's obtained using Compton-scatter intensities are generally closer to those obtained from major element computations. The Compton scatter peak for the rhodium tube is therefore generally defined in the trace element channel sets as Rh and appears in many measurement programs as the internal ratioing element.

The Link Program

The Measurement Program results in the production of intensity (count) data. The Link Program (LP) "links" this raw data to correction and calibration parameters in order to correct and convert the counts to concentration units. The LP, which is stored directly in the Data Bank (as opposed to the MP which is stored in the Parameter

Bank), is generally assigned the same number as the MP from which it receives the data.

EA6-24

Majors - XRF

In order to assemble a LP:

ALP l

NEW LIST (for LP not in memory)

NRCH = O : = 5

NRCH refers to the number of channels which are to have count correction applied to them. This count correction can be in the form of background substraction, internal standardization, etc. In the example of the creation of the MAJORS package, there are 5 channels which will receive offset correction.

OPTION 0:- 6.

OPTION refers to one of several data options including entry of external data, dilution factors, analysis totals etc. Option 6 gives the analysis total.

MBNR - O :- [R]

MBNR - Monitor Buffer number. If a monitor buffer is used to store counts, with or without ratioing to a buffer to correct for drift, the number should be entered here. If the buffer is not used, leave MBNR = 0. The application of monitor buffers is explained in the software instruction manual.

INT. RATIO - NO :- [R]

INT. RATIO - internal ratio to the counts appearing at one of the channels measured. If ratioing to background, the Compton-scatter peak or any true internal standard element is required, then Y is entered, and the system requests identification of the internal standard channels (up to three). It should be noted that only "real" channels can be used for internal ratio. Off-set channels, e.g. Ta- or Ga-f cannot be used.

MNR -O : - 1.

CPNR = calculation program (calibration coefficient file). Although any number can be used in this file, it is probably better to use a CP file of the same number as the LP and the MP.

DLST - O : - [R]

MODLST = moduli list s combination channel results and factors to provide some additional calculation or ratio or a specific application. In this example, the value is left at the default of 0.

CHAN RG IR

At this point in the assembly, the channels (CHAN), the rate correction number (RC) and the internal ratio

(standard) channel (IR) are specified. Since there can be up to 63 RC factors, the ones desired must be chosen carefully.

For example:

ELMNT RC

ELMNT RC

IR: :- Ti l [R]

IR: :- P+ 2 [R]...

If an internal ratio were required, the channel on which the ratioing would be calculated is entered following the

RC number. As soon as the number of entries is equal to the NRCH defined previously, the system returns to the cursor.

EA6-25

Majors - XRF

Calibration

Major elements are calibrated using a series of certified reference materials (CRM's) whose element concentra tion values cover an adequate range for the rock type being considered (i.e., silicate or carbonate rock). A regression file is prepared which contains the concentrations and observed intensities for the series. The following example illustrates the steps required. The measurement program is assumed to be MP 1.

Acquiring the Intensities

The following commands are made on the assumption that the appropriate measurement and link programs are available (along with the RC correction set).

SPR l - This command (start prepare regression) instructs the system to run samples and to store the data in a regression file.

"FILENAME" MAJORS - Regression file to be created = MAJORS.R01

"SAM POS" This command is common to all instructions operating the spectrometer. It asks the operator to identify the turret positions for the SRM's. The response is given as the turret numbers separated by spaces; in order to introduce sample (SRM) identifiers, the response is followed by an I.

12341

The sample holder (turret) then puts #1 into the loading position and opens the port. The system prints:

SAMPLE #1 : and awaits the entry of the sample name. The sample is placed in position, the name is typed at the terminal and upon receiving a carriage return, the port closes and the system puts #2 into the loading position. When all the positions specified earlierare filled, the system begins to measure X-ray intensities according to MP l. Intensities are stored automatically under file MAJORS.RO1. (XXXXXX.RYY where XXXXXX = file name, YY ~

Measurement Program Number). Additional reference materials can be run in a similar manner.

Entering Concentration Data

Once all intensities have been measured, the intensity file is opened and the accepted (literature) concentration values are carefully inserted into the file. (Every time the operator wishes to manipulate a data file, it must be opened using the instruction illustrated below).

OPFMAJORS.R01

The system responds with file information, ending message with cursor.

ACD ^ Assemble concentration data)

The system then asks the operator to input concentration data of the elements in the order specified in the MP/LP set for the standards (in the order stored in the regression data file). The Philips software is set-up generally to handle concentration units in percentages. Trace element values can be entered for convenience as ppm/1000.

Thus, for example, a value of 125 ppm would be entered as 0.125. Care must be taken with interpretation and reporting of data.

If the operator wishes to stop entry at any point, he/she types "/[R]" (slash/return). One can re-enter the ACD routine at any SRM, e.g., ACD 4 enters the routine at the fourth SRM.

To view the total concentration file, the operator types

EA6-26

Majors - XRF

PCD ^ Print concentration data) when finished, the file is closed,

CLF (s Close file)

Regression

After all intensities and concentrations are entered into the MAJORS.R01 file, the regression is performed. To do this, the operator leaves the X 14 software (by typing END), and enters the regression software (MCR REG).

The system requests the regression filename. The response is the data filename MAJORS.R01 (or file required).

For theories on the interpretation of XRF data the reader is referred to R.H. Jenkins and B. DeVries, "Worked

Examples in X-ray Spectrometry", Springer Verlag, New York, 1970.

The default calibration procedure is the De Jongh model (DJ) which is a concentration-based correction model.

The general formula is where

Cj = concentration of anal yte

R, = intensity of analyte

Ei = slope of calibration curve

Di = intercept of curve aij = alpha coefficient for interferent j

Cj = concentration of interferent j

This formula is similar to the Lachance-Trail model, but corrects for self-absorption. It is used extensively and has proved successful for most applications. Constants D and E are instrument dependent; alpha coefficients are fixed and known for any sample/spectrometer combination and can be input from tables or calculated during the regression procedure.

After signing on the regression software, the system waits for instruction

SRI s start regression input

MODEL = DJ : = [R] or DJ

LOAD ALL ELEMENTS = YE : = [R]

LOADED ELEMENTS: SI AL FE . . . . .

10 OF THE 10 SAMPLES LOADED(infomi operator of the number of standards in the file)

REGR. ELMNT : ~ SI (input element of interest)

To calculate the D and E values, the operator types:

SRA and the system responds:

ELEMENT = SI MODEL DJ SIGMA = 0.01039 K ~ 0.01358

D

-0.04110

E

1.25447

EA6-27

Majors - XRF

The SIGMA and K values indicate the quality of the fit. The SIGMA is the standard deviation of the regression and is equal to

SIGMA = fy^'^M l

n ~ k J

where (C.-CJ2 is the square of the deviation between observed (o) and accepted (a) concentration values; n s number of samples and k = (number of regression coefficients -f- the number of calculated influence coefficients

(alphas)). Thus, if no alphas are included, k = 2 for a straight line calibration. The K factor is related to the standard deviation and is inversely proportional to the overall concentration levels in the regression (SIGMA =

K C1/2). The regression program is arranged to reduce the K factor to a minimum.

If the k is too large for the number of standards being used in the regression, an appropriate message will be printed out.

The overall quality of fit can be observed by typing:

PRL ( ^ print results list)

The system will indicate the following:

INTENSITY, (observed)

CHEMICAL CONCENTRATION, (input manually through ACD)

CALCULATED CONCENTRATION,

CONCENTRATION DIFFERENCE,

CHEMICAL APPARENT CONCENTRATION

(Equal to the value C/0 -i- SUM aC) from the DJ correction formula. If no alphas are used, the chemical apparent concentration is equal to the chemical concentration),

CALCULATED APPARENT CONCENTRATION,

IDENTIFICATION.

The operator can introduce alpha corrections using the command

CAL XX (CAL = calculate alpha) where XX is the element influencing the result.

All alphas can be eliminated using command

RCF (reset)

Samples can be removed from the calibration list using command DSA and restored using ISA.

Plots can be observed on graphic terminal using command PLT.

Once the regression is deemed satisfactory, the D, E and alpha values can be stored at the appropriate Calculation

Parameter (CP) Set. The CP set is identified by number (as listed in the LP set). To place the regression parameters in the CP set, the operator enters:

SRR l (store regression results in CP 1)

If CP1 is a new list, the system responds with

EA6-28

Majors - XRF

NEW LIST

NAME : (can enter any name or title desired, e.g. Major Elements)

NDEC = 3: ^ (number of decimal places)

If the CP not new, then only the number of decimal places desired is requested.

If the same analyte (and influencing elements) is found in a different CP listing, and if upgrading is also required, the SRR instruction can be repeated, directing the D, E and alpha values to the other CP set. This is very useful in that small groups of elements can be calibrated separately (and more quickly) and new curves be installed in several CP sets.

The instruction PCP allows the operator to view the full calculation parameter set. The instruction ACP

(assemble calculation parameters) can be used to enter manually the D, E and alpha values if required.

To regress on another analyte, the operator types

NRE ^ next regression element)

REGR. ELMNT :

To go back to the main program, the operator exits from the regression program (END) and re-enters the X14 software (MCR X14). It is possible on the MICRO 11 computer to run the system from the printer and to leave the graphics terminal in the regression routine. It is important to remember however, that the same file cannot be open at both terminals.

There are several re-calibration routines available on the PW1400 software. The operator is referred to the operator's manual.

Scanning

As a sequential instrument, the PW1400 can produce excellent scan results. The graphics terminal can be used effectively to enhance scans of regions of interest and produce usable plots at the dedicated printer.

To obtain a scan with output in both numerical and graphical display, it is important to:

- be sure the system is in automatic mode

- have the sample in the correct measurement position (turret)

- assure the data is stored on disk - this can be done by typing

WDD XXXXXX (~ write data to disk) where XXXXXX = name of data file

Since the scan will likely be near an analyte of interest, the operator may want to use the parameters for that element for the scan. Thus, for example, if the scan is in the region corresponding to analyte Cu, the operator enters:

SCH CU

The operator then sets the scanning conditions:

- time for counting at each point:

PTS1 (l sec)

- initial (starting) angle:

SLF45 (slew to 2-theta = 450)

EA6-29

Majors - XRF

- final angle and scanning increment:

STS 50 0.05 (scan to 500 in 0.050 increments)

This final instruction will begin the scan. If the system is in automatic mode with a WDD instruction, the system will write the filename XXXXXX.YYY, which is to contain the scan information. The filename subscript

(.YYY) is assigned by the computer. As the scan proceeds, the numerical intensity value at each point will be listed at the printer (terminal) as it is obtained.

If the scan is performed in the manual mode, then the only presentation of the scan consists of the numerical output appearing at the printer. However, if the scan data is stored, then the file can be opened and a plot created.

To open file:

OFF XXXXXX.YYY

To plot scan:

PLS (the terminal should be the graphics terminal)

The system then responds with a series of messages allowing the operator select the mode of presentation (as curve or as points) as well as a choice of start and end angle. The full scale width of the intensity (count) axis is then printed (and either accepted or changed) and the plot put out on the screen. The screen can be "dumped" to the dedicated printer (LA50) for a hard copy output of the plot.

If a series of scans is to be obtained for the same angle range (and under the same conditions), the instructions can be set into a JOB stored in the parameter bank.

For example, to setup a job for a scan from 45 to 500 (increment 0.050) under the conditions normally used for measuring copper, the following steps are carried out:

AJB l Assemble job # l l EJB: = The first step is assumed to be the end of the job (default). The first instruction is then entered.

1 EJB : = SCH CU

2 EJB : = PTS 1.0

3 EJB : -

SLF 45.0

4 EJB : =

STS 50.0 0.05

5 EJB : - [R]

To perform the scan, one types:

WDD XXXXXX

SJB l (Start Job l)

There are many combinations of instructions which can be included in a jobfile (multiple scans, angle calibration, sample measurement). To delete a jobfile from the PB, type:

DJB l

Running the Sample

Once the Measurement Program (MP), Link Program (LP), rate correction set (RC), and the Calculation

Parameter (CP) sets have been defined, it is possible to run samples on a routine basis.

EA6-30

Majors - AA

DETERMINATION OF MAJOR ELEMENTS

(SiO2, AliO3, Fe2O3, MgO, CaO)

FLAME ATOMIC ABSORPTION SPECTROMETRY

Introduction:

These methods are used only where the sample matrix makes the constituents concerned unsuitable to be determined by XRF. This will arise, for example, when the sulphur or iron content of the rock is too high to allow a fused bead to be prepared for XRF deter mination. This method is also applied when the silica content is between 10 and 30 percent and existing programs for determination by XRF in this range are unsuitable.

Silicon - Silicon is the second most abundant element

in the lithosphere after oxygen. The simplest clas sification of silicate rocks is based upon silica content

(SiO2), which is normally in the range of 35 to 80 percent. Quartzites and sandstones may contain 90 to

95 percent silica. The carbonate rocks are a large and varied group of igneous (carbonatite), sedimentary and metamorphic rocks, with silica content ranging from

< l percent (calcite and certain marbles) to 30 percent in some sediments and carbonatites.

Silica occurs in crystalline forms (quartz, tridymite, crystoblite), and in the amorphous state (opal, onyx, chalcedony). More commonly, silica combines with magnesium, iron, aluminium, alkali and alkaline earth elements to form complex silicate minerals. These include rock-forming minerals such as: olivine, pyroxenes, amphiboles, micas, feldspars, feldspathoids ( e.g. nepheline, leucite, sodalite), and aluminosilicates (kyanite, kaolin, sillimanite). Other minerals such as zircon, tourmaline, and sphene also contain silica.

cryolite. Andalusite, sillimanite and kyanite may con tain up to 60 percent A1203. Rarer aluminium minerals include corundum, chrysoberyl, turquoise, alum and alunite. The main sources of aluminium metal are the bauxite and laterite ores which result from weathering, leaving aluminium as an insoluble residue.

Iron - Iron is the fourth most abundant element com

prising about 5 percent of the earth's crust. Basic rocks may contain 30 to 40 percent iron (as FeO), while many acidic rocks contain as little as l percent total iron.

Ferric iron is frequently associated with aluminium, and ferrous iron with magnesium.

Sulphide minerals such as pyrite, pyrrhotite and chal copyrite are common. Iron carbonate minerals are siderite or chalybite FeC03 and ankerite which is a mixed carbonate of iron, calcium and magnesium.

Calcium - Calcium ranks fifth in the order of elemental

abundance. Rocks which crystallize early, contain small amounts of calcium (periodotites, dunites).

Sucessive crystallization forms calcium-rich feldspars. Calcium containing silicate minerals in clude: augite, hornblende, amphiboles, and the plagioclase feldspars. Final crystallization occurs from a calcium depleted magma, resulting in rocks with less calcium (granite, rhyolite).

Calcium carbonate minerals include calcite and aragonite CaCO3 and dolomite CaMg(CO3)2. Non- silicate minerals of calcium include the sulphates gyp sum CaSO4.2H20 and anhydrite CaSO4, fluorite CaF2, perovskite CaTiO3 and scheelite CaWO4.

Aluminium - Aluminium follows silicon as the third

most abundant element in the earth's crust. The con centration of aluminium depends on the magmatic sequence of crystallization. The aluminium content in rocks is typically from 10 to 25 percent. Aluminium content decrease in order (dolerite, basalt, gabbro) >

(andesite, diorite)> (tonalite, granodioriteXgranite, rhyolite).

Some of the more important minerals of aluminium do not crystallize in the main stages of silicate differen tiation, but appear with rare element concentration at the pegmatite and other late stages of rock emplace ment. These include beryl, spodumene, topaz, and

Magnesium - Magnesium (seventh in abundance in

earth's crust) can constitute as much as 51 percent as

MgO in certain silicate rocks (dunite), or 30 to 40 percent in ultrabasic rocks (picrites and peridotites).

Magnesium content decreases in the order (dolerite, gabbro, basalt^ andesito (granodiorite, tonalite)> granite. During crystallization of ferromagnesian minerals, the first solid fractions are enriched in mag nesium relative to the composition of the magma, the last fractions are enriched in iron.

Magnesium carbonate minerals include magnesite

MgCO3 and dolomite CaMg(CO3)2.

EA7-1

Majors - AA

Manganese - Manganese has an average crustal abun

dance of 600 ppm. The highest manganese contents occur in the earliest rocks to crystallize (peridotites, basalts and gabbros). Granites and rhyolites contain only very small amounts of manganese. Clays and shales contain manganese in amounts similar to ig neous rocks.

Manganese carbonate rocks are variable, ranging from manganese-poor chalk and limestones, to the man ganese-rich ankeritic carbonates.

Reagents:

- Lithium metaborate, anhydrous, LiBO2

- Hydrofluoric acid, HF, 489fc

- Strontium nitrate, Sr(NO3)2

- Boric acid crystal, H3BO3

- Silica powder, SiO2

- Graphite powder

Safety advisory:

1. Exercise extreme care when using any acids and fluxes required for sample dissolution. Their use should only be attempted after the appropriate

MSDS sheets have been read and the safe handling

and first aid procedures understood. Acids should only be handled in a fumehood designated for their use and proper protective equipment worn. Proper ventilation is required when handling fluxes which create a dust control problem.

2. Before operating an atomic absorption spectrophotometer, ensure that the instructions found in the manufacturer's operator's manual are understood.

3. Cylinders of compressed gas used as fuel and oxidant for flame atomic absorption spectrometry must be securely fastened and have the proper regulator. Ensure that the system has no leaks and the gas hoses are in good condition. Review the section on compressed gases found in the Geos cience Laboratories' Safety Manual.

4. When using HF wear glasses and gloves, and be extremely careful. More information on HF is available in the Geoscience Laboratories' Safety

Manual.

Method:

The method consists of the following techniques.

1. Samples are fused with lithium metaborate and dissolved using HF

2. Analyte concentrations are measured by atomic absorption spectrometry

3. Calculation of results using calibration curves produced from reference materials

NOTE: For routine work, in-house reference materials

are used. In special cases, certified reference materials may be used. These reference solu tions may be store in polypropylene bottles and used for a number of sample batches.

Procedures:

1. Reagent preparation

1.1. Hydrofluoric acid (109fc) stock solution- Add

125 ml of 489k HFto 375 ml of distilled water in a 500 ml nalgene graduated cylinder. Store in a clean and empty polypropylene supplier container.

1.2. 30,000 ppm Strontium Buffer - Dissolve 72.0 g of strontium nitrate in distilled water and make to l liter volume with distilled water.

(see Note 1)

Apparatus:

- VarianAA775

- Nalgene ware, 250 ml beakers, 100 ml, 200 ml

volumetric flasks, 500 ml graduated cylinders

- Porcelain crucibles (Coors High-Form, 30 ml)

- Graphite crucibles (32x29 mm, 9 ml)

- Magnetic stirrers with teflon stirring bars

- Thermolyne Muffle Furnace

- Silica tray

- Crucible tongs

- Filtering funnels

- Whatman #41 filtering paper, 12.5 cm

2. Sample Decomposition

2.1. Weigh 0.200 g of sample into a porcelain crucible, (see Note 2)

2.2. Add 1.0 g of lithium metaborate and mix sample and flux with a small teflon coated spatula, (see Note 3)

2.3. Transfer quantitatively to a graphite crucible and place on a silica tray, (see Note 4)

EA7-2

Majors - AA

2.4. Fuse in a muffle furnace for 15 minutes at

10000C.

2.5. Open the oven door and pour the fused sample while hot; using crucible tongs and protective gloves, into a 250 ml nalgene beaker contain ing 60 ml of lO^c HF solution.

2.6. Add 4.5 g of boric acid crystal to the beaker and add 100 ml of distilled water.

2.7. Place a teflon-coated stirring bar in the beaker and place on a magnetic stirrer unit. Stir for

30 to 60 minutes, depending on the type of sample, or until dissolution is complete, (see

Note 5)

2.8. Filter the solution using Whatman #41 filter paper (12.5 cm) into a 200 ml nalgene volumetric flask. Wash filter paper and beaker several times with small amounts of distilled water. Make to volume with distilled water. This is the original stock sample solu tion A.

2.9. Prepare solution to be used for the AA deter mination of CaO, MgO, and Fe2O3.

2.9. l. Pipet 10 ml of sample stock solution A into a

100 ml nalgene volumetric flask.

2.9.2. Pipette 5 ml of 30,000 ppm Sr buffer solution to the flask and make to volume with distilled water.

2.10. Prepare solution to be used for the AA deter mination of SiO2 and A12O3.

2.10. l. Pipet a 50 ml aliquot of sample stock solution

A into a 100 ml nalgene volumetric flask.

2.10.2. Add 10 ml of 30,000 ppm Sr buffer solution to the flask and make to volume with distilled water.

2.11. Retain remainder of stock sample solution A for the determination of MnO.

NOTES:

1. The strontium nitrate should be checked for the presence of major elements as contaminants, in particular for calcium. In addition, a reagent blank should be prepared with each batch of samples.

2. If the samples are high in sulphur or organic material, weighed samples placed in a porcelain crucible should be roasted in a muffle furnace at

6500C for three to four hours, or preferably over night.

3. For samples known to contain >30% iron oxide, add about 15 mg of pure graphite powder to the porcelain crucible containing sample and flux.

This will aid when pouring the fused melt and prevent it from sticking to the bottom of the graphite crucible.

4. New graphite crucibles need to be conditioned by placing in a muffle furnace at 10000C for 15 minutes.

5. It is important that after fusion, the samples should be stirred until dissolved, filtered and made to volume (200 ml) on the same day to ensure a faster dissolution of calcium and magnesium fluorides and prevent precipitation of silica.

3. Calibration and measurement of analyte

The atomic absorption spectrometer is calibrated with in-house and certified reference materials prepared as solutions in the same manner as that used for samples.

Manganese content of geological materials is usually less than 2000 ppm. If the manganese is to be reported as ppm its determination can be made using pure aqueous standards containing manganese for calibra tion (see Traces - AA, page EA17-1). Table MA AS l

(page EA7-4) serves as a guide for chosing which certified or in-house reference materials to use. Once prepared, these solutions may be stored in polypropylene bottles and used for calibration at a later date.

NOTE: CRM are valuable and should only be used in

research work or for programs for certifica tion. In-house MRB reference materials should be utilized for routine work.

To calibrate samples which contain greater than 759fc

SiO2, silica standards can be prepared using 0.200g pure silica powder to obtain a stock solution of 1000 ppm SiO2. Prepare a solution as in procedure 2.10 to give a calibration standard equivalent to 1009fc Si02.

Smaller volumes of stock solution will produce calibration standards less than 100*^, i.e., 45 ml will yield a 909fc SiO2 calibration standard. Check the silica standards against a few reference materials to ensure consistent results for silica.

EA7-3

Majors - AA

CRM

GA

GH

BR

DTS

Mica-Fe

SY-2

SY-3

MRG-1

NBS-la

NBS-88a

PCC-1

TABLE MAASL CONCENTRATION OF CERTIFIED REFERENCE MATERIALS

To be used for Calibration when Determining

Major Elements by AAS

9fcSi02

69.8

75.8

38.5

40.6

34.4

60.07

59.7

39.22

14.1

1.20

41.9

7cA.203

14.5

12.5

10.2

0.29

19.6

12.15

11.7

8.51

4.16

0.19

0.73

^cCaO

2.45

0.65

13.74

0.14

0.45

8.03

8.30

14.68

41.32

30.1

0.53

?*MgO

0.95

0.03

13.28

49.75

4.65

2.66

2.63

13.49

2.19

21.3

43.37

7cFe2O3

2.86

1.33

12.92

8.59

25.76

6.33

6.45

17.9

1.63

0.28

8.23

Other CRM's can be substituted for those listed above.

A complete list of CRM available in the laboratory can be obtained from the Geostandards Coordinator.

A Varian AA775 is used for the determination of major elements in fluoborate solutions. See "Operation of the Atomic Absorption Spectrophotometer - Varian

AA775" on page EA4-1 for details.

Consult the element specific information sheets for the necessary details to perform the determination of each major element. Titanium, chromium and barium are possible additional elements which can be determined using sample stock solution A, Sensitivity limitations may restrict the usefulness of the range of concentra tion for these elements. Trace elements in rocks are generally not sufficiently concentrated in the fluoborate solution for determination by flame AAS.

Quality control:

The determination limits and precision for individually determined "major components" by AAS are listed below.

TABLE MAAS2.

Oxide Determination

Limit

(Percent)

Si02

A12O3

F62O3

CaO

MgO

1.0

0.2

0.2

0.1

0.1

Precision* atMRV

(Abs.Percent)

1.0

0.4

0.4

0.4

0.4

MRV

(Percent)

3

3

5

30

10

4. Calculation and reporting of results

After calibration, all readings are in percent oxide of the element. Readings are recorded, blank corrected and multiplied by the appropriate dilution factor.

Major l and Major 2 Analysis Worksheet forms are to be used for reporting analytical results. Results are reported as Percent Oxide (9fc Oxide). Three sig nificant figures are reported in the range 99.99fc to l.OO^o; two significant figures for Q.99% to 0.019fc.

Less than 0.01 ^o is indicated by Q.00%. A blank means that the element was not determined. N.D. (none detected) or *c(less than) are not allowed.

*Precision is the 957c confidence limit (2a). For ex ample CaO at the 3 percent level is determined to ±0.4 percent (absolute).

MRV Mid-range value.

EA7-4

Majors - AA

Silicon Dioxide SiO2 Aluminium Oxide A1203

STANDARD SOLUTIONS

Use certified reference materials and pure silica pow der prepared in the same manner as the samples.

INSTRUMENT PARAMETERS

Lamp Current (ma)

Wavelength (nm)

Spectral Band Pass (nm)

Flame Description

Nitrous oxide-acetylene

Fuel rich, red cone 10-15 mm high

INTERFERENCES

10

251.6

0.2

NOTES

1. The burner height position is very important as there is a very narrow region of the flame where

Si absorbs.

2. If single beam mode is used allow 30 minutes for lamp warm-up.

3. Scan samples for concentration range in order to bracket the sample with standards. This is neces sary as only a few samples can be run before burner blockage becomes excessive.

4. Before turning flame off, reduce the fuel control down on the instrument, particularly if fuel flow is greater than 100.

STANDARD SOLUTIONS

Use certified reference materials prepared in the same manner as the samples.

INSTRUMENT PARAMETERS

Lamp Current (ma)

Wavelength (nm)

Spectral Band Pass (nm)

Flame Description

Nitrous oxide-acetylene

Fuel rich, reducing, red cone.

10

309.3

0.5

INTERFERENCES

Aluminium is partially ionized in a nitrous oxide- acetylene flame. The added strontium acts as an ionization buffer.

Interferences are dependent on flame conditions and burner height. A fuel rich flame (red cone) decreases the effect.

Silicon depresses aluminium absorbance by the forma tion of an undissociated refractory complex. This is minimized by the addition of Sr as well as matching standards and samples with respect to the major matrix elements.

NOTES

1. Fuel-support ratio and burner height are critical in the determination at low levels of detection.

2. Burner blockage, although not as severe as with silica, will occur with fuel rich flames. Bracket samples and standards when reading absorbances.

EA7-5

Majors - AA

Calcium Oxide

CaO MagnesiumOxide

MgO

STANDARD SOLUTIONS

Use certified reference materials prepared in the same manner as the samples.

INSTRUMENT PARAMETERS

Lamp Current (ma)

Wavelength (nm)

Spectral Band Pass (nm)

Flame Description

Air-acetylene

Oxidizing, fuel lean, blue

INTERFERENCES

3.5

422.7

0.5

Silicon, aluminium, phosphate and sulphate depress the calcium absorbance. These interferences can be controlled by introducing a releasing agent such as strontium and by matching sample and standard solu tions to obviate combined interference effects.

The slight ionization interference which occurs with the air-acetylene flame, is controlled with the added strontium.

A nitrous oxide-acetylene flame causes severe ioniza tion of calcium. The addition of a readily ionizable substance will overcome this effect.

NOTES

1. The depression of calcium signals (air-acetylene flame) is caused by elements which give rise to stable oxy salts.

2. The presense of a cyanogen emission band at

421.5 nm may cause an increase in background noise if not completely separated by the monochromator.

3. The formation of oxysalts is virtually eliminated with the use of a nitrous oxide-acetylene flame.

To control the resulting ionization of calcium with this flame, Q.2% potassium salt should be added to both samples and standards.

4. Noise level was observed to be excessive when the nitrous oxide-acetylene flame was used and the air-acetylene flame is preferred.

STANDARD SOLUTIONS

Use certified reference materials prepared in the same manner as the samples.

INSTRUMENT PARAMETERS

Lamp Current (ma)

Wavelength (nm)

Spectral Band Pass (nm)

Flame Description

Nitrous oxide-acetylene

Fuel lean

Air-acetylene

Oxidizing, fuel lean, blue

INTERFERENCES

3.5

285.2

0.5

Interferences are similar to those which occur with calcium, i.e., silicon, aluminium, phosphate and sul phate will depress the magnesium signal in the air- acetylene flame. Strontium addition and matrix matching of samples and standards will correct for these interferences and any ionization of magnesium in the flame.

Ionization of magnesium in the nitrous oxide- acetylene flame is controlled with the added strontium to the solutions.

NOTES

1. Nitrous oxide-acetylene flame is preferred, al though the air-acetylene flame could be used in the same manner as with the determination of calcium.

2. The signal noise with the nitrous oxide-acetylene flame is not as severe as that observed when deter mining calcium with this flame.

3. The fluorborate solutions tend to diminish the chemical interferences in the flame attributed to the formation of "inter-oxide" compounds such as magnesium aluminate, calcium silicate, etc. A fluorborate solution binds interfering elements to fluorine, whereas sulphate and phosphate inter ferences would probably not be eliminated.

EA7-6

Iron Oxide (Total Iron) Fe2O3

STANDARD SOLUTIONS

Use certified reference materials prepared in the same manner as the samples.

INSTRUMENT PARAMETERS

Lamp Current (ma)

Wavelength (nm)

Spectral Band Pass (nm)

5.0

248.3

0.2

Flame Description

Air-acetylene

Oxidizing, fuel rich, blue

INTERFERENCES

Matrix matching of samples and standards is recom mended.

Majors - AA

EA7-7

Majors - Classical

DETERMINATION OF MAJOR ELEMENTS

(SiO2, A12O3, CaO 4 MgO)

CLASSICAL (GRAVIMETRIC) METHOD

Introduction:

For an introduction about the abundance and occur rence of the major elements, refer to, 'Determination of Major Elements by Flame Atomic Absorption

Spectometry', page EA7-1 of this manual.

Although largely superceded by more productive in strumental methods, classical methods for the deter mination of major elements have generally not been surpassed in terms of the quality of data produced.

The Geoscience Laboratories participates in projects with national and international laboratories and uses these methods for certification purposes.

These method are also used where the sample matrix makes the constituent concerned unsuitable to be determined by XRF. This will arise when the sulphur content of the rock is too high to allow a fused bead to be prepared by XRF.

Safety advisory:

1. When using HF wear glasses and gloves, and be extremely careful. More information on HF is available in the Geoscience Laboratories' Safety

Manual page IV-17.

2. Other acids are used throughout this method, and all staff must review the appropriate MSDS sheets.

Apparatus:

- Platinum crucibles, 25 ml with covers

- Glazed porcelain casseroles, 250 ml

- Watch glasses, 12 cm, supported by glass tri angles

- Glass rods

- Platinum tipped tongs

- Glass beakers, 250 ml, 400 ml, 600 ml, 800 ml and 1000 ml

- Glass funnel, 75 mm

- Filter papers, 12.5 cm Whatman No. 40,41, and

42

- Bunsen burner

- Meker burner

- Muffle furnace

- Hotplate

- Thermofab cloth

- Silica tray

- Graphite crucibles (32 X 29 mm, 9 ml)

- Teflon coated stirring bars

- Magnetic stirrers

- Nalgeneware, 100 ml and 200 ml volumetric flasks, 250 ml beakers, and 100 ml graduated cylinder

Reagents:

- Sodium carbonate, anhydrous, Na2CO3

- Lithium Metaborate, anhydrous, LiBO2

- Boric acid crystals, H3BO3

- Concentrated hydrochloric acid,HCl, 36.59fc -

- Concentrated nitric acid, HNO3, 69.09fc - 71.09fc

- Concentrated hydrofluoric acid, HF, 48*7c

- Concentrated sulphuric acid, H2SO4,

- Strontium nitrate, Sr(NO3)2

- Ammonium hydroxide, HN4OH, 28^ -

- Ammonium oxalate crystals, (NH4)2C2O4.H2O

- Ammonium chloride crystals, NH4C1

- Diammonium phosphate crystals, (NH4)2HPO4

- Ammonium nitrate, NH4.NO3

Method:

The method consists of the following techniques:

1. Fusion and decomposition of the sample by acid digestion

2. Precipitation of SiO2, R2O3, CaO and MgO

3. Separation of SiO2, R2O3, CaO and MgO

4. Weighing of SiO2, R2O3, CaO and MgO

5. Calculation of results

NOTE: R represents a group of elements - Al, Fe, Ti,

and P - which are precipitated as the hydrated oxide by means of ammonium hydroxide in the presence of ammonium chloride. Am monium chloride is added to prevent the precipitation of Mg(OH)2.

EA8-1

Majors - Classical

Procedures:

1. Reagent preparation

1.1.

1.2.

1.3.

ammonium oxalate solution

(NH4)2C2O4.H2O - Heat solution to dissolve, filter and make solution to volume with dis tilled water.

207c Diammonium phosphate solution

(NH4)2HPO4 - Dissolve in distilled water, fil ter and make to volume with distilled water.

29c Ammonium nitrate solution - Dissolve in distilled water, neutralize with ammonium hydroxide using a methyl red as an indicator, and filter.

1.4.

G.1% Ammonium oxalate solution - Dissolve in distilled water, if necessary by heating, and filter.

1.5.

5. 09fc Ammonium hydroxide solution v/v in distilled water.

1.6.

5^o Hydrochloric acid solution v/v in distilled water.

NOTE: Make solutions in quantities sufficient for the

number of samples to be analyzed.

2. Fusion and sample decomposition

2.1. Weigh exactly 0.800 g of sample in a clean 25 ml platinum crucible. Weigh 4 g (on top load ing balance) of Na2CO3 flux. Mix with a small teflon rod.

NOTE: If sample is very high in sulphides or ferrous

iron, a gentle preliminary roasting, in the crucible, of the weighed, unmixed sample at a dull red heat for several minutes is desirable to prevent any subsequent reduction of FeO to metallic iron, which then alloys with the platinum. By fusing with a basic flux the minerals of the rock are changed into com pounds which are soluble in HC1.

2.2. Return the covered crucible to the triangle sup port and heat the crucible with a low flame for approximately 5-10 minutes over a Meker burner. Then use full heat for 20 - 30 minutes.

At no time should the flame envelop the crucible, nor should the crucible bottom ever come into contact with the blue cone of the burner flame.

2.3. Heat the outer surface of the lid with another burner to fuse any particles that may have been spattered on the inner surface of the lid. It is good practice to move back the lid, grasp the crucible firmly with platinum tipped tongs and swirl it to incorporate any unfused material clinging to the sides of the crucible.

2.4. Remove the crucible cover and carefully place it, face up, on a clean watch glass. Grasp the crucible with platinum tipped tongs, remove it from the flame and rotate the crucible as it cools, so that the contents solidify around its walls.

2.5. Cool on a marble slab. Re-heat the crucible with the Meker burner until the inside becomes red but does not melt. Cool as before.

NOTE:The second heating helps break away the cake from the crucible.

2.6. Using a wash bottle, direct a jet of distilled water around the top edge of the cake. This will usually undermine and loosen it.

2.7. Place the crucible lid in a 250 ml glazed por celain casserole and cover with water. Add a few drops of HC1 and heat till all adhering particles are dissolved.

2.8. Remove the lid using platinum tipped tongs and a glass rod, and rinse into the casserole.

Carefully transfer the platinum crucible with the cake after wiping the bottom of the crucible with wet tissue. Tilt the crucible on one side, with the glass rod and add approximately 75 ml of distilled water.

2.9. Cover the casserole with a watch glass sup ported by glass triangle and carefully add 20 -

25 ml of concentrated HC1 from a graduated cylinder. Set the covered casserole on hot plate.

2.10. Heat on "medium" setting and gently press the cake with glass rod until the cake loosens and disintegrates. At this stage, if any unattacked gritty residue is noticed probing with the glass stirring rod, the sample should be discarded and the fusion repeated at a higher temperature.

2.11. Remove the platinum crucible using platinum tipped tongs.

EA8-2

Majors - Classical

2.12. Police (using a rubber policeman) the inside and outside of the crucible to remove adhering particles, adding all rinses to the casserole.

2.13. Rinse the policeman and glass rod into the casserole.

2.14. Set the casserole, with watch glass supported by glass triangle, on a hot plate covered with thermofab cloth. Set the temperature of hot plate between low and medium and evaporate to dryness.

dissolved and more water may be added if necessary in order to make fast dissolution of the salts.

3. l .4. Remove from the hot plate and filter into a 250 ml beaker using a 12.5 cm Whatman No. 41 filter paper.

3.1.5. Use a policeman to remove adhering silica.

3.1.6. Finally, wash the residue and filter paper 10 times with the hot SVc HC1 solution. Cover the funnel with ordinary filter paper and reserve.

3.1.7. Quantitatively transfer the contents of the beaker to the original casserole and evaporate the contents to dryness as before.

NOTE: On adding HC1 to the mixture, chlorides of all

metals present are formed, carbonates are decomposed, and silicic acid is formed. When a large amount of the latter is present, some of it will precipitate out as white solid. On evaporating the contents of the dish to dryness, the silicic acid present is partially dehydrated, and becomes almost insoluble in dilute HC1.

NOTE : Thermofab cloth on the hot plate and moderate

3.1.

temperature prevents splattering of the residue of silicic acid and other salts at the final stage of drying.

3. Determination of SiO2

Silica - first filtration

3.1.1. To the cool residue add 5 ml of concentrated

HC1, wetting all of the residue and particularly the ring marking the original level of the liquid in the casserole, from which it is difficult to remove particles of silica.

NOTE: If distilled water is added first, insoluble basic

salts of iron are apt to form, especially if the residue is warm.

NOTE:The first filtrate from the silica still contains a

small amount of silicic acid, which requires a second evaporation for recovery. This usually amounts to 10 mg SiO2.

3.1.2. Allow to stand 1-2 minutes and then add ap proximately 50 ml of distilled water, washing down the sides of the casserole and the stirring rod.

3.1.3. Cover and carefully heat the casserole on the hot plate until all soluble salts have dissolved as indicated by an absence of gritty particles.

This heating should not be prolonged in order that as little as possible of the silica will be

3.2. Silica - second filtration

3.2. l. Add 5 ml concentrated HG to the cool residue, wetting all of the residue.

3.2.2. Allow it to stand 1-2 minutes and then add approximately 50 ml distilled water. Heat the casserole on the hot plate carefully until all soluble salts have dissolved.

3.2.3. Filter without delay using a 12.5 cm Whatman

No. 40 filter paper in a 72 mm funnel as described previously, catching the filtrate in a

400 ml beaker.

3.2.4. Thoroughly police the inside of the casserole and the stirring rod, and add all washings to the filter funnel.

3.2.5. Wipe the stirring rod and lip of the casserole with a piece of filter paper and add it to the funnel.

3.2.6. Finally, wash the paper and the residue 10 times with hot 57c HC1 solution, paying par ticular attention to the upper edge of the paper.

Lift the paper to drain the stem of the funnel and rinse the tip of the funnel into the beaker.

3.2.7. Cover the beaker and reserve the filtrate for other determinations.

NOTE: A small amount of silica still escapes recovery

and goes into the filtrate from the second evaporation and filtration. This is recovered from the R2O3 precipitate and determined by

AA after fusing with LiBO2 in a platinum crucible.

EA8-3

Majors - Classical

3 J. Silica - ignition of precipitate

3.3.1. Carefully fold the papers in the funnels with platinum tipped forceps and transfer them to the original platinum crucible.

3.3.2. Wipe out the insides of the runnels with a piece of filter paper and add it to the crucible.

3.3.3. Place the crucible with the cover not quite in place, using a clean fused-silica small tray, in a cold electric muffle furnace and bum off the paper by allowing the temperature to rise slow ly.

3.3.4. Ignite the residue at 10000C for 30 minutes (the full heat of a Meker burner may be used but the furnace is preferred.)

3.3.5. Cool the covered crucible in a desiccator for 30 minutes and weigh it.

3.3.6. Return the covered crucible to the muffle fur nace for 20 minutes, cool for 30 minutes in a desiccator and again weigh it.

3.3.7. Continue to ignite for 20 minutes periods until two subsequent weights differ by no more than

0.0002 g. The final weight is that of the impure silica.

to drop into the crucible) and cautiously heat the outer surface of the cover with a low flame to volatilize condensed HF.

3.4.5. Evaporate the contents of the crucible to fumes of SO3 (raise the temperature of the hot plate as needed), then cautiously evaporate the excess

H2SO4 over a low flame, holding the open crucible with platinum tipped tongs.

NOTE: Care is necessary to prevent spattering, par

ticularly if there is much TiO2 and Zr present.

3.4.6. When the evolution of SO3 fumes has ceased, heat the crucible to dull redness to decompose the sulphates that are present.

3.4.7. Cool the crucible and gently wipe the outer surface with a damp cloth to remove adhering sand particles.

3.4.8. Ignite the covered crucible in an electric fur nace at about 8000C for 5 minutes, cool and weigh. The difference in weight is that of the pure silica.

3.5. Recovery of SiO2 from R2O3 precipitate and determination by Flame Atomic Absorption

3.4. Silica - Treatment with HF to determine impurities

3.4.1. Cautiously moisten the residue with ap proximately l ml of distilled water and add 6 drops of dilute H2SO4 (1:1).

3.4.2. Quickly pour 10 ml of HF (489fc) into the crucible and replace the cover.

3.4.3. Allow the crucible to stand for l or 2 minutes to permit the initial reaction to take place, then place the covered crucible on a hot plate, covered with a thermofab cloth, at about medium temperature in a fume hood and allow to stand covered for 5 minutes.

NOTE:The ignited silica residue is never pure, but

always contains small amounts of Fe, Al and

Ti oxides. If the washings were careless it may also contain NaCl. Silica from rocks high in Fe and Ti is usually the most impure.

3.4.4. Remove the cover with platinum tipped tongs

(allowing any condensate on the inner surface

Silica - decomposition

3.5.1. Add 1.0 g LiBO2 to the platinum crucible al ready having ignited R2O3 precipitate, mix well with a small spatula, and cover.

3.5.2. Fuse in a muffle furnace for 15 minutes at

10000C.

3.5.3. Place on a clean small silica tray.

3.5.4. Open the oven door and remove the crucible using platinum-tipped tongs and place on another clean fused silica tray and cool it.

3.5.5. Place crucibles in a clean 250 ml nalgene beaker.

3.5.6. Place a small teflon coated stirring bar in the crucible.

3.5.7. Add 60 ml of 109fc HF.

3.5.8. Add 100 ml of 4.57c boric acid solution.

3.5.9. Stir for l hour or more until dissolution is complete.

EA8-4

Majors - Classical

3.5.10. Filter the solution using Whatman No. 41 filter paper (12.5 cm) into a 200 ml nalgene volumetric flask.

NOTE: If iron is very high, it is difficult to obtain a

clear solution. SiO2 will be attacked and will be in solution.

3.5.11. Wash filter paper and beaker several times with small portions of distilled water. Make to volume with distilled water. This is your original solution.

3.5.12. Pipet 50 ml of original solution to a 100 ml nalgene volumetric flask.

3.5.13. Add 10 ml of 30,000 ppm Sr buffer solution to the flask and make to volume with distilled water.

3.5.14. Determine SiO2 as described in the method

Major Element Determinations by Flame

Atomic Absorption, page XXX.

3.5.15. Add this SiO2 value to the value of pure SiO2 to obtain total SiO2.

4.1.4. Heat to boiling, reduce the heat, and then add pure aqueous ammonia from a dropping bottle until a permanent precipitate forms, stirring vigorously at all times.

4.1.5. The precipitate will appear textureless until near the end point, at which stage it will coagu late. If an indicator was used, the supernatant liquid will be yellow at the end point.

4.1.6. If an indicator was not used, the nearness of approach to the end point can be tested for by adding a drop of the indicator to the quiescent solution and noting the colour of the drop as it strikes the surface.

4.1.7. Allow the precipitate to settle and confirm that the end point has been reached; the solution should smell faintly of ammonia.

4.1.8. Boil for one minute, let precipitate settle and filter through a well fitted 12.5 cm Whatman

No. 40 filter paper, in a 75 mm filter funnel.

NOTE: An excess of ammonia is avoided to prevent

the redissolving of A1(OH)3 An excess has no effect on Fe(OH)3 or Ti(OH)4.

4. Determination of R2O3 (All Fe, Al, Ti and P)

4.1. R2O3 precipitation

4.1.1. The filtrate from the second silica filtration is used which should be in a 400 ml beaker.

NOTE: All of the iron present in this filtrate is in the

ferric form, due to oxidation during fusions and evaporation.

4. l .2. Add 5-6 g of NH4C1, depending on the amount of Mg likely to be present, and dilute the solu tion to 200 ml.

4. l .9. Catch the first few milliliters of filtrate in a 150 ml beaker. If the filtrate is cloudy, re-filter, catching the solution in the original beaker.

Repeat these steps until a clear filtrate is ob tained.

NOTE: In this ammonia precipitation Fe, Al and Ti present are precipitated as hydroxides. Any small amount of phosphate present is also precipitated as Fe or Al phosphate.

NOTE:NH4C1 is added to prevent precipitation of

Mg(OH)2 with the Fe and Al. The more Mg present the greater the amount of NH4C1 needed.

4.1.3. Most of the free acid is neutralized by adding pure NH4OH from a small beaker. If the solu tion is dark-coloured, indicating the presence of much iron, there is no point in adding a pH indicator, if the solution is light-coloured, add

3 drops of G.2% methyl red (609fc alcoholic solution, pH 4.2 - 6.3).

4.1.10. When the filtrate is clear, replace the small beaker with one of 600 ml capacity and filter the remainder of the solution as readily as possible, keeping the bulk of the precipitate in the beaker.

NOTE: The filtrate from this precipitate contains am

monia and sodium salts, as well as Ca, Mg and most of the Mn orginally present.

4.1.11. Transfer the precipitate to the paper, wash the beaker and stirring rod twice with hot 296

NH4NO3 (neutralized) solution. Wash the paper and precipitate ten times, taking care to wash the precipitae away from the edges of the paper.

EA8-5

Majors - Classical

4.1.12. Wipe the lip of the beaker and the stirring rod with a small piece of filter paper and add this to the funnel.

4.2.9. Carefully police the beaker and stirring rod, using the NH4NO3 wash solution, and wash the paper and precipitate, 6-8 times with hot 29fc

NH4NO3 wash solution.

NOTE: It is difficult to wash this precipitate thorough

ly, due to its gelatinous nature. By re-dissolv ing in acid and re-precipitating, a much better separation is effected.

NOTE: It may be necessary, if there was much acid

retained by the paper, to add a few drops of aqueous ammonia to the solution.

4.2.8. Rinse down the sides of the beaker, and heat to boiling for l min. Allow the precipitate to settle and filter as before, combining the two filtrates in the 600 ml beaker.

4.2.10. Wipe the lip of the beaker and stirring rod with a small piece of filter paper.

4.2. RiOa re-precipitation

4.2.1. With the aid of platinum tipped forceps, remove the paper from the funnel and carefully spread it out on the inside wall of the original beaker.

4.2.2. Wash the precipitate from the paper with a jet of hot 5 9fc HC1 and rinse the walls of the beaker and stirring rod also.

4.2.3. Place the beaker under the funnel and rinse the funnel with 59fc HC1, then with distilled water.

4.2.4. Finally, wash the paper once or twice with distilled water, fold the paper into a triangle with forceps, and drape it over the rim of the beaker.

4.2.5. Heat the contents of the beaker until the precipitate has dissolved (add more 12M HC1 if required).

4.2.6. Dilute to about 150 ml and heat to boiling.

4.2.7. Repeat the precipitation as previously described but, when the end point has been reached, add 2 to 3 drops ammonia in excess, and add the filter paper to the solution, shred ding it with the stirring rod and forceps and stirring vigorously to macerate it thoroughly.

NOTE:The added pulp from the macerated paper aids

in the filtration to follow. Furthermore, when the hydroxide precipitate is ignited, the presence of the paper gives the ignited oxide a porous texture which aids in oxidation of iron to Fe2O3.

4.2.11. Cover funnel with filter paper and set aside until ready to ignite.

4.2.12. Reserve filtrate for determination of Ca and

4.3.

Mg.

Ignition of R2O3 precipitate

4.3. l. With the platinum-tipped forceps carefully lift the paper containing the bulk of the R2O3 precipitate from the funnel and place it in the platinum crucible used for the determination of silica.

4.3.2. Fold the upper edges of the paper over the precipitate; avoid soiling the sides of the crucible because, after ignition, it is difficult to remove these stains during the pyrosulphate fusion.

NOTE: As stated above, the residue from the silica

consists of Fe, Al and Ti oxides. These belong to the ammonia precipitate, and therefore the later is ignited in the crucible with the silica impurities.

4.3.3. Wipe the upper edge of the funnel with a small piece of filter paper and add it to the crucible also.

4.3.4. The bulky precipiate should be partly dried, or at least well drained, before being placed in the crucible; if not, there is danger that entrained liquid will boil and some precipitate will be lost by spurting.

4.3.5. Partly cover the crucible and place it in a cold electric muffle furnace.

4.3.6. Allow the temperature to rise slowly and en sure that there is free access of air to the furnace during the initial stages of the ignition.

4.3.7. Finally heat at just below 12000C for 40 minutes, cover the crucible, cool and weigh as usual. When much iron is present it is a useful precaution to transfer the crucible from the muffle furnace to the full heat of a Meker

EA8-6

Majors - Classical

burner for 5 minutes, with the lid displaced to allow free access of air to the crucible, to ensure that oxidation of iron is complete; there is little likelihood of the ignited A1203 absorbing water having been heated at 12000C.

4.3.8. Repeat ignition for 20 minutes until constant weight is obtained.

4.3.9. From the total percent of the R2O3 oxides, the precentage of A12O3 is obtained by subtracting the percentage of the other constituents of the group, in particular those for total iron (as

Fe2O3), P2O5, TiO2 and residual SiO2.

NOTE: Because of number of subtractions from the

R2O3 precipitate due to presence of Fe2O3,

TiO2, P2O5, residue SiO2 and other impurities, the A12O3 value will not be suitable for ac curate work. Therefore, A12O3 should be determined by Flame Atomic Absorption as described in the method 'Major Elements by

Flame Atomic Absorption Spectrometry', page EA7-1 of this manual.

5.1. Determination of CaO

5.1.1. The calcium is determined in the filtrate from the R2O3 precipitate.

5.1.2. Heat the solution in 600 ml beaker and evaporate to about 200 ml. Make it acidic with

12MHC1.

5.1.3. Add a few drops of methyl orange (G.02% aqueous solution) indicator.

5.1.4. Dissolve 3g ammonium oxalate

(NH4)2C2O4.H2O in 50 ml of distilled water, heat to 70-800C and filter through a 12.5 cm

Whatman No. 40 paper into the sample solution

(if several precipitations are to be done use 50 ml aliquots of hot, filtered 69fc solution of am monium oxalate.

NOTE:The amount of ammonium oxalate to take

depends on the amount of Ca to be expected from the nature of the rock, e.g., 3 g am monium oxalate is enough up to 509fc CaO based on a 0.8 g sample weight.

NOTE:The ammonium oxalate solution is filtered to

remove small amounts of impurities. Calcium oxalate is often present.

5.2.

5.2.2.

Calcium - first precipitation

5.2. l . Heat the solution to near the boiling point and add aqueous ammonia (1:1) dropwise, while stirring vigorously, until the indicator changes colour and then add about l to 2 ml in excess.

Heat to near boiling (but do not boil) and allow to stand without further heating 2-3 hours with occasional stirring.

5.2.3. When *^17o calcium is present the precipitate may not appear for several minutes and cal cium should not be judged to be absent until the solution has stood for the time recommended.

NOTE: The calcium oxalate precipitated is never pure but contains occluded magnesium oxalate as well as sodium salts. Hence, after filtering the precipitate is dissolved, re-precipitated and re-filtered.

NOTE: Be careful when heating to avoid bumping, especially where there is a heavy precipitate of calcium oxalate.

5.2.4. Filter the solution through 12.5 cm Whatman

No. 40 paper into an 800 ml beaker, retaining as much of the precipitate in the original beaker as is possible. Wash the precipitate 3 or 4 times, by decantation, with cold Q.1% am monium oxalate and pour the washings through the paper.

NOTE: Due to the common-ion effect, Ca2C2O4 is less

soluble in weak ammounium oxalate solution than in pure water.

5.2.5. Wash the paper 3 or 4 times with the am monium oxalate solution. Reserve the filtrate.

5.3.

Calcium - re-precipitation

5.3.1. Wash down the sides of the original beaker with hot 59k HC1, add 2 ml 12 M HC1 to the solution and heat to boiling to dissolve the precipitate.

5.3.2. Pour the hot solution through the paper, catch ing the filtrate in a 250 ml beaker.

5.3.3. Wash the beaker thoroughly with hot 57c HC1, pouring all washings through the paper, and then wash the paper thoroughly with the HG solution, making sure that all areas of the paper are reached (lift up the inside flap with

EA8-7

Majors - Classical

platinum tipped tongs and wash the area beneath it).

5.3.4. Finally, wash the beaker and paper once with water. The final volume should be about 100 to 150 ml.

5.3.5. Remove and discard the paper and rinse the funnel into the solution once with water.

5.3.6. Add approximately 0.5 g ammonium oxalate, dissolved in a few ml of distilled water, to the acid solution; add 2 drops of indicator (methyl orange, Q.02% aqueous solution) and heat the solution nearly to boiling.

5.3.7. Precipitate the calcium as described previously and allow to stand for at least 4 hours, or overnight.

5.3.8. Filter through 12.5 cm Whatman No. 40 paper, combining the filtrate with that obtained from the first filtration, and transferring the precipitate quantitatively to the paper.

