/ >> How to Locate Gas Chromatography Problems
1% and Solve Them Yourself....
EE ete By applying a systematic approach to trouble-
y Kar solain Source shooting, you can solve many of your own GC prob-
N | | a of Problem lems. The real task is identifying the cause of a prob-
lem in the shortest possible time. This guide outlines
N Win techniques that will enable you to troubleshoot your
ra T gas chromatography step-by-step. You'll reduce re-
\ — pair costs and instrument downtime.