5.3.9. Wash the precipitate and paper 10 times with the cold Q.1% ammonium oxalate solution.

Reserve the combined filtrates for the sub sequent separation and determination of mag nesium.

5.4. Calcium - Ignition of calcium oxalate

5.4.1. Place the loosely filter paper in a weighed

25 ml platinum crucible.

5.4.2. Wipe the inside of funnel near top with a small scrap of ashless paper to remove the small amount of precipitate which usually creep up from the edge of paper.

5.4.3. Add it to the crucible, partly cover it.

5.4.4. Burn off the paper at a low temperature in an electric muffle furnace, starting with a cold furnace.

5.4.5. Heat the crucible at 10000C for 30 minutes

(displace the cover slightly at the start to facilitate the escape of carbon dioxide and then cover it tightly), then cool for 30 minutes in a desiccator and weigh rapidly.

5.4.6. Re-heat at 10000C for 15 minutes, cool for 30 minutes and weigh as rapidly as possible (place proper weights on the balance in advance of the crucible).

5.4.7. Continue heating until constant weight is ob tained.

NOTE: When calcium oxalate is ignited it looses CO

and CO2, leaving a residue of CaO, which is weighed. The oxide should be white in colour but occasionally may be light brown or green because of manganese. If the successive separations have been made as described neither of these should be present at this junc ture. If a correction for manganese is con sidered necessary the determination may be made by AA on the solution of the ignited oxide. A knowledge of the strontium and barium content of the sample will indicate whether or not a correction must be applied to the ignited residue for the presence of these elements.

6.1. Determination of MgO

6.1.1. The magnesium determination is made on the filtrate from the calcium filiations which should be about 400 to 500 ml in a 800 ml beaker.

6.1.2. To the cold (about 100C) combined filtrates add a filtered solution of dibasic ammonium phos phate (NH4)2HPO4 to give approximately l g of the reagent per 100 ml of filtrate with l g in excess.

6.1.3. Stir and add with vigorous stirring sufficient concentrated aqueous ammonia to make the solution 107c by volume in aqueous ammonia and continue to stir until precipitation begins.

6. l .4. Allow to stand overnight in a fume hood.

NOTE: When any soluble ortho-phosphate salt is

added to an ammoniacal solution containing

Mg and ammonium salts, the Mg is precipitated as white, crystalline MgNH4PO4.

When this compound is precipitated the first time, from the calcium filtrate it is usually impure due to the presence of large amounts of ammonium and sodium salts in the solution.

This is rectified by dissolving the precipitate and re-precipitating under better conditions.

NOTE: Avoid striking the walls of the beaker with the stirring rod during stirring. The abrasion of the glass surface encourages the growth of fine

EA8-8

Majors - Classical

crystals of the precipitate on the walls of the beaker. When only a small amount of Mg is present, however, precipitation can some times be initiated in this fashion.

6.1.5.

Filter the solution through 12.5 cm Whatman

No. 42 filter paper into a 1000 ml beaker, preferably in front of a fume hood. The bulk of the precipitate should be kept in the beaker.

6. l .6.

Wash the beaker and precipitate twice with 59fc aqueous ammonia (v/v) and pour the washings through the filter, wash the precipitate and paper five times with this wash solution, adding all washings to the filtrate. Reserve the filtrate for subsequent examination.

6.2.

Magnesium - re-precipitation

6.2.1.

Dissolve the precipitate in the beaker in the minimum volume of hot 59fc HC1 and rinse down the walls of the beaker.

6.2.2.

Cover and heat to near the boiling point and pour the contents of the beaker through the filter, catching the filtrate in a 250 ml beaker.

quarters of the supernatant solution and filter the rest as previously described.

6.2.9. Transfer the second precipitate to the same paper. Wash beaker(s), precipitate and paper with cold 57e aqueous ammonia as before.

6.2.10. Police the beaker(s) if necessary. Wash the precipitate and paper ten times with aqueous ammonia solution, and once with water. Dis card the filtrate.

6.3. Magnesium - Ignition of magnesium am monium phosphate

6.3.1. Fold the paper lightly (do not make a tight fold because this makes the burning away of the carbon more difficult) and place it in weighed

25 ml platinum crucible.

6.3.2. Wipe the inside of the funnel with a small piece of filter paper and add this to the crucible.

6.3.3. Place the crucible with the cover drawn back slightly, in a cold electric muffle furnace.

6.3.4. Allow the temperature to rise to about 4500C and maintain it at this level until all of the carbon is burned off and the residue is greyish- white in colour. Do not allow the crucible to become even a dull red before this stage is reached, and at no time allow the contents of the crucible to catch fire.

6.2.3.

Wash the paper and funnel with a small amount of hot SVc HC1 (raise the inside flap of the paper and wash behind it to dissolve any trapped precipitate) then with water, and remove and discard the paper.

6.2.4.

Wash the inside of the funnel once with

HC1 and then with distilled water; rinse the tip of the funnel into the solution also.

6.2.5.

To the filtrate, in a 250 ml beaker and having a volume of about 100ml, add approximately 0. l g (NH4)2HPO4 and cool the solution to about

100C.

6.2.6.

Add pure aqueous ammonia dropwise until a precipitate forms; allow this to settle and con tinue to alternate addition of reagent and set tling of the precipitate until precipitation is complete as evidenced by no formation of a precipiate on the addition of a drop of aqueous ammonia.

6.2.7.

Add 10 ml of aqueous ammonia and allow to stand overnight.

6.2.8.

Examine the first filtrate for signs of precipitate. If none, discard the solution; if a small precipitate is present, decant about three

NOTE : When MgNH4PO4 is ignited, it losses H2O and

NH3 and leaves a residue of magnesium pyrophosphate, Mg2P2O7. Burning off the carbon at a low temperature is very important; otherwise the platinum crucible, if one is used, may be injured and furthermore, it will be impossible to get a residue free from carbon.

6.3.5. Heat the covered crucible and contents at ap proximately 11000C for 30 minutes, cool in a desiccator for 30 minutes and weigh as

Mg2P207.

6.3.6. Repeat the ignition, cooling and weighing until constant weight is obtained.

NOTE:The ignited residue may now be corrected for

co-precipitated manganese, if this was not removed previously (if it is present). The manganese found is subtracted as Mn2P2O7 from the total weight of the pyrophosphate.

EA8-9

Majors - Classical

NOTE: If Mn is determined in a separate sample using

AA, the Mg2P2O7 precipitate may be corrected for Mn as Mn2P2O7, assuming that all of the

Mn was precipitated with the Mg. Another alternative is to determine Mn by AA by dis solving Mg2P2O7 precipitate in HNO3. Im purity of Ca can also be checked in this solution if present by AA.

Conversion Factors:

Mg2PA

Mn2P2O7

Mn2P2O7

0.3623

2.7604

0.4998

2.0007

2.5831

± MgO

± MnO

Mn

3. All beakers, casseroles, flasks etc., containing solutions being analyzed must be labelled to prevent confusion and mixing of samples.

4. Two points on washing precipitates: (1) There is a danger of losing some of the precipitate if too vigorous a jet from the wash bottle is used; (2) In washing, use small quantities of wash liquid, and let each portion drain through before adding the next.

5. Preliminary heating of a crucible containing a wet paper and precipitate should be very slow until the water is removed, a very gradual increase of temperature follows, until the volatile matter from the paper has been smoked off. During this stage do not allow the gases from the paper to bum with a flame since this may result in drafts and loss of precipitate. Then increase the temperature to a dull red heat until all carbon has disappeared.

Never heat strongly while carbon is still present.

The crucible is kept only partially covered, to allow entrance of air. During final heating, the crucible is almost fully covered.

6. Hot platinum-ware should always be handled with platinum tipped tongs. Before picking up crucibles etc.,with tongs, be sure that the tips of the latter are clean.

Quality control:

The determination b'mit for this method is Q.50% for all oxides. This method is not used on a routine basis and only limited data are available to estimate precision and accuracy. However, a number of Stand ard Reference Materials (SRMs) have been analyzed by the Laboratories and the data contributed as part of the SRM certification program. In every case the accuracy and precision have been outstanding. Typi cally, the reported Geoscience Laboratories data have been within 0.2 of the certified value when considered as members of the population of contributed data.

Productivity:

A technician should be able to complete 4 samples per week.

6. l Most stains in platinum can be removed by fusion with little potassium pyrosulphate (K2S2O7).

After cooling, the pyrosulphate is dissolved in hot water.

6.2 Never allow the blue cone of a flame to touch the platinum crucible being heated. To do so causes formation of platinum carbide, which will even tually ruin the crucible.

6.3 Platinum-ware should be kept well burnished with fine sea-sand. This is important for crucibles which are subjected to long ignitions.

Additional Notes:

1. Beakers, funnels, flasks and other glassware should be cleaned well before use. If greasy, a rinse with chromic acid cleaning solution is desirable. Rinse vessels thoroughly with tap water, and follow with distilled water.

Bibliography:

Outline of methods for the Chemical Analysis of

Rocks and Minerals, Department of Mineralogy,

University of Toronto.

2. Wash solutions of various kinds are called for in the course of the analysis. To save time, special wash bottles containing 2^c ammonium nitrate,

5^c HC1,5^o ammonia and 0.19fc ammonium oxa- late solution are kept on hand.

Maxwell, J.A., Rock and Mineral Analysis, Wiley

Interscience Publishers, 1968.

Potts, J.P., A Handbook of Silicate Rock Analysis,

Blackie and Sons Limited, Glassglow, 1987.

EA8-10

Titanium

DETERMINATION OF TITANIUM AS (TiO2)

Introduction:

Titanium is recognised as one of the most universally distributed elements and the tenth most abundant in the crust of the earth.

It is present in all igneous, metamorphic, or sedimen tary rocks of a more or less siliceous character.

Though present even in the most siliceous rocks, it is more abundant in the so called basic silicate rocks.

Chief among its numerous minerals are rutile, oc tahedrite (anatase), ilmenite, titanite and pervoskite. It is also a component, in smaller amounts, of many pyroxenes, hornblendes, biotites, garnets, and other ferro-magnesian minerals, and is found in some mag netite and hematite.

Owing to the refractory nature of some of its com pounds, titanium tends to concentrate in the residual products of decomposition of many rocks. Typically, concentrations in rocks will not exceed l percent, but may rise to over 5 percent.

In this method the concentration of titanium is deter mined by measuring the absorbance of the yellow- coloured complex formed by titanium with hydrogen peroxide in an acidic solution. The measurement is taken by a photometer at wavelength 435 nm.

This method is used where the sample matrix makes titanium unsuitable for determination by XRF. This arises, for example, when the sulphur content of a rock is too high to allow a fused bead to be prepared for XRF determination.

This method is applied when TiO2 content is 3.07e, and above the existing XRF calibration range (O to S.0%).

This method is also applied for very accurate deter mination, e.g. where a standard reference material has to be analyzed for certification purposes.

Safety advisory;

l. When using HF wear glasses and gloves, and be extremely careful. More information on HF is available in the Geoscience Laboratories' Safety

Manual, page IV-17.

1. Sample decomposition by acid digestion, and fol lowed by residue fusion (if necessary).

2. Measurement of absorbance by developing colour.

3. Calculation of TiO2 concentration based on a calibration curve produced from known standard solutions

Apparatus:

- Platinum dishes, 50 ml

- Vycor brand glass crucibles, 30 ml

- Volumetric flasks, 100 ml

- Glass funnels, 75 mm

- Filter paper, Whatman No. 42, (12.5 cm)

- Burette O to 25 ml

- Baush and Lomb spectronic 501 colorimeter

Reagents:

- Hydrochloric acid, HC1, 36.57c - 38 9fc

- Sulphuric acid, H2SO4,967c

- Phosphoric acid, H3PO4, 85Ve

- Hydrofluoric acid, HF, 489fc

- Potassium pyrosulphate, K2S2O7

- Hydrogen peroxide (309fc), reagent grade

- Potassium titanium oxalate, K2TiO(C204).2H20

- Diammonium sulphate, (NH4)2SO4

Procedures:

1. Reagent preparation

1.1.

1.2.

Hydrogen peroxide (39fc) solution - Dilute 10 ml of 30^ reagent grade hydrogen peroxide to

100 ml with distilled water. Make fresh before use.

(l: 1) H2SO4 solution - Add, very carefully, 100 ml of concentrated H2SO4 using a graduated cylinder to a 200 ml volumetric flask already containing 75 ml of distilled water and mix.

1.2.1.

Immediately immerse the flask in a 800 ml beaker of cold tap water in order to dissipate the heat generated. When cool, make up to 200 ml with distilled water and mix well. Store in a glass bottle.

Method:

The method consists of the following techniques:

EA9-1

Titanium

2. Standard TiO2 solution preparation (0.5 mg

TKVml)

2. l. Weigh exactly 2.2163 g of potassium titanium oxalate K2TiO(C2O4).2H2O and transfer quan titatively to a 500 ml Erlenmeyer flask.

2.2. Add 8 g of diammonium sulphate (NH4)2SO4.

2.3. Add 100 ml of concentrated H2S04.

2.4. Place a short-stemmed glass funnel in the neck of the flask.

2.5. Gradually heat to boiling, and boil for 5-10 minutes very carefully.

2.6. Continue the heating until complete solution has been effected and no unattacked material remains on the walls of the flask.

2.7. Cool and transfer to a 1000 ml volumetric flask already containing 500 ml of distilled water.

2.8. Wash the Erlenmeyer flask with several por tions of 59c H2SO4 and finally with distilled water, and transfer all washings into the

1000 ml volumetric flask.

minutes). Fluoride ions interfere with the titanium determinations by bleaching the yel low colour. Therefore, it is important to expel fluoride by fuming with H2SO4.

3.9. Cool, add 25 ml of distilled water, cover and heat on hot plate until sample is dissolved (30 minutes approximately). If the sample is high in Ca, B a, and Mg, insoluble sulphates of these elements will be formed and will not be dis solved.

3.10. Filter using Whatman No. 42 (12.5 cm) filter paper in a 100 ml volumetric flask.

3.11. Wash the platinum dish several times with small portions of 59fc H2SO4 (warm) and trans fer all the washings to the filter paper.

3.12. Wash the filter paper several times with small portions of 57c H2SO4 wash solution.

3.13. Ensure volume in the 100 ml flask after wash ings does not exceed 75 ml.

3.14. This is your original solution to develop and measure colour.

2.9. When cool, make to 1000 ml with distilled water.

3. Sample Decomposition

3.1. Weigh exactly 0.500 to l .000 g (depending on the amount of TiO2 present) of sample and transfer to a 50 ml platinum dish.

3.2. Moisten with about l ml of distilled water.

3.3.

3.4.

Cover the dish and add 10 ml of concentrated

HC1.

Heat on a hot plate for about 30 minutes.

4. Measurement of TiO2 concentration

4.1. Add 5 ml of phosphoric acid (H3PO4) to the original solution in 100 ml volumetric flask and

4.2. shake well (the colour due to ferrie sulphate is bleached by the addition of phosphoric acid).

Add 10 ml of 39c solution (made fresh) of hydrogen peroxide (H202), mix and make up to

100 ml with distilled water.

4.3. Measure the absorbance due to the yellow- coloured complex formed by titanium with hydrogen peroxide in acid solution at 435 nm in a cell against a blank reagent containing

10mlof(l:l)H2SO4.

3.5. When effervescence has ceased, remove and rinse the cover with distilled water, adding washings to the dish.

3.6. Add 10ml ofHF(487c)and 10ml of 1:1 H2SO4.

3.7. Digest and evaporate to dense fumes of SO3 for

5 minutes.

3.8. Cool, wash the sides and evaporate again to dense fumes of S03 for few minutes (2 to 3

5. Calculation of the results

5.1 To a series of 100 ml volumetric flasks add 10 ml of (1:1) H2SO4 and dilute with distilled water to about 50 ml.

5.2. Add from a burette (0-25 ml), a series of ali- quots of the standard TiO2 solution (0.5 mg

TiCyml) to give 0.50, l .00,2.00,4.00,8.00 and

10.00 mg TiCVlOO ml and mix.

EA9-2

Titanium

5.3.

5.4.

5.5.

5.6.

5.7.

5.8.

5.9.

Add 5 ml of phosphoric acid (H3PO4) using a fast draining pipette and mix.

Add 10 ml of 37o H2O2 solution (made fresh) using a pipette mix and make it to 100 ml with distilled water and shake well.

Measure the absorbance of the yellow- coloured complex formed by the titanium with hydrogen peroxide in acid solution at 435 nm in a cell against a blank having all the reagents andlOmlof(l:l)H2SO4.

Make a calibration curve.

The curve should be linear.

The yellow-coloured complex is very stable and reproducible.

Use the relative absorbance of the sample at

435 nm to obtain the concentration of Ti02 per

100 ml of solution from the calibration curve.

mg ml

l 100

1000

sample weight

Quality Control:

The determination limit for this method is Q.02% TiO2.

There is not enough data available to establish precision and accuracy for this method. An estimate of precision at 959fc confidence limit (2o) for a mid range value is ± Q.1% (absolute), with a mid-range value of Q.5%. The accuracy would be similar to this value for SRMs.

Productivity:

A technician should be able to complete 15 determina tions per day.

Additional Notes:

l. Samples high in sulphur should be roasted first in a Vycor brand glass crucible (30 ml) in a furnace at 6500C for 3 hours or preferably overnight, then transfer the powder from the crucible to a platinum dish. Rinse the crucible with several portions of distilled water and finally with concentrated HC1 into the dish.

2. Titanium minerals are considerably more resistant to decomposition than the silicate matrix in which they occur. Care must be taken to ensure that the mineral grains are completely attacked.

3. Any residue (which could contain such minerals as rutile and ilmenite) should be fused with small amount of potassium pyrosulphate (K2S2O7) in

Vycor brand glass crucible (30 ml).

4. The residue, and the filter paper from filtration of the digested sample solution, should be ignited in a Vycor brand glass crucible using a muffle fur nace. Start with a cold furnace and bum off the paper allowing the temperature to rise slowly.

Ignite the residue for about 5 minutes at 6000C.

Then fuse with potassium pyrosulphate (100 to

300 mg depending upon the amount of residue present). After leaching the fused matter in 59fc

H2SO4 tranfer to a 100 ml volumetric flask and develop the colour as per procedure.

5. A number of other metals form coloured com plexes with hydrogen peroxide including vanadium, uranium, niobium, molybdenum, and, under certain circumstances, chromium.

6. Presence of alkali salts causes slight bleaching effect - the bleaching action of alkali salts is less in a IQfy (by volume) solution of H2SO4 than in one containing 5^c.

7. In titanium solutions a moderate amount (10-20*70) of acid must be present to prevent precipitation of titanium by hydrolysis.

Bibliography:

Hillebrand, W.F., and Lundell, G.E.F., Applied Inor ganic Analysis, Second edition, 1929,1034 pages.

Jeffery, P.G., Chemical Methods of Rock Analysis,

3rd Edition, 1981, 379 pages.

Maxwell, J.A., Rock and Mineral Analysis, Wiley

Interscience Publishers, 1968,584 pages.

Potts, P.J., A Handbook of Silicate Rock Analysis,

Blackie and Sons Ltd., Glasglow, 1987, 622 pages.

EA9-3

Phosphorus

DETERMINATION OF PHOSPHORUS (P2Os)

COLORIMETRIC METHOD

Introduction: Method:

The determination of phosphorus, usually present in the range Q.1% - G.5% P2O5i is a necessary part of any silicate rock analysis.

In silicate rocks and minerals the phosphorus is usually present as orthophosphate (apatite is the most common mineral). Phosphorus is also an essential constituent of living matter. The average natural crustal abun dance of phosphorus in igneous rock is approximately

1200 ppm. Limestone and sandstone contain rather less phosphorus than igneous rocks. A few hundred parts per million being typical. Deep-sea sediments contain up to a few thousand parts per million P2O5.

In this method, the concentration of phosphorus is determined by measuring the absorbance of the yel low-coloured molybdovanadophosphoric acid com plex, produced by adding ammonium vanadate and ammonium molybdate to the sample. The measure ment, by a photometer, is taken at wavelength 460 nm.

This method is used only where the sample matrix makes determination by XRF unsuitable e.g. when the sulphur content of the rock is too high to allow a fused bead to be prepared for XRF determination.

This method is also applied when the P2O5 content is

> l .09fc, and beyond the calibration range (O -1 .09c) of the current XRF method.

This method is used for accurate determinations such as where a standard reference material has to be analyzed for certification purposes.

A preliminary separation of phosphorus is sometimes made to eliminate the interference of elements such as copper, nickel and chromium, which form coloured solutions. Presence of titantium phosphate can cause dissolution problems.

The method consists of the following techniques:

1. Sample decomposition by acid digestion with residue fusion (if necessary).

2. Measurment of absorbance by developing colour.

3. Calculation of P2O5 concentration based on a calibration curve produced by known standard solutions.

Apparatus:

- Platinum or teflon dishes, 50 to 100 ml (teflon dishes are preferred due to lower cost, and for higher productivity)

- Glass volumetric flasks, 100 ml

- Glass funnels, 75 mm

- Filter paper, Whatman No. 40 and No. 42

(12.5cm)

- Baush and Lomb spectronic 501 colorimeter

- Micro-burette, O -10 ml

Reagents:

- Ammonium molydate, (NH4)6Mo7O24.4H2O

- Ammonium metavanadate, NH4VO3

- Boric acid crystals, H3BO3

- Nitric acid, HNO3, 69.0-71.07c

- Hydrofluoric acid, HF, 48?c

- Potassium phosphate monobasic, KH2PO4

Procedures:

1. Reagent preparation

1.1.1. Ammonium molybate (57c) stock solution -

Dissolve 50 g of (NH4)6Mo7O24.4H2O in ap proximately 500 ml of warm distilled water, and let stand for several hours.

Safety advisory:

l. When using HF wear glasses and gloves, and be extremely careful. More information on HF is available in the Geoscience Laboratories' Safety

Manual, page IV-17.

1.1.2. Filter through a 12.5 cm Whatman No. 42 filter paper.

1.1.3. Dilute to l liter with distilled water and store in a polyethylene bottle.

EA10-1

Phosphorus

1.2.1. Ammonium Metavanadate (Q.25%) stock solution - Dissolve 2.5 g of NH4VO3 in 500 ml of hot distilled water.

3.2. Moisten with distilled water.

3.3. Cover the dish and add 10 ml of concentrated

HN03.

l .2.2. Cool and add 20 ml concentrated HNO3.

1.2.3. Let stand for several hours, and filter if not clear.

3.4. When effervescence has ceased, remove and rinse off the cover.

l .2.4. Dilute to l liter with distilled water. l .2.5. Store in a glass bottle.

1.3.1. Boric Acid (H3BO3) 4.57c stock solution -

Dissolve 45 g in distilled water by heating slowly.

1.3.2. When dissolved, cool and make to l liter with distilled water.

3.5. AddlOmlof487cHF.

3.6.

3.7.

Place the dish on a hot plate and slowly evaporate the contents to dryness.

Cool, moisten with distilled water, add 5 ml of concentrated HNO3 and 5 ml of HF; and again evaporate to dryness.

1.3.3. Store in a polyethylene bottle.

2. Standard P2OS solution preparation (1.00 mg

2.1.

2.2.

Weigh exactly 0.959 g of potassium phos phate monobasic (KH2PO4) dried at 1000C.

Transfer to a 100 ml glass beaker.

2.3. Dissolve in distilled water. Potassium phospate monobasic should be soluble without heating.

3.8. Cool, add 20 ml of HNO3 (1:1) and again evaporate to dryness.

3.9. Heat the contents of the dish for a further 30 min. after the salts appear to be dry.

3.10. Cool, add 20 ml (1:1) HNO3 (previously boiled and cooled to remove oxides of nitrogen).

3.11. Add 10 ml of (4.5^c) boric acid solution.

3.12. Cover and digest the contents on a hot plate, until dissolution appears to be complete.

3.13. Filter through a 12.5 cm Whatman No. 40 paper into a 100 ml volumetric flask.

2.4. Transfer to a 500 ml volumetric flask quantita

2.5. tively, washing beaker with several small por tions of distilled water.

Make to volume with distilled water.

2.6. Store in a tightly capped polyethylene bottle.

(It is always a good practice to store standard solutions in a tightly capped bottle to minimize evaporation of the solution)

3.14. Police the dish and wash with distilled water having a few drops of colourless (1:1) HNO3 .

Transfer all washings to the filter paper.

3.15. Wash the filter paper several times with dis tilled water having few drops of colourless

(1:1)HNO3.

3.16. After filtration and washings, make to volume in a 100 ml volumetric flask with distilled water. This is your original solution.

3. Sample decomposition

3.1. Weigh exactly l .000 g of sample and transfer to a 50 ml teflon dish.

NOTE: Since the presence of P2O5 in rock samples is

usually in the range of 0.1 - Q.5%, a 1.000 g sample can be used. If P2O5 is in a greater concentration, less sample may have to be used.

4. Measurement of P2O5 concentration

4.1. Pipet a 50 ml aliquot into a 100 ml volumetric flask from the original solution. The splitting of the original solution (100 ml) may be done with a dry pipette or conveniently as follows:

4.1.1. Rinse a clean pipette with the original solution and transfer the rinsings to a 150 ml beaker.

EA10-2

Phosphorus

4. l .2. Pipet a 50 ml aliquot into a 100 ml volumetric flask, then rinse the pipette with water into the beaker.

4.1.3. Transfer the solution in the beaker to the 100 ml flask of the original solution.

4.1.4. Dilute to 100 ml and use for correcting the absorbance due to the presence of other coloured substances.

4.2. Add 10.0 ml of ammonium vanadate solution by pipette to the solution from step 4. l and mix.

4.3. Add 20.0 ml of ammonium molybdate solu tion by pipette and mix.

4.4. Mix, dilute to volume, and mix thoroughly again. Allow to stand for 30 minutes.

4.5. Measure the absorbance of the yellow colour complex (molybdovanadophosphoric acid) at

460 nm in a cell, against a blank consisting of the reagents and 10 ml of (l : 1) HN03 already boiled, cooled and colourless.

4.6. Use the other 50 ml portion of the original solution to correct the absorbance due to the presence of other coloured sustances as ex plained in step 4. 1.4.

5.8.

solution containing all the reagents with 10 ml colourless (l :1)HNO3.

Prepare a calibration curve using the absorp tion reading on each solution versus con centration.

5.9. The curve should be linear up to 2.5 mg

5.10. The yellow complex is stable for at least one day.

5.11.

Measure the absorbance of the samples at

460 nm and obtain the concentration of P205 per 100 ml of solution from the calibration curve.

Quality Control:

The determination limit of this method is G.02% P205 using a l .000 g sample.

The mid-range value is Q.5%, and an estimate of precision, at 95 ^c confidence limits (2a), at the mid range value is 0. 1 9fc (absolute).

Accuracy is similar to precision for this method.

5. Calculation of the results

5.1.

To a series of 100 ml volumetric flasks, add from a micro-burette, a series of aliquots of the standard phosphate solution (1.00 mg

P2(yml) to give 0.50, 1.00, 2.00 and 2.50 mg

Productivity:

A technician should be able to complete about 10 determinations per day.

5.2.

5.3.

5.4.

5.5.

5.6.

5.7.

Add 10 ml of (l : 1) HNO3, already boiled and colourless.

Dilute to about 50 ml with distilled water.

Add by pipette 10 ml of ammonium vanadate solution and mix.

Then add by pipette 20 ml of ammonium molybdate solution and mix.

Dilute to 100 ml and mix thoroughly. Allow to stand for 30 minutes.

Measure the absorbance of the yellow com plex at 460 nm for each solution against the

Additional notes:

l . Samples high in sulphur should be roasted first in a Vycor brand 30 ml glass crucible in a furnace at

6500C for 3 hours or preferably overnight. Then transfer the powder from the crucible to the platinum or teflon dish. Rinse the crucible with several portions of water and finally with HNO3 into the dish.

2. If the residue on the filter paper after filtration of the digested sample solution shows white - no gritty residue - the residue may be titanium phos phate.

If the residue amounts to more than a few grains, it should be ignited in a platinum crucible and fused with 0.5 g of anhydrous Na2C03. The cake is leached with water and HNO3, filtered and

EA 10-3

Phosphorus

added to the volumetric flask having the original Bibliography: solution.

3. Iron interference is negligible at 460 nm.

Jeffery, P.O., Chemical Methods of Rock Analysis,

First Edition, 1970,507 pages.

4. Sample blank absorbance correction is necessary Maxwell, J.A., Rock and Mineral Analysis, Wiley in the presence of other coloured substances. Interscience Publishers, 1968, 584 pages.

(Solutions should be white after decomposition).

Potts, P.J., A Handbook of Silicate Rock Analysis,

Blackie and Sons Ltd., Glasglow, 1987, 622 pages.

EA 10-4

Sodium! Potassium

DETERMINATION OF SODIUM AND POTASSIUM

FLAME PHOTOMETRIC METHOD

Introduction:

All silicate rocks and minerals contain both sodium and potassium in amounts varying from less than 100 ppm in some ultrabasic rocks such as dunite and peridotite, to as much as 10 percent K20 or 15 percent

Na2O in feldspar minerals.

Rocks containing large amounts of potassium or sodium are rare, and most silicate rock specimens contain both alkalies in somewhat similar amounts in the range of 1-6 percent Na2O and 0.5-6.0 percent K2O.

Both elements occur as major constituents of many rock-forming minerals, particularly the alkali feldspar group, and are always determined where a complete chemical analysis of a silicate rock or mineral is re quired.

This method is used only where the sample matrix makes the constituents concerned unsuitable to be determined by XRF. This will arise when the sulphur content of the rock is too high to allow a fused bead to be prepared by XRF.

This method is also applied for accurate determination where a standard reference material has to be analysed for certification purposes. The method is applicable to most rock samples where Li is present in trace amounts

(O - 100 ppm).

A compound heated in a flame dissociates and the excited ions emit light of a characteristic wavelength.

The intensity of a given wavelength is proportional to the concentration of the source ion. In the flame photometer, the light produced falls on narrow-band filters which reject all but the pertinent wavelengths.

These are transmitted to individual phototubes whose electrical outputs vary with light intensity, and hence provide a measure of ion concentration.

Absolute measurements of light intensity by this method are affected by variations in flame condition and aspiration rate. Sodium and potassium measure ments therefore require the introduction of another ion in a known concentration to provide a reference level.

Lithium is chosen because it emits light of a suitable wavelength and is not present in significant concentra tions in most rock samples.

The three electrical outputs are combined to produce two signals, one proportional to the ratio of concentra tion of Na/Li and the other K/Li. These signals are presented in digital form and the resulting readout indicates absolute concentrations of sodium and potas sium expressed in millequivalent per liter.

Safety advisory:

When using HF wear glasses and gloves, and be ex tremely careful. More information on HF is available in the Geoscience Laboratories' Safety Manual page

IV-17.

When using perchloric acid use fume hood designed for perchloric acid fumes.

Method:

The method consists of the following techniques:

1. Sample decomposition by acid digestion

2. Simultaneous measurement of Na and K using

Flame Photometer type FLM2.

3. Calculation of Na and K concentrations based on a calibration curve produced from known standard solutions

Apparatus:

- Platinum dishes, 50 -100 ml

- Glass funnels, 75 mm

- Filter paper, Whatman No. 40 (12.5 cm)

- Glass volumetric flasks, 250 ml

- Vycor brand glass crucibles, 30 ml

- Polypropylene beakers, 5 ml

- Flame photometer type FLM2, by Bach-Simpson

Ltd. London, Ontario

Reagents:

- Lithium carbonate, Li2CO3 reagent grade

- Sodium chloride, NaCl, reagent grade

- Potassium chloride, KC1, reagent grade

- Concentrated hydrochloric acid, HO, 36.5 - 389fc

- Concentrated perchloric acid, HGO4,

- Concentrated sulphuric acid, H2SO4i

- Concentrated hydrofluoric acid, HF,

EA11-1

Sodium/Potassium

- Sterox SE, non-ionic wetting agent (for use in

Flame Photometer)

Procedures:

1. Reagent preparation

1.1. 19fc Sterox SE (non-ionic wetting solution) -

Add 10 ml of Sterox SE solution to a l liter flask and make to volume with distilled water.

1.2. Li Stock Solution (1500 meq/1) - Weigh ac curately 55.4175 g of Li2CO3 and transfer to a clean l liter glass beaker. (There is no need to dry Li2CO3 since the standards and blanks are prepared in the same fashion)

1.2.1. Cover with a clean watch glass and add about

100 ml of distilled water.

1.2.2. Add, caustiously, small portions of 200 ml cone. HC1.

1.2.3. Once the effervescence has stopped and dis solution is complete, cool and transfer quan titatively to a l liter volumetric flask by washing the beaker with several small por tions of distilled water.

1.2.4. Add 10ml of 1 96 Sterox SE solution, mix and make to volume (l liter) with distilled water.

Store in a polyethylene bottle.

2.2. Weigh exactly 1.9069 g KC1 (dried at 1050C) and transfer quantitatively to the 250 ml beaker containing the NaCl.

2.3. Dissolve in about 150 ml distilled water.

2.4.

2.5.

Transfer to a l liter volumetric flask quantita tively by washing the 250 ml beaker several times with small portions of distilled water.

Make to volume (l liter) with distilled water.

Mix well and store in a polyethylene bottle.

3. Sample decomposition

3.1. Weigh exactly 0. 125 g of sample and transfer to a 50 ml platinum dish.

3.2. Moisten with about l ml of distilled water.

3.3. Coverthe dish with a watch glass (if carbonate is present) and add 5 ml of cone. HC1O4.

3.4. When effervescence has ceased, remove and rinse off the cover in the dish with distilled water.

3.5. Evaporate to complete dryness on a hot plate.

3.6.

Cool, add 10 ml of cone. HC1, 5 ml of HF

(489fc) using plastic graduated cylinder and 2 ml of l : l H2SO4 (using a pipette).

1.3.1. Preparation of (l : 1) H2SO4 solution - Add very carefully 100 ml of cone. H2SO4 using a graduated cylinder to a 200 ml volumetric flask already having approximately 75 ml of distilled water and mix.

3.7.

3.8.

1.3.2. Immediately immerse the flask in a 600 ml glass beaker or tray containing cold tap water 3.9. as the solution will be very hot due to the reaction.

3.10.

1.3.3. When cool, make to volume with distilled water and store in a glass bottle.

2. Standard Na and K solution preparation (1000

2.1.

ppm Na and K)

Weigh exactly 2.5423 g of Nad (dried at

1050C) and transfer quantitatively to a clean

250 ml beaker

Allow to go into solution at a moderate heat and take down to a low volume (approximate ly 10ml).

Add another 5 ml of cone. HC1 and 5 ml of HF

(489?;).

Evaporate to complete dryness (until no more fumes of SO3 are visible).

In order to remove the last traces of SO3, hold platinum dish using platinum tipped tongs and heat very carefully with a Meker burner until the reappearance of SO3 fumes and finally heat strongly with full flame until all fuming ceases. This helps in removing the last traces of HF and converting sulphates of many ele ments into oxides.

3.11. Cool the dish and add about 5 ml of cone. HC1.

EA11-2

Sodium/Potassium

3.12. Evaporate to dryness on a low heat. This helps in converting oxides to chlorides which are easily soluble in distilled water.

3.13. Cool and add about 5 ml of cone. HC1. Leave it for about 1-2 minutes to react.

3.14. Then add about 30 ml of distilled water and heat well until sample is in solution. It takes approximately 45 minutes to one hour depend ing upon the nature of the sample.

3.15. Filter into a 250 ml volumetric flask using a

Whatman No. 40 filter paper.

3.16. Wash the platinum dish several times with small portions of hot distilled water (using glass wash bottle) and transfer all washings to the filter paper.

4.2.2. Mix well and store in a polyethylene bottle.

4.2.3. This is the lithium blank/diluent which is used to centre lithium standard meter and also to set

Na and K, zero on the instrument.

5. Measurement of Na and K concentrations

5.1. Starting up

5.1.1. Set the 'Display' switch to 'Continuous'.

5. l .2. Check the drain line is filled with water. If in doubt, inject a few milliliters of distilled water into the drain tube outlet and allow the excess to drain off.

5.1.3. Open the valve at the top of the propane cylinder one full turn only.

3.17. Wash filter paper several times with small portions of hot distilled water (using glass wash bottle).

3.18. Finally, add 2.5 ml of Li stock solution (1500 meq/1) and make to 250 ml with distilled water in a volumetric flask. Now the solution is ready for measurement.

4. Addition of reference Li stock solution

4.1.1. Preparation of Na and K standard solution for calibration - Add O ml, 1.25 ml, 2.5 ml, 5.0 ml and 7.5 ml and 10.0 ml from microburette (O

-10 ml) of 1000 ppm Na and K stock solution into 500 ml volumetric flasks.

4.1.2. Add 5 ml (using a O -10 ml microburette) of

Li stock solution (1500 meq/1) to each of the

500 volumetric flasks.

4.1.3. Make to volume (500 ml) with distilled water.

Mix well, transfer to the polyethylene bottles and store.

4.1.4. This will give a concentration of 0,2.5, 5, 10,

15 and 20 ppm Na and K in the solution.

Calibration solutions are stable.

4.2. Preparation of Li blank/diluent

4.2.1. Add 5 ml (using a O - 10 ml microburette) of

Li stock solution (1500 meqA) into a 500 ml volumetric flask and make to volume with distilled water.

5. l .4. If a central air supply is being used, open the air valve.

5.1.5. If the air filter is fitted with a manual drain ensure this is closed.

5.1.6. Remove distilled water cup from the sample tray.

5.1.7. Switch on power.

5.1.8. Check gas, air and flame indicator lamps are lit in sequence. If any of the above fail to occur, gas and air supplies will be automat ically shut down and the ignition spark will stop after 10 seconds. Turn off the power, wait a few seconds, and turn on again. If the flame photometer again fails to ignite, refer to the installation and maintenance manual.

5.1.9. Place a plastic cup (30 ml) filled with lithium blank/diluent solution on the sample tray.

Raise the tray to the limit of its travel. Con tinue to aspirate lithium for 10 minutes or until the inside of the atomizer chamber is thoroughly wetted. Refill the cup as neces sary.

5.2. Zero setting and calibration

5.2.1. Set the "K" range to "200". Leave on "Dis play" switch set to "continuous".

5.2.2. Aspirate Li blank solution.

EA11-3

Sodium/Potassium

5.2.3. Unlock the "Set Li Standard" control at lower right by turning the base of the control knob counter clockwise.

5.2.4. Set the "lithium standard" meter to the centre of the green band.

5.2.5. Relock the control by turning the base careful ly clockwise as far as it will go.

5.2.6. Unlock the Na and K zero controls and adjust them until both indicators read "000". A read ing of "999" indicates that the setting is below zero, and the appropriate control should be carefully turned clockwise until the reading is

"000". Relock both controls.

5.2.7. Remove the lithium blank solution cup.

5.2.8. Insert a fresh sample of the desired Na-K standard.

5.2.9. Verify that the "lithium standard" meter reads in the centre of the green band. Readjust if necessary using "Set Li standard" control.

5.2.10. Using the Na and K "Cal" controls, adjust the two readouts to values corresponding to the standard being aspirated (first start with 15 ppm adjusting to read 150 on read out). Then adjust the values for other standards.

NOTE:

1. Insert the standard every tenth determination to check calibration.

2. Steps 5.2.1 to 5.2.10 should be repeated every 30 minutes or less as the user determines on the basis of experience.

5.3. Run samples

5.3.1. Remove the Na/K standard.

Alternatively, the operator can set the mode switch to

"Delayed Hold". In this condition, raising the sample tray initiates a timed sequence in which the readouts indicate freely for 15 seconds, allowing time for both

Na and K readings to stabilize. After this delay the readings are "held". Loss of reading will not occur until the sample tray is lowered and raised again.

6. Calculation of the results

6.1. Calculation of results is based on 0.125 g sample weight which is made to 250 ml.

Na2O = (\ig Na/ml) x

250

sample wt. " l O6

c K2O = (\ig K/ml) x

250 100

sample wt. l O6

x 1.2046

7. Shutting down

7.1.

Aspirate distilled water for ten minutes. It is important that this be done consistently be cause it will normally make it unnecessary to clean the atomizer chamber. If however, the chamber does become clouded, wash it with soap and water only

7.2. Close gas tank and wait for flame lamp to extinguish.

7.3. If a manual drain is fitted to the air filter, open this and allow condensed moisture to blow off.

7.4. Turn power off.

7.5. If central air supply is being used, close the air valve.

7.6. Place cup of distilled water on the sample tray and leave aspirator needle immersed.

5.3.2. Place a sample cup of unknown on the sample tray.

5.3.3. Raise the tray to the limit stop. The "display" switch can be left set to "continuous", which allows a sample to be monitored as long as sufficient fluid remained in the sample cup.

Readings can be taken when the operator feels that they have reached stability, usually 10-12 seconds after insertion of the sample.

8. Maintenance

8.1. Clean aspirator tube weekly. Use filament provided.

8.2. Clean glass chimney monthly. Wash with detergent. Check light filters at the back of the chimney compartment. Wipe with lense paper.

8.3. Check burner for encrustation. Clean when required.

EA11-4

Sodium/Potassium

8.4. Check aspirator rate frequently, 1.5 - 1.7 ml l min.

8.5. Bleed moisture off compressor periodically.

Productivity:

A technician should be able to complete 10 samples per day.

Quality control:

The determination limits are G.15% for Na2O and

G.12% for K2O using a 0.125 g sample and making to

250 ml.

Only limited quality control data are available for the estimation of accuracy and precision.

The Geoscience Laboratories has participated in a number of certification programs and SRMs - SY-2,

SY-3, MRG-1, AL-1, AN-G and BEN have been analysed using this method. The results (Table

NAK1) are consistent with an estimated precision, at the 95*7c confidence limit (2o), at a mid-range value, of Q.4% (absolute).

Additional Notes:

1. Samples high in sulphur should be roasted first in

Vycor brand glass crucibles, 30 ml, in a furnace at

6500C for 3 hours or preferably overnight. Then transfer the powder quantitatively from the crucible to the platinum dish. Rinse crucible with several portions of distilled water and finally with

HC1 into the dish.

2. This method is applicable to most of the rock samples where Li is present in trace amounts (O -

100 ppm). Rocks containing large amounts of Li are rare.

Table NAK1. NazO and KzO Values for SRMs

Obtained By Using Li as an

SRM

SY-2

SY-3

MRG-1

AN-G

BEN

AL-1

NBS-91

NBS-70A

Internal Standard

Na2O

GLOGS

4.34

4.20

0.77

1.73

3.36

10.66

8.47

2.48

Lit.

4.31

4.14

0.73

1.63

3.18

10.59

8.48

2.55

K2O

GLOGS

4.64

4.24

0.17

0.19

1.44

0.14

3.28

11.7

Ve

Lit.

4.45

4.20

0.18

0.13

1.39

0.14

3.25

11.8

Bibliography:

Jeffery, P.G., Chemical Methods of Rock Analysis,

First Edition, 1970

Maxwell, J.A., Rock and Mineral Analysis, Wiley

Interscience Publishers, 1968, pp. 405.

Instruction Manual for the Flame Photometer Model

FLM2 (December 1970) Bach-Simpson Ltd. 1255

Brydges St. London, Ontario

EA 11-5

Water

DETERMINATION OF WATER

COMBUSTION/INFRARED ABSORPTION METHOD

Introduction:

Waterexists in two main forms in rock samples: hydro scopic and structural. Hydroscopic (sorbed) water can be driven off by heating the powdered sample at 1050C to constant weight (2-3 hours).

Hydroscopic water is generally reported as H2O" and structural (also called crystalline) water is reported as

H2Cf.

A Leco RMC-100 water determinator is used to rapidly determine both hydroscopic and structural (crystal line) water. The sample is weighed and loaded into the instrument at 1050C. Free moisture (H2O~) is driven from the sample in a stream of nitrogen which is monitored in an infra red cell. Water vapour absorbs infra red radiation. The instrument collects the signal and converts it to appropriate units. The displayed signal is in units of weight percent.

A source of determinate error associated with the determination of ± water is that hydrated minerals like talc, topaz, staurolite, cordierite, and epidote do not decompose at temperatures below 12000C.

Safety advisory:

Protective gloves should be worn, or tongs used, to protect hands from hot sample vessels and when using magnesium perchlorate.

Method:

The method consists of the following techniques:

1. Separation of the water by heating

2. Detection of the water by monitoring the absorp tion of infra-red radiation by the H20 molecules

3. Calculation of final result (9fc H20) based on a calibration using standards of known composition.

Apparatus:

- Leco RMC-100 Moisture Determinator

- LB-80 Electronic Balance

- Cylinder of nitrogen with regulator capable of accurately delivering N2 at pressures of 30 to

50 psi

- Nickel boats

- Tongs

- Push rod

- Glass wool

- Hi vacuum grease

- Brush

- Spatula

Reagents:

- Anhydrone (magnesium perchlorate)

Procedures:

1. Instrumental Analysis

1.1. Turn 'gas on' switch to 'off.

l .2. Turn on nitrogen supply and set regulator to

40 psi.

1.3. Switch power to 'on'. The 'message centre' will display "POWERUP DIAGNOSTIC" while the instrument performs a diagnostic test.

1.4. After completing the test the message centre will display "ENTER TIME AND DATE

Hr/Min Mo/Dy/Yr". Using 24 hour time, enter the correct time and date. Use the num ber keys on the instrument and push "ENTER" after the information has been keyed in.

1.5. Switch on front panel gas and adjust the Purge flowmeter to 3 litres per minute.

1.6. Switch the Pump on and ensure that the

Analysis rotameter is set to l litre per minute.

l .7. Press System Update key; then press the YES key.

l .8. Allow 1/2 hour for warm-up.

1.9. Run 2 reference samples. Typical reference materials are MRB-7 and MRB-10 which con tain 2.31^c and 11.77c, respectively, H-H2O. If the results from these analyses are not within

±57c (i.e. 2.207e - 2.437c) of the expected values, re-standardize the instrument and rerun the reference materials.

EA12-1

Water

1.10. Run a blank.

1.11. Run the samples. The run should consist of:

Blank, Reference-1, Reference-2, l O samples,

Reference-1, Reference-2,.... Random dupli cates should be analyzed every 20 samples or at least once for sample batches of less than 20 samples.

2. Maintenance

2.1. Anhydrone should be replaced when signs of moisture are evident in glass cylinders.

2.2. After 8 hours use, the furnace and quartz wool filter should be cleaned by initiating the self- clean cycle. Cleaning takes about 1/2 hour, (cf

Leco Manual; Quick Access #4 p S10).

Quality Control:

Quality is controlled by running reference materials at the beginning of the run and after every 10 samples. If the results from these analyses are not within 59fc

(relative) of the expected values, re-standardize (Leco manual: Quick Access #3 p.28). Analysis of random duplicates is performed at least once during each run.

The method has been adopted directly from the Leco instrument manual. Additional information is avail able in the literature and in Potts.

The optimum working range is G.10% to 10*7c H2O.

The precision ('fcRSD) over this range is 2.57c - S.0%.

Based on 0.5 g sample weight the determination limit for hydrated and structural water are 0.05 9fc and Q.10% respectively.

The method should be accurate to within ±0.01 ^c or dLS.% of the observed value, whichever is greater. The sensitivity of the method is 0.0196.

Accuracy: Comparison with Established Method

Comparison of LECO RMC-100 with

Gravimetric Procedure (HiO*)

Sample Leco 96H2O* Gravimetric 96HiO:

132

158

182

199

200

01

22

214

104

117

118

119A

119B

121

124

125

215

218

219

104D

219D

1.70

1.80

2.07

3.99

2.80

3.09

2.79

2.86

3.13

4.40

4.79

4.16

0.90

2.52

1.55

3.21

2.74

2.71

3.21

3.32

3.07

3.06

1.76

2.00

3.76

2.82

3.06

2.61

2.71

2.90

4.14

4.61

4.00

0.77

2.60

1.54

3.13

1.58

2.56

2.91

3.16

3.03

Regression Output:

Constant

StdErrofYEst

R Squared

No. of Observations

Degrees of Freedom

X Coefficient(s)

StdErrofCoef.

0.039

0.103

0.988

21

19

0.952

0.024

Validity of Method

Data generated by this method (Leco) were compared with data produced by an established method

(Gravimetric). Regression analysis yielded a linear graph with slope close to unity and intercept close to zero. This indicates that the two methods are equivalent and supports the contention that significant sources of determinate error are not biasing the results.

These data are plotted on the accompanying graph.

Accuracy was further checked by analyzing standard reference materials.

EA 12-2

Water

Accuracy: Typical Analysis of SRM:

STD

NIM-L

NIM-L

MM-L

MRB-10

MRB-10

H2(X

0.19

0.26

0.27

0.46

0.45

2.24

2.24

2.28

11.74

11.85

Lit. Value (H2O+)

2.31

2.31

2.31

11.7

11.7

* Note: Accuracy statements do not apply to H2O~ as samples absorb variable amounts of this water.

Results indicate good agreement between observed and expected H2O* values.

Factors such as frequency of standardization, stability of standards and personal determinate error will deter mine the accuracy of particular determinations. The above data support the contention that accuracy can be expected to be within ±59fc of the true value, during routine analysis and at levels significantly above the determination limit.

The most common is "MGT TIMEOUT", which signifies that the analysis has exceeded 800 seconds (this is usually due to the presence of high levels of carbonates). If this occurs, repeat the analysis using V205. This procedure is similar to the procedure in the absence of vanadium pen- toxide, except that V2O5 powder is placed on top of the rock powder. A vanadium pentoxide blank should be carried through the analysis and the value obtained should be subtracted from the sample results. NOTE: Do not reuse nickel crucibles if V2O5 was employed. The reuse of these crucibles reduces the life of the combustion tube.

4.

The RMC-100 can be calibrated using the stand ards supplied by Leco; Ca(OH)2 and CaC2O4, which contain 24.0^o and 12.29& respectively

4-H2O. In-house reference rocks; MRB-10, 7, 14 which contain H.7%, 2.3196, 2.777c, 2.317c respectively +H2O, can also be used to standardize the instrument.

5.

Anhydrone should be sieved through 8 mesh screen.

Productivity:

A technician should be able to complete 30 samples per day.

Additional Notes:

1. While analyzing a sample, the instrument can store the weights of as many as three samples which are queued for analysis.

2. Leave analyzer power on at all times.

3. Alarms : refer to section P. S5 of the manual (l).

Bibliography:

Leco RME-100 Instruction Manual #601-300. Leco

Instruments 5151 Everest Drive Mississauga Ontario

L4W 2R2 (phone; 416 624-6933).

Abbey, S., 1980, Simultaneous Determination of

Water, CO2, and S in rocks by volatilization and non- dispersive Infra-red Absorptiometry, Can. Journal of

Spectroscopy V. 25, No. 5.

Potts, P., Handbook of Silicate Rock Analysis,

Blackie, Glasgow, 1987, pp 70 - 74.

Shapiro, L., Rapid Analysis of Silicate, Carbonate, and

Phosphate rocks; Rev. Ed., USGS Bulletin 1401,1975, pp 54-57.

EA12-3

a

3

V

-i p

(O

OD

1

w o"

C or bon/Sulphur

DETERMINATION OF CARBON AND SULPHUR

COMBUSTION/INFRARED ABSORPTION METHOD

Introduction:

In this method, the sample undergoes combustion in a stream of oxygen in a LECO induction furnace. The absorption of infrared (IR) radiation by the CO and

CO2 combustion products are measured by separate detectors with the outputs added electronically to yield a result for total carbon.

The IR absorption of the SO2 combustion product is measured by a third detector. Results are displayed directly on the instrument as 7cC and

Sample concentrations are calculated by the instru ment using the observed instrument response and the sample weight which is determined by a balance linked to the instrument.

This method determines the total carbon content of a sample (as CO2) and does not distinguish between carbonate-carbon, organic or graphitic carbon sour ces. Noncarbonate (graphite) carbon exists in negli gible quantities in most rock samples. For this reason

9cCO2 is generally considered equal to total carbon. It is convenient to consider this assumption valid, since

LECO determinations are substantially faster than other methods (oil displacement, gravimetric, volumetric, carbotrane, etc.). In cases where graphitic carbon is expected to be significant, both total (LECO) and carbonate -carbon (coulometric - Page EA28-1) can be determined and graphitic carbon calculated by difference. (Diamond is rarely encountered).

3. Calculation of results by reference to standards of known composition.

Apparatus:

- LECO CS-46 Carbon and Sulphur 748-600 Sys tem, including:

A. Determinator Model 770-200

B. Induction Furnace Model 768-100

C. EB-25 Electronic Balance

D. AWC automatic weight compensation control box

- cylinder of oxygen with 2-stage regulator

(2500/35 psi)

- crucibles, including re-usable crucible covers

- tray

- crucible tongs

- quartz combustion tube

- dust traps

- dust filter (10 micron mesh)

- ascarite anhydrone tube

- glass wool

- micro fibre filter tubes

Reagents:

- vanadium pentoxide

- iron chips

- tin pellets

- ascarite, NaOH on asbestos base

- anhydrone, anhydrous, Mg(GO4)2

Procedures:

Safety advisory:

l . Vanadium pentoxide and magnesium perchlorate can cause serious damage to the skin. Exercise caution and wear plastic gloves when handling these materials.

2. The induction furnace generates extreme heat. Al ways handle crucibles with tongs.

Method:

The method consists of the following techniques: l . Conversion and separation of the analytes by com bustion of the sample.

2. Detection of the analytes by an infrared detector.

1. Instrument Settings

1.1. Select the following settings for the operation of the CS-46 System:

Power

Gas

Pump

Loop flow

Automatic/manual

Data transmit/Inhibit

Function select

Weight compensator

High carbon blank

Low carbon blank

High carbon calibrate

On

On

On

7.5 LPM

Auto mode

Inhibit

Operate

1000

Do not adjust

Do not adjust

Do not adjust

EA13-1

Carbon/Sulphur

Low carbon calibrate

Sulphur blank

Sulphur calibrate

Identification code

Warm-up period

Range switch

Do not adjust

Do not adjust

Do not adjust

Do not adjust

3 hours minimum

Do not adjust

1.2.

Select the following settings for the operation of the CS-46 Induction Furnace:

Filament voltage

High voltage

Oxygen pressure

On (glow white)

On (glow green)

35 psi

1.3.

Select the following settings for the operation of the CS-46 Balance/AWC Control Box:

Power (balance) On

Auto/Manual switch (box) Auto the calibration dials adjusted to obtain the desired readings.

2.10. Reanalyze the reference material (at least three times) until the readings are consistent with values expected (see QC section).

2.11. Lock the calibration dials and rotate the func tion switch back to the OPERATE position; the instrument is now ready for sample analysis.

2.12. Samples are analyzed using the steps indicated above.

3. Calculation

Multiply the carbon reading by 3.66 to convert it to equivalent %CO2.

2. Calibration and Sample Analysis

Samples containing G-2.7% Carbon (Q-10% CO2) and

0-0.89fc Sulphur are analyzed by the following steps:

2.1. Turn the oxygen cylinder on and adjust the pressure to 35 psi.

2.2. Add l scoop of vanadium pentoxide ac celerator to a crucible.

2.3. Place the crucible on the balance pan and press the TARE switch on the AWC control box.

2.4. Add 0.3 g - 0.4 g of reference material to the crucible.

2.5. Press the ENTER switch on the AWC Control

Box; the red WEIGHT IN light will activate.

2.6. Add l scoop of iron accelerator and l tin pellet to the crucible.

2.7. Cover with a lid and place the loaded crucible on the pedestal of the induction furnace.

2.8. Set 'UP/DOWN' switch to 'UP' to begin the analysis cycle; while the sample is being heated, a second sample can be weighed.

2.9. When the CARBON and SULPHUR READ lights glow red, indicating complete analysis,

Record the 9fcC and 9fcS readings. If the values deviate from the standard (expected) value, the function switch of the determinator should be rotated to the CALIBRATE position and

4. Analysis of Samples with High C or S Con centrations

Samples containing ^.79fc C or r^.8% S are analyzed in a similar fashion; less weight (0.1 g or less) of sample is used, the AWC control is not employed and the weights are manually recorded. Results are ob tained by dividing the observed results by the sample weight.

5. Instrument Shutdown

5.1. At the end of the analysis remove any remain ing crucible from the furnace.

5.2. Turn the O2 gas supply off and allow any oxygen remaining in the lines to escape.

5.3. Turn the HIGH VOLTAGE and FILAMENT

VOLTAGE off.

5.4. Set the PUMP and GAS switches to the off position (down).

5.5. Leave the Determinator and Furnace in this mode for best instrument stability. If the sys tem is expected to be off for an extended period the POWER switch may be turned off.

6. Routine Maintenance

6.1. A system electronics check, as described in the

Leco manual (page 17), should be performed daily.

EA13-2

C arbonl Sulphur

6.2. A leak check, as described in the Leco manual

(page 16), should be performed weekly.

6.3. Linear Card Adjustment, (Leco Bulletin l, 2,

3,4), is carried out for C, CO2, or S if erratic readings are obtained during standardization or standardization checks. Typically this hap pens after two to four weeks of operation.

6.4. The Ascarite and Anhydrone in the oxygen inlet tube should be replaced weekly or when ever it appears moist. The anhydrone in the

Bubble Anhydrone tube should be replaced daily - or whenever it appears moist. The com bustion tube should be cleaned with a brush every 30 samples; the tube should be inspected from time to time (every 100 samples) for severe pitting and slag build-up. If the tube is deteriorating, it should be replaced.

6.5. The dust-trap on the furnace should be cleaned and the glass wool repacked and conditioned* every 20 samples, or whenever the Loop Pres sure reaches 6 psi. It was found that an addi tional dust-trap filled with Mg(ClO4)2 was needed, and the glass wool in this additional dust-trap should be replaced at the same time.

The 10-micron mesh dust filter should be cleaned daily using an ultrasonic cleaner.

About ten minutes is required for this opera tion.

* Glass Wool Conditioning: The glass wool should be conditioned by heating a crucible with 0.3 g to 0.4 g of a sample containing more than Q.3% sulphur. One scoop of iron chips and one tin pellet are also added to the crucible prior to heating. This operation is only necessary when the glass wool is changed.

Quality Control:

Quality control is carried out by analyzing one in- house reference material every ten samples; blanks are also run.

The optimum range for CO2 is 0.01 -107o and 0.01-

0.77cforS.

The instrumental detection limits of C and S are 0.001 and 0.0001 ^c respectively, under ideal conditions. The determination limit for both CO2 and S is Q.01%.

Precision is estimated at ±59fc at the mid-range value for both CO2 and S (relative).

Productivity:

A technician should be able to complete 30 samples per day.

Additional Notes:

1. The CS-46 system can be calibrated by standards other than reference rocks; synthetic standards can

Bibliography:

Instruction Manual for CS-46 Carbon and Sulphur

748-600 System, LECO Corporation, 1977.

Terashima, S., 1978, Anal. Chim. Acta, 101, pp 25.

Terashima, S., 1979, Geostandards Newsletter, 3, pp

195.

also be prepared to cover the appropriate ranges.

Synthetic standards must be calibrated against ref erence rocks prior to their use as standards.

EA 13-3

Total F e

DETERMINATION OF TOTAL FE

VOLUMETRIC (TITRIMETRIC) METHOD

Introduction:

Iron is the fourth most abundant element, comprising about 5 percent of the earths crust. Silicate minerals vary considerably in iron content, and the variation is reflected in the iron content of silicate rocks. Basic rocks may contain 30 to 40 percent iron (calculated as

Fe2O3), while many acidic rocks contain as little as l percent total iron. Ferric iron is frequently as sociated with aluminum, and ferrous iron with mag nesium.

The sulphide mineral pyrite, FeS, and associated minerals of composition FeSO4.7H2O in the oxidation zone surrounding pyrite, are common. Iron-contain ing carbonate minerals are siderite or chalybite FeCO3 and ankerite which is a mixed carbonate or iron, cal cium and magnesium.

In this method, the reduction of Fe3"1" to Fe2* is achieved by stannous chloride in a hot acid (HO) solution, the excess of which is removed by the addition of mercuric chloride prior to titration. The sample solution is then titrated with standard potassium permanganate solu tion. The concentration of Fe is calculated on the basis of the net volume after blank subtraction.

This method is used only where the sample matrix makes determination by XRF unsuitable, e.g. when the sulphur content of the rock is too high to allow a fused bead to be prepared for XRF determination.

This method is also applied when the total iron content, expressed as Fe2O3, is M59&; above the calibration range of the current XRF method.

This method is also applied for accurate determination where a standard reference material has to be analyzed for certification purposes.

Platinum, vanadium, copper, arsenic, antimony and molybdenum are reduced to a lower valence state like iron, with the addition of stannous chloride and titrated with KMnO4 solution. If present in appreciable amounts, the Fe concentration will appear higher than the true concentration.

Safety advisory:

1. When using HF wear glasses and gloves, and be extremely careful. More information on HF is available in the Geoscience Laboratories' Safety

Manual page IV-17.

Method:

The method consists of the following techniques:

1. Sample decomposition by acid digestion and residue fusion (if necessary).

2. Reduction of Fe3* to Fe2* by stannous chloride

3. Titration of the sample solution with standardized potassium permanganate solution

4. Calculation of Fe concentration based on Fe present in the net volume after blank subtraction

Apparatus:

- Teflon beakers, 50 ml

- Glass beakers, 250 ml, 600 ml, 2000 ml

- Vycor brand glass crucible, 30 ml

- Filter paper, Whatman No. 40 (12.5 cm)

- Glass funnels, 75 mm

- Glass burette, 0-50 ml

- Glass wool

Reagents:

- Hydrochloric acid, HC1, 36.5 - 38 Ve

- Sulphuric acid, H2SO4,967c

- Phosphoric acid, H3PO4, 85 Ve

- Stannous chloride, SnQ2.2H2O

- Mercuric chloride, HgG2

- Potassium permanganate crystals, KMnO4

- Manganese sulphate, MnSO4.4H2O

- Potassium pyrosulphate, K2S2O7

- Hydrofluoric acid, HF, 489fc

Procedures:

1. Reagent preparation

1.1. Stannous chloride, SnCl2.2H2O 59fc in We

HC1 - Dissolve 10 g SnCl2.2H2O in 10 ml of concentrated HC1.

1.1.1. Heat to effect the solution for few minutes if necessary.

EA14-1

Total F e

1.1.2. Once in solution make to 100ml with di stilled water. The solution should be prepared fresh.

1.2. Mercuric chloride, HgCl2 59fc in H2O - Dis solve 5 g of HgQ2 in 100 ml of distilled water.

1.3. Zimmermann - Reinhardt reagent - Dissolve

140 g of MnSO4.4H2O in l liter of distilled water in a 2 liter glass beaker.

1.3.1. Cautiously add 250 ml of concentrated H2SO4.

l .3.2. Add 250 ml of 859fc phosphoric acid (H3PO4).

l .3.3. Dilute to 2 liter with distilled water.

l .4. Potassium permanganate, KMnO4 0. IN solu tion - Weigh 32.0 g of potassium perman ganate crystals and transfer to approximately

2 liters of distilled water in a 2 liter glass beaker.

l .4. l . Heat to boiling and keep hot for one hour. l .4.2. Cover and let stand overnight.

l .4. 3. Filter the solution through glass wool into a l O litre dark brown glass bottle.

l .4.4. Add 8 liters of distilled water.

1.4.5. Mix the solution thoroughly, and let stand overnight before standardization.

2. Standardization of KMnO4 - Standardize in triplicate against a National Institute of Stand

ards and Technology (NIST) Reference

Material e.g. NIST 29(a) (Fe = 69.54*7c), or

2.1. Sample decomposition

2.1.1. Weigh 0.200 g of 29(a) (for example) and transfer to a 50 ml teflon dish.

2.1.2. Add 10 ml of HC1 and l O ml of distilled water.

2.1.3. Heat for a few minutes, make sure most of the sample is in solution.

2.1.4. Add 5 ml of HF.

2.1.5. Evaporate to dryness.

2.1.6. Add 5 ml of concentrated HC1 and leave it for l or 2 minutes. Then add about 25 ml of distilled water.

2.1.7. Heat well and make sure that the sample is completely dissolved.

2.1.8. If any black gritty particles are noticed, the solution should be filtered, using a Whatman

No. 40 filter paper, into a 250 ml beaker.

Wash a few times with small portions of hot distilled water. The paper with undissolved particles should be ignited in a Vycor glass crucible (30 ml). Fuse the ignited residue with a small amount of potassium pyrosulphate

(K2S2O7). Dissolve the fused matter in the crucible by adding distilled water and a small amount of HCL Heat the crucible on a hot plate until the residue is in solution. Transfer the solution to the original solution in a 250 ml beaker by washing the Vycor crucible a few times with small portions of distilled water.

2.2. Reduction with stannous chloride

2.2.1. Heat to boiling on a hot plate, the complete solution of the sample having a volume of about 50 ml or less.

2.2.2. Add stannous chloride drop by drop until the yellow Fe3"" colour disappears. Add one or two drops in excess.

2.2.3. Cover the reduced solution in the beaker with a watch glass and cool by placing in a tray having cold water.

2.2.4. Add rapidly 10 ml of 5?c HgCl2 measured in a graduated cylinder.

2.2.5. A small quantity of white precipitate should appear.

2.2.6. If no precipitate, or if the precipitate is grey, the sample should be discarded.

2J.Titration with KMnO4 (0.1N) solution

2.3.1. After 2 or 3 minutes, transfer the reduced solution quantitatively to a 600 ml beaker, containing 25 ml of Zimmerman - Reinhardt reagent and 300 ml of distilled water.

2.3.2. Titrate immediately with KMnO4 to the first faint pink end point that persists for 15 seconds.

EA14-2

Total F e

2.3.3. Do not titrate rapidly at any time.

2.3.4. Two blanks should be carried through the procedure.

2.3.5. Correct the volume of KMnO4 for the blank titration.

2.3.6. The standardization should be carried out in triplicate and the average titer taken for the normality.

2.3.7. Calculate the number of milligrams Fe and

Fe2O3 equivalent to l ml of KMnO4.

3. Sample decomposition

3.1. Weigh exactly 0.200 g of sample and transfer to a 50 ml teflon beaker.

3.2. Moisten with distilled water.

3.3. Cover the beaker and add 10 ml of cone. HC1.

3.4. When effervescence has ceased, heat on a hot plate for a few minutes.

3.5. Remove and rinse off the cover into a beaker.

3.6. Add5mlofHF(489fc).

3.7. Evaporate to dryness.

3.8. Add 5 ml of concentrated HC1 and leave it for l or 2 minutes. Add 25 ml of distilled water.

3.9. Heat well and ensure that the sample is com pletely dissolved.

3.10. If any black gritty particles are noticed, the solution should be filtered, using Whatman

No. 40 filter paper, into a 250 ml beaker.

3.11. Wash a few times with small portions of hot distilled water. The paper with undissolved particles should be ignited in a Vycor glass crucible (30 ml).

3.12. Fuse the ignited residue a with small amount of potassium pyrosulphate (K2S2O7).

3.13. Dissolve the fused matter in the crucible in distilled water and add a small amount of HC1 by heating until it is dissolved.

3.14. Transfer the solution to the original solution

4.4.

4.5.

4.6. in the 250 ml beaker by washing the Vycor crucible a few times with small portions of distilled water.

4. Reduction with stannous chloride

4.1. Heat to boiling on a hot plate, the complete solution of the sample having a volume of about 50 ml or less.

4.2. Add stannous chloride drop by drop until the yellow Fe3* colour disappears. Add one or two drops in excess.

4.3. Cover the reduced solution in the beaker with a watch glass and cool by placing in a tray having cold water.

Add rapidly 10 ml of 59fc HgCl2 measured in a graduated cylinder.

A small quantity of white precipitate should appear, after 2 to 3 minutes.

If no precipitate or if the precipitate is grey, the sample should be discarded.

5. Titration using KMnO4 (0.1N) solution

5.1. After 2 or 3 minutes, transfer the reduced solution quantitatively to a 600 ml beaker, containing 25 ml of Zimmerman - Reinhardt reagent and 300 ml of distilled water.

5.2. Titrate immediately with KMnO4 to the first faint pink end point that persists for 15 seconds.

5.3. Do not titrate rapidly at any time.

5.4. Two blanks should be carried through the procedure.

5.5. Correct the volume of KMnO4 for the blank titration.

6. Calculation of the results

VcFe = (mgFe/mlofKMnOt)

ml for sample - ml for blank}

1000

J

100

sample wt (g)

EA 14-3

Total F e

Quality Control:

The determination limit is Q.2% as Fe using a 1.0 g sample.

An estimate of precision, at95*7c confidence limit (2o), at mid-range value (5^o) is Q.2% (absolute).

Accuracy is similar to precision for this method.

Productivity:

A technician should be able to complete 15 determina tions per day.

The excess Sn2* is removed with the addition of mercuric chloride

2HgCl2-

(White)

** 4- 2cr

The insoluble mercurous chloride produced will not consume KMnO4 nor will the mercuric chloride reoxidize Fe2*

Care must be taken not to have a large excess of

Sn2* in solution or an alternate reaction with mer curic chloride occurs

Hg 4- Sn4* 4- 2CT

(Black)

Additional Notes:

1. Samples high in sulphur should be roasted first in

Vycor brand glass crucibles, 30 ml, in a furnace at

6500C for 3 hours or preferably overnight. Then transfer the powder from the crucible to the teflon beaker. Rinse the crucible with several portions of distilled water and finally with HC1 in the teflon beaker.

2. Potassium permanganate is a volumetric oxidizing agent which is used for all ferrous and ferric iron determinations. Its proper preparation and stand ardization is essential for accurate analysis.

5Fe

2-t5 Fe3* 4- 5 e"

5 e" 4- Mn04 + 8 H*Mn2* + 4 H2O

3. The freshly prepared KMn04 solution is allowed to stand overnight to allow it to react with con taminants such as dust, organic compounds and other oxidizable substances.

4. The KMnO4 solution is filtered to separate MnO2.

The presence of MnO2 accelerates the decompos- tion of KMnO4 in solution (autocatalytic process).

2Mn(L~4-4H**-

2Mn02 4- 2H204- 02

Glass wool is used since filter paper reacts with

KMnO4 to yield MnO2.

5. Stannous chloride is added to reduce Fe3* to Fe2*.

A slight excess is added to ensure complete con version to ferrous iron.

Sn2*4-2Fe3* 2Fe2* 4- Sn4*

Metallic mercury reacts with permanganate to cause a high result. A proper reduction is indi cated by the appearance of a white precipitate. A grey precipitate indicates the presence of mercury and hence discarding of the sample. The total absence of the white precipitate indicates that in sufficient amounts of SnCl2 was added.

6. The Zimmerman-Reinhardt reagent is added for the following reasons:

A. The manganous ion inhibits the oxidation of the chloride ion by KMnO4. The chloride oxidation is normally a slow reaction, but Fe2* accelerates it.

lOCT 4- 2MnO4" 4- 16H**-*2C12 + 2Mn2* 4- 8H2O

B. The phosphoric acid complexes the Fe3* produced in the titration and prevents the intense yellow color of the ferrie chloride complexes interfering with the end point. Phosphate ligands provide good oxygen donors for the ferrie ion

7. If the addition of KMnO4 is too rapid during titra tion, some MnO2 will be produced in addition to the Mn2*. Evidence of this is a faint brown dis coloration of the solution. If this occurs, the titra tion should be stopped until the solution becomes clear. The solution must be free of MnO2 at the equivalent point.

8. Platinum, vanadium, copper, arsenic, antimony and molybdenum interfere in KMnO4 titration in stannous chloride reduction. Platinum is usually introduced through operations carried out in platinum vessels.

EA 14-4

Total F e

Bibliography:

Hillebrand, W.F., and Lundell, G.E.F., Applied Inor ganic Analysis, Second Edition, 1929.

Jeffrey, P.O. and Hutchinson, D., Chemical Methods of Rock Analysis, Third Edition, 1981, pp. 192.

Potts, P.J., A Handbook of Silicate Rock Analysis,

Blackie and Sons Ltd., Glasglow, 1987, 622 pages.

Maxwell, J.A., Rock and Mineral Analysis, Wiley

Interscience Publishers, 1968, pp 421.

Skoog, D.A., and West, D.M., Fundamentals of

Analytical Chemistry, 1963, pp. 435-443.

EA 14-5

Ferrous

DETERMINATION OF FERROUS IRON

TITRIMETIC METHOD

Introduction:

Iron is an important industrial commodity. Iron-bear ing sedimentary rocks furnish the bulk of industrial iron ore. The main ferruginous groups are: carbonates, silicates, oxides and sulphides.

In addition to its commercial importance, chemical and mineralogical data relating to iron are used in studies researching the origins of rock deposition and metamorphism. (More information about iron is given on page EA 14-1).

Since most instrumental analytical techniques do not readily provide information about the oxidation state of constituents it is necessary to employ more classical volumetric techniques. Mossbaur spectroscopy has been used to distinguish iron oxidation states in rocks and minerals but is not suitable for routine analysis.

The rock sample is decomposed by heating with a mixture of sulphuric and hydrofluoric acids in a covered platinum crucible. Most of the iron bearing minerals are decomposed by this procedure forming soluble ferrous and ferrie sulphates. The crucible lid and contents are placed in a mixture of boric and sulphuric acids. The solution is titrated against stand ardized potassium permanganate which oxidizes the ferrous iron to ferrie iron.

5e

0.771V

5 e 4- MnO4 * 8H+ —— * Mn* 4- 4H2O

1.51V

MnO4 4- 5 Fez* + 8H* ——- Mn"* 4- 5 Fe 4- 4 H2O

Boric acid is used to remove excess fluoride ion:

B(OH)3 4- 2H20 ———- B(OH)4" 4- H3O*

H3O* 4- B(OH)4" 4- 4HF *———^BF4 * 3H2O 4^ H3O*

The method is similar to that described by Goldich. A number of potential interferences are highlighted in the

Quality Assurance portion of this method (below).

There are a number of alternative methods in the literature. Wilson's method is based on the reduction of ammonium metavanadate (NH4VO3) by ferrous iron.

Unfortunately all methods suffer from various sources of determinate error. The method outlined in this docu ment has the advantage of being relatively rapid and simple.

Safety Advisory:

1. When using HF wear glasses and gloves, and be extremely careful. More information on HF is available in the Geoscience Laboratories' Safety

Manual page IV-17.

2. Sulphuric acid, hydrochloric acid, phosphoric acid and potassium permanganate are corrosive sub stances.

3. Mercuric chloride is toxic.

4. Boric acid can cause skin irritation and may induce an allergic reaction. Avoid contact with any of these materials. Exercise caution when handling these materials. Wear protective clothing and use eye protection.

Method:

The method consists of the following techniques:

1. Sample decomposition using wet chemical proce dures.

2. Volumetric analysis using a titrimetric procedure employing potassium permanganate.

3. Calculation of the result.

Apparatus:

- 30 ml platinum crucible with tight fitting lid

- 25 ml polyethylene graduated cylinder

- glass burets, 10 ml and 50 ml

- glass beakers, 600 ml, l liter, 2 liter

- florence flask, l liter

- tongs for platinum crucible

- teflon beakers, 100 ml and 250 ml

- glass wool

- l O liter container for standardized potassium per

manganate

EA15-1

Ferrous

Reagents:

1.3.7. Allow to stand overnight before stand ardizing.

- boric acid crystals

- concentrated sulphuric acid

- concentrated hydrofluoric acid

- potassium permanganate crystals

- NBS 29 A Iron Ore Standard, 69.547C Fe

- stannous chloride, SnQ2.2H2O

- hydrochloric acid, 6.8 M

- phosphoric acid, 15 M

- manganese sulphate, MnSO4.4H2O

- mercuric chloride, HgQ2

Procedures:

1.3. N/10 (0.1 N) Standard Potassium Perman ganate - Weigh 32.0 g of potassium perman ganate crystals.

1.3.1. Transfer approximately 2 liters of recently boiled and cooled distilled water to a 2 liter beaker.

1.3.2. Add the potassium permanganate crystals to the contents of the 2 liter beaker.

1.3.3. Heat to boiling and keep hot for one hour, l .3.4. Cover and let stand overnight.

l .4. Stannous Chloride - Dissolve 37.5 g of iron- free stannous chloride in 250 ml of 6 N hydro chloric acid. This solution should be freshly prepared.

l .5. Mercuric Chloride - Dissolve 50 g of mercuric chloride in l liter of distilled water.

1.6. Zimmerman Reinhardt Reagent - Dissolve

140 g of manganese sulphate in l liter of distilled water.

1.6.1. Cautiously add, with stirring, 250 ml of con centrated sulphuric acid.

l .6.2. Cautiously add, with stirring, 250 ml of phos phoric acid.

1. Reagent Preparation

1.1. Saturated boric acid - Prepare a saturated solu tion by adding approximately 900 ml of

(recently boiled and cooled to room tempera ture) distilled water to 50 g of boric acid crystals in a flask. It is convenient to employ three, one liter beakers for the purpose of providing sufficient water (see Additional

Notes, below).

l .2. Decomposition Solution - Prepare a solution of (recently boiled and cooled to room temperature) distilled water, sulphuric and hydrofluoric acids in a 1:1:1 volume ratio using the teflon beaker. Approximately one hundred ml, total volume, is generally suffi cient for this purpose.

1.6.3. Dilute to 2 liters with (boiled and cooled) distilled water.

2. Standardization of Potassium Permanganate

Solution

2.1.1. Weigh 0.200 g of NBS 29a Iron Ore and place in a 250 ml teflon beaker. A blank should also be carried through the procedure.

2.1.2. Add l O ml of 6 N hydrochloric acid and 10ml ofHF.

2.1.3. Place on hotplate and evapourate to dryness.

2.1.4. Add l O ml concentrated HC1.

2. l .5. Add 50 ml distilled water.

2.1.6. Place on hotplate and heat to dissolve. Heat the solution nearly to boiling.

2.1.7. Add stannous chloride solution drop by drop until the yellow Fe3* colour disappears. Add one drop in excess.

2.1.8. Cool to room tempe rature.

1.3.5. Filter the solution through glass wool into the

10 liter permanganate container.

1.3.6. Add 8 liters of (recently boiled and cooled) distilled water and mix the solution thorough ly.

2.1.9. Transfer 25 ml of Zimmerman Reinhardt reagent to a clean 600 ml beaker. Add 300 ml

(boiled and cooled) distilled water.

2.1.10. Load buret with unstandardized potassium permanganate solution.

EA15-2

Ferrous

2.1.11. Rapidly add 10 ml of the mercuric chloride solution. A small quantity of white precipitate should appear. If no precipitate is observed or if the precipitate is grey the sample should be discarded.

2.1.12. After 2-3 minutes, transfer the reduced solu tion quantitatively to the 600ml beaker con taining 25ml of Zimmerman Reinhardt reagent and 300ml of distilled water.

2.1.13. Titrate immediately with KMnO4 to the first faint pink end point that persists for 15 seconds. Do not titrate rapidly at any time.

3.1.3. Add 15 ml of the decomposition solution and immediately cover the crucible with a tight fitting lid.

3.1.4. Place the crucible on a hot plate and bring the contents to a boil.

3.1.5. Adjust the temperature of the hot plate so that violent boiling does not occur.

2.1.14. Repeat steps 2.1.2. through 2.1.13. with the blank solution.

2.1.15. Correct the volume of KMnO4 for the blank titration by subtracting the volume of perman ganate reagent required to titrate the blank from that required to titrate the NBS standard.

3.1.6. Transfer 200 ml (boiled and cooled) distilled water to a clean 600 ml beaker.

3.1.7. Add 50 ml of saturated boric acid solution.

3.1.8. Add 5 ml sulphuric acid.

3.1.9. After 10 minutes, grasp the crucible around its centre with the tongs, hold the lid in place with a stirring rod, and transfer the crucible and lid into the 600 ml beaker containing 200 ml boiled distilled water, 50 ml of saturated boric acid and 5 ml of sulphuric acid.

2.1.16. Repeat steps 2.1.1. through 2.1.13. two more times.

2.1.17. Calculate the average titre.

2.1.18. Calculate the number of grams of Fe equivalent to l ml of KMnO4:

ml KMnO,

= F e equivalent = F e Eq

^ (Weight NBS Std)x(Mol Fraction NBS)

6 q ~ mlKMnO4

e.g., Weight NBS: ml KMn04:

- j,

0.20 g

Mol Fraction NBS: 0.6954

22.50

0.20*0.6954 nrv,, 10l

4. Measurement of Ferrous Iron

4.1.1. Fill the burette with standardized KMnO4 solution.

4.1.2. Immediately following dissolution, use the glass buret and titrate the contents of the 600 ml beaker with 1/1 ON (0. l N) potassium per manganate. Stir thoroughly throughout the titration to a faint pink end point that persists for several seconds.

4.1.3. Calculate the 9fcFe from the titre:

%Fe = mlKMnO4 xFeEq g Sample

x 100 e.g., ml KMn04:

FeEq: g Sample:

5.21

0.006 1 8 lg

0.223

3. Decomposition of Samples

3.1.1. Weigh 0.500 g aliquot of homogeneous pow dered sample and transfer to a 30 ml platinum crucible.

3.1.2. Moisten the sample with a small portion of

(boiled and cooled) distilled water.

5.21 X 0.006181 g

4.1.4. Calculate the ^cFeO, and 7eFe2O3 from the

Fe-*FeO 71.8/55.8

^oFeO s 1.286

^cFex

Fe ^Fe2O3 159.77(2 x 55.8^ 1.430

EA 15-3

Ferrous

Quality Control:

The quality of the analysis is controlled by running a standard reference material (MRB-11 T.37% Fe) and a blank. The blank should yield results of less than the reporting increment (Q.1%) and the SRM should be within ±8^0 (relative), (i.e. MRB-11 6.789fc - V.96%) with within run precision of better than 29fc.

The optimum range for analysis is G.1% to

The minimum readable volume of the titre should be not more than 0.025 ml. The determination limit is

0.1 O^o FeO with a 0.5 g aliquot.

Precision, at the 95^c confidence level (2o), is Q.2%

(relative) at the mid-range value of 5^c.

There are a number of sources of determinate error:

1. Iron can be introduced as a contaminant during the rock crushing operation where steel crushers are used.

2. Also, significant ferrous iron can be air oxidized while the sample is being ground since the surface exposed to the atmosphere is greatly increased.

Studies carried out by Fitton and Gill showed that the measured ferrous iron content of a basalt rock dropped from 6.8^^ to 6.09fc after only four minutes of grinding. They recommend no more than thirty seconds of grinding for rocks that are to be analyzed for ferrous iron. They do not com ment on the effect of incomplete grinding on the accuracy or precision of the method.

Langmhyr et al have reported difficulties with incom plete sample dissolution. Some iron-bearing minerals are refractory and even prolonged boiling will not decompose them. Tourmaline H4NaFe3Al6O31 ,

Staurolite Fe(OH)2(Al2SiO5)2, Ilmenite FeO.Ti02,

Magnetite FeO.Fe2O3 are examples. FeS2 is particular ly difficult to dissolve. Development of a method employing steel clad teflon lined bombs may yield a method that would open out these minerals.

The presence of sulphide minerals at levels greater than 896 in the sample may introduce determinate error and a resulting high bias in the results. Sulphur can reduce some of the Fe3*.

If accurate results are required it may be necessary to develop a method where the samples are ground in an inert atmosphere using tungsten carbide or other suitable grinding material.

Productivity:

A technician should be able to complete 30 samples per day.

Additional Notes:

1. Efficiency is improved if samples are run in sets of two. Two samples are weighed and queued for analysis. During the ten minute analysis time for these samples, the next two samples can be weighed and, if required, V2Os may be added.

2. Always keep plenty of hot distilled water on hand.

This water is used for making the saturated boric acid solution (50 ml per sample). It is convenient to use three separate one litre beakers for this purpose. The first is boiled and removed from the hotplate. While it is cooled, the second is placed on the hotplate to boil. When the first is cool, 50 g of boric acid are added. When the second has boiled it is removed from the hotplate and allowed to cool. The third follows the second through this process. When the first solution is used the first beaker follows the third and the cycle continues.

3. In the presence of HF, ferrous iron may oxidize to ferrie iron. The boric acid removes fluoride ions, forming BF62". The sequestering of fluoride ions prevents oxidation of the iron. The boric acid also aids in obtaining a lasting end point.

4. Pyrites are not dissolved by this procedure.

Bibliography:

Goldich, S.S., 1984, Determination of Ferrous Iron in

Silicate Rocks, Chem. Geol., 42, pp 343-347.

Fitton J.G., and Gill R.C.O., 1970, The Oxidation of

Ferrous Iron in Rocks During Mechanical Grinding,

Geochim. Cosmochim. Acta, 34, 518-524.

Langmhyr F.J., and Graff P.R., 1965, A Contribution to the Analytical Chemistry of Silicate Rocks: A

Scheme of Analysis for Eleven Main Constituents

Based on Decomposition by Hydrofluoric Acid, Nor- ges Geologiske Undersokelse No. 230, Univer- sitetsforlaget, Oslo.

Langmhyr F.J., and Kringstad K., 1966, An Investiga tion of the Composition of the Precipitates Formed by the Decomposition of Silicate Rocks in 38%-40%

Hydrofluoric Acid, Anal. Chim. Acta, 35, 131-135.

EA15-4

Potts, P., Handbook of Silicate Rock Analysis,

Blackie, Glasglow, 1987,622 pages.

Wilson A.D., 1955, Determination of Ferrous Iron in

Rocks and Minerals, Bull. Geol. Survey Great Britain,

9, 56-58.

Ferrous

EA15-5

Traces - XRF

DETERMINATION OF TRACE ELEMENTS (T3)

X-RAY FLUORESCENCE SPECTROSCOPY

Introduction:

The Philips PW1400 XRF spectrometer system is used for the quantitative determination of the trace elements and Th, Rb, Sr, Y, Zr, and Nb. This combination of elements is referred to as the Trace 3 (T3) package.

Cs, Ga, Ta, Ce, La and Nd can be determined as

"additions".

with the movement of bottled compressed gas. For more details refer to the Safety Advisory listed on page

EA6-6.

Method:

The method consists of the following techniques:

Rb, Sr: These elements are found in many minerals including micas, amphiboles, pyroxenes, feldspars and less common minerals such as apatite and carbonate minerals. One isotope of Rb (87) decays over a geologically useful time interval to 87Sr, thus the relative abundance ratio of this pair of isotopes can be used to date rock units.

Zr, Nb, Y, Rb, Ce, Ta: These elements, found in rock-forming minerals and accessory minerals, are used extensively to characterize the tectonic setting of basalts and granites. The processes that control the formation of magma at different regions within the earth also control the relative abundances of these elements, thus resulting in characteristic elemental signatures in different areas.

1. Matrix modification by pressing samples into pel lets

2. Irradiation and measurement of characteristic X- ray fluorescence using a sequential, wavelength- dispersive X-ray spectrometer

3. Calculation of final results using calibration cur ves stored in the instrument's computer

Samples are prepared as pellets in a similar manner to that described in method EA6, "Majors - XRF".

General theoretical considerations are presented in

Volume I of this manual.

Apparatus:

- Balance capable of weighing 1-20 g,± 0.001 g

- Ring press and die (40 mm)

- Aluminum "former sleeve"

- Plexiglass plunger

- Boric acid measuring vial (25 ml)

- Spex pellet press (Model 3624B)

Th, Cs, Ga: These elements substitute for more abun dant elements in rock or minerals (e.g. Th follows Zr and U, Cs follows Rb and Li, and Ga substitutes for

Al). Comparison of trace element/major element abundances can be used to track the interplay between formation or alteration process in rock units. Ga/Al ratios are used to understand unusual granites related to ore deposits; Cs is used to characterize fluid/magma/crystal interaction. Th is used either as a petrogenetic discriminator or XRF determinations can pre-screen samples high in Zr and Th that will be used for geochronology.

Ta, Cs, Y, Zr, Nb, Ta, La, Nd: These elements are usually only found at ppb-ppm levels in rock samples.

Several of the world's major economic producers of these elements are located in Canada (Tanco,

Manitoba, and Thor Lake, NWT.).

Safety advisory:

The main sources of danger from this method are from

X-radiation and from electric shock. Fire danger exists, as well as potential for eye damage during the sample preparation. There is some potential danger associated

Reagents:

- Boric acid (granular)

- Polyvinyl alcohol (2^c w/v in distilled water)

Procedures:

1. Preparation of Pressed Powder Pellets for XRF

Determination of Trace Elements in Rock Pow der

1.1.1. Wearing gloves, weigh 5.0 g (0. l g) of -200 mesh rock pulp into a 7 dram snap-cap plastic vial.

1.1.2. Add three drops of polyvinyl alcohol solution

(to serve as a binder) and blend into the pow der by stirring with a nickel spatula.

EA16-1

l races -

1.1.3. Transfer the sample to the 40 mm die to which an aluminum "former sleeve" has been added.

l. l .4. Pack the contents with a plexiglass plunger to form a compact puck.

1.1.5. Remove the sleeve and plunger and add the boric acid powder, placing it on top of the sample. A measuring vial in the boric acid container indicates the amount to be added.

1.1.6. Complete the assembly of the die and form the pellet by application of 15 tons pressure for 15 seconds using the pellet press.

1.1.7. Removed the pellet from the die and label it with a felt-tipped marker on the boric-acid side.

1.1.8. Leave the pellet face-down on a cellulose wipe for 24 hours to allow the binder to dry.

A well-made pellet should have no cracks, an even surface and the rock powder should be centered within the outer ring of boric acid. If a pellet is considered unsuitable, a new one is made from a fresh subsample of rock pulp.

It is important to keep the dies clean. Boric acid and rock powder can build up on the surfaces and con taminate subsequent samples. A thorough cleaning with a cellulose wipe after each pellet is removed will generally suffice. Methanol should be used on the wipe to give a more thorough cleaning. The surface in contact with the rock powder must be treated with care.

It can become pitted and must then be polished by gently rubbing die surface with a fine abrasive mois tened with methanol.

Application and release of pressure on the die-press should be even and slow. The die must be placed centrally and must be level. Failure to do so can result in uneven loading and mechanical failure.

2. Nondestructive Analysis

The use of the Philips PW1400 X-ray Fluorescence

Spectrometer System is described in the Section

'Majors - XRF', page EA6-1.

3. Calculation of Final Results

All necessary calculations are performed automat ically by the system computer.

Quality Control:

It is important to distinguish between short-term

(batch) precision and long-term (multi-year) precision.

Short-term precision on jobs and daily assays is monitored via duplicate analyses of 10*5?? of all samples submitted. These duplicates are subjected to identical preparation and analytical procedures. International reference materials (IRM) are used to monitor short- term precision and accuracy. One IRM similar in composition to the sample group being analysed is inserted into group of six samples (for both jobs and assays). These replicate and IRM data are assessed before the data for geological material are released.

The Ontario Geological Survey carries out many multi-year projects. The analytical data used in the resulting report may come from rocks collected and submitted for analysis over a period of several years.

For the geological interpretations to be significant, they must be based on a realistic evaluation of the long-term precision.

One of the Laboratories' blind-duplicate quality con trol programs involves the insertion of one subsample of three in-house reference materials (a granite, a basalt, and a syenite), in the routine analytical work on a monthly basis. The precision data from this program are used to derive the Laboratories' advertised analyti cal capabilities. The stringency of the test allows us to have confidence that the figures we quote are realistic, even when applied to multi-year projects. The rock- powder samples are introduced 'blind' into the routine workflow and are subject to no special procedures, as often happens when an analyst is aware that QC samples are being run.

Productivity:

A technician should be able to complete 40 samples per day.

Bibliography:

Health and Welfare Canada, Health Protection Branch,

Bureau of Radiation and Medical Devices, 775 Brook- field Rd., Ottawa, Ontario, K1A l CI.

Ontario Ministry of Labor, Occupational Health and

Safety Division, Occupational Health and Safety Act and WHMIS Regulations.

Abbey, S., 1983, Studies in "Standard Samples of

Silicate Rocks and Minerals", GSC Paper 83-15.

EA 16-2

Traces - XRF

Govindaraju, K., 1989,1989 Compilation of Working

Values and Sample Descriptions for 272 Geostan- dards, Geostandards Newsletter, Vol. 13, pp 1-113.

Nockolds, S., Average Chemical Composition of

Some Igneous Rocks, Bull. Geol. Soc. America, Vol.

65.

Potts, P.J., Handbook of Silicate Rock Analysis,

Blackie, Scotland, 1987,622 pages.

Determination Limits and Precision for Trace

Element Analysis

Element Determination Optimum Precision

Limit Range

(ppm) (ppm) (ppm)*

Nb

Rb

Sr

Th

Y

Zr

5

5

5

5

3

10

3-3000

5-1000

5-1000

10-5000

5-1000

5-1000

5

5

5

10

10

10

Additions

Ce

Cs

Ga

La

Nd

Ta

5

35

3

35

20

10

35-5000

3-3000

5-1000

35-1000

20-5000

10-5000

40

3

5

40

20

10

* Precision is quoted (in ppm) as the 95 percent con fidence limit (2o) for values at l OX the determination limit.

EA 16-3

Traces - XRF

APPENDIX A

PROGRAMS

Measurement Program for the Determination of T3 Elements

For this program, preliminary corrections (found in the RC set) include background correction factors and interference factors which are determined independently.

Corrections and Calibration

Background corrections: A fused disc of pure silica is used to evaluate background correction parameters.

These factors are determined from three background points in the scan across the analyte 2-theta angles, i.e., the

Y offset (Y+) for Sr, Y, Zr, Rb, U and Nb; TTI+ for Th, and PD+ for Pb. In order to avoid negative intensities, only 9596 of the background correction factor is used in the correction.

Interference corrections: Preliminary correction for interferences are introduced for U on Rb and for Rb on

U. Factors obtained from running samples with high interferent and negligible analyte are listed under L2 in the

RC table above. As these L2 factors are the fraction of the total counts of the interferent element which must be subtracted from the total counts of the interfered element, without regard to background correction, only partial correction is made at this stage of the data reduction. Mass Absorption (MA) correction is made by ratioing all intensities to the Compton line intensity. Final correction is made using the alpha influence factors.

Calibration: The standard reference materials used in the calibration include NIML, SY3, MicaFe, MAN, GH,

GA and MRG1. These represent a variety of silicate matrices and cover a wide range of concentrations of analytes.

Zr

Y

Sr

Rb

Th

U

Pb

Nb up to up to up to up to up to up to up to up to

1.19fc

Q.07%

G.46%

Q.36%

G.099%

Q.066%

Q.013%

Q.96%

(NIML)

(SYS)

(NIML)

(MAN)

(SYS)

(SYS)

(SYS)

(NIML)

The Measurement program (in OTHER) used for the Trace 3 package is MP5 and the parameters are listed below:

MP5

MASK : l

ABS : YE

CHAN PT PC

Rh

Zr

Y+

U

Th

Pb

20 GEO

100 GEO

40

100

100

100

OEO

OEO

OEO

OEO

CHAN PT PC

Nb

Y

Sr

Rb

TTI+

Pbn-

100 GEO

100 OEO

100 OEO

100 OEO

40

40

OEO

OEO

EA 16-4

Traces - XRF

This measurement program is also used for the determination of individual elements in this group (given the extent of mutual interference and alpha correction factors).

The Link program LP5 specifies the rate corrections and shows that all analyte channels are ratioed to the

Compton scatter peak (RH) channel Ousted as internal standard) for mass absorption corrections.

LPNR^5 NRCHsg CPM^5 MODLST^ MBM^O LLNR = O OPTION = O

INT^RH

ELMNT Zr Y Sr U Rb Th Pb Nb

RC 12345678

Accepted values (Abbey, 1983; Govindaraju, 1989) of element concentrations (in ppm) used in the calibrations are listed below:

MicaFe

NIM-L

GA

SY3

MAN

MRG-1

GH

BEN

Zr

800

11000

150

320

27

105

150

270

Y

25

25

21

740

1

16

70

30

Sr

5

4600

310

306

84

260

10

1350

U

60

14

4

650

12

ci

18

2.4

Rb

2200

190

175

208

3600

8

390

47

Th

150

65

17

990

90

11

1

1

Pb

13

43

30

130

29

10

45

4

Nb

270

960

10

130

175

20

85

100

Calibration curves obtained from the DJ (De Jongh) model are linear. Coefficients D, E and alpha values are stored in CP 5 and are listed below:

ELM ND NC

Pb

Th

Rb

U

Sr

Y

Nb

Zr 3

3

3

3

3

3

3

3

1

0

1

0

1

0

2

2

D

0.00636

0.00068

-0.00376

0.00036

0.00839

0.02234

-0.00257

0.02175

E

3.77756

3.04205

1.18923

2.93207

0.91437

1.10890

0.89649

0.77185

CEL

Rb

Rb

Rb

Zr

Sr

ALPHA

-9.54918

-19.07030

-28.09171

-30.95473

-35.05055

CEL

Th

Th

ALPHA

-40.7827

-0.84639

Agreement with expected and calculated values is generally good. For example, in an initial calibration of Zr, the following agreements were obtained (where Exp. = expected Zr values based on Govindaraju, 1989;

Calc1 = Zr determination based on calibration over full concentration range, Calc2 ^ Zr determination based on limited concentration range, i.e. excluding NIM-L).

EA 16-5

Traces - XRF

Standard

MicaFe

NIM-L

GA

SY3

MAN

MRG-1

GH

BEN

Exp.

(ppm)

800

11100

150

320

27

105

150

270

Calc 1

(ppm)

753

11080

133

330

29

106

119

348

Calc*

(ppm)

810

N/A

139

320

36

113

132

271

The agreement between Exp. and Calc2 indicates calibration over more limited ranges is preferable. The disagreement between Exp. and Calc1 could be assigned to various factors including:

- wide range of calibration with assumption of linearity over the total range, from 27 to 11,000 ppm;

- questionable accepted values;

- limited data for calculation of alphas;

- applicability of the utilization of the Rh Compton line for the mass absorption correction

The applicability of the Rh Compton line can be assessed using the program on page EA 16-9, which calculates the mass absorption based on background counts (MA(B)), Compton counts (MA(C)) and major elements

(MA(Majors)). Analysis of calibration intensity data results in the following data (based on the mass absorption of SY3 for Sr being 12.6).

Reference MA(B)

MAN

NIM-L

Mica Fe

GH

MRG-1

QTZ

SY2

Gl

DRN

GA

Mica Mg

11.2

10.0

13.3

11.7

13.7

13.2

13.0

12.0

13.5

12.1

13.4

MA(C) MA(Maj.)

7.22

16.5

21.1

1.11

18.2

8.22

11.9

8.11

12.6

8.50

13.3

7.81

12.8

19.7

9.40

17.6

7.34

12.0

9.33

12.8

9.58

13.1

In all cases, the Compton-computed value in the previous table (MA(C)) is closest to the MA from the major-element composition (MA(Major)). This indicates the validity for using the Compton line in the calculation of the corrected counts for calibration.

NIM-L is not a typical rock sample, because it contains percent level Zr and very high Sr, an interferent. The above calibration is carried out over a high range of concentrations. If a smaller range is used (i.e., NIM-L included), the agreements are indicated in Calc . Thus the NIM-L Zr determination is over-corrected for the extremely high level of Sr found in this sample (4600 ppm).

This example shows that the standards used in a calibration are often not representative of "normal" silicate rocks. In unusual cases, attention should be paid to the overall composition of the sample, especially unusually high levels of trace elements. XRF scans are often used to detect the latter. The agreement between expected and calculated values for other Trace 3 elements is generally good.

EA 16-6

Traces - XRF

Measurement Program for the Determination of Gallium

Gallium is determined as a separate measurement program, MP3, stored in DB/PB OTHER. Channel conditions are highlighted below:

Rh Ka NO F S l l 75 25 50 40 18.300

Ga Ka NO F S l l 75 25 50 40 38.845 .46

Ta Lp NO F S l l 75 25 50 40 38.465 .30

Tantalum (Ta L(3 line at 2-theta 38.4500) is not expected to be a significant interferent on the Ga analyte line

(Ka 38.8340) in the vast majority of rock samples. Because VSN is used as the highest concentration standard for the calibration of Ga, and as it contains 900 ppm Ta, the potential interference is accounted for in the program.

Measuring program parameters are listed below:

MP3

MASK : l

ABS : YE

CHAN

Rh

Ta-

Ta

Ga

Ga-H

PT

20

20

40

40

20

PC

GEO

OEO

OEO

OEO

OEO

Analysis of a quartz pellet over the four measured positions, Ta-, Ta, Ga and Ga-)-, indicated background correction factors:

Correction for Ta = 0.9260 Ta-

Correction for Ga = 1.1150 Ga-H

Analysis of a high level tantalum standard (Tal540) indicated an interference contribution at the Ga position of

0.111

Rate correction summary:

9 O Ta- 0.9260

10 O Ga-*- 1.1150

Link Program 3:

LP3

0.0000

TA 0.1111

0.0000

0.0000

0.0000

0.0000

= 3 NRO^2 CPM^3 MODLST^ MRBR^

OPTION = O

INT1

^

RH

ELEMENT Ta Ga

RC

INTR

9

RH

10

(The ratioing to the Compton line gave a good mass absorption correction for the Ga line, but not for the Ta line.)

EA16-7

Traces - XRF

Calibration factors were determined using the standards listed below. Expected and calculated values obtained in the original calibration are listed below. Calibration coefficient and alpha correction values (Ta on Ga and

Ga on Ta) are stored in CP3.

GALLIUM

NIM-S

SYS

MAN

VSN

Literature ppm

11

26

59

400

Observed ppm

96

29

51

401

Additional standards run as a program check indicate the following values:

NIM-L

GH

BEN

MRG-1

SY2

54

23

17

18

28

58

17

19

20

29

Counting statistics on the background of the VSN sample were used to calculate the detection limit at 3-4 ppm.

Determination limit was established at 5 ppm in the rock sample.

TANTALUM

NIM-S

Mica Fe

SY3

MAN

VSN

Literature Observed

9

34

7

310

900

-25

38

21

328

870

Tantalum agreement is poor at low levels, however approximation at higher levels is adequate for correction purposes. This program should not be used to determine Ta values, instead use that outlined overleaf.

EA16-8

Traces - XRF

Measurement program for the Determination of Tantalum

XRF is used to determine tantalum in rocks to a determination of limit of 10 ppm.

Analytical parameters are found in PB/DB GEORHO, with measurement and link programs MP8 and LP8 and calculation parameters in CP8.

Channel conditions for the elements are:

ELEMENT FLT

RhKct

Ta* Ka

Ta2 Ka

CuKa

NO

NO

NO

NO

COL

F

F

F

F

DET s s

S

S

XTL

1

1

1

1

ORD

1

1

1

1

UPL

75

65

65

65

LWL

25

35

35

35

KV

50

75

75

75

MA

40

40

40

40

ANGLE

+OFFS-OFFJ

18.300

44.375

45.520

44.995

The two tantalum channels correspond to peak (TA1) and background positions (TA2).

The tantalum line lies between the copper Ka and nickel Kp lines. These two elements rarely occur in tantalum-bearing rocks; however, they may be present in components of the X-ray spectrometer and are assumed to interfere with the Ta line. The resolving power of the PW1400 X-ray spectrometer is high enough to clearly separate the Ta and Ni lines, so copper is the only serious interferent in the determination.

Rate corrections used include: (1) the subtraction of Cu intensity due to the spectrometer (obtained from a wavelength scan of a quartz blank), and (2) background correction.

MP8

MASK : l

ABS : YE

CHAN

Rh

Ta 1

Cu

Ta2

PT

20

100

40

40

PC

OEO

OEO

OEO

OEO

In the development of the link program, it was observed that the best fit was obtained using Compton scatter

MA correction for the Cu and background MA correction for the tantalum line.

LP8

OPTION O

NRO^2 CPM^S MODLST^ MBNR^ LLNT^ O

INT^ RH

ELMNT Ta1 Cu

RC

INTR

14

Ta2

15

Rh

1NT2= TA2

EA 16-9

Traces - XRF

Rate corrections observed (and stored in the RC table under GEORHO) are listed:

NR

14

15

BKGR

0

40

CHL1 LI

Ta2 1.0126

Ta2 1.0000

CHL2 L2

0.0000

0.0000

CHL3 L3

0.0000

0.0000

CHL4 L4

0.0000

0.0000

There are no natural reference materials which have a high certified value for Ta. The synthetic glass VSN is used in the calibration of both Cu and Ta (concentration values: Ta = 900? ppm, Cu = 800 ppm).

Calibration standards used include:

TANTALUM

SRM

VSN

MAN

Mica Fe

Literature ppm

900

310

35

Observed ppm

900

310

36

COPPER

VSN

MAN

Mica Fe

N1M-S

800

140

4

19

801

132

^

30

The calculation parameter file indicates that tantalum is calibrated with an alpha correction for the copper interference.

As stated above, determination limit for tantalum was established to be 10 ppm in the rock sample.

EA16-10

Traces - XRF

Measurment program for the Determination of Ce, La and Nd

The rare earth elements Ce, La and Nd are generally determined in conjunction with the T4 package using ICP

Mass Spectrometry. In certain cases, where the concentrations of these elements are expected to be high, XRF can be used for the determinations.

There are two cerium lines in this wavelength region, the La 79.01 0 and the Lp at 71.700 (2-theta). The La line has a barium interference (Ba Lp 79.260) and is avoided in favour of the Ce Lp line.

MP13, LP13 and CP13 in the OTHER file are used for the determination.

Channel conditions and rate correction parameters are outlined below:

ELEMENT

CeLB

NdLA

La LA

FLT

NO

NO

NO

COL

F

F

F

DET

F

F

F

XTL

1

1

1

ORD

1

1

1

UPL

70

60

70

LWL

30

25

25

KV

65

65

65

MA

30

30

30

ANGLE+OFFS-OFFS

71.600

72.121

82.865

0.80

0.70

1.00

Background correction factors are obtained from the intensities observed at the peak and offset positions (Ce-,

Nd*, La-). In the original procedure reported in earlier manuals (for the PW1540), slope in the background was taken into account by calculating the background correction factor as

Bf = I(peak)7(Bl*B2)

To simplify calculations, and as this program is intended only for relatively high concentrations of these elements, background correction factors are obtained from single points, e.g.,

Bf(Ce) = I(Ce)7I(Ce-)

No correction is made for mass absorbance. This region of the spectrum does not lend itself to the method of

Feather and Willis (background) nor is the Compton scatter ratio meaningful. If the major-element data are available, MA corrections may be applied. However for most samples, MA variations are expected to be minimal and good values should be obtained without correction.

Interference factors may be calculated and inserted in the RC table. The program uses alpha factors exclusively for inter-element correction.

NR BKGR

11

12

13

0

0

0

CHL1 LI

Ce-

Nd*

La-

.9200

1.1105

1.0000

CHL2 L2

0.0000

0.0000

0.0000

CHL3 L3

0.0000

0.0000

0.0000

CHL4 L4

0.0000

0.0000

0.0000

EA16-11

Traces - XRF

MP13

MASK

ABS

: 1

: YE

CHAN

Ce-

Ce

Nd

NC1+

La-

La

PT PC

50 GEO

100 GEO

100 GEO

50 GEO

50 GEO

100 GEO

13 NRO^ 3 CPM^ 13 MOPDLSTrr O MBM^ O LLM^ O

OPTION = O

ELMNT Ce Nd La

RC 11 12 13

Best fit was obtained for Ce and La without alpha corrections. Nd uses an alpha correction for Ce.

Using a 50 second counting time at the background position, counting statistics indicated relatively poor detection limits for Ce, La and Nd. Determination limits were established as 35, 35 and 20 ppm respectively.

EA16-12

Traces - XRF

Measurement Program for the Determination of Cesium

The chromium X-ray tube is favoured for the determination of cesium in rock samples. Channel conditions stored in DB/PB GEORHO are listed below.

ELEMENT FLT COL DET XTL ORO UPL LWL KV MA ANGLE+OFFS-OFFS

CsLa NO F F l l 80 15 75 40 91.865 2.00

Measurement program and link program are stored under GEORHO as MP21 and LP21 with calibration coefficient information in CP21.

Background correction is applied as indicated below:

CHL4 L4

0.0000

NR BKGR CHL1 LI

16 O Cs-h 1.1165

CHL2 L2

0.0000

CHL3 L3

0.0000

Calibration without MA correction results in a good fit with a wide range of standards:

CESIUM

SRM

VSN

Mica Fe

GA

Mica Mg

GXR-3

NBS1633A

Determination limit is estimated to be 3 ppm.

Literature ppm

900

200

55

6

200

11

Observed ppm

900

205

6

59

178

14

EA16-13

Traces - XRF

APPENDIX B

MASS ABSORPTION CALCULATION PROGRAM FOR APPLE HE

10 D$ ~ CI^ (4)

20 FF$ - C^ (12)

30 DIM A$ (100) ,M1 (100) , M2 (100) ,M3 (100)31(100) ,B2 (100)

100 ^

200 HOME

205 PRINT : PRINT "THIS PROGRAM CALCULATES MASS ABSORPTION"

210 PRINT "COEFFICIENTS USING....": PRINT "BACKGROUND COUNTS":

PRINT"COMPTON SCATTER COUNTS": PRINT " MAJOR ELEMENT COMPOSITION"

215 PRINT : PRINT : PRINT "DO YOU WANT TO USE BKG. COUNTS . . . Y/N. . " ; :

INPUT A3

220 PRINT : PRINT "DO YOU WANT TO USE COMPTON COUNTS . . Y/N . . " ; : INPUT BS

230 PRINT : PRINT "DO YOU WANT TO USE MAJOR ELEMENTS . . Y/N. . " ; : INPUT C$

240 HOME

250 IF A$ = "Y" THEN 259

255 GOTO 300

For calculation on background....

259 HOME : PRINT : PRINT "FOR BACKGROUND...": PRINT : PRINT

260 INPUT "NAME OF MA REFERENCE ";NS3

265 INPUT "MA VALUE FOR REFERENCE ";MA

270 INPUT "BKG COUNTS FOR REFERENCE ";BC

280 KB - MA * BC

For calculation on Compton....

300 IF B$ - "Y" THEN 309

305 GOTO 350

309 HOME : PRINT "FOR COMPTON...": PRINT : PRINT

Initialization of 100 entries....

350 HOME

360 FOR I - l TO 100

365 NA3(I) - " ":M1(I) ~ 0:M2(I) = 0:M3(I) ~ 0:B1(I) ~ 0:B2(I) = O

370 NEXT I

310 INPUT "NAME OF MA REFERENCE tl ;NC$

315 INPUT "MA VALUE OF REFERENCE ";MN

320 INPUT "COMPTON COUNTS FOR REF . " ; CC

325 KC = MN * CC

Data Entry ....

400 1=0

405 1=1 + 1

410 PRINT : PRINT : INPUT "NAME OF SAMPLE . . " ; NA3 (I )

415 PRINT

420 IF A$ = "N" THEN 430

421 PRINT : INPUT "BKG COUNTS ";B1(I)

425 Ml (I) - KB X Bl (I)

426 M1(I) - INT (Ml (I) * 10000) l 10000

EA16-14

Traces - XRF

430 IF B$

~

"N" THEN 440

431 PRINT : INPUT "COMPTON COUNTS ";B2(I)

435 M2 (I) = KC

l

B2 (I)

436 M2(I) - INT (M2(I) * 10000)

l

10000

440 IF C$ = "N" THEN 750

For calculation using the major elements ...

450 HOME

451 T - O

455 PRINT "ENTER THE FOLLOWING CONCENTRATIONS AS": PRINT

460 PRINT : PRINT

OXIDE"

For each element, the operator inputs the oxide concentration as a percentage, which is then converted to a weight fraction. The MA contribution is calculated from both the analyte and oxygen components. For example in lines 465-471, the fraction Si in SiO2 is 0.4675 and the fraction of oxygen is (1-0.4675 = 0.5325). The MA values for Si and O at the Sr K-alpha line are 12.8 and 2.6 respectively; T - the accumulated total MA.

465 INPUT "SI02 ";SI

470 SI - SI

l

100

471 T - T

±

SI * .4675 * 12.8

SI * .5325 * 2.6

475 PRINT : INPUT "AL203 ";AL

476 AL = AL

l

100

477 T

~

T

j.

AL * .5291 * 10.3

+

AL * .4709 * 2.6

480 PRINT : INPUT "TOTAL FE203 ";FE

481 FE = FE

l

100

482 T = T + FE * .6994 * 66.8

+ FE *

.3006 * 2.6

485 PRINT : INPUT "MGO ";MG

486 MG - MG

l

100

487 T = T * MG * .6032 * 8.1 * MG * .3968 * 2.6

490 PRINT : INPUT "CAO ";CA

491 CA - CA

l

100

492 T

=

T

+

CA * .7147 * 33.7

+ CA *

.2853 * 2.6

495 PRINT : INPUT "NA20 ";NN

496 NN = NN

/

100

497 T = T + NN * .7919 * 6.3 * NN * .2081 * 2.6

500 PRINT : INPUT "K20

501 KK

=

KK

l

100

" ; KK

502 T = T -f KK * .8302 * 29.5

KK *

.1698

* 2

.6

505 PRINT : INPUT "TI02 ";TI

506 TI

=

TI

l

100

507 T ~ T

*

TI * .5995 * 43.3

510 PRINT : INPUT "P205 ";PP

511 PP - PP

l

100

512 T = T + PP * .4365 * 15.6

TI * pp *

.4005

.5635

* 2

* 2

515 PRINT : INPUT "MNO ";MM

516 MM

=

MM

l

100

517 T = T

+

MM * .7744 * 60.5

520 PRINT : INPUT "C02 ";CC

MM *

.2256

* 2

521 CC

~

CC

l

100

522 T = T + CC * .2720 * . 9

±

CC * .7280 * 2.6

525 PRINT : INPUT "H20 ";HH

526 T ~ (HH

l

100) * .8889 * 2.6

+

T

530 PRINT : INPUT "SULFUR ";SS

535 SS ~ SS

l

100

536 T - T

±

SS * 18.7

.6

.6

.6

EA16-15

Traces-XRF

540 HOME

545 PRINT "DO YOU WANT TO INCLUDE BA AND F?"

550 PRINT : PRINT "...Y/N...";: INPUT R3

555 IF R3 - "N" THEN 600

560 PRINT : PRINT : INPUT "PPM BA ";BA

565 BA - BA

l

1000000

566 T - T

*

BA * 76.3

570 PRINT : PRINT : INPUT "PPM F ";FF

571 FF - FF

l

1000000

572 T - T -t- FF * 3.4

600 T - INT (T * 10000) l 10000

605 M3(I) - T

Another sample.....

750 HOME : PRINT : PRINT "DO YOU WANT ANOTHER SAMPLE?": PRINT

" . . . .Y/N. .";: INPUT R$

755 IF R3

=

"Y" THEN 405

800 M ~ I

Printing sequence...

810 PRINT D3;"PR#2"

820 PRINT : PRINT "MASS ABSORBANCE CALCULATIONS"

830 PRINT : PRINT

840

850

IF A$

=

"Y" THEN PRINT "BKG. REFERENCE - ";NS3;"

IF B$ - "Y" THEN PRINT "COMPTON REF. ~ ";NC$;"

MA

=

";MA

MA

=

";MN

860 IF C$ ~ "Y" THEN PRINT "MAJOR ELEMENT CALCULATIONS INCLUDED"

870 PRINT : PRINT : PRINT

880 PRINT CHR$ (27)"D" CHR3 (20) CHR$ (40) CHR$ (60) CHR$ (0)

890 PRINT CHR$ (9); CHR$ (3)

900 G3

=

CHR$ (9)

910 PRINT "SAMPLE";G$"MA(BKG)";G$"MA(COMP)";GS"MA (MAJORS)"

920 PRINT : PRINT

930 FOR I ~ l TO M

940 IF A$

=

"Y" AND B$

=

"Y" AND C$

~

"Y" THEN 960

950 GOTO 970

960 PRINT NA3 (I) ;G3M1 (I) ,^M2 (I) ;G3M3 (I)

970 IF

AS **

"Y" AND B$ ~ "Y" AND C3 - "N" THEN 990

980 GOTO 1000

990 PRINT NA3 (I) ,^M1 (I) ,^M2 (I) ,^" "

1000 IF A$

~

"N" AND B3

=

"Y" AND C$

=

"Y" THEN 1020

1010 GOTO 1030

1020 PRINT NA$(I);GS" ";GSM2(I);G$M3(I)

1030 IF A3

=

"Y" AND B$ - "N" AND C3 = "Y" THEN 1050

1040 GOTO 1060

1050 PRINT NA3 (I) ;G$M1 (I) ;G$ II ",^M3(I)

1060 IF A$

=

"Y" AND B3 = "N" AND C$ = "N" THEN 1080

1070 GOTO 1090

1080 PRINT NA$ (I) ;G$M1(I) ;G$" ",^" "

1090 IF A3

=

"N" AND B$

=

"Y" AND C$

=

"N" THEN 1110

1100 GOTO 1120

1110 PRINT NA$(I);G3" ";G$M2(I);G$" "

1120 IF A$

~

"N" AND B$

=

"N" AND C$

=

"Y" THEN 1140

1130 GOTO 1150

1140 PRINT NA$ (I) ;G$" ";G3" "

EA16-16

1150 NEXT I

1155 PRINT FF$

1160 PRINT CRRS (27) "@"

1170 PRINT D$; II PR#0 II

1500 END

Traces - XRF

EA16-17

Traces - XRF

APPENDIX C

MASS ABSORBANCE VALUES FOR SELECTED SILICATE S.R.MS

SRM

MRB-7

MRB-8

MRB-9

MRB-10

MRB-11

MA(Sr-Ka)

14.5

8.7

9.0

9.5

10.9

NIM-G

NIM-D

NIM-L

NIM-S

NIM-P

NIM-N

G-2

W-l

GSP-1

BCR-1

SGR-1

SDC-1

RGM-1

STM-1

BHVO-1

SCO-1

MAG-1

QLO-1

G-l

9.5

13.8

10.4

14.5

8.2

11.1

9.1

10.3

14.5

10.1

11.3

10.4

8.8

9.0

13.8

13.2

10.7

13.5

12.8

GA

GH

VSN

UBN

Mica Fe

DTN

DRN

FKN

Mica Mg

GSN

BXN16.2

BR 15.3

QUARTZ

SR MMA(Sr-Ka)

SY-2

SY-3

MRG-1

12.0

12.6

17.5

9.5

8.7

12.7

9.9

19.4

7.5

12.7

9.5

12.8

10.0

ANG

MAN

BEN

7.4

11.3

7.8

8.2

EA16-18

APPENDIX D

MA FACTORS (SR K- a WAVELENGTH)

.5

.8

.5

.8

.0

.6

.8

.7

.7

.6

.8

.3

.5

.3

.6

.7

.3

.6

.5

.7

.0

.8

.6

.1

.3

.7

.3

104

113

122

132

73

81

88

96

143

21

48

54

60

66

29

33

38

43

15

18

22

25

10

12

8

6

2

1

3.4

4

.5

.6

.5

.9

.1

.2

Ga

Ge

As

Se

Co

Ni

Cu

Zn

Br

KR

K

Ca

Se

Ti

V

Cr

Mn

Fe

P

S

CI

Ar

Na

Mg

Al

Si

N

0

F

Ne

B

C

Li

Be

(Reference: Henrich's Tables)

.8

.5

.7

.3

.2

.0

.3

.2

.8

.4

.9

.7

.6

.9

.6

.7

52

55

58

61

41

44

.4

.0

47.0

49

31

33

36

38

23

25

27

29

.3

.8

.7

.8

.4

.5

.6

.4

97

100

105

110

115

120.0

.4

.1

124

129

135

140

80

84

88

92

65

68

72

76

.7

.5

.3

.7

Tb

Dy

Ho

Er

Pm

Sm

Eu

Gd

Tm

Yb

La

Ce

PT

Nd

I

Xe

Cs

Ba

In

Sn

Sb

Te

Rh

Pd

Ag

Cd

Rb

Sr

Y

Zr

Nb

Mo

Tc

Ru

67

20

73.9

77

.6

.7

81

85

90

.5

.1

.5

.2

96.0

134

138

61

64

164

122

125

130

169

175

181

186

145

151

156

163

.9

.6

.0

.2

.8

.2

.7

.2

.4

.7

.2

.1

.9

.5

.4

.3

Bi

Po

At

Rn

Au

Hg

Tl

Pb

Re

Os

Ir

Pt

Lu

Hf

Ta

W

Pa

U

Np

Pu

Fr

Ra

Ac

Th

Traces - XRF

EA16-19

Traces - XRF

APPENDIX E

XRF DETECTION LIMIT CALCULATION

10 D$ - CHR$ (4)

15 PRINT : PRINT "TURN ON PRINTER"

20 PRINT D$;"PR#2 II

30 PRINT : PRINT "DETECTION LIMIT

~

COUNTS XRF"

35 PRINT : PRINT

36 INPUT "ELEMENT NAME ";EL$

40 INPUT "BACKGROUND COUNTS ";B

50 INPUT "PEAK COUNTS

~

";P

55 N

=

P - B

60 PRINT "P-B

=

";N

65 INPUT "CONCENTRATION OF ELEMENT ";C

70 M = (P - B) 7 C

80 PRINT "CONCENTRATION FACTOR ~ "/M

85 INPUT "COUNT TIME AT BKG "; TB

90 PRINT : PRINT

91 DL = (3

l

M) * SQR (B

l

TB)

92 DL - INT (DL * 100)

l

100

100 PRINT "D.L.

=

(3/M)(SQRT B/TB) = ";DL

110 PRINT : PRINT

120 PRINT : PRINT "MORE....Y/N"

130 INPUT RS

140 IF R$ - "Y" THEN 35

149 PRINT D3;"PR#0"

150 END

EA16-20

Traces - XRF

BKGR

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

NR

13

14

16

17

9

10

11

12

7

8

5

6

3

4

2

1

APPENDIX F

RATE CORRECTION PARAMETERS

Trace-element determinations are divided into several groups or packages and divided within the software into several measurement programs. Measurement programs under file OTHER use a common rate correction file,

RC, which is listed below:

CHL1

Y+

Y+

Y+

Y+

Y+

TTI+

Pbf

Y+

Ta-

Ga-H

Ce-

Nd*

La-

Nb-

GS+

Nb-

Ll

.9260

1.1150

.9200

1.1105

1.0000

.9307

1.1165

.9000

1.3021

1.0855

.9090

.8262

.7967

1.2203

1.1717

1.5614

CHL2 L2

0.0000

Rb .2600

0.0000

0.0000

U .1400

0.0000

0.0000

0.0000

0.0000

Ta .1111

0.0000

0.0000

0.0000

0.0000

0.0000

0.0000

CHL3 L3

0.0000

0.0000

0.0000

0.0000

0.0000

0.0000

0.0000

0.0000

0.0000

0.0000

0.0000

0.0000

0.0000

0.0000

0.0000

0.0000

CHL4 L4

0.0000

0.0000

0.0000

0.0000

0.0000

0.0000

0.0000

0.0000

0.0000

0.0000

0.0000

0.0000

0.0000

0.0000

0.0000

0.0000

EA16-21

Traces - AA

DETERMINATION OF TRACE ELEMENTS

(Co, Cr, Cu, Ni, Pb, Zn, Ba, Li, Ag, Cd, Mn, Fe)

ATOMIC ABSORPTION SPECTROMETRY

Introduction:

The trace elements Co, Cr, Cu, Ni, Pb, Zn, Ba, Li, Mn, and Fe can be determined in all geological materials by atomic absorption spectrometry. Although man ganese and iron are not considered trace, they are determined in samples, such as soils and tills, known to contain less Mn and Fe than rocks. Silver and cadmium are determined on samples digested without

HC1 acid.

The average trace element content of various igneous rocks is listed in Table AAS 1 . Considerable variations in these values may occur, especially with samples which include specific minor accessory minerals which contain these elements as major components.

The following is a list of minerals containing these trace elements. The minerals are significant, not only because they can increase the normal expected abun dance in a rock, but because many of them are resistant to mineral acid attack and require fusion techniques to provide a complete dissolution. Ore grade materials may contain high percentages of these minerals thus producing analytical results in the percent range.

COBALT

Cobaltite

Skutterudite

Smaltite

Erythrite

(cobalt bloom)

CoAsS

(Co,Ni)As3

*

(Co,Ni)3(AsO4)2

* an arsenic deficient variety of skutterudite

CHROMIUM

Chromite chromic oxide.

FeCr2O4

Cr in rocks and soils is generally present as

COPPER

Native metal

Chalcopyrite

Bornite

Cuprite

Malachite

Cu

CuFeS2

Cu5FeS4

Cu2O

Cu2CO3

Cu is in most sulphide ores. There are 600 Cu minerals.

NICKEL

Braggite

Millerite

Pentlandite

Pyrrhotite

Kupfernickel

(Pt,Pd,Ni)S

NiS

(Fe,Ni)9S8

As mineral

Ontario has the world's nickel.

greatest abundance of

LEAD

Galena

Cerussite

Wulfenite

Anglesite

PbS

PbCO3

PbMoO4

PbSO4

Inorganic salts of lead are relatively insoluble.

ZINC

Sphalerite

Smithsonite

ZnS

ZnCO3

BARIUM

Barite

Witherite

BaS04

BaCO3

Barium in solution is precipitated by sulphate and carbonate ions.

LITHIUM

Spodumene

Petalite

Amblygonite

Hectorite

LiAlSiO6

(Li,Na)(Al,SUOio)

Li,Al PO4 (F,OH) complex silicate *

* can contain up to 1 9fc

Li compounds are very

Li2O soluble.

MANGANESE

Pyrolusite

Alabandite

Rhodocrosite

SILVER

Native metal

Argentite

Acanthite

MnO2

MnS

MnCO3

Ag

Ag2S

AgS

EA17-1

Traces - AA

TABLE AAS1. TRACE ELEMENT CONTENT (PPM) IN IGNEOUS ROCKS

Element Ultrabasic Gabbro Intermediate Granite

Pb

Zn

Ba

Li

Co

Cr

Cu

Ni

Mn

Ag(ppb)

Cd

100

3000

50

1000

0.3

50

20

3

1000

60

50

200

100

120

3

100

200

10

1400

110

6

70

1600

20

15

50

50

40

1400

60

0.2

25

3

15

10

20

50

950

40

400

40

Pegmatite

200?

700?

60

400

9

30

1

5

3

9

Most Ag compounds are relatively insoluble.

Silver is associated with Pb-Zn ores.

CADMIUM

Greenockite

Cd2SO4

Cadmium is associated with ZnS ores.

Safety advisory:

1. Exercise extreme care when using any acids and fluxes required for sample dissolution. Their use should only be attempted after the appropriate

MSDS sheets have been read and the safe handling

and first aid procedures understood. Acids should only be handled in a fume hood designated for their use and proper protective equipment worn. Proper ventilation is required when handling fluxes which create a dust control problem.

2. Before operating an atomic absorption spectrophotometer, ensure that the proper instruc tions found in the manufacturer's operator's manual are understood. Preliminary safety checks are noted in the section of this manual outlining the operation of an atomic absorption spectrophotometer.

3. Cylinders of compressed gas used as fuel and oxidant for flame atomic absorption spectrometry must be securely fastened and have the proper regulator. Ensure that the system has no leaks and the gas hoses are in good condition. Review the section on compressed gases found in the Safety

Manual for the Geoscience Laboratories.

Apparatus:

- Atomic Absorption Spectrometer

A. VarianAA775

B. Varian AA-5 (IM-6 update)

- Strip Recorder - Varian Model 9176 or Linear

Model 1200

- Pipettes - 5, 10, 20, 25, 50 ml

- Volumetric flasks - 50, 100, 200, 250, 500,

1000,2000ml

- Acid dispensers - Socorex Model 511 (10 ml)

- Various sized glass beakers

Reagents:

- Nitric acid, HNO3, 697c (w/w)

- Stock Standard Solutions (1000 ppm)

A. Custom Mixed Standard containing

Co, Cu, Ni, Pb, Zn, Mn, Fe, Ag.

B. Chromium

C. Lithium

D. Barium

E. Cadmium

- Potassium Chloride, KC1

- Lanthanum Oxide, La203, suitable for flame en hancement in atomic absorption spectrometry.

Method:

The method consists of the following techniques:

1. Sample decomposition by either acid digestion or fusion with a flux

2. Measurment of analyte signal by atomic absorp tion spectrometry

3. Calculation of results based on calibration curves produced from synthetic samples.

EA 17-2

Traces - AA

Procedures:

1. Reagent preparation

1.1. Lanthanum/Potassium Reagent for Barium

Determination (6000 ppm La 1 6000 ppm K*)

- Carefully add 200 ml of nitric acid to 14.07 g of lanthanum oxide placed in a 600 ml covered

1.2. glass beaker. CAUTION: Reaction is vigorous and generates heat.

To a 2 liter volumetric flask add 22.89 g of KQ and dissolve with 200 ml of distilled water.

1.3. Transfer the lanthanum solution when cool to the volumetric flask and dilute to 2 liters with distilled water.

2. Preparation of Calibration Working Standards

2. l .

2.2.

Working standards are prepared by dilution of the custom mixed stock solution, with 10% nitric acid. Typical concentrations of analytes in working standards range from 0.5 to 8 ppm.

Lithium standards: Add 20 ml of 10,000 ppm

K* for every 100 ml of working standard prepared. Final concentration of K"1" is 2000 ppm. Working standards are prepared by serial dilution of the stock lithium solution with 10% nitric acid.

2.3.

Chromium standards: Add 10 ml of 10,000 ppm K" for every 100 ml of working standard prepared. Final concentration of K* is 1000 ppm. Working standards are prepared by serial dilution of the stock chromium solution with nitric acid.

2.4.

Barium standards: Add 50 ml of 6000 ppm

La3VK* for every 100 ml of working standard prepared. Final concentration of LaVK* is

3000 ppm. Working standards are prepared by serial dilution of the stock barium solution with nitric acid.

3. Sample Decomposition

3.1. Samples are digested with mineral acid using either HN03-HC104-HF (Procedure 1), or

HC1-HF-HNO3 (Procedure 3), or HNO3-HF

(Procedure 5). Details are found in the section on sample dissolution by acid attack.

3.2. Samples are decomposed by fusion using lithium metaborate (Procedure 1) or sodium

3.3.

4. Measurement of Analyte Concentration by

Atomic Absorption Spectrometry

A Varian AA775 is used for the determination of trace elements in solution. See "Operation of the Atomic

Absorption Spectrophotometer - Varian AA775" on page EA4-1 in this manual for the steps required to obtain a concentration for each element of interest.

Consult the information sheets at the end of this section for the instrument parameters and necessary details to perform the determination of each trace element.

Samples with concentration of the element greater than the upper limit of its calibration curve are to be diluted to the appropriate level for remeasurement.

5. Calculation of results

Calibrate the instrument with the working standard solutions according to the instructions in the operator's manual. The spectrometer is constantly adjusted to distilled water zero, thus the number obtained from the digital readout is in concentration units, usually ppm.

This reading is converted to a result for the sample by the following: where peroxide (Procedure 2). Details are found in the section on sample dissolution by fusion.

Preparation of samples for Barium determina tion - Dilute l part of sample solution with l part of 6000 ppm La3VK* prior to determina tion by atomic absorption spectrometry.

R = (C - B) x DF

R = analyte concentration in the sample

C - concentration measured in solution

B = reagent blank concentration

DF = dilution factor from the preparation step

For a barium determination, a strip recorder is used to obtain a trace of the absorbance of the calibration standards and the sample. Peak heights are measured and concentration in solution for the sample is deter mined by simple ratio with the peak heights of the calibration standards. The above equation is used to calculate the result for the sample.

EA 17-3

Traces - AA

6. Reporting

Cobalt

Co

All results are reported on appropriate Trace Element

Analysis Forms. See the manual "Processing of

Whole Rock Chemical Data - Ontario Geological Sur vey" for all details.

Normally the trace element forms will be computer generated and complete with sample numbers, geologist's name and job number. If these are not available the appropriate identifying data and correctly sequenced sample numbers will be entered by the

Laboratories staff doing the analysis. Correctly se quenced sample numbers refer to the order as written in the Geoscience Laboratories Job Sheet. There are five column spaces for each analytical result. The first four are for the numerical value of the determined concentration, the fifth for the unit of concentration.

Concentration may be entered as percent (P), ppm (M), or ppb (B). The use of a decimal point is allowed in any of the first four columns. A blank in the unit column is assumed to be ppm (M). Amounts less than the detection limit will be recorded as the detection limit preceded by a "minus" sign (e.g. -5 indicates a value below the detection limit of 5). N.D. and are not allowed.

STANDARD SOLUTIONS

Use 100 ppm solution prepared from 1000 ppm mixed custom stock solution to prepare working standards of

0.5, 1.0,2.0, and 4.0 ppm.

INSTRUMENT PARAMETERS

Lamp Current (ma)

Wavelength (nm)

7.0

240.7

Spectral Band Pass (nm) 0.2

Background Correction ON

Flame Description

Air-acetylene, oxidizing, fuel lean, blue

WORKING CONDITIONS

Sensitivity l ppm Co solution reads 0.145 absorbance

Calibration

Set 1.00 ppm to read 0.50

Working range

0.5 to 4.0 ppm

The Trace Form l is used to report the Tl elements Co,

Cr, Cu, Ni, Pb, Zn, Ba, Li, and Ag. The Trace Form 5 is used to report Cd.

The Trace Form 7 is used to report "major" and

"minor" elements (Mn, Fe) when they occur at trace levels.

Assay samples are reported on Form 1117 - Assay

Work and Report Form.

Quality control and Additional notes:

Accuracy: In the absence of interfences, accuracy may approach precision but is dependent on the quality of the standard solution used. Blind duplicate and

SRM quality contol data confirm that this method can remain under good analytical control indefinitely.

This note applies to all of the following information sheets.

INTERFERENCES

Minimal interferences have been observed with the air-acetylene flame.

A cobalt response can be depressed in the presence of excess nickel (1500 ppm). Either dilute the sample, if possible, or matrix match the sample and standards.

NOTES

1. To reduce signal noise, the gain is reduced by setting working standards to one-half values during calibration. The subsequent readout for samples must be multiplied by 2 before reporting to obtain the correct concentration value.

ANALYTICAL CAPABILITIES

Determination Limit (ppm) 5

Precision, at the 95*7c confidence limit (2o), at 10X determination limit (50 ppm) is 5 ppm (absolute).

EA17-4

Traces - AA

Chromium

STANDARD SOLUTIONS

Use 100 ppm solution prepared from 1000 ppm stock

Cr solution to prepare 1.0,2.0,4.0 and 8.0 ppm work ing standards. Add 10 ml of 10,000 ppm K* for every

100 ml of working standard prepared, l ml = 1000 ppm K*

Cr Copper

Cu

STANDARD SOLUTIONS

Use 100 ppm solution prepared from 1000 ppm mixed custom stock solution to prepare working standards of

0.5, 1.0, 2.0 and 4.0 ppm.

INSTRUMENT PARAMETERS

INSTRUMENT PARAMETERS

Lamp Current (ma)

Wavelength (nm)

5.0

357.9

Spectral Band Pass (nm) 0.2

Background Correction OFF

Flame Description

Nitrous oxide-acetylene

Fuel rich, red cone

Lamp Current (ma)

Wavelength (nm)

Spectral Band Pass (nm) 0.5

Background Correction

OFF

Flame Description

3.5

324.8

Air-acetylene, oxidizing, fuel lean, blue

WORKING CONDITIONS

WORKING CONDITIONS

Sensitivity l ppm Cr solution reads 0.090 absorbance

Calibration

Set l ppm to read 0.50

Working range

1.0 to 8.0 ppm

INTERFERENCES

The nitrous oxide-acetylene flame overcomes most supressions observed with an air-acetylene flame.

Ionization suppression is eliminated by addition of

1000 ppm K* to the working standards. Most rock samples have sufficient alkali concentration to eliminate the need for addition of K* to the sample solution.

NOTES

1. To reduce signal noise, the gain is reduced by setting working standards to one-half values during calibration. The subsequent readout for samples must be multiplied by 2 before reporting to obtain the correct concentration value.

Sensitivity l ppm Cu solution reads 0.200 absorbance

Calibration

Set 1.00 ppm to read 1.00

Working range

0.5 to 4.0 ppm

INTERFERENCES

Few interferences with air-acetylene flame.

One percent Fe in solution, i.e., 50% Fe in the sample, can cause a lO^c suppression.

High Zn/Cu ratios can cause some suppression of the

Cu signal. Use a leaner air-acetylene flame to mini mize this effect.

ANALYTICAL CAPABILITIES

Determination Limit (ppm) 5

Precision, at the 95*^ confidence limit (2o), at l OX determination limit (50 ppm) is 4 ppm (absolute).

ANALYTICAL CAPABILITIES

Determination Limit (ppm) 10

Precision, at 95 7c confidence limit (2o), at 10X deter mination limit (100 ppm) is 20 ppm (absolute).

EA 17-5

Traces - AA

Nickel Ni Lead Pb

STANDARD SOLUTIONS

Use 100 ppm solution prepared from 1000 ppm mixed custom stock solution to prepare working standards of

0.5, 1.0, 2.0 and 4.0 ppm.

INSTRUMENT PARAMETERS

Lamp Current (ma)

Wavelength (nm)

5.0

232.0

Spectral Band Pass (nm) 0.2

Background Correction ON

Flame Description

Air-acetylene, oxidizing, fuel lean, blue

WORKING CONDITIONS

Sensitivity l ppm Ni solution reads 0.120 absorbance.

Calibration

Set 1.00 ppm to read 0.50

Working range

0.5 to 4.0 ppm

INTERFERENCES

Non-atomic species in the air-acetylene flame absorb strongly at 232.0 nm; background correction is neces sary.

NOTES

1. To reduce signal noise, the gain is reduced by setting working standards to one-half values during calibration. To obtain the correct con centration value for the samples, the readout must be multiplied by 2 before reporting the result.

2. A non-resonant line for Ni exists at 231.6 nm.

Make sure the correct wavelength is chosen when the instrument is being "peaked".

3. Calibrations with 232.0 nm are usually very curved because of a non-resonant line of Ni at

232.14 nm. This is of particular importance with samples of low Ni content-use a 0.5 ppm Ni calibration standard.

STANDARD SOLUTIONS

Use 100 ppm solution prepared from 1000 ppm mixed custom stock solution to prepare working standards of

0.5, 1.0, 2.0 and 4.0 ppm.

INSTRUMENT PARAMETERS

Lamp Current (ma)

Wavelength (nm)

5.0

217.0

Spectral Band Pass (nm) 1.0

Background Correction ON

Flame Description

Air-acetylene, oxidizing, fuel lean, blue

WORKING CONDITIONS

Sensitivity l ppm Pb solution reads 0.060 absorbance.

Calibration

Set 1.00 ppm to read 0.50

Working range

0.5 to 4.0 ppm

INTERFERENCES

Few serious interferences have been reported.

NOTES

1. The 283.3 nm line is sometimes preferred because of its better signal to noise ratio than the 217.0 nm line, although the sensitivity is one-half.

2. To reduce signal noise, the gain is reduced by setting working standards to one-half values during calibration. To obtain the correct con centration value for the samples, the readout must be multiplied by 2 before reporting.

ANALYTICAL CAPABILITIES

Determination Limit (ppm) 10

Precision, at the 959c confidence limit (2o), at l OX determination limit (100 ppm) is 8 ppm (absolute).

ANALYTICAL CAPABILITIES

Determination Limit (ppm) 5

Precision, at the 959fc confidence limit (2o), at l OX determination limit (50 ppm) is 6 ppm (absolute).

EA 17-6

Traces - AA

Zinc Zn Barium

STANDARD SOLUTIONS

Use 100 ppm solution prepared from 1000 ppm mixed custom stock solution to prepare working standards of

0.5,1.0,2.0 and 4.0 ppm.

INSTRUMENT PARAMETERS

STANDARD SOLUTIONS

Ba

Use 100 ppm solution prepared from 1000 ppm stock

Ba solution to prepare 1.0,2.0,4.0 and 8.0 ppm work ing standards. Add 50 ml of 6000 ppm La^/K* solu tion for every 100 ml of working standard prepared, l ml - 3000 ppm La3VK*

INSTRUMENT PARAMETERS

Lamp Current (ma)

Wavelength (nm)

5.0

213.9

Spectral Band Pass (nm) l .0

Background Correction ON

Flame Description

Air-acetylene, oxidizing, fuel lean, blue

WORKING CONDITIONS

Sensitivity l ppm Zn solution reads 0.325 absorbance

Calibration

Set 1.00 ppm to read 1.00

Working range

0.5 to 3.0 ppm

INTERFERENCES

Non-atomic species in the air-acetylene flame absorb strongly at 213.7 nm. Background correction is neces sary.

NOTES

1. Calibration curve will bend at ends. For samples with low Zn content calibration with a low stand ard (0.5 ppm) is necessary.

2. Dilute solutions containing more than 3 ppm Zn.

Linearity drops off sharply in this range. Attempts to calibrate with standards 3 ppm require excessive curve correction and cause the calibration to be unstable.

ANALYTICAL CAPABILITIES

Determination Limit (ppm) 10

Precision, at the 95*56 confidence limit (2a), at l OX determination limit (100 ppm) is 10 ppm (absolute).

Lamp Current (ma)

Wavelength (nm)

10

553.6

Spectral Band Pass (nm) l .0

AA5 (100)

Background Correction OFF

Flame Description

Nitrous oxide-acetylene

Fuel lean, small red cone

WORKING CONDITIONS

Use a strip recorder to trace absorbance readings.

For Varian Model 9176 set chart speed at l cm/min, mV/FS at 5, l ppm Ba signal for 3".

Working range

1.0 to 2.0 ppm- 5mV7FS

2.0to4.0ppm- lOmV/FS

4.0 to 8.0 ppm - 20 mV/FS

INTERFERENCES

Ionization in the nitrous oxide-acetylene flame is sup pressed with the addition of K* to samples, standards and blanks.

Stable compound interference from phosphate, sili cate, aluminate, etc. is minimized in a nitrous oxide- acetylene flame and/or by addition of La3*.

NOTES

1. In the region of 553.6 nm strong emissions from

B a and the nitrous oxide-acetylene flame con tribute to the noise of the absorbance signal. A strip recorder is used to record the signal. The noise may also be reduced by increasing the lamp current and thereby decreasing the PM voltage.

2. Run calibration standards frequently to ensure no loss in sensitivity. If this occurs, check fuel flow, burner clogging, or loss of peak alignment.

EA17-7

Traces - AA

3. Dilute samples outside of largest calibration range

BEFORE adding "buffer" solution.

ANALYTICAL CAPABILITIES

Determination Limit (ppm) 10

Precision, at the 95*^ confidence limit (2o), at l OX determination limit (100 ppm) is 16 ppm (absolute).

Lithium

STANDARD SOLUTIONS

Li

Use 100 ppm solution prepared from 1000 ppm stock

Li solution to prepare 0.5, l .0,2.0 and 4.0 ppm working standards. Add 20 ml of 10,000 ppm K* for every 100 ml of working standard prepared, l ml = 2000 ppm K*

INSTRUMENT PARAMETERS

Flame Emission - Single Beam

Wavelength (nm) 670.8

Spectral Band Pass (nm) 0.2

Background Correction OFF

Flame Description

Nitrous oxide-acetylene

Fuel lean, small red cone

WORKING CONDITIONS

While aspirating a 2 ppm Li solution, set absorbance to read approximately 0.750.

Calibration

Set 1.00 ppm to read 1.00

Working range

0.5 to 4.0 ppm

INTERFERENCES

Ionization in a nitrous oxide-acetylene flame is over come with the addition of 2000 ppm K*,

NOTES

1. Most rock samples have sufficient concentration of easily ionizable substances to eliminate the need for addition of K" to the sample solution.

2. The air-acetylene flame in the emission mode generates excessive signal to noise levels.

ANALYTICAL CAPABILITIES

Determination Limit (ppm) 3

Precision, at the 95^c confidence limit (2o), at l OX determination limit (30 ppm) is 3 ppm (absolute).

EA 17-8

Traces - AA

Manganese^^^^^^^^

Mn

STANDARD SOLUTIONS

Use 100 ppm solution prepared from 1000 ppm mixed custom stock solution to prepare working standards of

0.5, 1.0,2.0 and 4.0 ppm.

ANALYTICAL CAPABILITIES

Determination Limit (ppm) 5

Precision, at the 959fc confidence limit (2o), at l OX determination limit (50 ppm) is 10 ppm (absolute).

INSTRUMENT PARAMETERS

Lamp Current (ma)

Wavelength (nm)

5.0

279.5

Spectral Band Pass (nm) 0.2

Background Correction OFF

Flame Description

Air-acetylene, oxidizing, fuel lean, blue

WORKING CONDITIONS

Sensitivity l ppm Mn solution reads 0.300 absorbance

Calibration

Set 1.00 ppm to read 1.00

Working range

0.5 to 4.0 ppm

INTERFERENCES

The presence of phosphate, perchlorate, iron, nickel and cobalt will depress the Mn absorbance when a reducing air-acetylene flame is used. USE an oxidiz ing flame.

Silicon depresses the signal and is overcome by incor poration of G.2% CaCl2 in samples and standards.

NOTES

1. The presence of silica in samples fused with lithium metaborate may cause a depression in the

Mn absorbance if insufficient Ca is present in the sample. Use certified reference materials as calibration standards. See "Major Elements

Determined by Flame Atomic Absorption" in this manual.

2. The closely spaced triplet formed by the three strongest absorption lines of Mn may not be resolved with all monochromators.

3. Most soils and sediments require a l: l O dilution to be made on the solution prepared by acid diges tion.

EA 17-9

Traces - AA

Iron________________________Fe

STANDARD SOLUTIONS ANALYTICAL CAPABILITIES

Use 100 ppm solution prepared from 1000 ppm mixed Determination Limit (ppm) 5 custom stock solution to prepare working standards of

0.5, 1.0, 2.0 and 4.0 ppm. Precision, at the 959fc confidence limit (2a), at 10X determination limit (50 ppm) is 8 ppm (absolute).

INSTRUMENT PARAMETERS

Lamp Current (ma)

Wavelength (nm)

5.0

248.3

Spectral Band Pass (nm) 0.2

Background Correction OFF

Flame Description

Air-acetylene, oxidizing, fuel lean, blue

WORKING CONDITIONS

Sensitivity l ppm Fe solution reads 0.145 absorbance

Calibration

Set 1.00 ppm to read 0.50

Working range

0.5 to 4.0 ppm

INTERFERENCES

A reduction in sensitivity has been observed when iron is determined in the presence of nitric acid and nickel.

This effect is minimized by using a very lean flame.

All interferences can be removed with the use of a nitrous oxide-acetylene flame.

NOTES

1. To reduce signal noise, the gain is reduced by setting working standards to one-half values during calibration. To obtain the correct con centration value for the samples, the readout must be multiplied by 2 before reporting the result.

2. Most soils and sediments require a l: 100 dilution to be made on the solution prepared by acid diges tion. This has the added effect of reducing any concomitants which may cause interferences.

3. Before setting the AAS to zero, use a clean solu tion of distilled water. Change the water frequent ly, to avoid a buildup of iron in the rinse solution.

EA17-10

Traces - AA

Silver

Ag Cadmium Cd

STANDARD SOLUTIONS

STANDARD SOLUTIONS

Use 100 ppm solution prepared from 1000 ppm mixed Use 100 ppm solution prepared from 1000 ppm stock custom stock solution to prepare working standards of Cd solution to prepare 0.5, 1.0 and 2.0 ppm working

0.5, l .0, 2.0 and 4.0 ppm. standards.

INSTRUMENT PARAMETERS INSTRUMENT PARAMETERS

Lamp Current (ma)

Wavelength (nm)

2.5

328. l

Spectral Band Pass (nm) 0.5

Background Correction ON

Flame Description

Air-acetylene, oxidizing, fuel lean, blue

WORKING CONDITIONS

Sensitivity l ppm Ag solution reads 0.200 absorbance

Calibration

Set 1.00 ppm to read 1.00

Working range

0.5 to 4.0 ppm

NOTES

1. The higher salt content of the acid digested samples (dilution factor 25) may cause viscosity problems and/or burner clogging. Rinse well be tween samples.

Lamp Current (ma)

Wavelength (nm)

3.5

228.8

Spectral Band Pass (nm) 0.5

Background Correction ON

Flame Description

Air-acetylene, oxidizing, fuel lean, blue

WORKING CONDITIONS

Sensitivity l ppm Cd solution reads 0.540 absorbance

Calibration

Set 1.00 ppm to read 1.00

Working range

0.5 to 2.0 ppm

INTERFERENCES

Background correction is required because of low resonance wavelength.

2. Assay samples received as chemical leaches must not be diluted with HNO3 which may cause a precipitate to form. Use water for dilutions and determine concentration by flame AAS as soon as possible. Normal first dilution is 5/500.

ANALYTICAL CAPABILITIES

ANALYTICAL CAPABILITIES

Determination Limit (ppm) 2

Precision, at the 957c confidence level (2a), at 10X determination limit (20 ppm) is 2 ppm (absolute).

Determination Limit (ppm) 2

Precision, at the 959fc confidence limit (2a), at l OX determination limit (20 ppm) is l ppm (absolute).

EA17-11

Traces - ICP-OES

OVERVIEW OF TRACE ELEMENT DETERMINATION

BY ICP OPTICAL EMISSION SPECTROSCOPY

Introduction

Inductively coupled plasma-optical emission spectros copy, ICP-OES, is employed for the determination of a number of trace elements in geological materials.

The ICP-OES packages offered by the Geoscience

Laboratories (GLOGS) include:

Aqueous Scan: The Geoscience Laboratories is able to accept aqueous samples for analysis on the JY48P

ICP Spectrometer. No sample preparation is required and simultaneous determination of up to 36 elements is possible. The available elements are Sn, Mo, W, B,

Zn, P, Pb, Co, Ba, Se, Cr, Mg, V, Nb, Ca, Ag, Ti, Dy,

Y, Eu, Sr, Gd, Ni, Ta, Mn, Fe, Si, Al, Be, Cu, Yb, Zr,

Sm, La, Nd and Ce.

Trace 2 (T2): Rock pulps are prepared for analysis by acid digestion. The elements available in the Trace

2 package are: Be, Se, Nb, Mo, V, Sr, Y, Co, Cu, Ni and Zn. Tungsten is also offered (at relatively high determination limits). Ce, La and Nd are offered as options; Zr, normally determined by XRF, is also available. The acid resistance of most Zr-containing minerals make ICP-OES analysis of solutions un suitable for routine reporting of this element.

TSPA: The Tentative Spectroscopic Analysis pack age, run on rock pulp solutions prepared by acid diges tion, provides semiquantitative analysis of 26 elements including Al, Ba, Be, Ca, Ce, Cr, Co, Cu, Fe, La, Pb,

Mg, Mn, Mo, Ni, Nb, Nd, P, Sr, Ta, Ti, W, V, Y, Zn and Zr. Results are reported in terms of concentration ranges.

The operating concepts of inductively coupled plasma- optical emission spectroscopy are as follows:

(1) A high temperature excitation source is produced by the interaction between a radio-frequency field and a flow of argon gas. The interaction results in the ionization and excitation of the Ar to produce a plasma (or "flame"), which can attain tempera tures in the range 8000-100000C. At these temperatures, any sample which is introduced into the plasma will be vapourized, atomized, ionized and excited to emit radiation.

(2) The emitted light is dispersed by a grating in an optical system and the intensities of the spectral lines are measured, generally by photometric means, although photographic recording can be used. Instrumentation typically consists of either

(a) a scanning monochromator with a single detec tor placed behind an exit slit on the Rowland circle

(sequential measurement), or of (b) a polychrom- ator with a number of photomultiplier detectors positioned behind fixed slits cut along the optical axis providing simultaneous wavelength-selective detection (direct reader).

(3) The ICP-OES spectrometer in the Geoscience

Laboratories is a Jobin-Yvon JY48P direct read ing instrument, employing a holographic grating, with 36 channels dedicated to elements of geological interest. These elements, along with their "channel numbers" and first order analytical wavelengths are listed in Table ICP.

(4) The JY48P instrument employs a PlasmaTherm source, generator and torch, with maximum plas ma power of 2500 watts. Standard operating con ditions average 1800 watts. The source system is not under computer control, and all settings must be established and reproduced manually. Al though other nebulizer systems are available, the

Geoscience Laboratories has consistently used a concentric glass nebulizer (Meinhard C3 type) with high salt capability. This nebulizer, along with a glass spray chamber and quartz torch are used for sample introduction, aerosol formation and excitation.

The torch unit is mounted in a "torch box" whose position relative to the entrance slit of the polychromator can be adjusted with a universal stage. As with the plasma source, torch position and optimal signal collection are under manual control.

EA18-1

Traces - ICP-OES

TABLE ICP1. ANALYTICAL WAVELENGTHS

AND CHANNELS ON JY48P

ICP-OES SYSTEM

Fixed Channel Elements - Wavelengths (A)

CH#

33

34

35

36

26

27

28

29

22

23

24

25

30

31

32

17

19

20

21

13

14

15

16

9

10

11

12

7

8

5

6

3

4

2

1

Element

Magnesium

Vanadium

Aluminum

Niobium

Beryllium

Calcium

Copper

Silver

Ytterbium

Titanium

Zirconium

Dysprosium

Samarium

Yttrium

Lanthanum

Europium

Neodymium

Strontium

Cerium

Tin

Molybdenum

Tungsten

Boron

Zinc

Phosphorous

Lead

Gadolinium

Cobalt

Nickel

Barium

Tantalum

Scandium

Manganese

Chromium

Iron

Wavelength(A) Order

2798.06

2924.02

3082.15

3094.18

3130.42

3158.57

3247.54

3280.68

3289.37

3372.80

3438.23

3531.70

3592.60

3710.30

3794.78

3819.67

4061.09

4077.71

4137.65

1899.26

2020.30

2079.11

2089.59

2138.56

2149.14

2169.99

3422.46

2286.16

2316.04

2335.27

2400.63

2552.37

2576.10

2677.16

2739.55

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1

1 under specific conditions, regression analysis for calibration curve generation, plotting, background correction, matrix interference correction, entrance slit scans (profiles), blank subtraction, detection limit calculations and data acquisition, conversion and retrieval.

When not used to operate the spectrometer, the computer can be used as a general purpose device, running programs in FORTRAN, BASIC and

MACRO languages. It can also be used for general file manipulation and text-editing. Off-line programs have been written for report generation and the preparation of TSPA worksheets and cer tificates. These programs use the ASCII data files created during spectrometer operation.

(6) The magnitude of the signal reaching a particular channel detector is dependent on the line inten sity. This, in turn, is dependent on the concentra tion of analyte in the sample, as well as instrument conditions such as plasma power, rate of sample uptake, etc. The signal is also dependent on inten sity contributions from other elements present in the sample - from direct and partial line overlap, continuum emission and background effects. In strument software allows the operator to correct for many of these interferences through the use of background correction and interference correction coefficients.

The magnitude of the transduced photomultipler signal is controlled by the sensitivity of the detec tor. The JY48P permits separate control of the sensitivities for all channels by adjustment of switches on voltage control cards. Channel response is set in accordance with the expected abundance of the analyte in "normal" geological samples.

(7) The general operation of the system for the analysis of samples is described by the schematic diagram shown below. Features include:

(5) The instrument is equipped with a Digital 11/23 computer with 128 Kbytes of memory, dual RXO2 floppy disk drives, DEC WRITER III (L A120) and

Tektronix 4006-1 (video) terminals. The software, provided by the Instruments S.A.,

(Metuchen, N.J.), provides full control of the detection parameters (e.g. integration time, time between successive integrations, number of repli- cate integrations). It also employs specialized routines for establishing analytical programs

- source and sample delivery under manual con trol;

- spectrometer under partial computer control, i.e., entrance slit position (profile scanning), channel profile selection (for external scanning), polychromator data acquisition (at all channels).

- manual control of detector sensitivities.

- detection under computer control, i.e., replicates, integration time, flush and delay times, etc.;

- machine language raw data files (*.RAW) for all channels and at all slit positions (for background

EA 18-2

Traces - ICP-OES

correction) stored automatically in polychromator data acquisition routines;

- ASCII analytical data files (*.DAT) are program controlled "translations" of the raw data files.

The *.DAT files, obtained by processing the raw intensities through software TABLE files

(*.TAB), contain information as intensities

(counts) or concentrations for selected elements.

These data can also be corrected for background or interferences.

- computer-controlled access of *.DAT files for implementation of regression routines, detection limit calculations;

- all programs, routines, *.RAW, *.DAT files are stored on floppy disks.

Computer Management - Software Control

Introduction to the Computer

The DEC 11/23 system is used for instrument control and for data processing and management using software available from the manufacturer. Because the application software is written in FORTRAN lan guage, externally developed programs can be used to access the datafiles produced by the system and the computer used for the generation of analytical reports and certificates.

The system employs two floppy disk drives: the left- hand drive (LHD) is generally used for operating sys tem files and the right hand drive (RHD) is used for analytical programs (*.TAB files), data and text files.

The LHD is identified by name SY:, and the RHD by name DY1. Floppy disks are easily damaged, and the information stored on them cannot be retrieved if the disks are mishandled. Back-up copies should be made of all system disks every 6 months. When not in use, disks must be stored in the cases provided.

To use the computer, the system disk is placed in the

LHD. To activate (i.e., BOOT) the computer, all three switches at the main computer are switched off and on again. Alternatively, with all switches on, the instruc tion 173000G is typed at the Decwriter.

It is not possible to boot the system from the video terminal, although once activated, control can be passed to this terminal by typing:

R CHTERM [RI

This instruction can be used to transfer control back and forth between terminals as long as the computer is in the RT-11 operating system.

NOTE: In the following pages, all instructions to the computer are underlined. It is assumed that all commands are followed by a carriage return [R].

Once the computer has been booted, the system re quests the operator to input the date. Enter

XX-MON-YY where XX s day, MON = three letter abbreviation of the month and Y Y ^ last two digits of the current year, e.g. 17-AUG-87.

The time can also be entered in the format

HR:MN:SC where HR = hour, MN = minutes and SC = seconds, e.g. 14:30:20.

Formatting and Initializing a New Disk

Any new floppy disk to be used in the system must be formatted and initialized. To do this, the new disk is placed in the RHD and the system instructed:

R FORMAT

The system responds with a request for the location of the disk to be formatted by an asterick cursor (*), and operator answers:

DY1:

Since the operator may wish to re-format a disk which had been used previously, the computer asks for as surance that formatting is desired, i.e.,

ARE YOU SURE?

If sure, the answer Y is typed; if not sure, N or Ctrl-C

(press the control [Ctrl] key and depress the C). To double-check that an "old" disk has no files you wish to retain, the operator can check the disk directory by command DIRDY1: to obtain the directory listing.

When the disk has been formatted, the system returns to the routine cursor (*) and waits for further instruc tions. To exit from the FORMAT routine, the operator types Ctrl-C; the system returns to the RT-11 operat ing system which responses with the '.' type cursor at the left margin.

The disk is now initialized using one of several instruc tions:

EA 18-3

Traces - ICP-OES

INI/BADDY1:

INI7BAD7SEG:8DY1:

INI/BAD/VOLDY1:

INI/BADA^OL/SEGiS DY1:

The first is used to initialize a single density disk with no volume identifier; the second, also with no volume identity, is used for a double density disk. Inclusion of the VOL part of the command allows the operator to assign a volume number and owner identity to the disk. The SEG:8 permits a greater number of files to be stored on the double density disk. Including BAD into the instruction allows the system to seek out "bad" blocks (defects) on the disk and identify their location.

After the initialization command the system asks:

ARE YOU SURE?

and, after a Y response, carries out the procedure.

When initialization is complete, the system returns to the RT-11 operating system.

The Directory

After initialization, the operator can confirm the num ber of available storage "blocks" by typing

DIRorDIRDYl:

To obtain a directory of the system disk, the command is DIR SY:.

Alternate directory instructions include:

(2) Note files to be copied.

(3) Files may be copied one by one using the instruc tion:

COPY SY:XXXXXX.TYP DY1: where XXXXXX represents the 6-character al phanumeric name of the file, and TYP represents its type, i.e., RAW = raw data, DAT = ASCII data,

TAB = tables, etc. The operator must pay attention to the punctuation and spacing in the instruction, e.g. the period (.) between name and type (or ex tension).

The system then types out the name of the file and copies it in the direction indicated in the instruc tion. Reversing the SY: and DY1: addresses will reverse the direction of the copy.

(4) In order to copy many files of the same name, but different type, e.g., XXXXXX.RAW and

XXXXXX.DAT, a wild card instruction can be

used. An asterisk is used to represent the type or extension. For example,

COPY SY:XXXXXX.* DY1:

Similarly, to copy files of the same type, but dif ferent names, the wild card (*) can be used to represent the name. For example:

COPYSY:*.TYPDY1:

DIR/BADDY1:

DIR/VOLDY1:

DIR/BRIEF

The first searches out the bad blocks, the second gives volume number, owner as well as file inform atioa

The DIR/BRIEF command yields an abbreviated directory - useful for disks containing many files.

(5) All files, with the exception of systems files (type

= .SYS) can be copied with the use of a double wildcard, e.g.,

COPYSY:*.*DY1:

(6) The instructions given above assume that the name and type of the file will not be changed in the copy.

Files can be renamed in the copy command by the following instruction.

Disk Management: Copying, Deleting and Viewing

COPYING FILES

Copying files from the disk at drive S Y: to the a new initialized disk situated in drive DY1: is carried out as indicated below:

COPY SY:XXXXXX.TYP DY1:YYYYYY.TYP

A copy of a file can be made on the same disk in this fashion, e.g.,

COPY DY1:XXXXXX.TYP

DY1:YYYYYY.TYP

(1) Search the directory of the disk at SY:

DIR SY:

(7) The instruction given in (5) above, results in the indiscrimate copying of all files. In order to exer cise some control over the files to be transferred,

EA 18-4

Traces - ICP-OES

a special command can be appended to the copy instruction, i.e.,

COPYSY:*.*DY1:7Q

The name of each file to be copied appears at the terminal and the copy instruction accepted or ig nored on the basis of the Y or N response (a carriage return is equivalent to N).

(8) To copy systems files *.SYS, the instruction is:

COPY/SYS SY:*.SYS DY1: or

COPY/SYS SY:*.SYS DY1:7Q

To copy all files, including the systems files, as in the case of duplieating a systems disk, the instruc tion will be:

COPY/SYS SY:*.*DY1:

In the case of the former, all (non-system) files will be copied; in the latter each file will be listed and depending on the Y or [R] (no) responses, copied or not. Once all the files have been copied, the operator places the systems disk back into the

LHD and types

Ctrl-C

The system exits the PIP routine, returning to the system cursor Q.

DELETING FILES

In order to delete files from any disk, the disk name, as well as the name and type or appropriate wild card can be used, just as in the copy instructions. The /Q instruction can be used to avoid indiscriminate dele tion. Some of the command types are listed below (the address is S Y: but DY l: could also be used): or

COPY/SYS SY:*.* DY1:7Q

(9) In the case of duplicating a systems disk, it is important to copy over the BOOT routine. The following command sequence is used:

COPY/BOOT SY:DYMNSJ.SYS DY1:

COPY/BOOT DY1 :DYMNSJ.SYS DY1:

Another utility on the operating system disk allows for copying of large numbers of files, for example when archiving data files.

(1) A new disk is formatted and intialized (if re quired).

(2) The command is given R PIP, and the system responds with the asterisk cursor (*).

(3) The system disk is removed from the LHD and replaced with the disk to be copied (or from which files are to be copied). The new disk stays in the

RHD.

(4) The instruction is then given:

:*^SY:*.* or

DEL SY:XXXXXX.TYP

DEL SY:XXXXXX.* or DEL SY:XXXXXX.*7Q

DEL SY:*.TYP DEL SY:*.TYP7Q

DEL SY:*.*

DEL SY:*.*7Q

(safer with /Q)

In order to prevent all files from being erased, or to remind the operator what is being erased, the system types:

FILES DELETE

SY:XXXXXX.T^ and awaits a Y or [RI to complete the instruction.

RECOVERING A DELETED FILE

A deleted file can be recovered if, and only if, no further data have been written to disk since the deletion occurred.

Type the following commands to recover a file deleted from a disk in the right hand drive:

RDIR

DY l :7Q7B7O ... prints out the deleted file infor mation

Note the starting block number (76 used as an ex ample) for the particular file you are retrieving, thus:

EA 18-5

Traces - ICP-OES

B4.RPO 6 10-SEP-84 76

DY li/Q/B ... prints out deleted file information

Note the size (150 used as an example) of the file, thus:

DW:RAW 150 13-APR-84 104

CRTLC

R DUN

DY1: FILENAME.EXT s /CSTART BLOCK

NUMBER: FILE SIZE

CONTROL COMMANDS

In addition to the Ctrl-C command, which can be used to exit from any program and to return to the RT-11 monitor, there are a number of other Ctrl- commands which are useful.

Ctrl-O - stops generation of output and returns system to monitor or program initialization;

Ctrl-U - deletes a line of typing;

Ctrl-S - temporary halt to output, awaits restart;

The data file should now have been recovered.

TYPING FILES

Files written in ASCII can be viewed in "English" or the appropriate alphanumerics while the system is in the RT-11 operating system. These include such file- types as *.DAT, created by the JY48P software, and other files often created in the text editing mode, e.g.,

*.TXT, *.LST, *.BAK, etc. The file to be typed must be in the RHD to use the following series of instruc tions to work.

The instruction is:

TYPE XXXXXX.TYP

Ctrl-Q - output restart;

Ctrl-A - initialization of a subroutine, e.g., while in the spectrometer operation mode (program J Y will return the operator to the first line of the program,

CURRENT POSITION?

JY48P:AN?CC?DL?DR?EN?FP?PL?PR?RG?TB?

Ctrl-X - eliminates a string of instructions in the

RT-11 text editor.

If there is a long series of related files to be be typed, wild-cards can be used to avoid unecessary repetition of instructions:

For example, for a long series of data files with com mon alphanumerics, e.g., AB0001.DAT,

AB0002.DAT, AB0003.DAT, etc., the typed instruc tion would be:

TYPE AB*.DAT

If there is a series of files from AB0001.DAT to

AB0800.DAT, and one wished to print out only from

AB0500-AB0599, the instruction would read:

TYPEAB*5**.DAT

In some cases a code letter can be incorporated into a file name for convenience, i.e., B for blank, Q for quality control or S for standard. If all files of a specific code-type are required, then the instruction would read:

TYPE *S.DAT

and all .DAT files containing an "S" in the name will be typed.

The Analytical Method - Theoretical Aspects

The analytical signal for any spectroscope technique is generally made up of various contributions, i.e. the analyte signal, the background and matrix non-spectral interference and spectral interference. The latter con tribution is very important in a number of spectro- scopic techniques (i.e. optical emission, x-ray fluorescence and, to a lesser extent, mass spectrometry). The following discussion of inter ference corrections is centered around the problems observed in ICP-OES. The concepts, however, are applicable to all forms of spectral interference.

It is important to separate spectral interference from other forms of matrix interference. Spectral inter ference (line-overlap) results from the interaction of an interfering species with the analytical probe to produce a signal at the detector. This interaction is independent of the analyte identity and is only a result of analytical conditions. Non-spectral matrix interferences general ly result from a physical or chemical interaction be tween the analyte and the interferent or a physical or chemical interaction between the interferent and the analytical conditions which effectively changes the conditions and results in a different level of interaction between analyte and the probe. Thus the matrix ele ments may alter the number of analyte species which

EA18-6

Traces - ICP-OES

are in detectable form, e.g. upsetting the analyte ion/atom equilibrium and resulting in signal enhance ment or suppression. The matrix components may also cause physical flow problems, aerosol formation problems, etc.

If, for example, one examines the physical/chemical steps that occur in taking the analyte from the sample tube to the detection of radiation in an ICP experiment, it is possible to identify many sources of potential non-spectral interference.

(1) Transport of sample to nebulizer:

Whether one relies on the natural uptake rate of the solution or the use of a peristaltic pump, the vis cosity of the sample will have a bearing on the rate of analyte delivery to the plasma. Reduction in flow will generally result in signal depression.

(2) Nebulization of sample:

The rate of aerosol formation in the nebulizer will depend on the total salt content in the sample solution. Reduction of sample flow for high salt samples or nebulizer blockage due to precipitation will probably result in signal depression.

(3) Aerosol in the spray chamber:

The design of the spray chamber, the temperature of the spray chamber and the nature of the sample will determine the droplet size distribution in the aerosol. This in turn will determine the amount of sample which can be swept into the plasma.

bination (oxide formation) is generally minimized and ionization and excitation efficiencies are ex pected to be high. However, the presence of rela tively high concentrations of sample elements in the plasma can affect these efficiencies. The presence of large amounts of easily ionized species can upset the ion-atom equilibrium. The energy requirements for the processes may be high enough to lower plasma temperature and result in poorer efficiency of the processes. The additional excitations may contribute to the background.

A single element in a solution will produce a series of emission lines at different wavelengths. These result from the excitation of outershell (valence) electrons into unoccupied "upper" energy states (orbitals). The resulting atom/ion energy state is unstable, and the electrons lose the energy by returning to their original

"ground" states. The energy lost (emitted) will be equal to the difference between the upper and ground state energies and will appear as specific wavelengths of radiation. These in turn can be dispersed by an optical device (prism or grating) and the level of emis sion measured at a photo-sensitive detector such as a photomultiplier tube or diode array. The number of emission lines can be very large, especially for ele ments with relatively high atomic number, as there is generally a large number of accessible upper energy states available for the excitation.

The wavelength of the emitted light is related to the energy difference between states. The relative inten sities of the emitted lines (for a given, constant analyte concentration) is related to the energy state populations and transition probabilities). In the absence of spectral or non-spectral matrix effects, the intensity (I) of the emitted light for a defined analyte at a given wavelength will be directly proportional to the con centration of the analyte (C) in the sample.

(4) Processes in the Plasma:

- Vaporization of solvent;

- Vaporization of sample salts;

- Atomization of sample components;

- Ionization of sample atoms;

- Recombination of atoms to form transient

species;

- Excitation of atomic species;

- Excitation of ionic species;

- Excitation of molecular species;

- Oxide formation

The efficiency of these processes will depend on the flow rate of the sample through the instrument and the rate of delivery of sample vapor to the plasma, physical conditions of the plasma, e.g., temperature. Because of the very high tempera ture of the argon plasma compared to that achieved by a gas-oxidant flame in AAS, chemical recom

Furthermore in a system containing pure analyte, the signal (S) obtained from the detector will be linearly dependent on the intensity of the emitted light

S * k'(I) + B = k'k (C) 4- B = K(C) + B where B = background contribution from sample matrix, e.g. solvent, and from the instrument (stray light, electrical noise, etc.).

Since we use calibration curves to convert analytical signals to concentration, it is more convenient to rep resent C as a function of S, thus:

EA18-7

Traces - ICP-OES

where A l and AO are the appropriate slope and inter cept calibration coefficients respectively.

At high analyte concentrations, deviations from linearity can occur (due to detector effects, or due to nonspecific matrix effects caused by the analyte itself) and the calibration function is best described by a polynomial expression: linearly dependent on the concentration of the interfer ing element in the sample (CI).

SI - a(CI) 4- b

Substitution of this into the concentration expression above results in:

C s CA4-Al(a)(CI)4-Al(b)

~ CA 4- x(CI) 4- y or

C = A2(S2) + A1(S) + AO

C = A3 (S3) 4- A2 (S2) + A l (S) 4- AO

It can be shown that for all but the cases of direct overlap, the ability to separate the lines will depend on the band-width of the lines, the degree of separation, and the slit width of the exit slit (in the mask at the

Rowland circle of the spectrometer).

Quantifying Interferences

The contribution from the interferent will manifest itself as an "apparent" concentration of analyte higher than that expected (the true analyte concentration).

where x and y are defined as A l (a) and A l (b), the first and zero order interference constants respectively. In normal applications, only the first order constant is expected to have analytical significance.

In the case of multiple interferents, we can write:

In a complex system containing a large variety of emitter atoms and ions, there will be a large number of spectral lines. These lines have a very narrow natural band width, but do experience a degree of broadening

(Doppler effects, etc.). It is natural to expect that any given analyte line will have a relatively large number of lines due to concomitant elements nearby. It is also quite probable that a concomitant can have a line at exactly the same wavelength as that of the analyte

(direct overlap), or close enough to it to be virtually inseparable by the optical system of the spectrometer.

where xt and Cli represent the individual constants and concentrations of the i interfering elements.

Correcting Interferences

In the analysis of one or more elements and their interferents, the number of corrections that may need to be made can potentially be very large. Moreover, the presence of sequential interferences (A interferes on B which, in turn, interferes on C, etc.) and mutual interferences (A interferes on B, and B interferes on

A) will complicate the system. It is therefore impor tant to be able to select those interference corrections which will be significant for the determination of any analyte. This is especially important in cases where interference correction is done through the instrument software and where the number of permissible correc tions must be minimized (in order to be able to accom modate a maximum number of elements).

ST ~ SA 4- SI where ST ~ total signal, SA = analyte signal, SI = in terferent signal contribution.

C ~ Al(ST) -H AO = A1(SA 4- SI) -f AO

= Al(SA) + AO 4- A l (SI) = CA 4- A l (SI)

Interference coefficients are relatively simple to obtain by the following sequence of steps:

(1) Calibrate the spectrometer over the element

(analyte and interferent) concentration range of interest, using individual element standard solu tions. (Individual element standard solutions are preferred as mixed element standards can yield in unrecognized interferences in the original calibra tion.) where C = observed concentration and CA = analyte concentration.

The intensity contribution and therefore the signal contribution (SI) due to an interfering element at the analyte wavelength of interest will be directly and

(2) Nebulize intermediate and high concentration level individual element standard solutions.

(3) Convert intensity data to concentrations. The in terference coefficient may be calculated from the

EA 18-8

Traces - ICP-OES

slope of the curve of analyte concentration vs. interferent concentration. It is possible to use a single high level concentration solution and to estimate the first-order interference coefficient from the ratio

1

_ Apparent analyte concentration

Interferent concentration

The use of more than one solution will distinguish between true spectral interferences and back ground contributions. For example,

Depending on the value of CA, and the magnitudes of x l and x2, the final term in the correction may be significant, and should be included. Some software packages (e.g. that of the JY48P) do correct in se quence and rely on the order of entry of the correction coefficients (elements). Others use the original ob served concentrations only. Sequential interference corrections may need to be corrected off-line.

Mutual interferences are illustrated by the following set of equations (for the example of mutual inter ference by analytes A and B):

Interferent Cone.

blank lOOppmFe

1000 ppm Fe blank

100 ppm Fe

1000 ppm Fe

Apparent Analyte Cone.

0.0 ppm V

0.1 ppm V

0.2 ppm V

0.0 ppm Mo

0.1 ppm Mo

1.0 ppm Mo

The use of 100 ppm solely for the determination of an interference coefficient would predict coefficients of

0.001 for interference of Fe on Mo and V. However, use of both 100 and 1000 ppm Fe indicates that the effect of Fe on V is more background, whereas the effect on Mo is a true interference because the signal at Mo is proportional to the Fe concentration.

In the absence of sequential or mutual interferences, software (or manual) correction is straight forward.

The appropriate analyte and concentration data are obtained from the calibration curves and the interferent contribution calculated and subtracted from the ob served analyte concentration to give the "actual" analyte value.

In the case of sequential interferences, the correction is more complex. For example, consider a case where analyte A interferes on B (coefficient x 1) and B inter feres on C (coefficient x2). Therefore:

CA(obs) = CA(true)

CB(obs) = CB(true) * x l CA(true)

= CB(true) + xl CA(obs)

CC(obs) s CC(true) * x2 CB(true) and

CC(obs) = CC(true) -H x2{CB(obs) - xl CA(obs)}

CC(true) s CC(obs) - x2 CB(obs) * x l x2 CA(obs)

CA(obs) = CA(true) + x l CB(true)

CB(obs) = CB(true) * x2 CA(true)

This is a pair of linear equations:

CA(obs) = CA(true) + x l CB(true)

CB(obs) s x2 CA(true) + CB(true) with solutions

CA(true) ^

CA(obs)-x\CA(obs)

(\-x\x2)

and

CB(true) ~

CB(obs)-x2CB(obs)

(\-x\x2)

Where the term x 1x2 is very small, it can be ignored.

The system can also be handled by matrix algebra since the linear equations can be rewritten as:

[CO] = [A] [CT] where [CO] and [CT] are column matrices of observed and true concentrations respectively, and [A] is a square matrix of the type l xl

x2 l

To evaluate this matrix equation, one uses the inverse of the square matrix, i.e.,

[CO] = [A]x[CT]

[A-][CO] = [A-][A]x[CT]

= [CT]

In this example the inverse matrix [A-] is:

EA 18-9

Traces - ICP-OES l

l -xl

-x2 l

- xl x2 and the solution is found from the product: l -xl

CA(obs)

CA(true)

in exclusion is smaller than that obtained from in strumental precision, the correction can be ignored.

This is typically the case when either the magnitude of the coefficient is very small, or the concentration of interferent is very low. Corrections due to sequential or mutual interferences are generally found to be very low and are almost always ignored.

l

l-xlx2

-x2 l CB(obs)

CA(obs)-xl CB(obs)

x4, x5 x6, x?

x8

CB(true)

The advantage of the matrix approach becomes ap parent in highly complex systems, such as the follow ing theoretical example:

Five elements, A, B, G, D, E

Thus

A interferes on B, C, E

B interferes on D, E

C interferes on A, D

D interferes on C

E interferes on A, B, C coefficients x l x2 x3 x9,xlO, xll

CA(obs)

CB(obs)

CC(obs)

CD(obs)

CE(obs)

CA(true) * x6 CC(true) -i- x9 CE(true)

CB(true) + x l CA(true) * x 10 CE(true)

CC(true) * x2 CA(true)

*xllCE(true)

CD(true)-h x4 CB(true) +

CE(true) -i- x3 CA(true) + x8 CD(true) x7 CC(true) x5 CB(true) which in matrix form is

CA(obs)

CB(obs)

CC(obs)

CD(obs)

CE(obs)

l 0 x6 0 x9 xl 1 0 0 xlO

x2 0 1 *8 xll

0 x4 xl l 0

x3 x5 0 0 1

=

CA(true)

CB(true)

CC(true)

CD(true)

CE(true}

A programmable calculator or computer program can be used to solve the system. Uncorrected concentra tion data from the spectrometer are corrected off-line before reporting.

Although it is possible to correct for all of the possible interferences in a system, including the sequential and mutual corrections outlined above, it may not always be convenient or even significant to do so. If omission of a particular correction factor will not appreciably alter the final concentration and/or the error inherent

To Correct or Not to Correct

The relative standard deviation contributed to a result by the omission of a particular interference correction can be written as:

RSD = i Ci

CA(true)

i.e., the ratio of the interference contribution to the

"actual" or "true" analyte concentration.

It is possible to predefine a value of RSD, related to expected experimental precision, below which the in terference is considered to be insignificant. Thus, for example, if we define significance at RSD = 0.01 (19&) then for a coefficient of x; = 0.01 we can write:

00.1 =

0.01

CA(true)

The limiting ratio of interferent to analyte concentra tion is therefore 1. Conversely if the analyte con centration is expected to be more than the concentration of the interferent, the interferent can be ignored.

For example: Vanadium shows an ICP interference on aluminum with coefficient 0.15. Under normal cir cumstances this is a relatively high value for an inter ference coefficient and would not be ignored.

From the RSD criterion, we can estimate limiting concentration ratios for a series of RSD values.

RSD= l

0.5

0.25

0.1

0.05

0.025

0.01

C(V)7C(AL)~ 6.7

3.3

1.7

0.7

0.3

0.2

0.1

For levels of aluminum in silicate rocks of S-15%, and acceptable precision levels of 5*7c (RSD s 0.05), the concentration of vanadium would need to be between l .5 and 4.59fc for the vanadium interference to have any significant contribution. As this concentration is not likely to occur in most silicate rocks, the vanadium

EA18-10

Traces - ICP-OES

correction can be ignored in the determination of Al in and the error, [E+], will be

Oil 1^'lt'OC

Interferences and the Determination Limit

[E+] = [CA(solM - [CA(sol)]

= A1(0.05SA)*0.005 CM

The significance of a potential interference can also be evaluated in terms of: (1) the determination limit or (2) practical detection limit for an analyte. Detection limits

(instrumental) are generally based on the noise ob served when a large number of replicate integrations are obtained during a single nebulization and represent instrument stability over a relatively short period of time. On the other hand, the determination limit, a much more realistic estimate of the sensitivity of the method, is based on 3 times the standard deviation obtained for replicate determinations of a low-level sample and may be degraded further by allowance for matrix effects.

Similarly the lower concentration limit, [C A(sol)-] can be evaluated and the error [E-] defined as

[E-] s [CA(sol)-] - [CA(sol)]

* -A1(0.05SA)-0.05CM

The practical detection limit of the analyte, PDL, can be defined as twice the error in CA(sol), or more simply:

PDL ^ [E+] - [E-]

33 [E+l + IfE-]!

= 0.1(A1)(SA)*0.1(CM)

As emission intensities are found to reproduce within

S-10% RSD from one nebulization to the next or from one day to the next a "practical detection limit" (PDL) can be defined as 1 096 (RSD) blank/background inten sity, i.e., approximately 2 times the error equivalent to a 596 RSD in intensity. The determination limit

(generally used in quoting analytical capabilities) and the PDL are often found to be approximately equivalent.

In the absence of any interference (CMsQ). the analyte contribution to the practical detection limit of the analyte, PDL(A) can be defined as

0.1(A1)(SA)

In the absence of analyte, the matrix contribution to the practical determination limit of the analyte, PDL(M), can be defined as

PDL(M) s 0.1 (CM) Since the acceptability criteria for a particular result can be expressed in terms of the determination limit

(or PDL), we must account for matrix corrections which will affect or contribute to the determination limit.

We can write:

CA(sol) - A l S A * AO * CM(sol)

Error in the determination of the matrix interferents and in the subsequent contribution of this error to the error in the observed analyte concentration will affect the overall PDL or determination limit of the method.

Example:

Consider niobium in a matrix containing 596 Fe, 496

Al, 496 Ca, 296 Mg and Q.2% Ti.

CA(sol) = solution apparent analyte concentration; A l ,

AO = calibration coefficients; S A s signal contribution from analyte; CM(sol) = concentration contribution to analyte from matrix =SUM(xi Ci)). Rock-equivalent concentrations are obtained by

CA(rock) = CA(sol) X DF

Calibration curve coefficient (Al) for Nb ^ 5.6"6

Calibration curve coefficient (AO) for Nb = -0.5

Measured interference coefficients, xi? for matrix ele ments on Nb: where DF - dilution factor (normally DF= 1 00 for 0.5 g sample in final volume of 50 ml).

If we accept a realistic error of 59fc RSD in the analyte and interferent intensities, then the upper concentra tion limit in the apparent solution concentration,

[CA(solHL will be equal to x(Fe) = 0.000027

x(Al) = 0.00022

x(Ca) = 0.000030

x(Mg) = 0.00012

x(Ti) s 0.00018

Niobium blank intensity ^ 90000 counts

[CA(sol)*] s 0.05)SA * AO

0.05)CM

EA18-11

Traces - ICP-OES

PDL(A) = 0.1 (90000)(5.6 ^

= 0.05 ppm

= 5 ppm-rock

From the interferent element concentrations expected in solution, Q, i.e., C(Fe) = 500 ppm, C(A1) s 400 ppm,

C(Ca) = 400 ppm, C(Mg) = 200 ppm, C(Ti) = 20 ppm, we can calculate the total matrix contribution, CM

(where CM ^ SUM(x1Ci).

Contributions:

Consider an ICP-OES result at the Nb channel =

280000 counts

Apparent Nb solution concentration = 1.07 ppm.

Since we do not know the true analyte concentration, we assume that it is approximately equal to the ap parent concentration. We can therefore estimate ob served C/CA values and compare them to the the limiting interferent concentrations: x(Fe)C(Fe) x(Al)C(Al) x(Ca)C(Ca) x(Mg)C(Mg) x(Ti)C(Ti)

CM

0.0135 ppm solution

0.0880 ppm

0.0120 ppm

0.0240 ppm

0.0036 ppm

0.1411 ppm solution

14 ppm-rock

Fe

Al

Ca

Mg

Ti

Ci/CA

Limiting Observed

370

45

330

85

55

470

370

370

190

20

Correction

yes yes yes yes ignore and

PDL(M) = 0.1 (CM)

= 0.0141 ppm

= 1.4 ppm-rock

We can examine the effect of ignoring interferences in greater detail. For example, if we correct the result for all five interferences:

PDL

= 0.05 + 0.0141

~ 0.06-0.07 ppm

= 6-7 ppm-rock.

We can use the expression derived previously to deter mine the relative importance of the matrix corrections, i.e.,

[Nb] apparent

[Nb] corrected.

= 1.07 ppm solution

= 1.07-0.14

= 0.93 ppm solution.

Thus, the contribution of error in the determination of matrix components results in an effective PDL of

-6-7 ppm-rock in the matrix of interest.

The error in ignoring the total matrix contribution would be 159fc RSD which is greater than would be allowed by conventional acceptability criteria.

Ignoring the contribution from Ti alone would result in a concentration of 0.93(±0.36) ppm with an error of

Q.4%, i.e., less than l 9fc RSD as predicted by the critical ratio method.

RSD = and if we accept a value of RSD s 1 9fc as the limiting error contribution we can establish critical ratios of

Fe

Al

Ca

Mg

Ti for the various interferents.

Interferent Critical Cj/CA

370

45

330

85

55

Interferent Ci/CA

Similarly, ignoring the contribution from Mg, Ca, Al and Fe would result in errors of 2.59fc, 1.39fc, 9.5Ve and

1.5^o RSD respectively.

In some cases, the observed analyte concentration may not be a suitable approximation of the true analyte concentration and the limiting ratio method will not be suitable to judge the need for a correction. Another approach can be considered. It is possible to assign an arbitrary "cut-off point below which interferences can be ignored. As a first approximation, if x, Q 0.5 PDL(A) the contribution can be ignored with minor contribu tion to total analytical error.

EA18-12

Traces - ICP-OES

In the Nb example, 0.5(PDL(A)) = 0.025 and all but the Al could be ignored. Mg is a borderline case and the correction is included. The subsequent, partially corrected, Nb value would be 0.956 ppm (total error ~

3.19b RSD). To reduce the error the "cut-off value could be set at an even lower fraction of the analyte

PDL(A) if desired.

Summary

In the determination of elements by ICP-OES, the potential interferences can be summarized into four principal categories:

From this limiting criterion, it is possible to estimate critical interferent concentrations for any analyte based on the PDL(A) and interference coefficient:

PDL-A

2

PDL-A

Major Interferent

Major Interferent

Trace Interferent

Trace Interferent

Major Analyte

Trace Analyte

Major Analyte

Trace Analyte

Given the relative concentration levels of most major and trace elements in geological samples, we are generally only interested in the interference of trace elements by major elements. To a limited extent we are also concerned with Trace-Trace interferences where the coefficient is very high.

Thus, for the Nb example, PDL(A) = 0.05 ppm solu tion (5 ppm-rock) the critical concentrations for a series of potential interferences are listed below:

Element

Ti

Ni

V

Cu

Nd

Fe

Mg

Ca

Al

0.000027

0.00012

0.000030

0.00022

0.00018

0.00018

0.10

0.00021

0.0012

Ci(crit)-sol Ci(crit)-rock

925 ppm

210

830

114

140

140

0.25

120

21

9.25 Vo

2.1

8.3

1.1

1.4

1.4

25 ppm

1.2 Ve

2100 ppm

(G.2%)

It is important, however, to evaluate all the interferen ces critically for any given matrix type. The samples arriving at the Laboratories are varied and complex.

What may be acceptable for a relatively barren silicate rock may not be acceptable for an ore or a mineral.

Each case must be examined and where necessary, additional interference corrections used. The use of concentration ratios or PDL criteria become very im portant. A complete list of interference coefficients is on file.

EA18-13

Traces - ICP-OES

THE TRACE 2 (T2) PACKAGE

INDUCTIVELY COUPLED PLASMA

OPTICAL EMISSION SPECTROSCOPY

Introduction:

Given the extensive analytical potential and flexibility of ICP-OES, a wide variety of elements can be deter mined using this instrument. At the Geoscience

Laboratories we have grouped many geologically im portant elements into the Trace 2 package. Usually these elements are at the ppm to ppb levels, but in some samples, from specific geological environments, they can be much higher.

Geologists use trace elements to monitor the genesis of rocks and evaluate process that affect rocks after formation. Elements such as Ni, Se, Y, Nb, REE (Ce,

La, Nd), Sr and V usually occur at trace levels in rocks and are used as discriminators to determine origin, tectonic setting or degree of alteration. Other elements such as W, Co, Be, Cu and Zn are used more specifi cally to assess rocks for ore potential.

An inductively coupled plasma source emission spectrometer (JY48+) is used to perform elemental analysis on samples which have been decomposed using a wet chemical (acid digestion) procedure.

The TRACE 2 (T2) package elements, their deter mination limits, optimum range, and precision are listed in Table OES l.

Safety Advisory:

UNTRAINED OPERATORS SHOULD NOT AT

TEMPT TO START OR USE THE EQUIPMENT

WITHOUT PRIOR TRAINING.

There are five types of hazard associated with this apparatus: electrical, radiation, high temperature, noxious gases and physical hazard.

1. Electrical hazards can be encountered in three areas: the plasma power generator, the plasma torch compartment the spectrometer tank com partment and associated with various peripheral equipment like pumps, autosamplers and other special apparatus which may be set up to facilitate the analysis. The power generator is of particular concern because high voltages are retained on capacitors even when the power has been com-

TABLE OESL

Element Determination Optimum Precision*

Limit Range

(ppm) (ppm)

Se

V

Y

Zn

Be

Co

Cu

Mo

Ni

10

5

2

5

5

5

5

5

1

1-

5-

5-

10-

5-

2-

5-

5-

5-

100

1000

1000

4000

1000

100

500

1000

1000

Options

Ce

La

Nb

Nd

Sr

W

35

25

5

5

5

50

35-

5-

5-

25-

5-

50-

3000

2000

500

2500

10000

500

70

10

10

50

15

100

* Precision is quoted as the 95^c Confidence Limit

(ppm) for a value at l OX the determination limit.

pletely disconnected from the instrument. DO

NOT ATTEMPT TO GAIN ENTRY TO THE

POWER GENERATOR WITHOUT

PROPER TRAINING. Other areas are protected

by interlocks. DO NOT OVERRIDE INTER-

LOCKS.

2. Two types of electromagnetic radiation are of con cern: radio frequency and ultraviolet. Radio fre quency radiation has no known teratogenic effect, but can cause somatic damage and could affect eyesight by the absorption of radiation by the fluids of the eye causing a damaging increase in pressure within the eye. It is important to ensure that the plasma torch compartment is well shielded by inspecting the viewing and access potts.

Operate the instrument with the door closed to make sure that the RF radiation cannot escape. The power transfer cable should be inspected to ensure that it is a tight fit and that no corrosion has occurred at the connecting points on both the in-

5

10

20

10

10

10

30

10

1

EA18-14

Traces - ICP-OES stmment and on the plasma power generator (EX

PERIENCED TECHNOLOGISTS ONLY!).

Ultraviolet radiation is emitted by the plasma.

THE PLASMA SHOULD NEVER BE

VIEWED DIRECTLY. Even light from the plas

ma which is reflected from the source housing has a high level of ultraviolet radiation and direct viewing of this light should be avoided.

3. The plasma operates at extremely high tempera tures. The interlocks protecting the plasma should never be overridden. The torch itself can become quite hot and sufficient time should be allowed, after extinguishing the plasma before disassem bling the torch for routine inspection and cleaning.

The plasma power tube runs at about 3000C and should be cooled at least 20 minutes prior to shut ting the generator down at the end of the day or for inspection.

4. The exhaust gases from both the plasma compart ment and the generator can be a source of both heat and noxious fumes. Both beryllium and ozone are emitted from the generator. The exhaust from the plasma can contain toxic materials from the samples. THE EXHAUST FROM BOTH THE

GENERATOR AND THE PLASMA MUST

BE VENTED FROM THE ROOM AT ALL

TIMES. Remember to check the damper posi

tions to ensure that sufficient exhaust velocity is being maintained.

5. Physical hazards can be encountered when moving or working around bottled gases. Exercise appropriate caution: secure the cylinders, use the valve caps, use properly designed carts to move the cylinders and use appropriate footwear.

Method:

The method consists of the following techniques:

1. Dissolution of sample using acid digestion techni que.

2. Measurement of analyte concentration using the

Jobin-Yvon JY48P ICP-OES instrument.

3. Calculation of results.

- Gilson Minipuls-3 peristaltic pump

- Gilson autosampler

Reagents:

- Argon gas

- Nitric acid, HNO3,

Procedures:

1. JY48P Start-up

1.1. l .2. l .3.

Turn on the RF generator. A few seconds later, a "click" will be heard and that signifies the cooling water is being turned ON.

Switch the reflected power switch to 'OFF'.

Turn on the argon gas supply.

l .4. Connect the nebulizer tubing to the peristaltic pump and secure peristaltic pump tubing clamps.

1.5. Switch on Plasma, Nebulizer and Auxiliary l .6.

1.7.

1.8. Switch the nebulizer gas flow off.

1.9. Set tuning control in the front panel to

MANUAL and obtain a reading of ap proximately 164 by adjusting with the white buttons at the top left comer.

1.10. gas flows by depressing buttons marked ac cordingly on the ICP front panel.

Nebulizer water and allow the system to purge argon for two or three minutes.

Set

- Plasma gas at 14 1pm

- Nebulizer gas at 0.75 1pm

- Auxiliary gas at 0.52 1pm

- Uptake rate of nebulization at l ml/min

(pump setting @ 19)

RESET THE TUNING CONTROL TO

AUTO.

Apparatus:

- Jobin-Yvon JY48P simultaneous ICP source emission spectrometer

- Meinhard TR-30-C3 nebulizer

1.11. Press the RF power burton of the torch box

ON.

1.12. Dial the forward power up to about l k W and ignite plasma by briefly pressing the ICP IG

NITE button. STOP as soon as the plasma has ignited. After ignition is achieved, immedi-

EA18-15

Traces - ICP-OES

ately switch the NEBULIZER gas flow on and increase the incident power to 1.8 k W.

Switch the reflected power switch at the RF generator to the right (on).

1.13. If unsuccessful in igniting the plasma, purge the system with argon for two to three minutes and repeat the procedures starting at step 1.8.

l. 14. If necessary, adjust reflected power to a min imum with the fine knob located on the right end-plate of the box.

1.15. BE PREPARED TO IMMEDIATELY

PRESS RF OFF IN THE EVENT THE

PLASMA EXTINGUISHES!

1.16. Let the system warm up for at least 45 minutes or until the inside chamber temperature remains fairly constant (monitored by a ther mometer at the upper part of the front panel of the plasma box).

NOTE: Refer to diagram 13 in JY48P spectroanalyzer user's manual

3. JY48P Digital PDP-11/23 Computer Start-up

3.1. Insert JY48P system disk into left system drive (S Y: or DYO:) and a data disk into the right drive (DY l:)

3.2. Boot up the system at the keyboard terminal by typing 173000G and press return. Addi tional information on the computer system is given in the Overview Section - page EA 18-1 toEA18-13.

3.3. Enter Date (e.g. 12-Sep-89) and press

RETURN.

3.4. Verify the date entered by Y (yes) or N (no).

3.5. Enter the Time (08:32:56) and press

RETURN.

3.6. Verify the time entered by Y (yes) or N (no).

3.7. A "Dot" prompt will appear and that signifies the system is ready.

2. JY48P Shutdown procedure

2.1. Dial the forward POWER slowly down to

2.7.

2.8. approximately 6 on the dial.

2.2. Press the RF OFF button on the front panel.

2.3. Depress the plasma and auxiliary gas flows

(off).

2.4. Dial the forward POWER down to 0.

2.5. Flush the system by nebulizing deionized water for a few minutes.

2.6. Switch off the Nebulizer gas flow.

Let the RF generator to cool off for at least 15 minutes before switching the main supply off.

Turn off the peristaltic pump.

2.9. Disengage the clamps of the peristaltic pump and release the tubing.

2.10. Close the valve of the Argon gas supply.

4. Routine Analysis

NOTE:Ensure that all plasma conditions, resistance settings for all analytes of interest and the parameter settings in the Table are correct.

4.1. At the "dot" (.) prompt, enter R JY48P and press return.

4.2. Enter current position (read from slit position dial on the upper pan of the JY48P).

4.3. Enter Peak position. The "JY Main Menu" will now appear.

4.4. Enter "AN" for analysis.

4.5. Enter Table file name in the format of

XXXXXX where X represents alphanumeri- cal characters. ( A *.TAB file is created pre viously and has to be on the data disk for execution. See Section 7 for Parameter-Table set-up).

4.6. Enter CHG to ensure the analytical parameters are right. Changes can be entered at this time.

EA18-16

Traces - ICP-OES

4.7.

Manual Analysis: enter ANA for sample

analysis.

Example of typical routine run protocols (maximum number of samples for a batch of samples is 50): a. Warm-up solutions: Do not enter sample ID's for these samples to avoid saving data on disk. This can be achieved by simply hitting the RETURN key.

b. Recalibration standards (if restandardization is necessary and requested as the mode). These ID should begin with CAL###, where ### must cor respond to a LOW or HIGH standard as defined in the Table routine.

c. Rinse(107cHNO3). (Normally sample ID is not entered so as not to store data on disk).

d. Blank solution (l09fcHNO3). The sample ID has to begin with the letters BLK if blank subtraction is requested as the mode. (All samples following this

BLK will be automatically blank corrected).

e. Check solution (normally its ID is CK###).

f. 10-15 routine samples.

g. Check solution.

h. 10-15 routine samples (if spaces are available).

i. Check solution (if spaces are available).

j. .... (check, samples, check, samples, check and so on).

k. l or 2 reference standards and/or calibration stand ards.

4.7.1. Enter CNCODFBSSD (see JY48P software manual for explanations).

4.7.2. Enter XX where XX = # of samples (maxi mum # of samples is 50).

4.7.3. Enter the sample identifiers according to the protocol required by the data reduction software programmes. (Refer to the JY48P software manual).

4.7.4. When all sample ID's have been entered and at the "ready?" prompt, put the nebulizer tubing into the first sample tube. Wait until the sample has been taken into the plasma and allow 20-30 seconds for equilibrium/stabiliza tion, hit the RETURN key to begin the sample analysis.

4.7.5. When the sample analysis is finished, transfer the tubing (probe) to the rinse bottle. If the measurement is OK, enter Y. Otherwise enter

N for a not OK answer and the prompt will prompt for "REMEASUREMENT". Simply hit RETURN and analysis will continue.

4.7.6. At each "ready?" prompt repeat the process until all of the samples have been analysed and the prompt "ANA?CHG?END?" appears.

4.7.7. Enter END to end the analysis and return to beginning of JY48P routine.

4.8. Use of the Autosampler

Enter ANA for sample analysis and include AS in analytical command string.

a. Connect rinse solution tubing to peristaltic pump and secure clamps.

b. Enter command string (include AS) and press return.

c. Enter l at the prompt "# samples?".

d. Hit return at the prompt" l sample name?".

e. Enter E to exit at the prompt "Ready?". This will

'move' the light from ACQ to RESET on the upper front panel of the JY48P.

f. Check the adjustment of the pipette tube of the autosampler by slowly lowering it into the snap tube with the Auto-Adj switch. Check both the depth to which the probe reaches in the Snap tube and the centering of the probe as it enters the Snap tube.

g. At the base of the autosampler unit, set:

- power switch to ON

- function switch to EXT

- autosampler pipetting unit

- Auto-Adj switch to AUTO (up) position

- Op-Load switch to OP (up) position h. Enter ANA at" ANA?CHG?END?".

i. Enter CNCODFASBS (see JY48P software manual for explanations).

EA18-17

Traces - ICP-OES

j. Enter XX where XX = # of samples. MAXIMUM

# of samples is 50.

k. At the prompt "X sample name?" where X s sequence of sample number, start entering the sample identifiers. (Refer to the JY48P software manual).

1. At the last sample ID entry, ensure all of the above have been done before hitting RETURN. At this point, the pipet of the autosampler will go into the first sample tube and the analysis will begin.

5. System Calibration

5. l. Set/check all plasma conditions and resistance settings for the analytes of interest.

5.2. Run the calibration standards:

5.3. In the JY48P mode, enter RG for regression routine.

5.4. The computer first asks for a table filename

(there is no need to answer this question; just hit RETURN) and then requests a standards filename. See JY48P software manual for details.

5.5. Enter INT at the RG mode.

5.6.

5.7.

- blank (deionied water)

- blank (10*26 HNO3)

- calibration standards

(see additional notes)

- enter the standard numbers (l to 20) and the corresponding data filenames.

- when all standard files have been entered and "standard #?" prompt is displayed hit

RETURN to end intensity entry and the computer will return to the RG mode.

Enter CON:

- channel by channel, enter the standards numbers (defined at 5.5) and the corres ponding concentrations (units of concen trations used here determine units output in analysis).

Enter CAL at RG mode and l at the prompt

"DEGREE?" for first order curve.

5.8. Enter N (RETURN) for normal weighing

(equal weighing of all points minimizes ab solute error).

5.9. Enter Y (yes) to save curve in memory or N

(no) not to save curve.

5.10. Enter VER or MOD for verify or modify the calibration results (see JY48P software manual for further explanation). At the end of

VER or MOD mode, make sure to CAL

(calibrate) and accept (Y for yes) the calibra tion curves.

5.11. Repeat steps 7-8 until calibration is complete.

5.12. When all calibration curves have been ac cepted, enter EN to exit the RG routine.

5.13. Enter Y (yes) to save the standard file. If this is not done, all files will be lost and you must repeat the entire calibration process.

5.14. Enter XXXXXX to store the standards file on disk for later use (XXXXXX = alphanumeri- cal characters for the standards filename).

5.15. Enter Y (yes) to save the recent changes (up dated curves) in the Table on disk.

5.16. Enter XXXXXX (the name) to store Table parameters on disk for later use. If no name is given, parameters are not saved on disk, but are still retained in memory.

NOTE: In the Table file, a low and a high standard are

defined for each analyte of interest under

CAL### naming convention (e.g. CAL002,

CAL004). The standards on the following page are used.

When system restandardization is required, the option

CA has to be included in the AN (analysis) mode.

With the same plasma source conditions, the gains and offsets associated with the restandardization calcula tions should change by less than 109fc (generally within

5^c). However, if the change is greater than 109fc, the following operating parameters should be considered and checked: nebulizer clogging, changing plasma gas flows, contamination of standards, incorrect measure ment conditions (integration time etc.), peristaltic pump problems, incorrect PMT resistance settings, exhaust fan problems, incorrect slit position, misiden- tified calibration standards, uncleaned torch, etc.

EA18-18

Traces - ICP-OES

Element Low st d # Cone. Highstd Cone. 8. Report Generation

Tungsten

Beryllium

Cobalt

Nickel

Copper

Vanadium

Strontium

Scandium

Yttrium

Molybdenum

Zinc

Barium

Chromium

Niobium

Zirconium

Cerium

Magnesium

Phosphorus

Iron

Calcium

Aluminum

Titanium

Manganese

Tantalum

Lead

Lanthanum

Neodymium

2,0

2,0

2,0

2,0

2,0

2,0

2,0

2,0

2,0

2,0

2,0

2,0

2,0

2,0

2,0

2,0

2,0

2,0

2,0

2,0

2,0

2,0

2,0

1,0

2,0

2,0

2,0

8,10

3,0.1

4,10

4,10

4,10

4,10

4,10

4,1

4,1

4,1

5,10

5,10

5,10

5,1

5,1

5,1

6,500

6,50

7,1000

7,1000

7,80

7,80

10,10

9,10

9,10

9,10

8.1.

8.2.

NOTE: If semi-quantitative analysis is required to be

To report results from the calcium and magnesium package (water samples), run the analysis under

CAMG.TAB to generate the Ca and Mg concentra tions on the data disk. Then use the program

"WATER.SAV" to generate the report.

9. Digital PDP-11/23 Computer Shutdown

9.1.

For the T2 package, refer to details listed in

"GENERATION OF T-2 FINAL REPORT", page EA 18-21.

For TSPA package, refer to details listed in

"GENERATION OF THE TSPA CERTIFI

CATES AND/OR WORKSHEETS", page

EA 18-25.

reported in ppm instead of in percentage use the program JANET in place of WORCER.

Exit the JY48P mode by entering EN at the prompt JY48P routine. A 'Dot' prompt will appear.

7. Parameter-Table Set-up

7. l. Enter TB at the "JY main menu".

9.2. Remove both the JY48P system and the 'Data collection' disks from the two drives. Leave the doors open and the switches at ON posi tions.

To monitor the drift of the instrument, check solutions are usually run every 10-15 samples within a batch analysis. When a particular analyte drift is greater than

3 times the standard deviation of its statistical data, restandardization of the system is necessary or the above mentioned operating parameters should be checked.

7.2. When setting new parameters, simply hit

RETURN at the next prompt. Otherwise, enter the Table filename of the routine (stored on system or data disk as *****.TAB).

7.3. For daily routine T2 and TSPA packages, see the additional notes for the parameters that are currently used in this laboratory. Details of setting up these parameters are listed in the

JY48P software manual (ref. 12.0).

9.3. Press the button "BREAK" on the keyboard terminal.

Quality Control:

The overall quality of analytical data for the Trace 2 method is as quoted in Table OES1. This has been devised from the Laboratories blind duplicate program over a 5 year period.

Quality control for the techniques which comprise the method are described below.

Preparation of check solution: After each batch of

samples has been analysed, the remaining T2 rock solutions are collected.

When a sufficient volume (approximately 3 to 4 liters) has been collected, the solution is filtered and becomes a check solution for the instrument quality control.

The following is the list of the concentration means and the associated standard deviations of that present solu tion.

EA18-19

Traces - ICP-OES

Concentration Mean (ppm) Standard Deviation

Ni

Se

Sr

V

Be

Co

Cu

Mo

Y

Zn

0.89

36

99

105

24

303

199

22

102

-

0.052

0.51

-

1.33

1.15

0.47

3.06

1.95

0.34

1.64

The above results are calculated on 60 observations, with the exception of Be which is based on 26 obser vations.

Aliquots of the check solution are analysed with each batch of samples using the routine run protocol and the results are compared against those in the above table.

The number of standard deviations of the check solu tion away from the mean is calculated using the for mula:

(observed check - solution cone. - mean)/ s.d.

If the analytical error is normally distributed, 95 9fc of the results should be between 2 times s.d. When one or more analytes in the check solutions fall outside 3 times s.d., the operating parameters outlined in section

5 should be checked and restandarizationof the system is necessary.

If reagent blanks are less than the method determina tion limits, no further action is necessary. However, if one or more elements in the reagent blanks are greater than the instrument detection limits, data will then be submitted to the Supervisor or ICP Spectroscopist for further decision.

Numerous in-house (MRB) and international SRMs of varying geological compositions including GA,

BHVO1, MRG-1, SY2, SY3, NIMG, NIMD, NIMP,

NIML, NIMM, NIMS, GSS1-8 etc., are used to monitor accuracy and precision within the ICP-OES laboratory.

Productivity:

A technician should be able to complete 150 samples per day.

Additional Notes:

l. The computer manufacturer recommends that the system boot be through the keyboard terminal rather than through the system switches. General ly, the system is left with the two drive doors open and all the switches at the ON (UP) positions.

2. To minimize the interelement spectral interferen ces, a series of calibration standards is prepared from a serial dilution of stock standards. There are six stock calibration standards prepared for the daily routine analysis.

a. 100W 100 ppm W b. A 100 ppm Co, Ni, V, Cu, Sr l O ppm Se, Y, Mo c. B 100 ppm Zn, Ba, Cr lOppmNb, Be, Zr, Ce d. MN (minor)500 ppm Mg

50 ppm P e. MJ (major) 1000 ppm Fe, AL, Ca

80 ppm Mn, Ti f. C g. Ta

100ppmPb,La,Nd

100 ppm Ta

3. To maintain optimal analytical precision, the fol lowing procedures are recommended: a. clean the torch.

b. clean and unclog nebulizer.

c. analyte profiles (top and bottom) should be check ed periodically.

d. ensure the "wick" (for guiding the drain) in the spray chamber is in the correct position. Small fluctuations in the pressure within the spray cham ber substantially alter the emitted signal. (The drain bucket should be emptied periodically - never allow it to get more than half full).

e. gas lines to the torch and nebulizer should be clamped with snapper hose clamps (critical for good RSD).

f. periodically monitor the gas flow rate to ensure constant flow.

g. change the sample delivery tubings periodically and check the uptake rate (l ml/min distilled water).

h. there should be no cracks in any of the rotometers, gas lines (located inside the torch box, left hand side).

EA18-20

Traces - ICP-OES

APPENDIX A

GENERATION OF T-2 FINAL REPORTS

Introduction:

The T2FORM programs produce the final T-2 job report forms. To do this, three programs must be run in succession; the programs T2CON, JOBFIL and T2FORM. The following describes how to use these programs.

The program T2CON creates a concentration file containing T-2 concentration data on the system disk. A maximum of 150 samples can be processed at one time.

The user runs the program JOBFIL to create two files: a file containing the sample identifiers and a file containing the geologists' names and job numbers. A maximum of 150 samples and 10 jobs can be processed at one time.

The program T2FORM prints out a signature-ready T2 form. All data are rounded to the nearest l ppm (rock).

Any results less than the method determination limit are reported as the "minus" of the determination limit. Any results greater than or equal to one percent are reported to the nearest decimal with a 'P' appended to them.

(1) Insert the system disk forT2 report forms into drive 1.

NOTE:

The T2 Report programs must be run on system disks that have been formatted to maximize disk space. Use ONLY these specially formatted disks to run the T2 programs.

(2) Insert the disk that has the JY*.DAT files into DY1: drive. In this report, the symbol **' denotes a user- entered filename.

(3) Check the directory of the system disk. The names of the two files *.DAT and *.CON that are generated by the T2CON and JOBFIL programs have to be unique. Note: If necessary delete any unnecessary *.DAT and *.CON files to maximize disk space on the system disk. The file TTIO.DAT should never be deleted.

(4) Enter R T2CON- For details see T2CON.

(5) Enter R JOBFIL- For details see JOBFIL.

(6) Enter R T2FORM- For details see T2FORM.

Details:

T2CON: The T2CON program was written to store concentration data for the T2 elements determined by the

JY-48P ICP.OES. The program will ask for the concentration filename in which the concentration data are going to be stored. This file will appear on the system disk as *.CON.

The program will now ask the user to enter

(1) The J Y file name to start

(2) The JY file name to end the sequence.

(3) The duplicates and/or SRM filenames if there are any.

EA18-21

Traces - ICP-OES

NOTE: To ensure the proper execution of the T2FORM program, the sample identifiers for the JY data must have the form LLNNN.DAT where LL are letters and NNN are numbers. When a batch of samples are run on the JY, the sample identifiers are numbered in sequence; e.g.PW125 to PW175 inclusive for a batch of 51 samples. A batch of samples is a group of samples from one to ten different jobs.

It is important that NO duplicates are allowed within the above sequence. The duplicates should be numbered with a 'D' appended to their JY filenames. For examples PW125 and PW170 will have their duplicates labelled as PW125D and PW170D respectively. SRM identifiers do not have to follow this rule. Refer to the example dialogue included with this report.

The data are stored into the concentration file *.CON (which was created at the beginning of the execution of this program) and is ready for the final reporting program T2FORM.

JOBFIL JOBFIL creates a 'sample identifier' file (file (1)) and a 'job number with geologist name' file (file

(2)). Both files will be stored on the system disk with the file extension ".DAT" files must be deleted if the same filenames are going to be used again. The program will process a maximum of 10 jobs and a maximum total number of 150 samples. Note: Do not delete the TTIO.DAT file on the system disk under any circumstances.

The user is then requested to enter the total number of jobs to be reported. This number will be stored in the first record of file (2). Then for each job, you will be prompted to enter information in the following order:

(1) The geologist's name.

(2) The corresponding job number or date received.

(3) The corresponding sample identifiers in the order that they will appear on the report form. This also corresponds to the order in which they were processed by T2CON.

If there is a common prefix for the sample identifiers, the user is allowed to enter the prefix once prior to the entries of the sample numbers. The maximum number of characters allowed for either the prefix or the sample numbers is less than or equal to 5 (maximum number of characters for a sample identifier is 10). For example, the prefix and the sample number for the sample identifier 87AAA-0001 are 87 A AA and -0001 respectively.

Once the required information has been entered, the program allows the user to go back and alter any of the entered information. The total number of samples for all of the jobs to be reported is stored in the first record of file (1). Refer to the example dialogue included with this report.

T2FORM The T2FORM program produces the final T2 reports from the secondary corrected concentration data. The program will ask for the name of the file that contains the concentration data ( *.CON created by the T2CON program), the file name for the sample identifiers and the file name for the geologists names and job numbers (the two *.DAT which were created by the JOBFIL program).

These files should appear on the system disk as a *.CON file and two *.DAT files. The concentration data are then rounded to the nearest one ppm rock and compared with the method detection limit for each element. Any results that are less than the method detection limits are printed as the minus of the method detection limits. Results that are greater or equal to one percent are rounded to the nearest decimal with a 'P' (percentage) appended to them. The program will then print out the final concentration file. The record number (*) associated with each sample is printed out at the left of the page (refer to the example outputs).

For each job to be reported, the program will prompt for the following information:

1) The record number (*) associated with the first and last samples in the job.

2) Are there any duplicates?

EA18-22

Traces - ICP-OES

If so, then enter the record numbers associated with the first and last duplicates. Changes are allowed at this point.

3) The number of T2 elements to be reported.

If less than 11T2-elements are requested for a job, the user will be asked to enter the required number of elements for that job. If W is included in the T2 package, enter 11 for the number of T2 elements, otherwise enter 10.

4) The name of the analyst.

5) The report date.

Change the paper to 14 7/8" x 11 "and position the print head at the last line position of the page. Vertical and horizontal pitches on the DEC terminal have to be changed to 6 and 8 respectively. When all this has been done, press the RETURN key and the program will printout all of the jobs on a T2 form that is ready to be reported.

EA 18-23

Traces - ICP-OES

AQUEOUS SAMPLES ANALYSIS PACKAGE

INDUCTIVELY COUPLED PLASMA

SOURCE EMISSION SPECTROSCOPY

Introduction:

A general introduction to the principles of ICP source emission spectroscopy is given in the Overview to

Trace Element Determination by ICP Optical Emis sion Spectroscopy. An explanation of potential inter ferences and procedures for dealing with these is also presented.

The Geoscience Laboratories is able to accept aqueous samples for analysis on the JY48P ICPOES

Spectrometer. No sample preparation is required and simultaneous determination of up to 36 elements is possible. The available elements are Sn, Mo, W, B, Zn,

P, Pb, Co, Ba, Se, Cr, Mg, V, Nb, Ca, Ag, Ti, Dy, Y,

Eu, Sr, Gd, Ni, Ta, Mn, Fe, Si, Al, Be, Cu, Yb, Zr, Sm,

La, Nd and Ce.

The method can be applied to a wide variety of aqueous samples. Data may be biased due to the presence of numerous spectral interferences. All of these inter ferences are quantifiable, but may degrade the deter mination limits. It is advisable to discuss the potential problems inherent in a specific sample matrix with the

Supervisor, Spectroscopy Subsection, prior to the sub mission of aqueous samples.

Safety advisory:

Refer to the Trace 2 Package.

Method:

Refer to the Trace 2 Package. The one major dif ference between these samples and the method as described earlier is that the dissolution (digestion) technique can be ignored.

Apparatus:

Refer to the Trace 2 Package.

Reagents:

Refer to the Trace 2 Package.

Procedures:

Refer to the Trace 2 Package.

Quality Control:

Determination Limits and in Aqueous Samples

Ranges for Elements

Element

Sn

W

Zn

Pb

Se

Cr

Mg* v*

Ag

Ti*

Dy

Y

Eu

Sr*

Range

(ppm)

- 200

- 40

- 30

- 500

3

- 45

3

1

-3000

- 4

- 2

-1200

- 3

- 100

- 3

1

1

- 150

0.15

0.01

0.02

0.07

0.025

0.01

0.02

0.004

0.005

0.01

0.04

0.03

0.01

0.40

0.01

0.003

0.01

0.01

Mo

P*

Gd

Ni

Mn*

Fe*

Si*

Element Range

(ppm)

Be*

Cu*

Yb*

Zr

Sm

La

Nd

Ce

1.0

0.25

0.001

0.004

0.002

0.015

0.06

0.025

0.15

0.4

0.02

0.02

0.30

0.04

0.02

0.04

0.003

0.07

- 35

- 30

-5000

- 15

- 35

- 10

- 100

-2800

- 1200

-7000

0.1

- 100

0.5

- 15

- 15

7

- 30

- 40

* These elements are routinely determined at less than optimum sensitivity due to their high natural abundances. Lower determination limits may be possible if requested.

Productivity:

Refer to the Trace 2 Package.

Additional Notes:

Refer to the Trace 2 Package.

EA18-24

Traces - ICP-OES

TENTATIVE SPECTROSCOPIC ANALYIS PACKAGE (TSPA)

INDUCTIVELY COUPLED PLASMA

OPTICAL EMISSION SPECTROSCOPY

Introduction:

A general introduction to the principles of ICP optical emission spectroscopy is given in Overview to Trace

Element Determination by ICP Optical Emission

Spectroscopy. An explanation of potential interferen ces and procedures for dealing with these is given.

The Tentative SPectroscopic Analysis (TSPA) pack age provides semiquantitative analysis of 26 elements including Al, Ba, Be, Ca, Ce, Cr, Co, Cu, Fe, La, Pb,

Mg, Mn, Mo, Ni, Nb, Nd, P, Sr, Ta, Ti, W, V, Y, Zn and Zr. Results are reported in terms of percentage concentration ranges.

Apparatus:

Refer to the Trace 2 Package.

Reagents:

Refer to the Trace 2 Package.

Procedures:

Refer to the Trace 2 Package.

Quality Control:

Refer to the Trace 2 Package.

Safety advisory:

Refer to the Trace 2 Package.

Productivity:

Refer to the Trace 2 Package.

Method:

Refer to the Trace 2 Package.

Additional Notes:

Refer to the Trace 2 Package.

EA 18-25

Traces - ICP-OES

APPENDIX A

GENERATION OF THE TSPA CERTIFICATES

Introduction:

A series of programs have been written to allow a user to generate signature-ready TSPA certificates and/or worksheets. There are three programs that must be run to accomplish this task: INFORM, ELEFIL and

WORCER. This report describes the use of the programs.

The INFORM program sets up a client database file that can be easily updated with new information as required.

Each time TSPA samples are run, new client information will be added to the database. This file will appear on the system disk as *.DAT file. In this report, the symbol **' denotes a user-supplied file name.

The ELEFIL program allows the user to select the elements of interest. At the present time, the TSPA program for the JY48 is fixed in this regard, so this utility program need only be run once to create the necessary element data file. If the number of elements in the JY48 TSPA program is changed, this program must be run again.

The element identifier file is stored on the system disk as *.DAT.

The WORCER program is used to produce the signature-ready certificates and/or worksheets.

Procedures:

1. Insert the system disk for the TSPA certificate and/or worksheet into drive l.

NOTE: The TSPA certificate and/or worksheet program must be run on the system disks that have been formatted to maximize disk space. Use ONLY these specially formatted disks to run the TSPA programs.

2. Insert the disk that contains the data into drive 2, i.e., the original JY48 *.DAT TSPA files (in percent concentrations format).

3. Check the directory of the system disk and make sure the client information file and element identifier file are present.

4. Enter R INFORM. For details see INFORM.

5. Enter R ELEFIL. For details see ELEFIL.

6. Enter R WORCER. For details see WORCER

INFORM

The INFORM program allows a user to enter, delete, verify or save client information (names, addresses, etc.) into a *.DAT file on the system disk.

The program will first prompt for the client information file name. The first time this program is run, the file will be created on the system disk.

The program will then present the following options: l. Verify the client information on file

EA18-26

Traces - ICP-OES

2. Enter new information

3. Change information in a record

Option l

Under option l, the user can obtain a hardcopy of part or all of the client information on file. The program will tell the user how many clients (i.e., the total number of records) are on file and then prompt for the number of records to be printed.

If the whole file is to be displayed, then enter the total number of records on the file. If only part of the file is going to be displayed, the user will then enter in the particular record numbers of interest. The client information will be printed out and the program will terminate execution. This option should be used whenever new information has been entered since an updated hardcopy of this file is required when the WORCER program is used.

Option 2

Option 2 is run when new information is to be saved into the client information file. If the client file exists, the new information is appended. If the client file is new, then the file will be created using the name the user entered at the beginning of the program. In the case of a new file, the program will initialize the file records and print out a message to this effect. There is programmed protection against overwriting an existing file, but no protection is fool-proof, so pay close attention to any messages that may be printed out at this point in the program.

The program will then prompt for the following information: a. Cli ent's name?

b. Address?

c. City, province and postal code?

d. More information to be entered?

The maximum number of records in each client file is 100. The program will print out the starting record number before it prompts you for the new information so the user can act accordingly (i.e., if a client file is filled, then a new file can be started).

When no further information is to be entered, the data is written to the client file and an appropriate message is printed. The total number of clients on file will be written to the last record of the client file. The user can now produce a hardcopy of part or all of the file.

Option 3

Option 3 allows the user to modify names, addresses etc. in a particular record. The program prompts for the number of records to be altered (or ^ 100) and then prompts for the individual record numbers. The information in these records is printed for verification. The program then prompts for the new client information as per the procedure in Option 2 and also gives the user the option to print out part or all of the client file.

ELEFIL

ELEFIL creates a *.DAT on the system disk. This file contains all the elemental identifiers associated with the JY48 TSPA analysis. The maximum number of elements allowed is 30. At present,/the number of elements used for the certificate or worksheet is 27 including total radioactivity. The TSPA data from the

JY48P also includes arsenic, silver and tin. These have to be included, even though they are not reported.

Therefore, the total number of elements in the file is 30.

EA18-27

Traces - ICP-OES

The program will prompt for a filename to store the element identifiers. NOTE: The element identifiers must be entered in the same order as they appear in the original JY48 data file. Once the required information has been entered, the program allows the user to correct the entered information.

WORCER

This program generates the signature-ready TSPA certificates and/or worksheets. It prompts for the names of the client information file (created with the INFORM program) and the element file name (created by the

ELEFIL). The program will prompt the user to enter the total number of clients and then, for each client in turn, asks: a. the record number corresponding to the client's name and address b. the date the samples were received c. the total number of samples for this client

For each sample, the program will ask for the sample identifier that is to appear on the certificate, the associated

JY48 *.DAT file name and the total radioactivity for the sample.

Once the client information has been entered, the user has the option of producing the worksheet (one or two copies), the certificate or both. The program will prompt the user to change vertical and horizontal pitches on the DEC terminal to 6 and 12 respectively.

The user is also prompted to line up the print head to the appropriate position and insert carbon paper in the case of the TSPA certificate. NOTE: If both the worksheet and certificate are required, the worksheet is printed first, then, after changing paper, the corresponding certificate is printed.

EA 18-28

Traces - ICP-OES

SAMPLE

SIGNAL

DEC 11/23

COMPUTER

Figure ICP1. Schematic Diagram.

JY48P

SOFTWARE

GLOGS

SOFTWARE

RAW DATA

ANALYTICAL

DATA

ANALYTICAL

REPORTS

EA 18-29

Traces - ICP-MS

SAMPLE PREPARATION FOR THE T4 AND T5 ICP-MS PACKAGES

Introduction:

The ICP-MS technique is not well suited to the direct analysis of solid geological samples because wet chemical preparation procedures are preferred to decompose the samples. The techniques outlined below are suitable for the determination of the follow ing elements in most rocks:

- Y and the lanthanide elements (La, Ce, Pr, Nd,

Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb, Lu) i.e., the complete T4 package.

- Hf, Ta i.e., part of the T5 package (Th, U, Sn, Pb and Tl are also included in the T5 package, but the HF stabilization technique has not been validated for the determination of these ele ments).

- Rb, Cs, Nb, Zr and Sr. The concentrations of these elements are generally determined by XRF as pan of T3 package. The concentrations of Nb and Sr can also be determined by ICP-OES.

Analysis for these elements by ICP-MS is primarily intended for those situations where use of the other two techniques is inappropriate (e.g., small sample sizes, efficient use of resources or determination limit requirements).

The technique below calls for the addition of a small amount of hydrofluoric acid to the final solution. This addition keeps traces of Hf, Ta, Nb and Zr in solution while avoiding the precipitation of the lanthanides as fluorides. If analysis for Hf, Ta, Nb and Zr is not required, then it is not necessary to add HF to the final solution.

Safety advisory:

1. Two different acid mixtures are used for sample preparation. For safety reasons, the composition of each must be recorded on the dispensing bottles.

2. When using HF, wear glasses and gloves, and be extremely careful. More information on HF is available in the Geoscience Laboratories' Safety

Manual.

3. Nitric acid, hydrochloric acid, and perchloric acid are corrosive substances. Wear protective cloth ing and exercise caution when handling these sub stances. Additional information on perchloric acid is available on pages EA2-1 and EA2-2.

Apparatus:

- 50 ml Teflon beakers

- plastic pipettes

- two 500 ml graduated plastic acid-dispensing bottles with graduated dispensing reservoir (to be used for the T4 A and T4B acid mixtures) contain ing HF or HQO4.

- two 500 ml low actinic glass acid-dispensing bottles for dispensing concentrated HNO3 and

HC1

- one 500 ml low actinic glass acid-dispensing bottle with calibrated dispensing barrel for use with the Ru/Re internal standard mixture

Reagents:

- Hydrofluoric acid, HF, 487c - 51 Ve

- Hydrochloric acid, HC1, 369fc -

- Nitric acid, HNO3, 69*70-7196

- Perchloric acid, HC1O4, 629fc -

Procedure:

1. Reagent Preparation

Two acid mixtures are used for sample preparation.

The composition of each must be recorded on the plastic dispensing bottles. The dispensers are also labeled T4A and T4B.

1.1. T4A solution is prepared by measuring con centrated 400 ml HF, 40 ml HC1, and 40 ml

HC1O4 and mixing them in a 500 ml graduated, plastic dispensing bottle. When preparing these solutions, always add the acid to water.

l .2. T4B solution is prepared by measuring 380 ml deionized distilled water, 70 ml HC1, and 30 ml HC104 and mixing them in a 500 ml graduated plastic dispensing bottle.

2. Standard Preparation

2.1.1. Pipette 5 ml of 1000 g/ml Ru, followed by 5 ml of 1000 g/ml Re into a 500 ml volumetric flask. Make to volume with 109fc HNO3. Use class A volumetric glassware.

2.1.2. Transfer to a 500 ml glass dispensing bottle.

EA19-1

Traces - ICP-MS

2.1.3. Verify that the Ru and Re concentrations in the old and new standard solutions are the same.

Comparison is done by dispensing (using the

500 ml internal standard dispensing bottle) l ml of the standard into a 100 ml class A volumetric flask and making to volume with

10*^ HNO3. This solution is then be com pared against the old standard by examining the signal trace for each standard (and each analyte) on the ICP-MS using the SPECJ3IS program. You should not be able to discern any differences between the signal traces for the two standard solutions.

3. Sample Dissolution

3.1.1 Weigh 200 mg of sample into a Teflon beaker and add 15 ml of T4A solution. Using the T4A solution, wash down the sides of the beaker.

3.1.2. Prepare and carry at least 2 blanks and 2 ali- quots of the digestion control material through the entire procedure (see Quality Control Sec tion).

3.1.3. Place sample and acid on a hotplate which has been set to 1200C. Leave overnight. Con tamination of the sample can occur from debris falling from structural members of the fumehood above the hotplate, so be sure that all necessary precautions have been taken to avoid this.

3.1.4. The next morning, remove the beaker from the hotplate and gently tap the beaker to dislodge the cake.

3.1.5. Add 15 ml of the T4B solution, replace on hotplate and leave overnight again.

3.1.6. To the hot dry sample cake, add 8 drops of concentrated HC1, allow to cool for about l minute and then add l ml of concentrated

HNO3 and swirl the solution gently.

3.1.7. Using the plastic dispensing pipette, add 4 drops of concentrated HF and swirl the solu tion gently.

3.1.8. Add approximately 15 ml of deionized dis tilled water and replace the beaker on the hotplate and reduce the volume to about 10 ml.

THE SOLUTION MUST NOT BE AL

LOWED TO GO TO DRYNESS. If it does, the aliquot must be discarded and the proce dure repeated.

3.1.9. Remove the beaker from the hotplate and cool to room temperature.

3.1.10. If the samples are not going to be transferred within a few hours, add about 25 ml 109fc

HNO3.

3.1.11. Transfer l .0 ml of the internal standard solu tion to a clean 100 ml volumetric flask. Using

1096 HNO3, transfer the contents of the beaker to this flask and make to mark with 1096

HNO3.

3.1.12. Use Parafilm to stopper the flask and shake thoroughly.

3.1.13. Transfer an aliquot of the solution to a 15 ml snap-cap polystyrene test-tube.

3.1.14. The solutions are ready for determination by

ICP-MS.

Quality Control:

The sample digestion procedure is monitored by in cluding reagent blanks and in-house reference materials in the batch of client samples. Three hotplates are used for the sample digestion. Each hotplate can accommodate 20 of the 50 ml Teflon beakers.

When a batch of samples is prepared, include one reagent blank and one in-house reference material on each hotplate used. A bottle (Bottle #55) of the in- house basalt reference material (a basalt collected by

P. C. Lightfoot in 1988) has been characterized for quality control purposes for the following analytes:

Rb, Sr, Y, Zr, Nb, Cs, the lanthanides (except Pm), Hf andTa.

Since about 60^c of the samples listed on the

PETROCH database are basalts, KIWI is appropriate reference material. During the certification of this reference material, three people prepared a total of 24 aliquots and the solutions were analyzed once by the procedures outlined in this manual (Table MSI).

Means and associated standard deviations were calcu lated from these data and the values are used to monitor the digestion step of the T4 and T5 procedures.

EA 19-2

Traces - ICP-MS

TABLE MSI. DIGESTION CONTROL DATA

(IN-HOUSE REFERENCE BAS

ALT (1988) MATERIAL,

BOTTLE #55, ^24)

Element

Sm

Eu

Gd

Tb

La

Ce

PT

Nd

Cs

Hf

Ta

Y

Rb

Sr

Zr

Nb

Dy

Ho

Er

Tm

Yb

Lu

Mean

14.374

307.3

173.84

13.6

0.229

4.578

0.823

25.025

21.342

50.273

6.104

27.247

6.105

1.873

5.658

0.838

5.037

0.994

2.581

0.363

2.366

0.351

Std. Dev.

0.655

1.484

0.17

0.65

0.156

0.045

0.152

0.023

0.348

6.64

4.431

0.20

0.009

0.145

0.028

0.662

0.20

0.022

0.090

0.014

0.072

0.015

The digestion-control standard data for a particular batch is compared against the above values by calculat ing the number of standard deviations the control solution data differ from the mean. These calculations are printed at the end of the interim report produced by the Apple II concentration calibration/calculation software. This interim report is filed along with the original bench sheets for the job. The run number associated with the data is also included on the bench sheets. The run numbers on a particular data disk are written on the disk label. The run numbers on a particular data disk can also be identified by using the

Apple II utility program 'FIXRUNQUE'.

With all this information, data can be retrieved as required. The reagent blank data for Zr and Hf and the average value used in calculating the results for report ing are hand written on the bench sheets.

Productivity:

Sixty samples (2 RBLK, 2 digestion check preps and

16 samples per hot plate) can be prepared in three days.

With assistance, two batches per week could be prepared if the first batch is started on Monday. As sistance during the final transfer step is required if a second batch of samples is to be started on the Wed nesday of the work week.

Reagent blanks from each batch must always be check ed. However, after three years of using the T4 proce dure, a reagent blank (RBLK) problem has yet to be encountered for Y and the REE, although Zr has a substantial reagent blank contamination (typically

2 ppm in the rock) as does Hf (typically 0.05 ppm in the rock). These RBLK values for Zr and Hf are established by using the Elan 'QUANT' program and the run protocol:

HN03, RBLK1, 107o HNO3, RBLK2, ..., o HNO3, 80 ppb calibration standard

Results are calculated manually. The average analyte signal measured for the 109fc HN03 solutions, which bracket the RBLK, is subtracted from the RBLK solu tion signal. A sensitivity factor is calculated from 80 ppb/cpSna where cpsnet is the analyte signal corrected for the 109fc HN03 signal obtained immediately prior to the 80 ppb calibration solution standard.

The instrument must have warmed-up sufficiently so that signal drift is minimal (approximately 3/4 of an hour).

Additional Notes:

1. Concerning the final steps of the digestion proce dure (addition of 8 drops concentrated HC1 fol lowed by l ml concentrated HNO3). It is

imperative that the HC1 be added first. Failure to

do so will result in irreproducible Hf, Ta, Nb and

Zr analyses and the samples will have to be reprepared from the beginning.

2. The gloves have a talcum coating on the inside.

Wash the outside of the gloves before tranferring samples to avoid getting talcum from inside the gloves in the sample.

3. The beakers are cleaned in tap-water solution of

"Sparkleen". Rinse this soap solution with hot water. Otherwise, a thin soap film is left on the beaker.

4. The Ru/Re dispensing bottle has a screw-on cap which connects the bottle with the dispensing device. DO NOT SCREW the cap on too tightly, otherwise, a vacuum is produced and air bubbles end up in the dispensing barrel. Occasionally un screw the cap to avoid this.

EA 19-3

Traces - ICP-MS

5. Maintenance - Work areas in and near the

fumehood and work bench should be cleaned whenever required. The fumehood has internal plumbing for washing down the stacks and the inaccessible walls of the fumehood. This should be used at the finish of each batch. Bytac protec tive Teflon covering has been installed and should be wet mopped between batches.

EA 19-4

Traces - ICP-MS

THE ELAN-250

Introduction:

This document is intended for use as a bench procedure for the routine operation of the Sciex 250 Elan Induc tively Coupled Plasma Mass spectrometer.

Method:

The instrumental method consists of the following procedures:

1. Instrument start-up

2. Mass calibration

3. Signal optimization

4. Setting up the run

5. Performing analysis

6. Analyzing the run

7. Performing calculations

8. Instrument shutdown - normal operation

9. Instrument shutdown - emergency procedure

Procedures:

1. Instrument Start-up

Become familiar with the location of all switches and controls on the instrument before operation. Many of the steps require a sequence to be performed quickly.

If things go wrong, don't panic, but move immediately.

If you are not confident about any operation do not start the procedure, or abort the procedure and seek assis tance.

1.1.1. The outer sleeve of the torch should extend 37 mm past the tip of the aerosol injector (i.e., long torch).

1.1.2. The torch should be positioned in the work coil so that the distance between the aerosol injec tor and the bottom turn of the load coil is

2mm.

1.1.3. The torch stand should be positioned so that when it is at the extreme end of its travel towards the torch, the injector tube is 43 mm away from the sampling orifice and the top turn of the load coil is 23 mm from the sam pling orifice. After this distance has been checked ("by eye") move the box away from the sampling orifice until it rests against the back stop. The box will be returned to the stop position after the plasma is ignited.

1.1.4. Visually ensure that the load-coil centre-tap solder join is intact and the interface/torch cooling water is on.

1.2. Starting The Computer

l .2. l. Activate the printer using the toggle switch at the back left-hand corner.

l .2.2. Activate the Envision monitor/terminal using the switch at the back left-hand corner.

1.2.3. Wait 10-15 seconds.

1.2.4. Activate the Elan computer using the switch labelled 'POWER' on the face of the com puter.

1.2.5. Wait 30 seconds.

l .2.6. Press the switch labelled 'RESET' on the Elan computer panel.

1.2.7. Wait (3-4 minutes) until the following mes sage appears at the top of the monitor:

"XXXX Files XXXXX Blocks XXXX Free"

If there are more than 3000 files, or less than

700 free, DO NOT PROCEED. Contact the person responsible for the ELAN.

NOTE: The Elan software relies on the use of a color

coded numeric keyboard. In the procedures described here, the following short forms for references to operator responses using this keypad; the keys are numbered left to right l ,2, and 3, G ~ Grey, R = Red, Y - Yellow, B -

Blue and Gn - Green.

1.2.8. At the "Login: " prompt type Elan [R]. Then type in Y [R] to the prompt "Does the printer need initialization (y/n)".

1.2.9. Press B3. The screen should now display text titled "ECL Commands". Make sure caps- lock is off. The computer is now operational.

l J. Starting the Plasma Source

1.3.1. Turn on the RF Generator by depressing the

CKTS push button.

EA19-5

Traces - ICP-MS

l .3.2. Turn the reflected power switch on the APCS-

3 panel to OFF.

1.3.3. Turn the power pot on the R.F. generator to a setting of about 8.2.

1.3.4. Check that the voltage to the channel electron multiplier (GEM) detector is switched off by checking that the left-hand toggle switch on the fourth board from the top in the signal and ion lens electronics rack is set to the left.

NOTE : One lead of the R l O resistor on the GEM board

has been disconnected. This is to ensure that the GEM on/off toggle switch will always turn off volts to the detector independent of the computer status.

l .3.5. If the lens/quadrupole power supplies are not energized, replace relay Kl. Check that the yellow relay below K l is seated properly.

NOTE:This yellow relay controls the valves between

the mechanical pump and the mass spectrometer chamber. Improperly seated, this relay could leave one of the valves open and the mass spectrometer will not get out of stage l pumpdown.

1.3.6. IMPORTANT: Turn on the interface and

load coil water supply located at the sink beside the distilled deionized water-still in the ICP lab (far right tap).

1.3.7. Ensure that the torch stand has been moved away from the interface plate and is resting against the back stop.

1.3.8. Turn on the argon gas supply and nebulize deionized water (plasma gas 16 1/min, auxil- liary 2.2 1/min, nebulizer gas 1.0 1/min) and allow the system to flush for 3-4 minutes.

1.3.9. Plug in the autosampler.

1.3.10. Turn off the nebulizer gas flow using the needle valve control knob. When the nebulizer pressure gauge registers minimum, bleed in about 0.051/min.

1.3.11. Set the drum capacitor (toggle switch on the front upper left comer of the torch-box) to

MANUAL and adjust the drum capacitor reading to 82 using the white buttons located to the left of the toggle switch.

1.3.12. After the start-up value has reached 82, set the toggle switch to AUTOMATIC.

1.3.13. Set the RF torch-box power pot to zero (fully counter clockwise) and push the RF ON but ton.

1.3.14. Rotate the RF torch-box power dial in a clock wise direction so that 0.5 kW of forward power is indicated on the forward power meter. Maintain contact with the power dial; it may be necessary to adjust it for more or less power during the ensuing operations. Activate the tesla coil by depressing the white "ICP

IGNITE" button located on the torch stand control panel. It may be necessary to adjust the forward power up to initiate the plasma or down to stop arcing.

NOTE: Arcing from the coil to the torch (accom panied by the green Cu emission, sparks and the sound of RF arcing) occurs occasionally,

(particularly when 1.4 kW power are used instead of 0.5 kW). If this happens, QUICK

LY rotate the power knob COUNTER

CLOCKWISE i.e., to OFF. Push the RF

POWER OFF button. A check should then be made to ensure the load coil centre tap solder join is intact and the interface/torch cooling water is on.

1.3.15. At this point, the plasma should light. If it does not ignite after a few seconds, press the RF

POWER OFF burton and consult the senior technician or Supervisor.

1.3.16. Rotate the power pot to select a forward power setting of l .5 kW, as indicated on the Forward

Power Meter.

1.3.17. The following operations should be per formed quickly: depress the 'Operation' button ON.

switch the 'RF Reflected' toggle switch ON.

turn the power knob to 'MAX' (fully clockwise).

open the nebulizer gas line via the needle value to 1.01/min.

1.3.18. Wait for the interface gate to begin dropping, then move the torch stand forward (up to the interface plate), until it reaches the end of its travel. Lock it in place using black knobs.

EA 19-6

Traces - ICP-MS

l .3.19. Place the ultraviolet shield in place.

1.3.20. Check the mass spectrometer vacuum (8-

12 x 10'6 torr) and temperature (17-19^). If the gauges do not indicate these values consult the senior technician or Supervisor.

1.3.21. Turn on the GEM.

2.1.3. Press RI. The second page of the Spectrum

Display software, (Measurement Settings), will appear.

2.1.4. Move the cursor and change the displayed fields to match those listed in Table MS3.

2.1.5. Press RI again. The Mass Selection page should appear.

1.4. Daily Checkout Routine:

The operating conditions used for the daily checkout and sample analysis are listed in Table MS2. Lens voltage settings are not fixed. Consult with a senior technician or Supervisor if there are any questions on this matter.

TABLE MS2. ICP-MS OPERATING CONDI

TIONS

Plasma Conditions

Torch

RF power

Reflected power

Aux. flow (intermediate)

Nebulizerflow

Plasma gas flow (outer)

Solution uptake rate

1.4.1. Allow the plasma, mass spectrometer and detector to warm up for about 20 minutes.

1.4.2. Ensure that all instrument operating parameters are set to the values indicated in

Table MS2.

NOTE:The Elan software relies on the use of a color

coded numeric keyboard. In the procedures described here, the following short forms for references to operator responses using this keypad; the keys are numbered left to right l ,2, and 3, G - Grey, R = Red, Y - Yellow, B =

Blue and Gn = Green.

Scanning:

2.1.6. Enter 139 [R] ...Enter 139 [R] again. The Elan will scan over this mass range (the 139La peak).

Extended outer sleeve style

1.5 k W

^W

2.21/min

1.01/min

16 1/min

0.85 ml/min

2. Mass Calibration

This procedure must be carried out at the beginning of every work day. It updates the mass calibration, and tests instrument sensitivity by providing intensity data which can be compared with standard (previously acquired) data.

2.1.1. Enter SPEC PIS [R]. The first page of the

Spectrum Display program will show up on the screen. The cursor can be moved through the various fields by entering [R],

2. l .2. Nebulize a solution containing 80 ppb REEA.

2. l .7. Press G3 to start the scan.

2. l .8. Press [R] in response to the prompt "Press [R] or Abort to continue".

2. l .9. Press G l to get into Graphics mode.

2.1.10. Press and hold the 'FUNCT key down. While this key is depressed, press Gn3. This will engage the crosshair markers.

2.1.11. Continue holding the 'FUNCT' key and press

Y3. Move the vertical crosshair to the middle of the peak.

2. l. 12. Continue holding the 'FUNCT' key and press

R2. Move the horizontal crosshair to the apex of the peak.

2.1.13. Release the 'FUNCT' and the R2 keys.

2.1.14. Press RI to zoom in on the image and use

'FUNCT' R2, and 'FUNCT' Y3 to adjust the markers so that they intersect at the centre of the peak.

2.1.15. Press R3 to zoom back to original image size.

Record the peak position (to 2 decimal places).

2.1.16. Press and hold the 'FUNCT' key down.

While this key is depressed, press Gn3 twice.

This should remove the crosshair markers.

EA19-7

Traces - ICP-MS

2.1.17. Press G l to get out of Graphics mode.

2.1.18. Press Gn3 to stop scanning (a bell should sound).

2.1.19. Press Gnl to get the Mass Scan Range selec tion screen.

2.1.20. Repeat the above scan procedure for 89Y,

175Lu, 151Eu. Record all four observed masses.

2.1.21. Press Gnl three times to return to the "ECL

Commands screen".

Mass Calibration

2.1.22. Enter Meal [R]

2.1.23. Enter REEMCC [RI

2.1.24. Enter m [R]

2. l .25. Press G l to get to the mass-entry screen.

2.1.26. Enter the observed mass for Y [R].

2. l .27. Enter the observed mass for La [R].

2.1.28. Enter the observed mass for Eu [R].

2.1.29. Enter the observed mass for Lu [R].

TABLE MS3. MEASUREMENT

METERS

PARA

Measurement mode Sequential, l point per analyte peak

Measurement time

No. integrations

Resolution

0.5 s/point

6

0.9 amu at 109fc valley

Analyte

Er

Tm

Yb

Lu

Re

Tb

Gd

Dy

Ho

Pr

Nd

Sm

Eu

Y

Zr

Ru

La

Ce

Mass (daltons)

151

159

160

163

165

167

169

174

175

185

89

90

99

139

140

141

143

147

Oxide/isobar int.*

13^BaO

149 SmO

158GdO

2.1.30. Press G3 to obtain print out of the day's mass calibration. These are filed for l - 2 months.

2.1.31. Press Gnl to get back to the first screen of mass calibration.

2.1.32. Press G3 to save the new mass calibration table.

2.1.33. Press Gnl to get back to the ECL command screen.

3. Signal Optimization

If the sampler and/or skimmer has been removed, then the signals must be optimized. The operating condi tions and measurement parameters are listed in Tables

MS2 and MS3.

* This column represents the oxide and isobaric corrections considered significant. Other species may be formed but their contributions are con sidered negligible.

3.1.1. Ensure the plasma, mass spectrometer, and detector have warmed-up for about 20 minutes.

3.1.2. Ensure that all instrument operating parameters are set to the values indicated in

Tables MS2 and MS3.

3.1.3. Set the Photon Stop and Plate digipots to zero.

3. l .4. Set the Einzel and Barrel digipots to 70 and 50 respectively.

3.1.5. Select the "Mulelem" program (ECL com mand). Select the measurement conditions to monitor the following masses m Cd and 187Os.

Use 1/2 s/point, l point/peak. Toggle the isobaric correction for 187Os off.

EA 19-8

Traces - ICP-MS

3.1.6. While monitoring the Os background sig nal, increase the Plate digipot setting until just before the background signal starts to rise.

3.1.9. Monitor the signal from 80 ppb Y, La, and Lu

(REEA solution) and adjust the Barrel voltage until the Y and La signal are about equal. This adjustment is normally very small. The Y signal must not be significantly lower (maxi mum 107c difference) than the La signal.

NOTE:The La counts should be about 70 to 110 kHz depending on the condition of the sampler and skimmer. A newly Ca conditioned sampler and skimmer give 70-80 kHz. During sample analysis, as material is deposited on the sampler and skimmer, the counts typically drift up over part or all of the REE mass range.

4. Setting Up the Run

4.1.1. Load the autosampler according to the follow ing protocol:

3. l .7. Monitor the 139La signal while nebulizing a 80 ppb REEA solution.

3.1.8. If the copper sampler and skimmer have just been cleaned, the 1000 ppm Ca solution must be nebulized for about 10 minutes to deposit a layer of Ca oxide on sampler and skimmer to protect them from inadvertent exposure to hydrochloric acid.

NOTE: Using kHz as a unit for an ion signal is a good way of reminding one that the number which is presented as the intensity is a measure of frequency (i.e., counts per second) with which ions strike the detector. For example, the noise power spectrum of an instrument can be experimentally determined, possibly giving insight to the source of signal noise in the instrument and perhaps uncovering ways to improve analytical precision.

3. l. 10. Adjust the Barrel lens so that the 89Y and 175Lu signals are maximized.

NOTE:This procedure typically results in the signal strengths being skewed in favour of the lighter

REE. Adjusting the Barrel and Einzel voltages to make the 89Y and 175Lu signal approximate ly equal generally results in an unacceptable overall loss of signal across the mass region of interest.

3.1.11. If the observed intensities are lower than ex pected, check the torch position.

Run Protocol

Sample solutions are run with the autosampler in batches using the following scheme:

ACDBLK

INTERFERENCE STANDARDS

DRIFTSTD

JUNK

CHK

7-10 samples

CHK

7-10 samples.....

CHK

DRIFTSTD

BLK nitric acid

REEA/REEB

REESTD reagent blank

NOTE:The sample IDs ACDBLK and BLK must be used. CHK is the within-run control standard whose values are established against those of an SRM. Currently a rock composite solution has been calibrated for REE against BHVO-1 and is used as the CHK solution. JUNK is a sample primarily used to minimize memory effects from the drift standards.

4. l .2. Position the first sample in the run so that it is in front of the sample sipper.

4.1.3. Initiate analysis as described below.

5. Performing Analysis

Quantitative analysis is performed with the aid of two computers: the instrument computer and a microcom puter (Apple Ile). The Apple computer collects and stores the raw intensities during analysis. At the con clusion of the run, it processes the data to determine final concentrations. The operation of these computers is described below.

This procedure consists of two parts;

1. Setting up the run i.e., entering sample id's, dilu tion factors etc.,

2. Running the samples.

EA19-9

Traces - ICP-MS

Setting up the run

5.1.1. Confirm that the "ECL Command" display is on the screen. If not press Gn2 until the display no longer changes and then press B3.

5.1.2. Enter QUANT [R]. The cursor can be moved to the various fields by using [R].

5.1.3. Type the name of the appropriate parameter set followed by [R]. (NOTE: The Calibration

Field must ALWAYS show "STDJ1). If it doesn't then change it to the correct value.

5.1.4. Enter [R] twice.

5. l .5. Enter a run number followed by [R].

5.1.6. Enter a run description followed by [R].

5.1.7. Enter n [R] twice.

5.1.8. Enter s [R].

5. l .9. Press G l to get to the Sample Entry screen.

5.1.10. Enter sample identification and description

(usually the OGS job number). Use [R] to

. select fields.

5.1.11. Press G l to store the sample information.

5.1.12. Continue entering sample information and pressing G l until all samples identifiers have been entered.

5.1.13. When finished press Gnl to get back to the first screen of 'QUANT'.

6. Analysing the Run

6.1.1. Use [R] to move the cursor to the "Run Mode" field and enter a [R] to indicate that the autosampler is to be used.

6. l .2. Place the Elan Apple System Disk in the left- hand drive.

6.1.3. Place the Intensity Data Disk (appropriate to the analysis to be performed) in the right-hand drive.

6.1.4. Activate the monitor (front top right-hand corner of the video terminal).

6.1.5. Activate the computer (back lower left-hand corner). After about 30 seconds, the display will show the Elan Apple main menu.

6.1.6. Select option l. This will initiate the intensity data collection program.

6.1.7. Respond to the first prompt by entering the appropriate parameter set name.

6.1.8. The program will prompt for; 1) Run ID, 2)

Autosampler (y/n), 3) Number of tubes in the run. Enter this information following each entry with [R].

At this point, the screen will display the or dinal number associated with the sample/tube position in the run. The record number into which the data will be written in the Apple disk file will also be displayed.

NOTE:The Apple II keyboard will be 'live' at this time. The Apple cannot distinguish between a keyboard entry from data comming from the

Elan. Consequently, inadvertent keyboard entry will corrupt the data and probably cause termination of the program. So don't touch the Apple keyboard.

6. l .9. After information has been entered, return to the Elan and press G l to begin the analysis.

The program running on the Apple II will monitor the data coming from the Elan. Upon completion of the set of samples, the Elan will return to the first screen of the 'QUANT' program. The Apple will require additional input.

6.1.10. Respond to the following Apple prompts:

* Do more analysis using the same parameter set?

If answer is no then:

* Do analysis using a different para meter set?

If answer is still no, the program returns to the main menu.

If answer is yes, the program will proceed and will ask for parameter set name etc., as before.

NOTE: Be certain to place appropriate data disk (for new parameter set) in the right-hand drive

BEFORE ENTERING THE NEW

PARAMETER SET NAME.

EA19-10

Traces - ICP-MS

7. Performing Calculations

There are two calculation algorithms in use. An inter nal standard technique has been developed for Y and the Rare Earth Elements (REE), Hf, Ta, Nb, Rb, Cs,

Sr, and Zr. All other analyses utilize external calibra tion with signal drift compensation (linear interpola tion).

7.1.1. Select option 2 from the Apple main menu.

The 'CONCAL' menu will be displayed.

Place the intensity data disk in the right-hand drive.

7. l .2. Select the appropriate calculation algorithm.

7. l .3. Enter the slot number for the printer card.

7. l .4. Enter the parameter set name for the set used to collect the data.

7. l .5. After 30 seconds, the program will display the identifiers of unprocessed runs and prompt for the run number to be processed. Select the desired run number.

7.1.6. The program will prompt for the filename to be used to store instrument and preparation

QC data. Enter the appropriate filenames. If this function is not being used enter 'S' to bypass these functions.

7.1.7. The program will read the intensity data. Wait until the program is finished doing this and has displayed the prompt requesting a "Default

Dilution Factor". Enter the factor.

7.1.8. The program will prompt for any factors that are different from the default. Enter these fac tors.

The program will perform the intensity to concentration calculations. No additional input is required until after the final results have been printed.

7.1.9. The program will prompt for the number of the instrument QC solution identifiers and their positions in the run. Enter the correct information.

The program will calculate and produce a report summarizing the instrument QC data.

This information is presented as the difference between the observed and the expected values divided by the standard deviation associated with the expected value. If the error is normal ly distributed, then 669fc of the results should be between O and one standard deviation.

Less than G.5% of the results will be greater than 3 standard deviations when there is no determinate error associated with the meas urement. If such an occurrence is observed, then there is good reason to believe that the measurement is biased. A run (or data between check samples) will be rejected if more than two of the check samples lie outside 3 standard deviations.

The program will calculate and produce a report summarizing the preparation QC data.

This information is presented in a manner similar to that produced for the instrument QC.

The program will then ask whether the con centration data is to be stored. Respond ap propriately. If the data is not to be stored, the program will return to the 'CONCAL' main menu.

If the data is to be stored, three options will be presented: 1) Routine Sample Data Storage 2)

International Reference Material 3) In-house

Reference Material. Some or all of these op tions may be used for one set of data.

7.1.10. The program will display each sample-iden tifier and prompt for the proper run-identifier.

Routine sample data should be stored using the four digit run-number assigned by the lab

(i.e., job number).

7.1.11. The program will prompt the operator to place the appropriate data disk in the right-hand drive. Place the disk in the correct drive.

When all data has been stored, the program returns to the main 'CONCAL' menu.

8. Instrument Shut-Down - Normal Operation

8.1.1. Turn off the detector.

8. l .2. Unlock the torch-stand.

8.1.3. Depress the 'Operation' button OFF and, as the gate begins to rise, steadily move the torch stand away from the interface plate.

8.1.4. Slowly, but steadily, decrease the forward power until the plasma extinguishes and then depress the 'RF OFF' button.

EA19-11

Traces - ICP-MS

8.1.5. Turn off the water to the interface and the load coil.

8. l .6. Allow the RF generator to cool for 10 minutes before shutting it off.

8.1.7. Exit the ELAN control programs (software shutdown).

8.1.8. Turn off the 'Envision' terminal, printer, depress the 'DISCS' button and turn the com puter power off.

NOTE:This must be done at the end of every day otherwise there is a chance of damage to the computer.

9. Instrument Shut-down - Emergency Shutdown

Emergency situations previously encountered include broken water/gas lines and grinding noises from the cryopump.

9.1.1. Depress the big RED button on the front of the torch- stand or RF generator.

9.1.2. Depress the 'Operation' button OFF.

9. l .3. Turn off the detector.

9.1.4. Turn off the water to the interface and the load coil.

In cases where the situation is considered serious, quickly shut down the power to the instrument using the breaker switch located outside room 1019 (Panel

C, Switch #2) or turn the power burton OFF located on the front panel of the instrument.

NOTE: When powering back up, the emergency

'lock' clasp on the red button must be released or power will not be restored using the normal start-up sequence. Push the red button in and turn slightly - it will pop back out into the normal position.

Temporary Loss of Cooling Water: Occasionally the building water supply is turned off. Since the cryopump must have cooling water to run, the instru ment is best put into cryoclean (cryopump is not re quired) for the duration of the interrupted service. A frill pumpdown from atmospheric will be required after (2-3 hours).

EA19-12

Traces - ICP-MS

TRACE 4 (T4) ELEMENTS

INDUCTIVELY COUPLED PLASMA SOURCE

MASS SPECTROSCOSPY

Introduction:

This suite of elements comprises a geochemically co herent group that has many geological applications.

Minerals such as apatite, zircon, monazite and other accessory phases usually concentrate the REE in a sample, although these elements can substitute for Ca and Sr in rock-forming minerals such as feldspar, amphibole or pyroxene. The REE are typically reported normalized to a chondritic meteorite com position on a semi-log diagram which is automatically included in the T4 data report.

The REE provide evidence concerning the formation of the solar system, the evolution of the crust from the mantle, generation of magma within the crust. All such data are used in petrogenetic modelling and to understand ore formation and the interplay of fluids of various compositions with pre-existing rocks. Also, the radioactive decay of La and Sm to Ce and Nd

(respectively) over geologically significant time periods can yield further insights into crustal proces ses.

The speed, precision and elemental coverage of the determination of REE abundances by ICP-MS is une qualled by any other analytical technique. However, although isotopic ratios can be determined by ICP-MS, the precision on the isotopic ratios is not competitive with thermal ionization mass spectrometry (TIMS), except for screening purposes.

The Trace 4 (T4) package consists of the elements listed in Table MS4, along with the typical determina tion limit, optimum range and precision quoted as the

957c Confidence Limit (ppm) for a value at 10 times the determination limit.

Safety advisory:

There are a number of potential safety hazards as sociated with the use of this equipment. UN-

TRALNED OPERATORS MUST NOT USE THE

EQUIPMENT. The OGS ICP-MS has had many modifications and consequently starting procedures are very different from a standard factory delivered instrument.

TABLE MS4. DETERMINATION LIMITS

AND PRECISION FOR THE

T4 ELEMENTS

Element Determination Optimum Precision

Limit Range

(ppm)

Y

La

Ce

Pr

Nd

Sm

Eu

Gd

Tb

Dy

Ho

Er

Tm

Yb

Lu

0.02

0.05

0.05

0.05

0.18

0.15

0.07

0.14

0.03

0.13

0.03

0.10

0.03

0.11

0.04

0.02-200

0.05-200

0.05-300

0.05-200

0.18-200

0.15-100

0.07-20

0.14-100

0.03-20

0.13-20

0.03-20

0.10-20

0.03-20

0.11-20

0.04-20

(Chondrite plots are automatically included with

T4 Data Reports)

There are five types of hazard associated with this apparatus; electrical, radiation, high temperature, noxious gases and physical hazard.

1. Electrical hazards can be encountered in three areas; the plasma power generator, the plasma torch compartment, the spectrometer tank com partment. Such hazards are also associated with various peripheral equipment like pumps, autosamplers and other special apparatus which may be set up to facilitate the analysis. The RF power generator is of particular concern because high voltages are retained on capacitors even when the power has been completely disconnected from the instrument. DO NOT ATTEMPT TO GAIN

ENTRY TO THE RF POWER GENERATOR

WITHOUT PROPER TRAINING. Other areas are protected by interlocks. DO NOT OVERRIDE

INTERLOCKS.

0.02

0.05

0.05

0.05

0.18

0.15

0.07

0.14

0.03

0.13

0.03

0.10

0.03

0.11

0.04

EA19-13

Traces - ICP-MS

2. High voltages should be marked stating the ap proximate voltage. Wear safety glasses where sparks or arcing may occur. Remove rings and metal watchbands when working with circuits or control devices. Never handle electrical equip ment with wet or perspiring hands. Some high voltage danger points are:

- Transformer terminals

- Rectifier-tube plate caps

- Filter capacitor terminals

- Filter choke

- RF tuning capacitors and coils

- Fuse panels

3. Two types of electromagnetic radiation are of con cern, radio frequency and ultraviolet. Radio fre quency radiation has no known teratogenic effect, but can cause somatic damage. Such radiation could affect eyesight as the eye fluids will absorb the radiation which causes a damaging increase in pressure within the eye. Ensure that the plasma torch compartment is well shielded by inspecting the viewing and access ports to make sure that the

RF radiation cannot escape. The 50 ohm coaxial cable should be inspected to ensure that it is a tight fit and that no corrosion has occurred at the con necting points on both the instrument and on the plasma power generator, (EXPERIENCED

TECHNOLOGISTS ONLY!). Ultraviolet radia tion is emitted by the plasma. THE PLASMA

SHOULD NEVER BE VIEWED DIRECTLY.

Even light from the plasma which is reflected from the source housing has a high level of ultraviolet radiation and direct viewing of this light should be avoided. If it is necessary to view the unshielded plasma, wear fully sealed No. 12 welders goggles.

AND THE PLASMA MUST BE VENTED

FROM THE ROOM AT ALL TIMES. Remember to check the damper positions to ensure that suffi cient exhaust velocity is being maintained.

6. Physical hazards can be encountered when moving or working around bottled gases. Exercise appropriate caution: secure the cylinders, use the valve caps, use properly designed carts to move the cylinders, use appropriate footwear, etc.

Method:

The method consists of the following techniques:

1. Acid digestion of the rock sample (see sample preparation - page EA 1 9-1)

2. Determination by ICP-MS

3. Calculation of final results using (a) an internal standard or (b) from external calibration data

Apparatus:

- Inductively coupled plasma source-mass spectrometer: Perkin-Elmer Sciex ELAN 250

- Ultra-violet shield

- Microcomputer: Apple Ile with dual floppy disk drives and Epson FX printer

4. The interlocks protecting the plasma should never be overridden. The torch itself can become quite hot and sufficient time should be allowed, after extinguishing the plasma, before disassembling the torch for routine inspection and cleaning. The plasma power tube runs at about 3000C and should be cooled at least 20 minutes prior to shutting the generator down at the end of the day or for inspec tion.

5. The exhaust gases from both the plasma compart ment and the generator can be a source of both heat and noxious fumes. Both beryllium (a vesicant) and ozone are emitted from the generator.

The exhaust from the plasma can contain toxic materials from the samples. THEREFORE, THE

EXHAUST FROM BOTH THE GENERATOR

Reagents:

- Purified Argon gas

- Distilled deionized water

- Single element 1000 ppm stock solutions of Y,

La, Ce, Nd, Pr, Sm, Eu, Gd, Dy, Yb, Ru and Re obtained from Spex Industries (Metuchen NJ)

Procedures:

1. Reagent and Standard Preparation

1.1.1.

1.1.2.

1.2.

200 ml of 1000 ppm single element stock solutions of Tb, Ho Er, Tm and Lu prepared by dissolving the appropriate amount of rare earth oxide powder (obtained from Alpha

Products, Danvers, MA) in about 2 ml of concentrated HNO3 and 6 to 10 drops of con centrated HC1.

For all serial dilution operations described below, use lO^c HN03 (v/v) for the diluent.

200 ml of a 100 ppm multielement standard

(called REEA') containing Y, La, Pr, Nd, Sm,

EA19-14

Traces - ICP-MS

Eu and Yb prepared by serial dilution of the

1000 ppm single element standards.

1.3. 200 ml of a 100 ppm multielement standard

(called REEA') containing Ce and Lu is prepared by serial dilution of the appropriate

1000 ppm single element standards.

1.4. 200 ml of a 100 ppm multielement standard

(called REEB) containing Gd, Dy, Tb, Ho, Er and Tm is prepared by serial dilution of the appropriate 1000 ppm single element stand ards.

1.5. 200 ml of a 100 ppm solution containing Ru and Re is prepared by serial dilution of the appropriate 1000 ppm single element stand ards.

1.6. 200 ml of a l O ppm solution containing Y, La,

Ce, PT, Nd, Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm,

Yb and Lu (called REEA') is prepared by serial dilution of the 100 ppm solutions

REEA'and REEA".

1.7. The REESTD calibration standard (contain ing Y, the REE, Ru and Re) is prepared by serial of the appropriate 10 ppm calibration standards.

1.8. The REEB and REEA' interference solutions are prepared by weighing 0.200 g of the ap propriate 100 ppm stock solution, transferring to a 250 ml volumetric and making up to volume.

1.9. The 2 ppm B a interference solution is prepared by weighing 0.50 g of 1000 ppm Ba stock solution and transferring to a 250 ml volumetric.

1.10. Instrument control solution identified as

CHK2 was prepared by collecting the remain ing solution after analysis until about 3 liters was collected. The solution was allowed to sit for about two weeks and then filtered through a 45 micron filter.

2. Measurement of Analyte Concentration

The operation of the Elan 250, and the measurement of the analyte concentration is outlined quite exten sively on pages EA 19-5 to EA 19-12.

3. Performing Calculations

There are two calculation algorithms in use. An inter nal standard technique has been developed for Y and the Rare Earth Elements (REE), Hf, Ta, Nb, Rb, Cs,

Sr, and Zr. All other analyses utilize external calibra tion with signal drift compensation (linear interpola tion).

3.1. Select option 2 from the Apple II main menu.

The 'CONCAL' menu will be displayed.

Place the intensity data disk in the right-hand drive.

3.2. Select the appropriate calculation algorithm.

3.3. Enter the slot number for the printer card.

3.4. Enter the parameter set name for the set used to collect the data.

3.5. After 30 seconds or so, the program will dis play the identifiers of all the runs saved to disk and prompt for the run number to be processed. Select the desired run number.

3.6. The program will prompt for the filename to be used to store instrument and preparation of

QC data. Enter the appropriate filenames. If this function is not being used, enter 'S' to bypass these functions.

3.7. The program will read the intensity data. Wait until the program is finished doing this and has displayed the prompt requesting a "Default

Dilution Factor". Enter the factor.

3.8. The program will prompt for any factors that are different from the default. Enter these factors. The program will perform the inten sity-to-concentration calculations. No addi tional input is required until the final results are printed.

3.9. The program will prompt for the number of the instrument QC solutions that were run and their positions in the run. Enter the correct information.

The program will calculate and produce a report summarizing the instrument QC data.

This information is presented as the difference between the observed and the expected values divided by the standard deviation associated with the expected value. If the error is normal ly distributed, then 669fc of the results should

EA19-15

Traces - ICP-MS

be between O and one standard deviation.

Less than Q.5% of the results will be greater than 3 standard deviations when there is no analytical bias associated with the measure ment. Thus if such an occurrence is observed, there is good reason to believe that the meas urement is biased. A run (or data between check samples) will be rejected if more than two of the check samples lie outside 3 standard deviations.

The program will calculate and produce a report summarizing the preparation QC data.

This information is presented in a manner similar to that produced f or the instrument QC.

The program will ask whether the concentra tion data are to be stored. Respond ap propriately. If the data are not to be stored the program will return to the 'CONCAL' main menu.

If the data are to be stored, four options will be presented: 1) Routine Sample Data Storage

2) International Reference Material 3) In- house Reference Material 4) Exit Menu.

Some or all of these options may be used for a set of data.

3.10. The program will display each sample iden tifier and prompt for the proper run identifier.

Routine sample data should be stored using the four digit run number assigned by the lab

(job number).

3.11. The program will prompt the operator to place the appropriate data disk in the right-hand drive. Place the disk in the correct drive.

When all data has been stored the program returns to the main 'CONCAL' menu.

Quality Control:

Precision:The CHK2 solution is used for instrument quality control purposes. The following table lists the means and associated standard deviations (established over a l month period) 0^=50). These values are used by the QC functions of the Apple II software.

TABLE MSS. COMPOSITION OF CHK2.

Element

Tb

Dy

Ho

Er

Tm

Yb

Lu

Nd

Sm

Eu

Gd

Y

La

Ce

Pr

The instrument check solution was run in batches of international reference materials used in the develop ment of the current ICP-MS T4/T5 method. Accuracy is assured in two ways: (1) by using international reference materials and (2) by matrix effect ex perimentation. The mean values established in this way for the CHK2 solution can be used to assess accuracy. Determination for a wider range of interna tional reference materials (and a larger number of analytes) has been carried out (in press). For research grade jobs appropriate reference materials are run

(e.g., SY-2, MRG-1, AC-E). Accuracy is about ±5^o.

Quality assurance:

Mean

(ug/g)

18.32

20.535

44.164

5.013

19.347

3.962

1.121

3.629

0.552

3.447

0.703

2.017

0.284

1.892

0.279

Std. Dev

(n-1)

0.479

0.422

1.14

0.147

0.535

0.128

0.038

0.118

0.022

0.123

0.031

0.089

0.018

0.075

0.016

The digestion and instrument control data for a par ticular batch are compared with the above values by calculating the number of standard deviations the con trol solution data differ from the mean. These calcula tions are printed at the end of the interim report produced by the Apple concentration calibration/cal culation software. This interim report is filed along with the original bench sheets for the job. The run number associated with the data is also included on the bench sheets.

The run numbers on a particular data disk are written on the disk label. The run numbers on a particular data disk can also be identified by using the Apple utility program 'FIXRUNQUE'. With all this information, data can be retrieved as required.

EA19-16

Traces - ICP-MS

Productivity:

Ideally, a technician should be able to complete about

60 solutions per day. Data reduction and reporting are not included.

Additional Notes:

1. If a clean sampler and skimmer are being used, a

1000 ppm Ca solution must be nebulized to protect them. This Ca standard must be prepared from the pure Johnson Matthey CaCO3 powder. The ap propriate weight of CaCO3 is dissolved in a mini mum volume of concentrated HNO3 and then made up to volume with distilled deionized water.

If work has been done in the vacuum chamber or if a new sampler or skimmer is being used for the first time, use appropriate lens optimization proce dures (Doherty, 1989).

If work has been done on the interface plates, it is

critical that the plates are centered properly on the vacuum port. The validity of the Ru-Re internal standard scheme depends on this.

2. Under no circumstances should the deposits on the shadow stop (at the base of the skimmer) be removed as instrument drift will become uncon trollable. It takes about five working days nebuliz ing 1000 ppm Ca almost continuously to build up the required electrically insulating layer.

3. The air filters at the base of the RF generator and the filters at the head of the Elan (computer end) should be cleaned as required (about every 2-3 months).

4. The mechanical pump oil and alumina balls should be changed about every 5 to 7 weeks.

5. Samplers and skimmers should be cleaned at the end of every second work day. See senior staff if you have never done this before.

6. Empty the slops buckets (located behind the Elan) every three days.

Bibliography:

Doherty, W., 1989, Spectrochimica Acta, Vol. 44B, page 263.

EA19-17

Traces - ICP-MS

THE TRACE 5 (T5) PACKAGE

INDUCTIVELY COUPLED PLASMA SOURCE

MASS SPECTROSCOPY

Introduction:

This suite of elements (Hf, Ta, Th, U; options: Pb, Sn,

Tl) has been traditionally difficult to determine in geological materials because the host minerals are often refractory (difficult to dissolve and treat analyti cally). Futhermore, these elements usually occur in discrete accessory mineral phases thus meaning spe cial attention must be paid to sample collection and preparation. Finally, these elements are usually found only at trace levels in most geological materials. ICP-

MS provides the low determination limits required by geologists.

Elements in the T5 package have varied geological significance and, although determined as a group, these elements are not used as a single suite. Ta and

Hf, found in rock-forming and accessory minerals, are used extensively to characterize the tectonic setting of basalt and granite. Ta, U, Sn, Th are used to monitor or evaluate processes that control the formation and differentiation of magma within magma chambers.

They are also used to discern the possible ore potential in granitic rocks. Pb is incorporated into either galena

(PbS) as a major constituent or at trace levels into potassium feldspars (common in granites). Because three Pb isotopes are produced by the radioactive decay of U orTh, the assessment of the amount of lead present in a sample can be a preselection criteria for geochronology samples. Tl is a rare element that is often associated with Au, Ag and PGE minerals. Other elements such as Rb, Cs, Sr, Zr, and Nb can be deter mined in the T5 package.

TABLE MS6. DETERMINATION LIMITS

AND PRECISION FOR TRACE

ELEMENTS (T5)

Element Determination Optimum Precision*

Limit Range

(ppm) (ppm)

Hf

Ta

Th

U

Options

Pb

Sn

Tl

Rb )

Cs )

Sr )

Zr )

Nb )

0.05

0.05

0.10

0.10

0.10

0.50

0.10

0.20 - 100

0.05- 20

0.10-100

0.10-100

0.10- 100

0.50- 100

0.10- 100

Under Development

0.05

0.05

0.10

0.10

0.10

0.50

0.10

* Precision is quoted as the 959fc Confidence Limit

(M-g/g) for a value at l Ox the determination limit

(i.e. absolute, not relative).

Reagents:

Refer to the Trace 4 method.

Safety advisory:

Refer to the Safety Advisory Section in the Trace 4 elements page EA 19-13.

Procedures:

Refer to the Trace 4 method.

Method:

The method is similar to that used for the Trace 4 elements, page EA 19-14.

Quality Control:

Refer to the Trace 4 method.

Table MS7 contains the data for the instrument check solution. The validation of the current ICP-MS method for the other T5 elements is in progress.

Apparatus:

Refer to the Trace 4 method.

EA19-18

Traces - ICP-MS

TABLE MS7. COMPOSITION OF T5 CHECK

SOLUTION

Element Mean (ppm)

Std. Dev.

Rb

Sr

Zr

Nb

Cs

Hf

Ta

Accuracy is about S^c.

41.406

306.34

94.86

7.691

1.349

2.479

0.349

0.958

6.507

2.146

0.187

0.036

0.081

0.012

Productivity:

A technician should be able to complete 60 samples a day. This does not include data reduction and reporting time.

Additional Notes:

1. The T5 package requires three separate solution with the T5 elements separated as follows: a. Hf, Ta, Sr, Zr, Nb, Rb, Cs b. Pb,Th, U,T1 c. Sn

The 80 ppb working solution is prepared by serial dilution of the 1000 ppm stock solutions.

2. The Ru-Re calibration scheme is used for Rb, Cs,

Nb, Sr, Zr, Hf and Ta. All others use the external calibration procedure. Thus, separate runs on the

Elan 250 are required to accomplish a full T5 analysis.

3. Sn fusion solutions received from the Chemistry

Subsection are further diluted by a factor of 5 for a final dilution factor of 2500.

EA19-19

Traces - NAA

DETERMINATION OF THE RARE EARTH ELEMENTS

(La, Ce, Nd, Sm, Eu, Yb, Lu), THORIUM, TANTALUM, HAFNIUM,

URANIUM AND SCANDIUM BY

INSTRUMENTAL NEUTRON ACTIVATION ANALYSIS

Introduction:

Instrumental neutron activation analysis (INAA) provides non-destructive determination of the REE,

Th, Ta, Hf, U and Se. These elements are extremely useful as tracers in a wide variety of geological proces ses (refer to Volume I Chapter 16).

An advantage with this method is that it does not involve dissolution. A disadvantage is that it does not allow for the determination of all rare earth elements.

Safety advisory:

1. Sample containers should be air-tight and handled with extreme caution to avoid the leakage of radioactive rock dust after the samples are removed from the reactor and attached to the aluminium counting plates with masking tape.

2. Avoid irradiating samples which become very

'hot' during long irradiations. If the samples are not common silicate rocks, then check to ensure that long-lived radioisotopes are not produced during irradiation. Tantalum ores are an example of samples which become 'hot'.

3. Ensure that samples > l % in volatile elements such as chlorine are not introduced into the reactor.

Users must take an official University of Toronto safety course before commencing work, and the user must be signed-in to the reactor facility in the presence of the supervising staff (R.G.V. Hancock).

Method:

The method consists of the following techniques:

1. Sample preparation - bagging and accurate weigh ing of the samples and standards

2. Irradiation of samples - placing them into the nuclear reactor for a 16 hour time period

3. Measuring the concentration - counting the

'cooled' samples after 7 and 40 days, with on-line data reduction

4. Calculation of the results based on a comparison of counts from the samples with counts from standards

Apparatus:

- Bag sealing unit

- Slowpoke II nuclear reactor

- Gamma counter and multichannel analyzer equipped with computer for on-line data reduc tion

Reagents:

None required.

Procedure:

1. Sample preparation

A wide range of geological samples can be analyzed by neutron activation, and there are many different preparation techniques. The following procedure may be followed:

1.1. The weight of the rock powder is recorded to four decimal places. Pulverized rock powder weighing 300-400 mg is sealed in plastic using a bag-sealing unit. The container is marked using a indelible marker pen. UTB-1 is used to make up a reference standard, and either

WHIN SILL (the Open University irradiation standard) or BHVO-1 are used to monitor quality within and between batches of samples.

1.2. Up to twenty small 'baggies' are placed in each cylinder for irradiation. Normally, there is no need for flux corrections when irradia tion is completed in a small reactor of the

SLOWPOKE type.

EA20-1

Traces - NAA

1.3. Samples are irradiated for 16 hours, and counted 7 and 40 days after they are removed from the reactor.

examined and regions of interest established whilst data acquisition continues.

2. Determination of element concentrations

The sample is affixed to the center of an aluminium plate so that it is flat and free of wrinkles. The plate is inserted into a rack on the nose of the detector. Data are acquired using a Canberra multi-channel analyser

(MCA) with the spectral regions of interest defined in

Table NAA1.

Each sample is counted for 5000 or more live seconds to achieve a 2a error of 10%, or better, for the peaks of interest. Dead time must be < 159fc or an alternative sample with lower dead time should be counted first.

The multi-channel analyzer is equipped with two detectors (ADC#1 and ADC#2) and collects counts simultaneously from both detectors. Peaks may be

Quality Control:

UTB-1 (University of Toronto Basalt Standard) is used in establishing a primary calibration. Secondary stand ards are used to ensure precision and accuracy are within acceptable limits (Refer to NAA2); various external standards are available for this purpose.

Productivity:

This technique is not automated, and is extremely labour intensive. It may take about eight days over a

60 day period to produce data for 50 samples.

Bibliography:

Frogatt, P., (unpublished), Analyzing for Rare Earth

Elements by Neutron Activation at the University of

Toronto Slowpoke Reactor.

EA20-2

Traces - NAA

TABLE NAA1. TABLE OF ELEMENTS AND INTERFERENCES PEAKS FOR INAA COUNTING

Element

Ho

Nd

Sm

U

Mo

Lu

Ba*

Yb

Tb

Th

Yb*

Au

As

La

Se

Rb

Fe

La*

7 day counts

Peak

(eV)

80.6

91.0

396.1

411.8

559.1

815.6

889.3

1076.6

1099.3

1596.2

103.2

106.4

140.3

208.3

216.0

282.6

298.6

311.9

Interferences

TEMP

[email protected] 92.3 &L possible Br

U,Th

U @ 209.7

[email protected] 215.6

Th (o) 300 (poor)

Th

Element

Ta*

Cr

Hf

Cs

Ni

Tb

Se

Fe

Zn

Tm

Eu

Ce

Ba

Tb*

Th*

Ta

Eu*

40 day counts

Peak

(eV)

Interferences

67.8

795.8

810.8

879.3

889.3

1099.3

1115.5

1221.3

1332.5

1408.1

84.3

122.0

145.4

216.0

298.6

311.9

320.1

482.0

[email protected]

[email protected]

Th

[email protected] 215.6

[email protected]

Tb

* Preferred peak for most rocks

TABLE NAA2 . ANALYTICAL DATA FOR ROCK STANDARDS DETERMINED

BY INAA DURING A ROUTINE ANALYTICAL SCHEME

Element

La

Ce

Nd

Sm

Th

Ta

Hf

U

Eu

Yb

Lu

Se

Mean

UTB1

(ppm)

4.21

0.96

5.06

1.02

26.4

62.1

33.6

7.98

2.30

4.05

0.64

40.0

o

0.06

0.06

0.96

0.18

1.6

2.8

2.0

0.18

0.22

0.42

0.06

1.3

Expected

Value

(ppm)

(4.3)

(1.02)

(4.6)

(1.0)

(26.7)

(60.5 )

(32.0)

(8.0)

(2.4)

(4.0)

(0.58) n.d.

Mean

WHIN SILL

2.75

1.6

4.92

0.45

24.7

60.4

28.4

7.02

2.04

2.5

0.35

30.4

a

0.26

0.9

0.48

0.14

0.5

3.8

5.2

0.26

0.10

10.14

0.02

0.8

n.d. not determined

Epected

Value

(ppm)

(3.1)

(1.26)

(4.9)

(0.90)

(22.5 )

(57.5)

(32.9 )

(7.3)

(2.3)

(2.54)

(0.39) n.d.

EA20-3

Au, Pt, Pd

DETERMINATION OF GOLD, PLATINUM AND PALLADIUM

GRAPHITE FURNACE ATOMIC ABSORPTION METHOD

Introduction:

A renewed interest in the search for deposits of Au, Pt and Pd has been stimulated by recent economic and political events. This, in turn, has led to improvements in the methodologies by which laboratories determine the levels of these elements in geochemical samples.

It is well known that Au, Pt and Pd are distributed in rocks in a heterogeneous manner, occurring as discrete particles and minerals and existing in solid solutions in sulfides, silicates, and spinels. Because of this so- called 'nugget effect', a minimum of 10 g of rock powder is used.

There has been a significant increase in the number of analytical techniques used to determine Au, Pt and Pd in the past decade. Of these techniques, there is a general consensus that graphite furnace atomic absorp tion spectrophotometry is well suited in the routine determination of these elements.

The natural abundance level for platinum is about 0.05 ppb, which is approximately eighty-fold lower than for gold. Platinum tends be be slightly more abundant than palladium.

Gold, platinum and palladium are determined in geological materials by graphite furnace atomic ab sorption spectrophotometry utilizing an autosampling system.

These elements are concentrated into a silver bead of approximately 15 mg by the classical lead fire-assay method (page M14-5). Dissolution of the silver bead is accomplished with nitric acid in a 10 x 75 mm test tube placed in an aluminum block and set on a hot plate. The silver is precipitated as the chloride with hydrochloric acid whereby the gold, platinum and palladium are dissolved in the aqua regia.

The acid mixture is diluted with distilled water, mixed and the AgCl allowed to settle on the bottom of the test tube. An aliquot of the supernatant liquid is atomized in a graphite furnace and the atomic absorption signal observed as a recorder trace.

Safety advisory:

1. The method involves the use of specific acids to dissolve the silver bead. Personnel should review the MSDS sheets for these acids prior to use.

Method:

This method consists of the following techniques:

1. Preparation of silver bead (sample fusion and ex traction)

2. Sample decomposition by acid digestion

3. Measurement using graphite furnace atomic ab sorption

4. Calculation of Au, Pt, and Pd concentrations based on a calibration curve produced from known standard solutions

Apparatus:

- Perkin Elmer model 603 atomic absorption spectrophotometer equipped with a model 56 chart recorder.

- HG A 500 graphite furnace and programmer with microcomputer controlled power supply.

- AS-1 auto sampler - includes sample table with provision for 30 samples; sampling dipper as sembly with separate electronic control unit con taining precision pumps for sampling and rinsing.

- Test tubes, borosilicate, 10 X 75 mm

- Aluminum heating block

Reagents:

- Matrix acid mixture

- Nitric acid, HNO3, (1:3)

- Hydrochloric acid, HC1, (1:3)

- 1000 ppm Au - such as supplied by Fisher Scien tific Company (gold chloride in distilled water)

- 100 ppm Au standard solution

- 10 ppm Au standard solution

- 1000 ppm Pt - such as supplied by BDH Chemical

- 100 ppm Pt standard solution

- 10 ppm Pt standard solution

- 1000 ppm Pd - such as supplied by BDH Chemi cal

EA21-1

Au, Pt, P d

- 100 ppm Pd standard solution

- 10 ppm Pd standard solution

- Working Standard Solutions microburette to 100 ml volumetric flasks.

Make to volume with the matrix acid mixture.

Procedures:

1. Reagent preparation

1.1. Nitric acid (l: 3) - To a 500 ml volumetric flask containing 200 ml of distilled water add 125 ml of concentrated nitric acid. Cool and make to volume with distilled water.

1.2. Hydrochloric acid (1:3) - To a 500 ml volumetric flask containing 200 ml of distilled water add 125 ml of concentrated hydrochloric acid. Cool and make to volume with distilled water.

1.3. Matrix acid mixture - To a 1000 ml volumetric flask containing approximately 500 ml of dis tilled water, add 62.5 ml of concentrated nitric acid and 62.5 ml of concentrated hydrochloric acid. Cool and make to volume with distilled water.

2. Standard Au, Pt, and Pd solution preparation

2.1. 100 ppm Au solution - Pipet 20 ml of 1000 ppm Au into a 200 ml volumetric flask and make to volume with 0.5 N HC1.

3. Sample decomposition (acid digestion)

3.1. Add by microburette 0.5 ml of (l :3) nitric acid to a 10 x 75 mm test tube containing the silver bead obtained from the fire-assay technique.

3.2. After dissolution of the silver, 0.5 ml of (l :3) hydrochloric acid is added and the contents mixed. The AgCl precipitate is digested in the test tube and placed in an aluminum block on a hot plate.

3.3. Heat until the supernatant solution is clear and the AgCl has coagulated as a 'lump' on the bottom of the test tube.

NOTE: Do not heat so long that a loss of volume occurs through evaporation.

3.4. To the acid mixture, add 1.0 ml of distilled water using a microburette. The test tube is then shaken and placed in the aluminum block to cool.

3.5. Transfer a portion of the supernatant solution to a sample cup to determine gold, platinum and palladium.

2.2. 10 ppm Au solution - Pipet 20 ml of 100 ppm

2.4.

Au into a 200 ml volumetric flask and make to volume with 0.5 N HC1.

2.3. 100 ppm Pt solution - Pipet 20 ml of 1000 ppm

Pt into a 200 ml volumetric flask and make to volume with matrix acid mixture.

10 ppm Pt solution - Pipet 20 ml of 100 ppm

Pt into a 200 ml volumetric flask and make to volume with matrix acid mixture.

4. Measurement of Au, Pt and Pd concentrations

Set up the Perkin-Elmer 603 atomic absorption spectrophotometer equipped with an HGA 500

Programmer and Graphite Furnace and utilizing the

AS-1 auto sampling system according to the procedure described in the manufacturer's operation manual.

The atomic absorption signal is measured with a

Perkin-Elmer Model 56 recorder. The instrument parameters for each element are listed in the following table.

2.5. 100 ppm Pd solution-Pipet 20 ml of 1000 ppm

Pd into a 200 ml volumetric flask and make to volume with matrix acid mixture.

2.6. 10 ppm Pd solution - Pipet 20 ml of 100 ppm

Pd into a 200 ml volumetric flask and make to volume with matrix acid mixture.

2.7. Working standard solutions - Prepare 0.5,

0.10, 0.20 and 0.30 ppm Au, Pt and Pd solu tions by adding 0.5, 1.0, 2.0 and 3.0 ml of 10 ppm Au, Pt or Pd solution with a 10 ml

Instrument Parameters

Au Pt

Pd

Lamp Current

Wavelength

Slit Width (setting)

Background Corrector

8mA 12mA 16mA

242.8nm 265.9nm 247.6nm

0.7nm(4) 0.7nm(4) 0.2nm(3)

On On On

EA21-2

Au, Pt, Pd

The following parameters are common for Au,Pt and

Pd:

CV

W

C s net concentration of Au, Pt or Pd in g/ml in solution

Water Coolant rate

Sample Volume

Recorder Power

Chart Speed

Range

2.5 1/min

20^1

Servo

40 mm/min

V = Volume in ml of the sample solution

(usually 2ml), and

W = weight of sample in grams (usually 10 g)

5 mv (for 0.05 to 0.30 ppm

Au and Pd)

2 mv (for < 0.05 ppm Au and 0.05 to 0.20 ppm Pd) l mv (for 0.05 to 0.20 ppm

Pt and < 0.05 ppm Pd)

Graphite furnace program steps

Quality Control:

The optimum working range for Au is 2 to 40 ppb, and l to 40 ppb for Pt and Pd (in rock).

The determination limit is 2 ppb Au in rock, and l ppb for Pt or Pd.

The estimated precision, at 959c confidence limit (2o), for a value at 10 times the determination limit, is ± 5 ppb for Au and ± 3 ppb for Pt and Pd (relative).

A. Drying Temperature

Ramp time

Hold time

B. Charting Temperature

Ramp time

Hold time

C. Atomization Temperature

Ramp time*

Hold time

1200C

10 sec

25 sec

8000C

15 sec

15 sec

27000C

O sec **

5 sec

Accuracy is comparable to precision based on data collected from standard reference materials.

Productivity:

A technician should be able to complete 20 samples per day if there are no usually high concentrations.

Additional Notes:

Enter - 5 and press REC key, READ key and INT.

GAS key

1. Standards must be prepared with the same acid concentrations contained in the samples, that is, nitric acid (1:15) and hydrochloric (1:15).

* Must be entered to automatically apply maximum power heating. Calibration of the optical tempera ture control is necessary.

** For Pt use 5 sec.

2. All dilutions must be made with an acid mixture containing these same concentrations.

3. Standards should be prepared every 2 weeks and stored in polypropylene containers.

5. Calculation of the results

The concentrations of gold, platinum and palladium in the sample solution are read with the aid of a calibra tion graph. The net concentration is obtained by sub tracting the average overall-blank value. The concentration in the sample is calculated according to the formula:

4. Concentrations of 50 ppm of platinum, palladium, and silver have no interfering effect on a gold absorbance represented by 0.2 ppm.

5. Concentrations of 50 ppm silver, gold and platinum have no interferring effect on a palladium absorbance represented by 0.2 ppm.

6. The following observations on the absorbance of a 0.2 ppm Pt solution are: where :

M = ppb of Au, Pt or Pd in rock

50 ppm Ag reduced the absorbance by 59fc,

50 ppm Au and 5 ppm Pd had no effect, and

50 ppm Pd reduced the peak absorbance by

EA21-3

Au, Pt, Pd

Bibliography:

Mologhney, P.E., 1980, A fire-assay and Wet Chemi cal Method for the Determination of Palladium,

"Analytical Methods for Atomic Absorption Spectros- Platinum, Gold, and Silver in Ores and Concentrates, copy Using the HGA Graphite Furnace", Revised Talanta, Vol. 27, No. 4, pp 365-367.

March 1977, Perkin Elmer Corporation, Norwalk,

Connecticut U.S.A.

"The AS-1 Automatic Sampling System", Revised Absorption Spectrophotometry, Anal.Chem., Vol.

October 1978, Perkin Elmer Corporaton, Norwalk, 246, pp 122-124.

Connecticut, U.S.A.

Van Loon, J.C., 1969, Determination of Platinum,

Palladium, and Gold in a Silver Assay Bead by Atomic

EA21-4

Gold

DETERMINATION OF GOLD IN NATURAL WATERS

SOLVENT EXTRACTION AND ELECTROTHERMAL

ATOMIZATION METHOD

Introduction:

A renewed interest in the search for deposits of gold has been stimulated by recent economic and political events. This, in turn, has led to improvements in the methodologies by which laboratories determine the levels of these elements in geochemical samples.

The determination of gold in water is based on the method described by McHugh (1984) in which a one liter sample is evaporated to dryness, the residue is dissolved in bromine-hydrobromic acid, gold is ex tracted into methyl-isobutyl ketone, and determined by graphite furnace atomic absorption spectrophotometry

(GFAAS).

The gold content in the MIBK portion (or in the l liter of water) is calculated based on the concentration of gold in MIBK and the volume of MIBK. The extrac tion procedure includes back washing with dilute HBr to remove unwanted iron in the MIBK phase, thereby optimizing the accuracy of measurement.

Safety advisory:

1. Reagent preparation and sample extractions are to be done in a fume hood to avoid inhaling and exposure to MIBK vapor and acid fumes.

2. Great care should be exercised in pipetting bromine, to avoid dripping or spilling of the ob noxious liquid. (Never pipet by mouth, always use a pipette bulb)

Apparatus:

- Perkin Elmer Model 603 atomic absorption spectrophotometer equipped with a model 56 chart recorder.

- HGA 500 graphite furnace and programmer with microcomputer controlled power supply.

- AS-1 autosampler

- Pyrolytically coated graphite tubes

- Hotplate

- Glass beakers, 400 ml

- Glass filtering funnels and Whatman No. 40 filter paper

- Glass separatory funnels, 125 ml

- Glass test tubes, 10 X 75 mm

Reagents:

- Hydrobromic acid, HBr, 48^

- Hydrochloric acid, HC1, 389fc

- Nitric acid, HNO3, 709fc

- Bromine

- Q.5% bromine in hydrobromic acid

- Cleaning solution

- Methyl-isobutyl ketone, MIBK

- 0. IN HBr solution

- 59fc (v/v) Br2-HCl solution

- 1000 ppm Au standard solution - (gold chloride

in distilled water)

- l ppm Au in 89fc HC1 solution

- 100 ng Au/ml in S9c HC1 solution

- 5 ng Au/ml in 0.5 ^c bromine in hydrobromic acid

Procedures:

Method:

The method consists of the following techniques:

1. Extraction and preconcentration of the gold

2. Measurement of the gold absorption signal by

GFAAS

3. Calculation of the gold concentration based on a calibration curve produced from known standard solutions

1. Reagent preparation

1.1. Q.5% bromine in hydrobromic acid - pipet l ml of bromine into a 200 ml volumetric flask and make to volume with hydrobromic acid.

1.2. 0.1N HBr solution - Add 6 ml HBr to 500 ml of distilled water in a one liter separatory fun nel. Equilibrate with 25 ml of MIBK. Discard excess MIBK. Store in a stoppered glass con tainer.

EA22-1

Gold

1.3. 59fc (v/v) Br2-HCl solution - pipet 10 ml of bromine into a 200 ml glass volumetric flask and make to volume with concentrated hydrochloric acid.

1.4. Cleaning solution - mix nitric acid, hydrobromic acid and distilled water in the ratio of 1:3:10.

3.8. Add 2 ml MIBK from a 10 ml microburette to the separatory funnel.

3.9. Shake for five minutes.

3.10. Allow layers to separate (about 15 minutes) and drain off the aqueous layer (centrifuge if emulsion forms).

2. Standard Au solution preparation

2.1. l ppm Au in 89fc HQ - prepare accurately by serial dilution of the 1000 ppm Au standard solution with 87o HC1.

2.2. 100 ng/ml of Au in 89fc HC1 - pipet 5 ml aliquot of the l ppm Au solution into a 50 ml glass volumetric flask and make to volume with 89fc

HC1.

2.3. 5 ng/ml of Au in Q.5% bromine in hydrobromic acid - pipet 2.5 ml aliquot of the 100 ng/ml Au solution into a 50 ml glass volumetric flask.

Make to volume with Q.5% bromine in hydrobromic acid. (This solution is stable for l month).

3.11. Add 8 ml 0. IN hydrobromic acid solution from a burette.

3.12. Shake for one minute, allow to stand for 15 minutes and drain off aqueous layer.

3.13. Transfer the MIBK extractant solution to a test tube using a disposable pipette.

3.14. Determine absorbance by graphite furnace

A AS technique.

3.15. Prepare 5, 10, and 20 ng standards daily by pipetting 1,2, and 4 ml of 5 ng/ml Au in Q.5%

Br2-HBr solution into three separatory funnels followed by 6, 5, and 3 ml of Q.5% Br2-HBr solution respectively. Add 7 ml of distilled water to each of the funnels. Continue as from step 3.8 to 3.14.

3. Extraction and preconcentration of the gold

3.1. Add 10 ml of 5^o (v/v) Br2-HCl solution to the water sample in the polyethylene container to desorb Au from the container walls. Allow to stand for 2 days with occasional shaking.

3.2. Filter one liter of water sample using a What man No. 40 filter paper.

3.3. Evaporate the water sample in aliquots on a hot plate using a 400 ml beaker. It will take about

15 hours to evaporate a one liter sample to near dryness. Do not allow the sample to boil, par- ticularily toward the end of the evaporation.

3.4. Add 7 ml Q.5% bromine in hydrobromic acid from a burette.

3.5. Warm gently and alio w to stand for 15 minutes.

3.6. Transfer the solution to a clean 125 ml separatory funnel.

3.7. Rinse the beaker with 7 ml distilled water added from a burette and add the rinsings to the separatory funnel.

4. Measurement of Au concentration

4.1. Set-up the Perkin-Elmer 603 atomic absorption spectrophotometer equipped with a HGA 500 programmer and graphite furnace and utilizing the AS-1 auto sampler according to the proce dure described in the manufacturer's operation manual. The atomic absorption signal is measured as peak height with a Perkin-Elmer

Model 56 chan recorder. The instrument parameters are listed in the following table.

Instrument Parameters

Wavelength

Lamp Current

Slit with setting

Deuterium arc back ground corrector

Signal

242.8 nm

8mA

4 (0.7 nm) on absorbance

EA22-2

Gold

Recorder

Power

Chart Speed

Range servo

40 mm/min.

ImV

P-E Model HGA 500 graphite furnace and AS-1 auto sampler

Purge gas

Sample Volume

Rinsings argon

20 ul water

Quality Control;

The optimum working range is 5 ng - 50 ng Au/1.

The determination limit, expressed as three times the standard deviation of the reagent blank, is 3 ng/1.

The precision at 959fc confidence limit (2a), at the 10 ng/1 level is ± 2 ng/1, based on 5 separate measurements carried out on spiked bulk control samples.

The accuracy is similar to the precision for this method.

Productivity:

A technician should be able to complete 18 determina tions per week or about 4 determinations per day.

Graphite furnace program steps

A. Drying Step

Temperature

Ramp time

Hold time

B. Charring Step

Temperature

Ramp time

Hold time

1200C

10 sec

10 sec

8000C

10 sec

5 sec

C. Atomization Step

Temperature

Ramp time

Hold time

27000C

3 sec

3 sec

Enter -5 and press REC key, READ key, and INT

FLO W key.

4.2. Record the absorption signals on chart paper.

Measure the peak height of the standards and draw a calibration graph.

Additional Notes:

1. To avoid gold contamination, glassware, such as separatory funnels, must be cleaned thoroughly after being used. An effective procedure is as follows: Fill glassware with cleaning solution and allow to stand overnight. Wash with a stream of tap water and rinse with distilled water.

2. GFAAS absorbance signals vary from day to day.

Calibration should be performed on a daily basis.

3. MIBK waste is not to be disposed of in the sink.

It should be collected in a well labelled bottle, and the bottle should be tightly covered. A profes sional disposal firm will dispose of the bottle.

4. Avoid draining the MIBK layer through the stem of the separatory funnel. Instead, the MIBK por tion should be picked up by a disposable pipette, and tranferred to a 10 X 75 mm test tube.

5. Calculation of the results

The absolute quantity (ng) of Au in the sample solu tion is read with the aid of a calibration graph estab lished using 5, 10, and 20 ng standards. The net quantity is obtained by subtracting the average of three or more blank values for a run. The gold con tent in water is expressed as total soluble Au in ng per liter of water.

Bibliography:

Hall, G.E.M., Vaive, J.E., and Ballantyne, S.B., 1986,

Field and Laboratory Procedures for Determining

Gold in Natural Waters: Relative Merits of Precon- centrations with Activated Charcoal. Journal of

Geochemical Exploration, Vol. 26, pp 191-202.

McHugh, J.B., 1984, Gold in Natural Water: A Method of Determination by Solvent Extraction and

Electrothermal Atomization., Journal of Geochemical

Exploration, Vol. 20, pp 303-310.

EA22-3

Flouride

DETERMINATION OF FLUORIDE

AUTOMATED COLORIMETRIC METHOD

Introduction:

The crustal abundance of fluorine is about 0.08 9fc.

Fluorine has an ionic radius just slightly smaller than oxygen and as a result can substitute for oxygen in oxy-minerals, if valency compensations are allowed.

In some hydrous minerals such as amphiboles, mica and apatite, the replacement is quite easy. Fluorine also occurs in the following minerals: fluorite (CaF2), topaz (Al2F2SiO4), villiaumite (NaF) and amblygonite

(LiAlFPO4).

This automated procedure for the determination of fluoride in fused rock solution samples is based upon the distillation of hydrogen fluoride and subsequent reaction of the distillate with alizarin fluorine blue-lan thanum reagent to form a lilac-blue complex which is measured colorimetrically at 620 nm. The fusion and distillation processes are represented by the following chemical equations:

2NaOH

2NaF * Ca(OH)2

2 NaF 4- H2S04 distillation 2HF

Safety advisory:

Method:

1650C

This method consists of the following techniques: l. Fusion and sample decomposition

Na2S04

1. This method makes use of various acids. Make sure that all pertinent MSDS sheets are reviewed before starting.

2. In preparation of the (1:1) H2SO4, slowly add sulphuric acid in portions to 500 ml of distilled water and mix well. Cool the solution in a water trough, and ensure that the temperature of the solution is kept below its boiling point during preparation.

3. Preparation of the acetate buffer solution and the alizarin fluorine blue stock solution should be carried out in a fume hood to avoid inhalation of vapors of glacial acetic acid and ammonium hydroxide.

2. Distillation for separation of analyte

3. Colorimetric measurement

4. Calculation of F concentration based on calibra tion curve produced from known standard solu tions

Apparatus:

- Nickel crucibles, 40 ml

- Hotplate

- Plastic beakers, 50 ml graduated

- Lachat colorimeter with flow cell and 620 nm interference filter

- Recorder, Linear Model 1200

- Technicon Auto Analyzer II for fluoride analysis consisting of the following components:

A. Sampler IV with 5 ml plastic sample cups

B. Proportioning Pump III with pump tubes required for fluoride analysis. A flow diagram of the analytical system is shown in Figure FI.

C. Microdistillation Apparatus (shown sche matically in Figure F2) The major com ponents include the following:

- Heating bath, equipped with electric heating bar, thermo-regulator and motor-driven stirrer

- 15 ft. Teflon coil, 1/8" i.d., immersed in the silicone oil contained in the heating bath

- rotary vacuum pump with vacuum gauge and regulator

- fractionation column of borosilicate glass

- water-jacketed condenser

- distillate collector

- waste condenser and trap assembly

D. Cartridge consisting of housing which supports the reagents tubing, manifold, mixing coils, connectors and fittings.

E. Voltage stabilizer.

Reagents:

- Sodium hydroxide pellets, NaOH, < fluoride

- Perchloric acid, HC1O4, (1:1)

- Sulphuric acid, H2SO4, (1:1)

- Acetone buffer, pH 4.0

- Acetone l ppm

EA23-1

Flouride

- Alizarin fluorine blue

- Lanthanum nitrate stock solution

- Brij 35 wetting agent, 309fc solution

- Alizarin reagent

- EDTA reagent, 19fc w/v

- Stock standard flouride solution

- Working standard solutions

Procedures:

1. Reagent preparation

1.1. Perchloric acid, HC1O4, (l: l) - Mix 500 ml of perchloric acid (707c) with 500 ml distilled water.

1.2. Sulphuric acid, H2SO4, (1:1) - Carefully add

500 ml of concentrated sulphuric acid to 500 ml of distilled water. Mix and cool to room temperature before use.

1.3.1. Acetate buffer, pH 4.0 - Dissolve 60 g of sodium acetate trihydrate in 500 ml of distilled water.

1.3.2. Add 100 ml of glacial acetic acid and dilute to l liter with distilled water.

1.7. EDTA reagent, l*7c w/v - Dissolve 10 g of tetrasodium ethylenediamine tetracetic acid, along with a few pellets of sodium hydroxide in distilled water and dilute to l liter.

2. Standard F solution preparation

2.1. Stock standard solution, 100 u.g F/ml - Dis solve 0.2207 g of sodium fluoride in distilled water and dilute to l liter. Store in a polyethylene bottle.

2.2. Working standard solutions - Prepare 0. l, 0.2,

0.3,0.5, and 1.0 |J.g F/ml standard solutions by serial dilution of the stock standard solution with distilled water. The standard solutions should contain 4 g of sodium hydoxide and 24 ml (l: 1) perchloric acid per 100 ml in order to match the amounts of these substances used in alkali fusion of the rock samples. Store in polyethylene bottles.

2.3. Diluent - Dilute 25 ml of 100 fig F/ml stock standard solution to l liter with distilled water and mix.

3. Fusion and sample decomposition

3.1. Weigh out 0.050 g of rock sample and transfer into a 40 ml nickel crucible.

l .4. l. Alizarin fluorine blue (3-amino-ethylalizarin-

N, N-diacetic acid) stock solution, 0.01M -

Suspend 0.963 g of alizarin fluorine blue in about 100 ml of distilled water.

1.4.2. Add 2 ml of concentrated ammonium hydroxide and shake until the dye has com pletely dissolved.

l .4.3. Add 2 ml of glacial acetic acid and dilute to 250 ml with distilled water. Store in an amber bottle at 40C.

l .5. Lanthanum nitrate stock solution, 0.01M - Dis solve 1.082 g of lanthanum nitrate,

La(NO3)3.6H2O in 100 ml of distilled water and dilute to 250 ml with distilled water.

1.6.1. Alizarin reagent - Mix the solutions in the following order: 300 ml of acetate buffer, 150 ml of acetone, 50 ml of t-butyl alcohol, 36 ml of alizarin fluorine blue stock solution, 40 ml of lanthanum nitrate and 2 ml of Brij-35.

3.2. Add 0.2 g of sodium hydroxide pellets (ap proximately 12), and fuse the sample in a muf fle furnace at 6000 C for 5 minutes.

3.3. Remove the crucible and swirl to suspend all paniculate matter.

3.4. Cool and dissolve the melt with water by heat ing the contents on a hot plate at low heat for half an hour.

3.5. Transfer the contents into a plastic beaker graduated at 50 ml.

3.6. Rinse the crucible with distilled water and add to the beaker. Add 12 ml (l: l) perchloric acid to dissolve the suspended particulates.

1.6.2. Dilute to l liter with distilled water. The reagent is stable for at least 2 days. Store at 40C when not in use.

3.7. Dilute the sample to 50 ml with distilled water.

Mix the solution thoroughly with a plastic stir ring rod.

3.8. The solution is ready for the determination of fluoride by an Auto Analyzer. One fusion blank is normally prepared for each 12 samples.

EA23-2

Flouride

4. Measurement of F concentration

4. l. Start Up Procedure for AutoAnalyzer

4.1.1. Inspect tubing connections between distillation apparatus, sampler and proportioning pump and confirm that the analytical system is set up properly.

4. l .2. Turn on the heating bath. It takes about l hour to heat up the silicone oil to the preset tempera ture of 1650C.

4.1.3. Turn on colorimeter (warm-up time is 20 min.).

4. l .4. Fill the sample cups and place them in order on the sample tray. Mount the loaded sample tray onto the sampler.

4.1.5. Engage the pump platen and switch on the proportioning pump with all the reagent tubes dipped in distilled water.

4.1.6. Turn on the vacuum pump.

4.1.7. Gently turn on the cold water tap, and let the cold water flow steadily through the con densers and drain into the cup sink.

4.1.8. Switch on the recorder.

4.1.9. Check the vacuum system and the reagents flow system to ensure no leaks or spills. Dis tillate should now fill the collector.

4.1.10. Insert the reagent tubes into their correspond ing solutions (Figure FI) and establish a base line tracing on the recorder chart.

4.1.11. Press POWER push button to turn on the sampler.

4.1.12. The sample is pumped, together with (1:1)

H2SO4 and air, through the teflon coil which is immersed in the heating bath. The fluoride in the sample reacts with sulphuric acid in the teflon coil to generate HF and water vapor which distill and condense into the collector, while the non-volatile solutions, including

H2SO4 and HGO4, are separated and drained off under vacuum. The distillate is pumped through the resample tube and combined with the alizarin fluorine blue reagent to develop a colored stream. The stream then passes through a flow cell and the absorbance is measured at 620 nm. Results appear as peaks tion: where : traced on a chart recorder. The lag time from sampling to the appearance of a peak on the strip chart is about 3 minutes.

5. Calculation of the results

The fluoride concentration in the sample solution is read with the aid of a calibration curve, subtracting the blank value. The fluoride content of the rock sample is calculated in ppm according to the equa

F =

(usually 50 ml)

(usually 0.05 g)

CV

W

F = ppm of fluoride in rock

C = jig of F/ml of the sample solution

V = volume in ml of sample solution

W = weight of sample in grams

6. Shutdown

6.1. Place all reagent tubes in distilled water and run water through the system for approximately 10 minutes.

6.2. Switch off the sampler when the sample probe is in the water reservoir position.

6.3. Switch off the recorder.

6.4. Switch off the colorimeter.

6.5. Switch off the proportioning pump and h'ft off the pump platen.

6.6. Turn off the vacuum pump.

6.7. Turn off the tap water.

6.8. Turn off the heating bath.

6.9. Lift off sample tray and discard sample solu tions. Sample cup can be reused after rinsing with distilled water.

6.10. Discard the waste liquid in the liquid trap.

EA23-3

Flouride

7. Maintenance

Cleaning of teflon distillation coil - This is ac complished by inserting the tygon tubes connected to the air inlet line of the distillation apparatus (Figure F2) briefly into l ^c Na^DTA solution. A quantity of the wash fluid will be swept through the teflon coil and will remove deposited material. When all paniculate matter is removed from the coil, wash briefly with distilled water several times.

Pump tubes should be replaced after 200 working hours or prior to that if they become inflexible or flattened. They should always be left in a relaxed position when not in use.

The nickel crucible should be cleaned as soon as possible after use. The most effective way is to place them in hot 3N HC1 for 10-15 min; transfer to a 59fc

NaOH solution and boil for half an hour; wash with

Alconox solution; rinse with tap water and finally rinse with distilled water.

Quality control:

The optimum working range for the analytical system is 0.2 to 2.0 u.g F/ml, which is equivalent to 200 - 2000 ppm in rock.

The determination limit is 40 ppm in rock.

The precision, expressed at the 95*7o confidence limit

(2o), is 107c (relative), e.g. 400 ppm ± 40 ppm.

The accuracy is comparable to the precision based on data collected from standard reference materials.

4. The quality of alizarin fluorine blue is crucial to good performance. It has been necessary to check several suppliers to find an acceptable grade of reagent. Using good grade reagent, an intense colored complex formed with fluoride can be developed. The colour intensity or signal response can be monitored with a standard fluoride solution in order to determine whether the quality of the reagent is satisfactory.

5. Refer to Technicon operation manual for trouble shooting should instrumental problems arise.

6. Chloride and nitrate ions in percent concentrations can be distilled with fluoride ion and will interfere with the analysis by bleaching the alizarin fluorine blue - lanthanum reagent. Such high concentra tions of chloride and nitrate are not expected in rock samples.

7. The alizarin working reagent is stable at 40C for at least 7 days.

8. Samples with high fluoride concentration should be diluted to an appropriate level to fall within the optimum working range of 0.2 to 2.0 |ig F/ml, with a solution containing 4g NaOH and 24 ml (1:1)

HClO4 perlOOml.

9. By introducing diluent into the reacting stream as indicated in Figure FI l, a linear calibration curve can be maintained, even at low concentrations of F.

Productivity:

A technician should be able to complete 10 deter minations per day.

Additional notes:

1. Do not allow the platen to remain down on the pump tubes when the pump in not operating. The fixed pressure will damage the pump tubes.

2. The vacuum gauge reading is set at 200-300 mm

Hg. Adjustment is normally not required.

3. Occasionally, recovery of fluoride added to samples should be determined. Low values indi cate a loss of fluoride possibly during pretreat- ment; high values indicate contamination.

Bibliography:

Operation Manual for the Technicon Auto Analyzer II,

Technical Publications, No. TAO-0159-10, No. TA1-

0257-10, No. TAO-0219-20 and No. TD2-0170-00,

1972

"Fluoride in Plant Tissues", Industrial Method No.

206-72A, Revised August 1978, Technicon Industrial

System, Tarrytown, N.Y.

Chan, C., 1983, Semiautomated Determination of

Fluoride in Rocks, American Laboratory, Vol. 15, No.

10, pp 32-41.

Cralley, L.V., and Weinstein, L.H., 1969, Tentative

Method of Analysis for Fluoride content of the Atmos phere and Plant Tissues (Semiautomated Method),

Health Laboratory Science, Vol. 6, No. 2, pp 84-101.

Fuge, R., 1981, Determination of Fluorine and

Chlorine in eight USGS Reference Samples Using

EA23-4

Automated Photometric Analysis, Geostandards

Newsletter, Vol. V, No. 2.

Kesler, S.E., and Van Loon, J.C., 1973, Analysis of

Fluoride in Rocks and an Application to Exploration,

Journal of Geochemical Exploration, Vol. 2, pp 11-17.

Flowride

EA23-5

r luunae

O n

S 5*

—3 to

o 2.

PIT,

•••i

•M

•M

**—

M rt-

CP

^

3 a o

*

ro

*

S

O

M

(O

ea l 6

O l

*

O

to

O

M t*

O o

I o

99

K

0000

0

3

^

*

4k.

ro

4

* ^ i x"

000

V

W

CO

^^ k

H

*

O\

O s o

\ tr1

C

•d

C

•d

G to

4 o o *d

^ ^^

S t*

O OJ

( *

^ to

10 O

•^ ^^

N* d

3 t

5

90

D k

S

^v

s

V

\ t

t

^

w

C

)

A

O

O s o

O

S p*

J o

H

A

Figure FI. Flow diagram of analytical system for fluoride (Range: 0-2 jig/ml).

EA23-6

Flouride

(a) Jacket of heating bath

(b) Immersion electric heater

(c) Thermometer - thermoregulator, range 0-2000C

(d) Motor driven stirrer

(e) Flexible Teflon tubing, 15ft long coiled on a rigid support

(f) Silicone oil

(g) Fractionation column of boro-silicate glass

(h) Water-jacketed condenser

(i) Distillate collector

(j) Waste liquid condenser

(k) Air inlet

(1) Sample inlet

(m) Sulfuric acid (1+1) inlet c. w. refers to "cold water" t.v. refers to "to vacuum"

Figure F2. Schematic drawing of microdistillation apparatus.

EA23-7

Chloride

DETERMINATION OF CHLORIDE

AUTOMATED COLORIMETRIC FLOW INJECTION ANALYSIS

METHOD

Introduction:

Chlorine is a strong hydrophile element and as a result concentrations of chlorine reported in rocks and minerals are erratic.

Chlorine has a larger ionic radius than oxygen and is not easily accommodated in the structure of most oxy-minerals. It has a similar ionic radius to sulphur, but does not readily substitute for sulphur in sulphide minerals.

Chlorine is precipitated in such minerals as apatite, mica, amphiboles, the sodalite family of minerals in nepheline syenite and related pegmatites, and in the scapolite family of minerals in granite pegmatites and contact metasomatic deposits.

Most chlorine removed from magmas by hydrothermal derivatives finds its way to the sea where it accumu lates.

Chloride is determined by an automated colorimetric method using flow injection analysis (FIA) technique.

The chemical principle of the method is that chloride reacts with mercuric thiocyanate to form mercuric chloride and liberate thiocyanate, which in turn reacts with the ferrie ion to form a highly coloured complex that is stable and suitable for colorimetry. The absor- bance of the coloured complex, which is proportional to the concentration of the chloride is measured at 480 nm. The chemical reactions involved are as follows: ng^^iM;2 -i1 - z^i

' ng^i2 -i- , 2(SCN)'

3(SCN)' 4

1"^^. /O /TWT\

(coloured complex)

2. Lithium metaborate flux should be handled with care. Avoid inhalation and contact with the skin.

In the process of mixing lithium metaborate with the sample, stir the contents gently to avoid the fine powder of flux being stirred up in the air.

Method:

The method consists of the following techniques:

1. Fusion and sample decomposition

2. Measurement of chloride signal using a Model

1000-300 Colorimeter, making use of the flow injection analysis technique

3. Calculation of chloride concentration based on a calibration curve produced from known standard solutions

Apparatus:

- A Lachat flow injection analysis system consist ing of:

A. Colorimeter (Model 1000-300)

B. A single valve injection module (Model 1600-

000)

C. A sample/valve controller (Model 1000-150)

- Technicon Sampler II

- Technicon Proportioning Pump I with pump tubes required for chloride analysis. (A flow diagram of the analytical system is shown in

Figure Cll.)

- Strip chart recorder

- Graphite crucibles, 7.5 ml

- Teflon beakers, 50 ml

- Test tubes, calibrated at 15 ml

Safety advisory:

l. The following operations should be performed in a fume hood:

A. Preparation of the 69fc and solution.

nitric acid

B. Preparation of ferrie nitrate solution.

Reagents:

- Lithium metaborate, such as supplied by Spex

Industries, Cat. No. L170

- Concentrated nitric acid, HNO3, 12 N

- Nitric acid, HNO3,97c

- Nitric acid, HNO3, 67c

- Ferric nitrate solution, Fe(N03)3.9H2O

- Mercuric thiocyanate solution, Hg(SCN)2

EA24-1

Chloride

- Stock standard solutions

- Working standard solutions

3.5. Prepare a reagent blank simultaneously.

Decant a portion of the solution to a sample cup for chloride determination as described below.

Procedures:

1. Reagent preparation

1.1. 99fc Nitric acid - Dilute 90 ml of concentrated nitric acid and make up to l liter with distilled water.

l .2. 6*7c Nitric acid - Dilute 60 ml of concentrated nitric acid and make up to l liter with distilled water.

1.3. Ferrie nitrate solution, 0.5 M. In a one liter volumetric flask, dissolve 202g of ferrie nitrate in approximately 800 ml of distilled water.

Add 25 ml of concentrated nitric acid and dilute to the mark. Mix thoroughly. Filter the solu tion if necessary.

1.4. Mercuric thiocyanate saturated solution - Dis solve 5 g of mercuric thiocyanate in l liter of distilled water and allow to stand for 4 hours.

4. Measurement of CI concentration

4. l. Turn on the power on all modules, except the sampler.

4.2. Place the reagent feedlines into the proper con tainers (see manifold in Figure Cll).

4.3. Set the sampling time at 30 sec. and the wash ing time at 90 sec. on the sample/valve control ler.

4.4. Place the calibration standards and sample solutions in the sample tray in order.

4.5. As soon as the system has stabilized and the baseline established, turn on the sampler The standard and the sample solutions can then be run sequentially.

The mechanisms of flow injection analysis:

2. Standard CI solution preparation

2.1. Stock standard solution, 1000 u,g Cl/ml - Dis solve 1.6485 g of sodium chloride, which has been dried in an oven at 1050C for several hours, in l liter of distilled water.

2.2. Working standard solution - Prepare l, 2, 3,4, and 5 u,g Cl/ml standard solutions by serial dilution of the stock standard solution with 69k nitric acid.

3. Fusion and decomposition of sample

3.1. Weigh out 0.150 g of powdered rock sample and 0.300 g of anhydrous lithium metaborate into a graphite crucible. Mix thoroughly.

3.2. Fuse the sample at 8600C in a muffle furnace for 15 minutes.

3.3. Quickly transfer the molten sample into a 50 ml teflon beaker containing 10 ml of 99c nitric acid. Stir the solution using a magnetic stirrer for at least l hour to dissolve the fusion cake.

3.4. Transfer the contents to a test tube calibrated at

15 ml, and make to volume with distilled water.

Seal the test tube with a piece of Parafilm, and mix the solution thoroughly.

In performing flow injection analysis, samples and carrier are alternatively pumped into and flushed out of the sample loop which is mounted on the valve of the flow injection module. The sample loop is 56 cm long and 0.97 mm in diameter. Sampling time is 30 sec. and flushing time is 90 sec. The connection of these components and the analytical system are shown in Figure Cll.

In position l (sampling) of Figure Q l, the sample solution is picked up from the sample cup and is drawn by the action of the proportioning pump. It enters through port A of the injection valve filling the sample loop BC, and the excess exits through port D and drains out as waste. The sample flow path follows AB CD direction, while the flow path of the carrier, 69c HNO3, is EF in order. The rotor of the injection valve then rotates 60 degrees and turns into position 2 (injection), changing the flow paths to AD and ECBF respectively.

This permits the sample solution which has been col lected in the loop BC to be propelled by the 69fc HNO3 carrier into the reacting stream. At the same time, the sample probe swings into the sampler reservoir in a synchronized motion with the rotor. The solution in the sample line, which is now 69c HNO3, enters port A and exists through port D. This flow flushes out the tailings of the sample solution in preparation for a new cycle of sampling.

EA24-2

Chloride

5. Calculation of the results

The chloride concentration, u.g Cl/ml, in the sample solution is read with the aid of a calibration curve

(blank being subtracted). Multiplying this value by a factor of 100, (15 ml/0.15 g), to obtain the chloride content in rock in ppm.

Quality Control:

The optimum working range for this method is 0.5 to

5.0 jig/ml in solution, equivalent to 50 to 500 ppm in rock.

The determination limit is 30 ppm.

The precision, expressed at the 95*7c confidence limit

(2o), is ± 109fc relative, e.g. 250 ppm ± 25 ppm.

The accuracy is comparable to the precision, based on data collected from standard reference materials.

Productivity:

A technician should be able to complete 15 determina tions per day.

Additional notes:

1. In order to maintain a minimum blank value, a good grade of lithium metaborate (such as sup plied by Spex Industries, Cat. No. L170) and nitric acid (such as the "Baker Analyzed" reagent) should be used.

2. If samples are low in chloride (less than 300 ppm), production rate can be increased by cutting down the washing time of the FIA system from 90 seconds to 60 seconds.

3. At the end of the analysis, place all feedlines in distilled water to flush system before turning off the power on all modules.

Bibliography:

Ruzicka, J., and Hansen, E.H., Flow Injection

Analysis, Wiley Interscience Publishers, New York,

1981.

Technicon Auto Analyzer Methodology, Chloride Pro cedure, Published by Technicon Instruments Corpora tion, Ardsley, New York, 1966.

EA24-3

Chloride

CU

(X*

(O

-t o.

o

.rf.

(9

CD

O

o

3

<

O*

O

VI t/J

Ol

01

CO rv*

m

ro

en

o*

ft

H

A

*

p-

H

Figure CL1. Analytical System and Reagent Manifold for the Determination of Chloride

EA24-4

M o

CO

As, Sb, Bi

THE DETERMINATION OF ARSENIC, ANTIMONY AND BISMUTH

HYDRIDE GENERATION - ATOMIC ABSORPTION METHOD

Introduction:

Arsenic appears to follow sulphur quite closely, and as a result readily substitutes for sulphur in many sulphide minerals. Perhaps the best known arsenic bearing mineral is arsenopyrite. In ultrabasic rocks arsenic concentrations are typically 3 ppm while this decreases to about l ppm in intermediate rocks.

Antimony is less abundant in the earths crust than arsenic. It too is closely associated with sulphide minerals, but often shows higher concentrations in intermediate and acidic rocks.

Bismuth is generally less abundant than antimony.

Bismuth occurs as a minor component of massive sulphide segregations, and in greater abundance in their derived hydrothermal deposits. It is also reported to be present in apatite of igneous rocks, and in many pematitic metamict minerals containing rare earth ele ments -possibly in the ferromagnesian minerals, or in plagioclase to some extent.

Arsenic, antimony and bismuth are individually deter mined by a hydride generation/atomic absorption method following sample decomposition.

This decomposition is accomplished in the following manner: for arsenic, the sample is fused with sodium hydroxide and brought into solution with hydrochloric acid; for antimony, the sample is digested with a mix ture of sulphuric and hydrofluoric acids; and for bis muth, the sample is digested in a mixture of perchloric and hydrofluoric acids.

Sodium borohydride solution is introduced to the sample solution by means of an automatic sampler and a proportioning pump. The element of interest is con verted to its hydride which is separated in a gas-liquid separator, then swept by a stream of argon into an electrically heated quartz tube atomizer. The atomic absorption of the element is measured.

Safety advisory:

1. When using HF wear glasses and gloves, and be extremely careful. More information on HF is available in the Geoscience Laboratories' Safety

Manual page IV-17.

The sample tray and the waste bottles which col lect the acidic drains from the hydride generator, should be covered to prevent the acid vapour from escaping to the ambient air. Good ventilation in the working area is essential.

Method:

The method consists of the following techniques:

1. Fusion and sample decomposition

A. Arsenic - fusion with sodium hydroxide and dissolution by hydrochloric acid

B. Antimony - decomposition by sulphuric and hydrofluoric acid

C. Bismuth - decomposition by perchloric and hydrofluoric acid

2. The elements are converted to hydride form by sodium borohydride

3. Measurement by atomic absorption spectrometry

4. Calculation of As, Sb and Bi concentrations based on a calibration curve produced from known standard solutions

Apparatus:

- Varian Model AA-6 atomic absorption spectrophotometer equipped with a model 9176, l-100 M V recorder

- Technicon sampler H

- Technicon proportioning pump I

- Gas-liquid separator

- Gasimpinger

- Quartz tube (16 cm long, l O mm i .d. with an inlet tube fused into the centre), wound with a 22- gauge chromel A heating wire and insulated with a layer of wrapped Thermofab string

- Staco variable transformer

- Hotplate

- Nickel crucibles, 40 ml

- Stoppered graduated cylinder, 50 ml

- Glass beakers, 100 ml

- Teflon beakers, 30 ml

- Test rubes, borosilicate, 18 X 150 mm, graduated at 15 ml

- Test tubes, polystyrene, 16 X 150 mm

A25-1

As, Sb, Bi

Reagents:

- Hydrochloric acid, HC1,

- Borohydride solution

- Argon gas

- Sodium hydroxide pellets, NaOH

- Magnesium oxide

- Hydrofluoric acid, HF, 49.3 9c

- Sulphuric acid, H2SO4, 96*7c

- Potassium iodide solution

- Perchloric acid, HC1O4, 607c

- Digestion mixture

- Masking agent

- Stock arsenic standard solution (1000 M-g/ml)

- Working arsenic standard solutions

- Stock antimony standard solution (1000 M-g/ml)

- Working antimony standard solutions

- Stock Bismuth standard solution (1000 u-g/ml)

- Working bismuth standard solutions

Procedures:

2.1.2. Working arsenic standard solutions - Prepare

0.010, 0.020, 0.050 and 0.075 ^ig/rnl standard solutions by serial dilution of the stock stand ard solution with lO^o (v/v) HC1.

2.2. l . Stock antimony standard solution, 1000 (ig/ml

- Supplied by Harleco, specially prepared from

SbCl3 and standardized for use in atomic ab sorption.

2.2.2. Working standard solutions - Prepare 0.005,

0.010, 0.020, and 0.030 jig Sb/ml standard solutions by serial dilution of the stock stand ard solution with a mixture of 10% sulphuric acid and 207c hydrochloric acid.

2.3.1. Stock bismuth standard solution, 1000 U-g/ml -

Dissolve 1.1 148 g of Bi2O3 in 207c (v/v) HC1, and dilute to l liter with the same acid.

2.3.2. Working standard solutions - Prepare 0.001,

0.002, 0.003, 0.005, and 0.010 [0-g Bi/ml stand ard solutions by serial dilution of the stock standard solution with 20% (v/v) HC1.

1. Reagent preparation

1.1.1. Borohydride solution - Dissolve 3 g of sodium borohydride in 300 ml of distilled water.

1.1.2. Add 3 pellets of sodium hydroxide. Store in a refrigerator when not in use.

1.2. Digestion mixture l - Mix equal volumes of hydrofluoric acid, sulphuric acid, and distilled water. Store in a polyethylene bottle.

1.3.1. Potassium iodide solution - Dissolve 10 g of potassium iodide in 100 ml of distilled water.

1.3.2. Add 2 pellets of sodium hydroxide.

1.4. Digestion mixture 2 - Mix hydrofluoric acid, perchloric acid and distilled water in the ratio of 2:2:1. Store in a polyethylene bottle.

1.5. Masking reagent - Dissolve 0.5 g each of thiosemicarbazide and 1,10-phenanthroline in

100 ml of 0.1 M HC1 solution.

2. Standard As, Sb, and Bi solution preparation

2.1.1. Stock arsenic standard solution (1000 jig/ml)

- Dissolve 0.132 g of arsenic oxide (As2O3) in

2 ml of 1M NaOH, acidify with l ml of 107o

(v/v) HC1, and dilute to 100 ml with distilled water.

3. Fusion and sample decomposition

3.1.1. For As analysis - Weigh out 0.250 g of rock sample and transfer to a 40 ml nickel crucible.

Add 0.1 g of magnesium oxide and 2 g (ap proximately 15 pellets) sodium hydroxide, and mix the dry contents of the crucible.

3.1.2. Fuse the sample in a muffle furnace at 5500C for 15 minutes.

3.1.3. Cool and leach the fusion cake with distilled water.

3.1.4. Transfer the solution with rinsing to a 100 ml glass beaker. Allow the contents to cool. Add

10 ml of concentrated HC1 to the beaker.

3.1.5. Transfer the solution to a 50 ml graduated cylinder, and make to volume with distilled water. Stopper the cylinder and mix the solu tion thoroughly.

3.1.6. Prepare a reagent blank simultaneously.

3.2.1. For Sb analysis - Weigh out 0.100 g of rock sample and transfer to a 30 ml teflon beaker.

3.2.2. Digest the sample with 5 ml of acid digestion mixture l on a hot plate at low heat for ap proximately one hour (until white fumes of

A25-2

As, Sb, Bi

sulphuric acid appear and the volume is reduced to 1-2 ml). Avoid heating the contents to dryness.

3.2.3. Cool and dilute the contents with 10 ml of

3.2.4. Heat the contents to near boiling and cool again to room temperature.

3.2.5. Reheat if any white precipitate is present.

3.2.6. Transfer the contents with rinsings to a borosilicate test tube graduated at 15 ml. Make to volume with (1:4)HC1. Seal the test tube with a piece of Parafilm and mix the solution thoroughly.

3.2.7. Prepare a reagent blank simultaneously.

3.3.1. For Bi analysis - Weigh out 0.100 g of rock sample and transfer to a 30 ml teflon beaker.

3.3.2. Digest the sample with 5 ml of the acid diges tion mixture 2 on a hot plate at low heat for approximately one hour (until white fumes of perchloric acid appear and the volume is reduced to 1-2 ml).

3.3.3. Avoid heating the contents to dryness.

3.3.4. Cool and dilute the contents with about 5ml of

3.3.5. Transfer the contents with (1:4)HC1, rinsing into a plastic test tube calibrated at 15 ml.

Make to volume with (l :4)HC1. Mix the solu tion thoroughly and allow the residue to settle.

3.3.6. Prepare a reagent blank simultaneously.

4. Measurement of As, Sb, and Bi concentrations

Wavelength, run

Lamp Current, mA

Slit Width, \L

Damping

Expansion

Recorder span, mV

Recorder chart speed, cm/min

Variable transformer dial

7

300

B

5

193.7

20

0.5

40

217.6

5

10

100

C

10

1.0

40

223.2

8

50

C

6

10

1.0

40

Flowmeter reading

4 4 4

Sample time/wash time 40s760s 45s745s 45s745s

4.3. Load the sample solutions which are prepared on a batch basis, in the sample cups and place them in order into the slots of a sample tray held on an automatic sampler. After the hollow cathode lamp has warmed up, align the quartz tube with the light beam to allow maximum radiation to reach the detector.

4.4. Obtain the required temperature

(8500C ± 200C) of the quartz tube atomizer by switching on the pre-set variable transformer.

Turn on the proportioning pump with all the reagent tubes dipped in the water. Introduce argon immediately with its flow rate regulated by a flowmeter. As soon as the system has stabilized, insert the reagent tubes into their corresponding solutions. Establish a baseline signal. Switch on the automatic sampler. The standard, sample and blank solutions will then be analysed continuously. Record the absorp tion signals on chart paper. Measure the peak heights of the standards and draw a calibration graph.

5. Calculation of the results

As Sb Bi

The concentration of the element in the sample solu tion is read from the calibration graph. Subtract the blank to obtain the net concentration. Calculate the content of the element in rock according to the cor responding formula as shown in the following table: 4.1. Setting up the hydride generator - Set up the hydride generation equipment as depicted in

Figure As I., using appropriate tube manifold.

Mount the quartz tube on a burner head, with its inlet connected to a tygon tubing leading to the hydride generator.

4.2. Turn on the power of the Varian AA-6 atomic absorption spectrophotometer. Select the in strumental parameters according to the order given in the following table:

As

Sample wt/vol

M-g/g (ppro) in rock

Sb

Sample wt/vol jj-g/g (ppm) in rock

0.25g750ml fig/ml x 200mVg

O.lOg/lSml

H-g/ml x 150mVg

A25-3

As, Sb, Bi

Bi

Sample wt/vol jig/g (ppm) in rock

Productivity:

per day for each element.

Additional Notes:

O.lOg/lSml jig/ml x ISOmVg

Quality Control:

Optimum working range is 0.01-0.05 u,gAnl for As,

0.005-0.030 ^ig/rnl for Sb and 0.002-0.010 jig/ml for

B i (where no excessive interferences were noted).

The determination limits (expressed as concentrations in rock samples) for the three elements are l .0 ppm for

As, 0.1 ppm for Sb, and 0.05 ppm for Bi.

The precision for each element, expressed at the 95 9fc confidence limit (2o), is ± Wo for As, ± 147c for Sb, and ± 127o for Bi (relative).

The accuracy for each element is comparable to their respective precisions based on data collected from standard reference materials.

A technician should be able to complete 20 samples

A. General comments about all three elements:

1. The effect of argon flow rate on absorption signal is significant, and the flow rate should not be changed or readjusted in the midst of an analysis.

2. Accumulation of deposit of heavy metals in the glass fittings may occur after running a large num ber of samples. The fittings can be easily cleaned by pumping 19fc (v/v) hydrogen peroxide and 5^c

EDTA solutions in sequence.

3. The impinger filled with concentrated sulphuric acid acts as a gas mixer and moisture absorber. It has the effect of homogenizing the gas mixture and hence reducing the signal noise.

4. It should be noted that mutual interference effects between the hydride-forming elements exist. If the concentration ratio of the other hydride-form ing element or elements to the analyte is higher than 100, standard addition technique should be employed to ensure correct results. The tolerance limits of the hydride-forming elements can also be served as a guide line for determining the problem.

These tolerance limits can be found in the litera ture such as some of those quoted in the bibliog raphy.

B. Comments specific to As determination:

5. Magnesium oxide is used in the fusion of sample to capture any arsenic which may otherwise escape due to volatilization.

6. The fusion cake in the nickel crucible can be dissolved in distilled water more readily if it is warmed on a hot plate at low heat for half an hour.

7. Cu and Ni in concentrations of 20 and 10 fig/ml interfere with the analysis.

8. For samples containing organic materials, addition of approximately l ml or more of concentrated nitric acid during digestion is recommended.

C. Comments specific to Sb determination:

9. Avoid digesting the sample to dryness which will result in loss of Sb.

10. Potassium iodide is an essential component of the analysis system. It quantitatively pre-reduces Sb*5 in the presence of HC1 to Sb*3 prior to the reaction with sodium borohydride, and hence improves the sensitivity as well as the precision of the deter mination.

11. When the system is ready to run, it is good practice to condition it first by repeatedly analyzing a stan dard solution until a constant signal is attained prior to the actual analysis.

12. Cu, Ni, and As in high concentrations interfere with Sb analysis. The tolerance limits are 3000,

1500 and 750 ppm in rock respectively

13. For samples containing organic materials, addition of approximately l ml of concentrated nitric acid during digestion is recommended.

D. Comments specific to Bi determination:

14. Avoid digesting the sample to dryness which will result in loss of Bi.

15. A batch of 40 samples can be digested simul taneously on a hot plate with ease.

16. The digested solution should be kept in plastic test tubes.

A25-4

As, Sb, Bi

17. Thiosemicarbazide and 1,10-phenanthroline are used as masking agents for Bi to minimize inter ferences from Cu and Ni.

18. When the system is ready to run, it is a good practice to condition it first by repeatedly analyz ing a standard solution until a constant signal is attained prior to the actual analysis.

Bibliography:

Chan, C.Y., and Vijan, P.N., 1978, Semi-automated

Determination of Antimony in Rocks, Analytica

Chimica Acta, Vol. 101, pp 33-43.

Chan, C.Y., Baig, M.W.A., and Pitts, A.E., 1979,

Semi-automated Method for the Determination of Bis muth in Rocks, Analytica Chimica Acta, Vol. 111, pp

169-176.

Fernandez, F.J., 1973, Atomic Absorption Determina tion of Gaseous Hydrides Utilizing Sodium

Borohydride Reduction, Atomic Absorption Newslet ter, Vol. 12, No. 4, pp 93-97.

Kirkbright, G.F. and Taddia, M., 1978, Analytica

Chimica Acta., Vol. 100, pp 145.

Rubeska, I., and Hlavinkova, V., 1979, Determination of Arsenic in Rocks and Soils by Atomic Absorption

Spectrophotometry usingMHS-1 Automated Hydride

System, Atomic Absorption Newsletter, Vol. 18,

No. l,pp5-7.

Smith, A.E., 1975, Interferences in the Determination of Elements That form Volatile Hydrides with Sodium

Borohydride Using Atomic Absorption

Spectrophotometry and the Argon-Hydrogen Flame,

Analyst, Vol. 100, pp 300-306.

Thompson, K.C., 1974, Atomic Absorption Studies on the Determination of Antimony, Arsenic, Bismuth,

Germanium, Lead, Selenium, Tellurium and Tin by

Utilizing the Generation of Covalent Hydrides,

Analyst, Vol. 99, pp 595-601.

A25-5

As, Sb, B i

a

ft

(D

Oi

t^

Hr h

H no l' 1, 1 ';',!

iger

H2S04

Waste -t —

VJ en

*parat(

Oi

(O

— r

ro

VO

O

Figure Sbl. Auto Analyzer-A AS Row System for Determination of Antimony

U)

10

A25-6

H

O

As, Sb, Bi

i

0)

W e*

O iH

O ffi

^

J*

•H

C

s

-H

4J

M l

04

x

2

M

•o

M

O

O

Q)

3

A

to

o

(d

Q

0)

CO

O

CO

(N

X

H

Figure As l. Auto Analyzer-AAS Flow System for Determination of Arsenic.

N

4^

id

•C

Q)

4^

10

O

A25-7

As, Sb, Bi

ft

(D

o

ft

N cr

(D

H-

ifl

** r

O

O

O*

(D

H

CO

fo

01

H

O* ft

O

H l

H

H*

iQ

(D

H to

o

ft

(D

*^ ^— i

iQ ^}

n D

fi

K

O

S

Figure Bi l. AutoAnalyzer-AAS Flow System for Determination of Bismuth.

A25-8

Selenium

DETERMINATION OF SELENIUM

HYDRIDE AAS - AUTOMATED FLOW INJECTION ANALYSIS

Introduction:

In igneous rocks, selenium follows sulphur very close ly. Therefore, selenium is usually present in detectable amounts in all sulphide minerals in massive segrega tions, contact metasomatic deposits, veins, pegmatites and in desseminated accessory form in igneous rocks.

Selenium substitutes for sulphur in these environments and occasionally forms separate selenides of Cu, Pb,

Ag etc.

In high temperature deposits, selenium concentrations can be in the range of 15 to 60 ppm, while in medium to low temperature deposits, it may occur in the l to

5 ppm level.

Safety advisory:

1. When using HF wear glasses and gloves, and be extremely careful. More information on HF is available in the Geoscience Laboratories' Safety

Manual page IV-17.

Method:

The method consists of the following techniques:

1. Sample decomposition

2. Separation and concentration by hydride reduction

3. Measurement of selenium signal using Atomic

Absorption Spectrometry

4. Calculation of selenium concentration based on a calibraton curve produced from known standard solutions

The rock sample is digested with a mixture of hydrofluoric, nitric and perchloric acids. The acidified sample solution is introduced into a carrier stream by a flow injection module in conjunction with an autosampler and a proportioning pump. The carrier stream is merged with sodium borohydride solution to bring about a redox reaction between the reductant and the selenium in the sample.

Selenium is converted to its hydride which is separated in a gas-liquid separator and swept by a stream of argon into an electrically heated quartz tube atomizer. The atomic absorption of selenium is measured at

196.0 nm.

The concentration of selenium in the sample solution is read from a calibration graph, and the content of selenium in rock is calculated according to the sample weight and total volume of the sample solution.

Apparatus:

- Varian Model AA-6 atomic absorption spectrometer equipped with a model 9176 strip- chart recorder.

- Technicon Sampler II

- Technicon Proportioning Pump I

- Flow injection module (Model No. 1000-600,

Lachat Instruments)

- Gas-liquid separator

- Gasimpinger

- Quartz tube (16 cm long, 10 mm i.d. with an inlet tube fused into the centre), wound with a 22- gauge Chromel A heating wire and insulated with a layer of wrapped Thermofab string

- Staco variable transformer

- Teflon beakers, 30 ml

- Hotplate

- 18 x 150 mm test tubes, graduated at 15 ml

2. The sample tray and the waste bottles which col lect the acidic drains should be covered to prevent the acid vapour from escaping to the ambient air.

To ensure a clean air working environment, good ventilation is recommended.

Reagents:

- Hydrofluoric acid, HF, 487c

- Perchloric acid, HC1O4, 607o

- Hydrochloric acid, HC1, 387o

- Nitric acid, HNO3, 707c

- Digestion mixture

- Reducing solution

- Masking agent

- Stock selenium standard solutions, 1000 jj-g/ml

EA26-1

Selenium

Procedures:

3.8. Decant a portion of the solution to a sample cup for subsequent AAS determination.

1. Reagent preparation

1.1,

Digestion mixture - Mix hydrofluoric acid, perchloric acid, nitric acid, and distilled water in the ratio of 4:4: l: l respectively and store in a polyethylene bottle.

1.2.

Reducing solution - Dissolve 5 g of sodium borohydride and five pellets of sodium hydroxide in 500 ml of distilled water. Store in a refrigerator when not in use. The solution is stable at 40C for at least a week.

1.3.

Masking reagent - Dissolve l g of phenanthroline in 100 ml of 0.1 M HC1.

1,10-

2. Standard Se solution preparation

2.1. Stock Se Standard Solution, 1 000 (ig/ml - Dis solve 0.100 g of powdered selenium in 100 ml of 109fc nitric acid.

2.2. Working standard solutions - Prepare 0.25,

0.50, 1.0, 2.0 and 4.0 ng/ml solutions by serial dilution of the stock standard solution with

3.6N HC1.

4. Measurement of Se concentration

Instrument parameters

Atomic absorption spectrometer

Wavelength 196.0nm

Lamp current 8 mA

Slit width

Damping

Expansion

300 jo.

C (Maximum)

6

Flow injection module

Sample time 30 sec

Washing time 40 sec

Variable transformer

Dial 50

(set to produce 850 +A 200C in the atomizer)

Recorder

Span 10 mV

Chart speed l cm/min.

Argon

How rate 300 ml/min.

3. Sample decomposition

3.1.

Digest 0.200 g of rock sample with 5 ml of digestion mixture in a 30 ml teflon beaker on a hot plate for about one hour or until white fumes of perchloric acid appear and the volume of the contents reduces to approximately l ml.

3.2.

3.3.

3.4.

3.5.

3.6.

Cool and add about 2 ml of distilled water, and

4.5 ml of concentrated HC1.

Heat the contents to just under boiling for several minutes to reduce Se that is in Se6 oxidation state to Se

4-t-

Cool and transfer the contents to a test tube calibrated at 15 ml.

Make to volume with distilled water.

Seal the test tube with a piece of Parafilm and mix the solution thoroughly. The concentra tion of HC1 in the sample solution is now 3.6N v/v).

3.7. Prepare a reagent blank simultaneously.

Set up the hydride generation equipment and connect the tubing that leads to and from the injection valve of the flow injection module according to the layout shown in Figure Sel. Mount the quartz tube on the burner with its side arm connected to the hydride generator with tygon tubing. Align the quartz tube with the light beam to allow maximum radiation to reach the detector. Switch on the preset variable trans former to provide the desired temperature in the atomizer. Turn on the proportioning pump with all the reagent tubes dipped in water. Introduce the argon immediately with its flow rate regulated at 300 ml/min.

Turn on the flow injection module which is interfaced with the sampler. The motions of the injection valve and the sample probe are synchronized. Set the sam pling time to 30 sec. and the washing time for 40 sec. on the flow injection module. Insert the reagent tubes into the corresponding solutions (see arrangement in

Figure Sel). As soon as the system has stabilized and the baseline is established, the standard and sample solutions which have been loaded in the sampler can then be run sequentially.

In position l of Figure Sel, the sample solution is picked up from the sample cup and is drawn through

EA26-2

Selenium

the sample tubes (line 4a and 4 in Figure Se l) by the action of the peristaltic pump. It enters port A of the injection valve, filling up the sample loop AB, and the excess exits through port B. The rotor of the injection valve then rotates a quarter turn so that ports A and B interchange places with ports C and D (position 2,

Figure Sel), permitting the sample solution in the loop

AB to be propelled by the 5*7c HC1 carrier into the reacting stream. At the same time, the sample probe swings into the sample reservoir in a synchronized motion with the rotor. The solution in sample line (4a), which is now 5*7c HC1, enters port C and exits through port D. This flow will flush out the tailings of the sample solution, in preparation for a new cycle of sampling.

5. Calculation of the results

When the analysis is complete, calculate the Se con centration from a calibration plot of peak height vs. concentration which is linear up to 4 ng Se/ml.

Read the concentration of Se in the sample solution from the calibration graph. Subtract the blank value to obtain the net concentration (ng Se/ml). Calculate the selenium content in the rock sample according to the formula:

Productivity:

A technician should be able to complete 20 deter minations per day.

Additional Notes:

1. Avoid digesting the sample to dryness which will result in loss of Se.

2. Nitric acid is an essential ingredient for sample digestion.

3. Because of the absorbance of Se4* is much stronger than that of Se6*; Se in the digested sample solu tion, mainly in the form of Se *, must be reduced to Se by hot HC1 (8-12M) prior to determination.

4.

5.

1,10 phenanthroline is used as a masking agent to minimize interferences from Cu and Ni.

Cu and Ni in high concentrations will interfere with the analysis. The tolerance limits are 0.3 and

4.09fc in rock respectively.

Quality control:

= ng Se /ml x 75 ml /g

The optimum concentration range for selenium is 50 to 300 ng/g.

The determination limit of this method is 20 ppb in rock samples.

Precision, expressed at the 95 9fc confidence limit (2o), is 12^o (relative).

Accuracy is comparable to the precision based on data collected from standard reference materials.

Bibliography:

Agemian, H., and Bedek E., 1980, A Semi-automated

Method for the Determination of Total Arsenic and

Selenium in Soils and Sediments, Anal. Chim. Acta,

Vol. 119, pp 323-330.

Astrom, O., 1982, Flow Injection Analysis for the determination of Bismuth by Atomic Absorption

Spectrometry with Hydride Generation, Anal. Chem.,

Vol. 54, pp 190-193.

Chan, C.Y., 1985, Semiautomated Method for Deter mination of Selenium in Geological Materials Using a

How Injection Analysis Technique, Anal. Chem., Vol.

57, pp 1482-1485.

Chan, C Y., and Baig, M. W. A., 1984, Semi-automated

Method for De termination of Selenium in Rocks, Anal.

Letter, 17, pp 143-155.

Ruzicka, J., and Hansen, E.H., Flow Injection

Analysis, Wiley Interscience Publishers, New York,

1981.

EA26-3

Selenium

Q,

'l.

g c* 3

O

Figure Sel. Analytical Manifold for the Determination of Se by Automated Flow Injection and Hydride

AA Techniques.

EA26-4

Mercury

DETERMINATION OF MERCURY

COLD VAPOR FLAMELESS ATOMIC ABSORPTION METHOD

Introduction:

A majorchallenge in Hg analysis lies in the preparation and storage of geological samples prior to analysis.

This is due to the high vapor pressure of mercury, its presence in the atmosphere, and the generally low levels in naturally occurring materials. A number of procedures developed to allow for the above factors are detailed below.

the lamp cover while the instrument is on without suitable eye protection.

Installation of a mercury scrubber at the vent of the cell chamber of the mercury monitor is recommended.

The scrubber is made of a small tube loosely packed with dry moss (sphagnum) and a few gold chips. It will serve as a mercury trap, preventing the hazardous vapor from escaping to the ambient air.

Soil samples are normally collected at a depth of 4 to

6 inches. As the Hg concentration in soils often varies considerably within a soil profile, a standardized depth of collection is important. In areas with a thick organic horizon, such as British Columbia, samples should be collected below the obvious organic horizon.

If the soil samples are dry they should be sieved to -80 mesh in the field, using a stainless steel sieve, and stored in air-tight screw top glass vials.

Method:

The method consists of the following techniques:

1. Sample decomposition

2. Measurement of mercury absorption signal by a

LDC/Milton Roy Mercury monitor

Wet samples should be dried at room temperature or in the shade before sieving. Heat lamps should not be used as this could result in volatilization of Hg from the sample.

Rock samples should not be ground in a ceramic plate mill. The heat generated by such grinding tends to volatilize a significant portion of the included Hg, especially in samples containing sulphides.

Instead, rock samples should be coarsely crushed in a jaw crusher and the +20 mesh fraction removed. This fraction should then be hand ground to -80 mesh.

Sample pulps should then be stored in air-tight glass vials.

The sample is digested at low heat with nitric and hydrochloric acids. The digested sample and stannous chloride solution are continuously pumped through a mixing coil and into a gas-liquid separator. Mercuric ions present in the solution are reduced to elemental mercury. The volatilized mercury is separated and swept by a stream of argon into the absorption cell of a LDC/Milton Roy Mercury Monitor where the atomic absorption at 253.7 nm is measured.

Safety advisory:

l. Ultraviolet radiation is emitted from the mercury lamp when the instrument is on. Do not remove

3. Calculation of mercury concentration based on a calibration curve produced from known standard solutions

Apparatus:

- LDC/Milton Roy Mercury Monitor

- Technicon Sampler II

- Technicon Proportioning Pump I

- Gas-liquid separator

- Drying tube filled with magnesium perchlorate

- Flowmeter

- Test tubes, Pyrex No. 7900, graduated at 25 and

50ml

- Hot plate

- Thermometer

- Aluminum heating block, 2" x 4" x 6" with 12 wells to hold test tubes and one small well to hold thermometer

- Vortex-Genie mixer

- Linear strip-chart recorder

Reagents:

- The following reagents should be free of mercury and reagent grade:

- Hydrochloric acid, HC1, 38*7c, such as Ultrex

Ultrapure grade supplied by Baker Chemical

Company

- Nitric acid, HNO3, 7096

- Potassium dichromate solution (29fc w/v)

- Stannous Chloride solution (309fc w/v)

EA27-1

Mercury

- Hydroxylamine hydrochloride solution (5 ft w/v)

- Mercury standard stock solution (1000 jig/ml

Hg), such as supplied by Baker Chemical Co.

- Mercury standard solutions

- Mercury working standard solutions

Procedures:

1. Reagent preparation

1.1.

107c Nitric acid (v/v) - Prepare 2 liters of solution by diluting 200 ml of concentrated nitric acid with 1.8 liters of distilled water.

1.2.

l .3.

1.4.

209fc Nitric acid (v/v) for sample dilution pur poses - Prepare 2 liters of solution by diluting

400 ml of concentrated nitric acid with 1.6 liters of distilled water.

Potassium dichromate solution (29fc w/v) - Dis solve 4 g of K2Cr2O7 in 200 ml of distilled water.

Stannous chloride solution (309fc w/v) - Dis solve 60 g of stannous chloride in 60 ml of concentrated HC1, and dilute to 200 ml with distilled water.

1.5. Hydroxylamine hydrochloride solution w/v) - Dissolve 10 g of hydroxylamine hydrochloride in distilled water and make up to

200ml.

3. Sample decomposition

3.1. Weigh out 0.250 g of sample and transfer into a Pyrex No. 7900 test tube.

3.2. Add 5 ml of concentrated nitric acid and 0.5 ml of concentrated hydrochloric acid to the sample. Mix the contents. Place the test tube in a well in the aluminum block.

3.3. When a batch of samples, including the blank, have been prepared in the same way, place the aluminum block on the hot plate and maintain a temperature of 1100C.

3.4. Insert a thermometer in the small well in the aluminum block to monitor the temperature.

3.5. Digest the sample for two hours.

3.6. Add the 27c K2Cr2O7 solution dropwise to the sample with shaking until the orange color persists.

3.7. Place the test tube into a test tube rack (or into another aluminum block) and let cool at room temperature.

3.8. Add distilled water to the test tube and make to the 25 ml mark. Mix the solution thoroughly using a vortex mixer.

3.9. Allow the residues to settle (about one hour).

Transfer the supernatent sample solution to a sample cup for Hg determination.

2. Standard Hg solution preparation

2.1. Mercury standard solutions

2.1.1. l jig/ml Hg solution - Prepare by serial dilution of the 1000 (ig/ml Hg stock solution with 20^c nitric acid.

2.1.2. 25 ng/ml Hg solution - Prepare by serial dilu tion of the l jig/ml Hg stock solution with 20^o nitric acid.

2.1.3. Store solutions in a glass container.

2.2. Mercury working standard solutions

2.2.1. 0.5 and 1.0 ng/ml Hg standard solutions are prepared by serial dilution of the 25 ng/ml Hg solution with 207o nitric acid. These solutions are stable for at least two months.

2.2.2. Store solutions in a glass container.

4. Measurement of Hg concentration

4. l. Turn on the power switch of the LDC/Milton

Roy Mercury Monitor, and check that the ener gy meter settles in the 0.2 to 0.5 milliamperes range.

4.2. Select the optimum absorbance range (normal ly at 0.08 position).

4.3. Turn on the recorder. Set the span at l mV and chart speed at l cm/min.

4.4. Set up the auto analyzer system using the tube manifold as depicted in Figure Hgl.

4.5. Load the sample cups containing the sample solutions, including the standard and blank solutions, into the slots of a sample tray held on the automatic sampler.

EA27-2

Mercury

4.6. Turn on the proportioning pump with all the reagent tubes dipped in distilled water.

4.7. Turn on the argon cylinder and introduce argon to the reacting stream with the argon flow-rate regulated by the flowmeter at dial setting 30.

4.8. Once the system has stabilized, insert the reagent tubes into their corresponding solu tions.

4.9. Re-stabilize the system and establish a baseline on the recorder chart.

4.10. Switch on the automatic sampler. The stand ard, sample, and blank solutions will then be run sequentially.

4.11. Record the absorption signals on a recorder chart.

4.12. Measure the peak heights and draw a calibra tion graph.

Read the concentration of Hg in the sample solution from the calibration graph. Subtract the blank value to obtain the net concentration (ng Hg/ml). Calculate the

Hg concentration in the rock sample according to the formula:

ng H g/g in rock = ng Hg/ml x

25ml

0.25 g

= ngHg/ml x lOOml/g

Quality Control:

The optimum working range is 0.1 to l ng/ml in solution, equivalent to 10 -100 ppb in rock.

The determination limit of this method is 5 ppb in rock.

Precision, expressed at the 95 Ve confidence limit (2o), is ±1296 (relative).

The accuracy is comparable to the precision, based on data collected from standard reference material.

Productivity:

A technician should be able to complete 15 to 20 determinations per day.

Additional Notes:

1. Standard solutions and reacting reagents should be kept in glass containers. Avoid using plastic bot tles, since most plastic materials contain mercury compound ingredients as plasticizers and will con taminate the solutions.

2. Glassware, such as test tubes, should be cleaned with nitric acid, followed by a rinse with distilled water. (Soaking overnight in lO^o nitric acid is quite effective). Glassware known to have been used for samples with high Hg concentrations

(M ppm) should be cleaned with cone, nitric acid, followed by a rinse with tap water and then dis tilled water. (Tap water usually contains negli gible amounts of Hg. It can safely substitute distilled water for cleaning, although this is not recommended conventionally).

3. Gold, platinum, palladium, arsenic, antimony, bis muth and selenium are known interferents, but are rarely present in concentrations sufficient to cause a significant problems.

Bibliography:

Instruction Manual, Mercury Monitor elemental mer cury detector 920404, LDC/Milton Roy, Riviera

Beach, Fla., (1985).

Chan, C. and Bina, S., 1989, A Sensitive Automated

Method for Determination of Mercury in Geological

Materials by Cold Vapor Atomic Absorption, Geos- tandards Newsletter, Vol. XIII, No. 1.

Hatch, W.R., 1972, Flameless Atomic Absorption

Methods for the Determination of Mercury, Canadian

Research and Development, Nov/Dec., pp. 13-26.

Johnson, W., and Maxwell, J., Rock and Mineral

Analysis, Wiley Interscience Publishers, 1981, pp.

302-303.

EA27-3

Mercury

Figure Hg l. Automated cold vapour flameless AA system for determination of mercury.

EA27-4

Carbonate Carbon

DETERMINATION OF CARBONATE-CARBON (CO2)

COULOMETRIC METHOD

Introduction: Method:

Carbonate-carbon is the main component of total carb on in the majority of geological samples. It is conven tional, therefore, to report total carbon as C02, as if all the carbon is from carbonate-carbon. In some samples, however, carbonate-carbon can represent only a small part of the total carbon. The remainder of the total carbon is either graphitic, organic carbon, diamond and/or carbide.

The method consists of the following techniques:

1. Decomposition of the sample

2. Measurement of CO2 signal by the Coulometrics

C02 Determinator

3. Calculation of the results based on weight

In studies of geological processes, it is sometimes necessary to know not only the carbonate-carbon or the total carbon content, but also the non-carbonate carbon in the samples. Total carbon can be determined by inductive furnace-infra-red method (refer to EA 12-1).

Non-carbonate carbon cannot be determined directly, however, by determining carbonate-carbon, non-car bonate carbon can be calculated by subtracting car bonate-carbon from total carbon.

The determination of carbonate-carbon on the

Coulometrics CO2 Determinator is initiated by heating a known weight (0.05 - 0.50 g depending on the carbonate content) of sample with 2 ml of 2N perchloric acid in the heated reaction tube. The evolved C02 is swept by a stream of filtered air into the Coulometrics cell which is filled with a partially aqueous medium containing ethanolamine and a colorimetric indicator.

When CO2 is passed through the cell solution, it is quantitatively absorbed and converted to a strong

(titratable) acid by the ethanolamine causing the colour of the indicator to fade. As soon as this happens, the coulometer electrically generates base to restore the colour.

Apparatus:

- Coulometrics CO2 Determinator which consists of:

A. Model 5010 Coulometrics CO2 Coulometer

B. Model 5030 Carbonate Carbon Apparatus with reaction tubes

Reagents:

- Perchloric acid, HC1O4,2N

- Potassium hydroxide solution, KOH, 45*^

- Potassium Iodide solution, KI, 50*8; KI solution

atpH3

- Coulometer solution

- Anode solution

- Potassium iodide crystals, KI

Procedures:

1. Measurement of carbonate-carbon concentra tion

The accumulated charge (coulombs) is displayed on a digital readout as micrograms of carbon. When the accumulation ceases, the end point is reached. The percent of carbon in a sample is then mathematically converted to percent of C02.

Safety advisory:

l. Due to the odor of ethanolamine coming from the coulometer cell, the analytical operation should be performed in a well ventilated area or in a fume hood.

1.1. Coulometer set-up

1.1.1. Place approximately 100 ml of coulometer solution in the cell body. (See Figure CCL).

1.1.2. Add stirring bar.

1.1.3. Place a few crystals of KI inside the anode compartment.

1.1.4. Place the cell cap on the cell body, and add anode solution to the anode compartment so that when the silver electrode is in place, the anode solution level will be the same as the solution level in the cell body.

EA28-1

Carbonate Carbon

1.1.5. Place the cell in the coulometer cell holder with the electrodes to the back.

1.1.6. Be sure the cell current switch is off, and turn on coulometer.

1.1.7. The anode compartment frit, and gas inlet tube should be positioned so that they are in the back of the light beam. This may be checked by rotating the cell until maximum ^cT is ob served. The Pt electrode should project out and away from the light beam for highest sen sitivity.

1.1.8. Plug in the electrodes noting polarity (cathode

- black, anode - red).

1.3.10. Turn on the heater and adjust heating as desired.

1.3.11. Allow sufficient time (l 0-30 minutes) for com plete reaction, and record carbon reading when coulometer gives a steady maximum reading.

1.3.12. Run a blank by heating 2 ml of 2N HC1O4 similarly, but omitting the sample.

2. Calculation of the results

The CO2 content of the sample is calculated according to the following formula:

::: —TT—————————;—

Micrograms sample

1.2. Carbonate-car bon apparatus set-up

Figure CC l shows the assembled CO2 apparatus. (For tubing connections, follow the procedure shown in the manufacturer's instruction manual.)

1.2.1. Fill air scrubber with approximately 12 ml of

457o KOR solution.

1.2.2. Fill sample scrubber tube with 50*^ KI solu tion.

1.3. Operation

1.3.1. Weigh and transfer sample (0.05 g to 0.50 g depending on the carbonate content) into a sample tube.

l .3.2. Moisten the sample with approximately 0.5 ml of di stilled water.

1.3.3. Place the sample tube on the heater of the CO2 apparatus.

l .3.4. Place the condenser on top of the sample tube.

1.3.5. Be sure system is leak free.

1.3.6. Switch on the air pump and purge the system.

1.3.7. Connect the scrubber outlet of the C02 ap paratus to the gas inlet of the coulometer cell.

1.3.8. Press the reset button on the coulometer to zero carbon reading.

l .3.9. Press the plunger on the acid dispenser to dis pense about 2 ml of acid.

3. Shutdown

3.1.

3.2.

3.3.

Switch off the air pump of the CO2 apparatus.

Disconnect the scrubber outlet line.

Turn off the heater.

3.4. Turn off the cell current and coulometer power.

3.5. Unplug the electrodes and remove the cell from the holder.

3.6. Dispose the coulometer solution and rinse the cell body.

3.7. Remove the silver electrode and rinse with water.

3.8. Rinse the outside of anode compartment, gas inlet tube and Pt electrode with distilled water.

3.9. Rinse out the anode compartment with acetone.

Remove KI and deposits on top of the glass frit.

4. Maintenance

4.1. The cell solutions should be replaced if over

100 mg of carbon have been titrated.

4.2. If the deposits stick in the glass frit, pull acetone through frit with aspirator or other vacuum source, or clean it with ultrasonic cleaner.

Anode deposits can be removed with a saturated KI solution.

EA28-2

Carbonate Carbon

4.3. If a deposit is evident on the platinum electrode, this should be dissolved in 1:1

HN03.

4.4. The cell current switch should be turned off whenever the cell is disconnected.

4.5. A high blank result is usually the result of exhaust KOH solution in air scrubber. Replace the solution when this happens.

Quality control:

The optimum working range for this method is 0.01 -

The determination limit for CO2 is 0.01 9fc in rock, although the system is capable of detecting as low as

G.003%.

Precision, expressed at the 95 tfc confidence limit (2a), for the mid-range value is ±49fc (relative), e.g. 10*^? ihO.4%.

The accuracy is comparable to the precision based on data collected from standard reference materials.

Productivity:

A technician should be able to complete 20 determina tions per day.

3. Any volatile acid or base not removed by the scrubbing solution will interfere with the analysis.

Potentially interfering gases which the scrubber removes include: H2S, C12, HI, HF, SO2 and SO3.

Another scrubbing solution which is effective for removal of chlorine and sulphur compounds is a saturated Ag2SO4 solution containing 3^c H2O2 at pH3.

4. Although running standard samples is not required for calibration, it is desirable to confirm proper operation of the apparatus and coulometerby run ning standards such as primary standard grade calcium carbonate periodically.

5. Potassium iodide crystals should appear in the anode compartment at all times.

6. A variety of mineral acids can be used for evolu tion of CO2. The following acids are most com mon: a. 2N HC1O4 b. 2N H2SO4 c. 2N HC1 - care must be taken to insure the

HC1 gas does not overload the scrubber.

Additional Notes:

1. Typical blank values are 5-10 u,g C/10 min. If a higher value is observed, the air scrubber solution should be changed.

2. Do not wash the anode with water, since the precipitated Agl will plug the frits.

Bibliography:

Instruction Manual for Coulometrics CO2 Deter- minator, Coulometrics Incorporated.

Chan, C. Y., 1986, Determination of Carbonate Carbon in 41 International Geochemical Reference Samples by Coulometric Method, Geostandards Newsletter,

Vol. X, No. 2, pp 131-134.

Huffman, E.W.D., 1977, Performance of a New Auto matic Carbon Dioxide Coulometer, Microchemical

Journal, Vol. 22, pp 567-573.

EA28-3

Carbonate Carbon

A. Cell body

B. Gas inlet

C. Gas outlet

D. Anode compartment

E. Frit

F. Stirrer

G. Platinum electrode (cathode)

H. Cathode terminal

I. Silver electrode (anode)

J. Anode terminal

Figure CC l. Coulometer cell.

EA28-4

Light source

Moisture

Introduction:

Minus water (H2O~), or moisture, is the portion of water in rocks held by surface forces such as adsorption and capillarity. It can be driven off at 100-1100C. The

H2O" content is determined by measuring the loss of weight of a l g sample when it is dried overnight at

1100C.

Safety advisory:

1. No specific safety advisory is required for this method.

1.4. Fully cover, then remove the crucible and quickly place it in a desiccator. Let cool to room temperature.

1.5. Weigh the dried sample and the crucible with the lid as quickly as possible to avoid moisture being re-absorbed by the sample.

2. Calculation of the results

The difference in weight, before and after drying, represents the amount of moisture, or minus water, in the sample.

Method:

1. Direct gravimetric determination of loss on heat ing

2. Calculation based on sample weight loss

Percent of minus water =

w

where, w = original wt. of the sample

W! = wt. of original sample

+ crucible with lid w2 = wt. of dried sample

-i- crucible with lid

x 100

Apparatus:

- Drying oven

- Porcelain crucible with lid

- Desiccator

Reagents:

- No reagents are required for this method.

Quality Control:

The determination limit for this method is 0.01 9fc.

Accuracy and precision have not been established.

Procedures:

1. Direct gravimetric determination of H2O"

1.1. Weigh l .0000 g of sample and transfer into a pre-dried porcelain crucible.

1.2. Record the total weight of the sample and crucible with cover.

1.3. Heat the sample in an oven at 1050 - 1100C for at least one hour (preferably overnight). Par tially cover the crucible by tilting the lid to the side.

Productivity:

A technician should be able to complete 40 determina tions per day.

Bibliography:

Maxwell, J. A., Rock and Mineral Analysis, Inter- science Publishers, 1968, 584 pages.

Potts, P.J., A Handbook of Silicate Rock Analysis,

Blackie and Sons Limited, Glasglow, 1987, 622 pages.

EA29-1

Insoluble Residue

Introduction:

This method is used as a relative measure of impurities in carbonate samples. In most samples silica will predominate in the acid insoluble residue. The amount of residue obtained depends upon the acid used, the time and temperature of the digestion and the grain- size of the sample. The conditions under which the determinations are made should be consistent and well defined.

Method:

The method consists of the following techniques:

1. Sample decomposition and ignition of residue.

2. Direct gravimetric determination of the residue remaining after the weighed sample is digested with hydrochloric acid and the residue ignited.

3. Calculation based on the sample weight loss.

- Hydrochloric acid, HC1, (5:95)

Procedures:

1. Reagent preparation

1.1. (1:3) HC1 solution - Prepare l liter of this solution by mixing 250 ml of concentrated HG with 750 ml of distilled water.

1.2. (5:95) HC1 solution - Prepare l liter of this solution by mixing 50 ml of concentrated HG with 950 ml of distilled water.

Safety advisory:

1. The following operations should be performed in a fume hood:

A. Preparation of HC1 solution

B. Addition of the (1:3) HC1 solution to the samples

C. Heating and digesting on a hot plate

2. The MSDS sheet pertaining to the use of hydrochloric acid must be reviewed before proceeding. Safety procedures as outlined in the

Geoscience Laboratories' Safety Manual should also be reviewed.

2. Sample decomposition and ignition of residue

2. l . Weigh exactly 0.500 g of sample and transfer to a 1 50 ml beaker.

2.2. Slowlyadd50ml(l:3)HG. Swirl occasional ly.

2.3. When vigorous effervescence has ceased, heat the contents on a hot plate to near boiling.

Digest for 15 minutes at a temperature just below boiling.

2.4. Filter at once through a 1 1 cm Whatman No.

40 filter paper and wash the paper and residue thoroughly with hot dilute HC1 (5:95).

2.5. Transfer the paper and residue to a weighed porcelain crucible.

2.6. Burn off the paper at 2500 to 3000C, and then ignite the residue at 9000 - 10000C for 15 minutes.

2.7. Cool in a desiccator for half an hour and weigh.

Apparatus:

- Glass beakers, 150ml

- Hotplate

- Porcelain crucibles, 30 ml, with lids

- Glass filtering funnels

- Funnel supports

- Filter papers, Whatman No. 40, 11 cm

Reagents:

- Hydrochloric acid, HC1, (1:3)

3. Calculation of the results

The content of acid insoluble residue, or acid in- solubles, is calculated according to the equation:

Acid Insolubles = 2(w2 - w^ x where, w j = weight of crucible in g, w2 = weight of crucible and residue in g after ignition

EA30-1

insoluble Residue

Quality control:

The determination limit for this method is Q.2%.

The precision, at the 95 Ve confidence limit (2o) is ±57o

(relative).

Accuracy has not been determined since a suitable standard reference material is not available.

Productivity:

A technician should be able to complete 20 determina tions per day.

Additional Notes:

1. Samples are analyzed simultaneously in a batch of

12 or more, depending on the availability of work ing space.

2. The method does not determine organic insoluble residues as these would be destroyed in the com bustion stage.

Bibliography:

Maxwell, J.A., Rock and Mineral Analysis, Wiley

Interscience Publishers, 1968, pp 304-305 and pp487-

488.

Potts, P.J., A Handbook of Silicate Rock Analysis,

Blackie and Sons Limited, Glasglow, 1987,622 pages.

EA30-2

Chittick

DETERMINATION OF THE CALCITE TO DOLOMITE RATIO

CHITTICK APPARATUS METHOD

Introduction:

The determination of the carbonate content and cal- cite/dolomite ratio of sedimentary rocks and tills, provides valuable information enabling the geologist to classify the materials and determine their suitability as materials for use in the construction industry. The concentrations of these minerals are highly variable in rocks and tills.

The method is based on the different reaction rates of calcite and dolomite in hydrochloric acid. The ratio of calcite to dolomite in a finely ground sample (minus

200 mesh) can be determined by taking two readings of the volume of evolved CO2 in a Chittick gasometric apparatus. The first reading records gas evolved prin cipally from the faster reacting calcite, the second reading records the gas evolved from both minerals.

Reagents:

- Hydrochloric acid, HC1, (1:1)

- Methyl orange indicator

- Displacement solution

Procedures:

1. Reagent preparation

1.1. (1:1) HC1 solution - Mix 500 ml of con centrated HC1 with 500 ml of distilled water.

1.2. Methyl orange indicator - Dissolve 0.05 g methyl orange in distilled water and dilute to

100 ml with distilled water.

1.3.1. Displacement solution - Dissolve 100 g NaCl in 350 ml distilled water.

Safety advisory:

1. This method involves the specific use of hydrochloric acid. Staff must read the MSDS sheet on hydrochloric acid, and review all safety procedures required for this chemical. The (1:1)

HC1 should be prepared in a fume hood to avoid inhaling acid fumes.

1.3.2. Add l g NaHCO3 and 2 ml methyl orange indicator, and then sufficient HC1 to make just acid (decided pink). Stir until all CO2 is removed.

l .3.3. This solution is used in the gas measuring tube and levelling bulb and seldom needs to be replaced.

Method:

The method consists of the following techniques:

1. Sample decomposition by HC1 attack

2. Measurement of C02 gas evolved

3. Calculation of the results based on an accepted formula

Apparatus:

- Chittick gasometric apparatus (Fig. Chl) which consists of the following components:

A. Decomposition flask, 250 ml

B. Graduated 20 ml pipette with stopcock at the base

C. Stopcock

D. Gas measuring tube

E. Levelling bulb

F. Magnetic stirrer and stirring bar (4 cm in length)

2. Sample decomposition

2.1. Weigh 1.700gofthe-200meshsample(0.850 g in case the sample contains more than 40 percent carbonates), and place it in the decom position flask (A in Figure CH1).

2.2. Connect the flask to the Chittick apparatus.

Open the stopcock (C) and bring the displace ment solution in the measuring tube (D) to the

20 ml mark above the zero, by raising or lower ing the levelling bulb (E).

2.3. The pipette (B) is filled to the zero mark with

2.4.

Close the stopcock (C) and lower the levelling bulb somewhat to reduce the pressure within the apparatus.

EA31-1

-iChitiick

Open the stopcock of the pipette and let 20 ml of the (1:1) HC1 run into die decomposition flask as quickly as possible.

Switch on the magnetic stirrer to secure in timate mixture of the contents. At the same time lower the levelling bulb.

E. First reading:

(a) volume of CO2

(b) corrected first reading (obtained by multiply ing the first reading by its correction factor)

F. Second reading:

2.7.

During decomposition, prevent the liberated

CO2 from escaping through the pipette by keeping the displacement solution in the levell ing bulb at a lower level (about an inch lower) than that in the gas measuring tube at all times.

2.8. After having dispensed 20 ml of (l: l) HC1 into the decomposition flask, close the stopcock of the pipette.

(a) volume of CO2

(b) corrected second reading (obtained by multi plying the second reading by its correction factor)

G. Calulation: Corrected second reading minus cor rected first reading, or Fb - Eb.

H. Corrected first reading minus 4 percent of the value of (G).

3. Measurement of Evolved CO2 Gas

3.2. Then lower the levelling bulb again until the evolution of CO2 has ceased completely (20 minutes is usually required for complete decomposition).

C. Room temperature

D. Correction factor for the temperature

I. 4 percent of the value of (G) is added to (G).

3. L Exactly 30 seconds after the delivery started, equalize the levels of the displacement solution in the levelling bulb and the measuring tube, and take the first reading of the volume of CO2.

3.3. After 30 minutes take the second reading of the volume of CO2.

Quality Control:

3.4. Observe the temperature of the air surrounding the apparatus and also the barometric pressure at the time of the reading.

3.5. Find the correction factor for temperature and pressure from the table (see 'Association of

Official Agricultural Chemists', 1955, pp 945-

951.).

The optimum working range for this method is 5 96 to total carbonate.

The determination limit for both calcite and dolomite is approximately Q.5%.

The precision at mid-range is 1.5*7e at the 95^o con fidence level (2o), e.g. 209fc ± 1.57o.

4. Recording and Calculation of the results

Productivity:

A record (Form CH l) is kept of the following readings, correction factors, and the computed results for each sample.

A technician should be able to complete 25 determina tions per day working on two sets of Chittick apparatus simultaneously.

A. Sample Number

Additional Notes:

B. Weight of sample analyzed (usually 0.850 g or

1.700 g)

One tenth of the values of (H) and (I) represent the percent CO2 from calcite and dolomite respectively when the 1.700 g sample weight is used (see Associa tion of Official Agricultural Chemists, 1955, pp 127-

128). The values are doubled if 0.850 g of sample is used. Convert percent of CO2 (from calcite) to calcite by multiplying by a factor of 2.275, and percent CO2

(from dolomite) to dolomite by multiplying by a factor of 2.095. The sum of the carbonates and the ratio of calcite to dolomite are also calculated.

l . The method is simple, but lacks accuracy and has poor precision. There are many sources of error which cannot be eliminated such as the presence of water vapor in the Chittick apparatus, the presence of carbonates other than calcite or

EA31-2

Ghittick

3.

dolomite in the sample, and the inconsistency of speed of reaction.

Although the determination limit is expected to be about Q.2% calcite or dolomite (equivalent to l ml of evolved CO^ the Laboratories has found that for samples with less than 4^o total carbonate the reproducibility of results is poor.

When one mineral (calcite or dolomite) predominates, the ratio is liable to considerable error even though the determination of the major component may be quite precise. If the concentra tion of either mineral is below the determination limit the ratio can only be expressed as > or < the value obtained using the limit.

In calculating the percent of CO2 from dolomite, 4 percent of the difference between the corrected first and second readings is added in order to compensate the C02 that is derived from dolomite, yet liberated in the initial 30 seconds.

In analyzing a large number of samples, two Chit- tick apparati may be operated simultaneously.

This would increase production substantially.

4. Barometric pressure change is normally very small, hence a pressure correction step may be omitted.

5. Care must be taken to rinse the decomposition flask thoroughly with water after each analysis, since the slightest trace of acid left in the flask will cause some decomposition before the operator has a chance to connect the flask with the rest of the apparatus. Moreover, the upper part of the flask should be dried after washing to prevent any of the powdered sample adhering to the sides or the neck of the flask.

Bibliography:

Official Methods of Analysis of the Association of

Official Agricultural Chemists, 8th Edition, 1955, pp 127-128 and page 947-951.

Dreimanis, A., 1962, Quantitative Gasometric Deter mination of Calcite and Dolomite by Using Chittick

Apparatus; Journal of Sedimentary Petrology, Vol. 32,

No. 3, pp 520-529.

EA31-3

A - Decomposition Flask

B - Pipet

C - Stopcock

D L Gas-measuring Tube

E - Levelling Bulb

F - Magnetic Stirrer and Magnet

Figure CH l. Chittick Gasometric Apparatus

•EA31-4

D

X^'\

\

E x

20

200

v

J

V

)

6

1 i s e

0.

s a

G o

.3.2

E "*i o

1

1

I

*

^ j.

u in

S o

•c- o o

9

M

i

"y

(9

U

18

fi

1 c

*

JC

— O

IE ra

® 0) c w it GC

5

"S to ^ li

3

1 ffe

M xo

FormCHl.

05

.8

"m

J

3

9

0 3

•5

g

C

|

0)

Z tJ

'•5

?

K

3 fi

1

I i i

1

•o

s

5 u.

j

O

U

*

B

3 o

1

S

^

E

3

i

O)

6

I

E

8

H

-

M

T

,-

"EA31-5

Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertisement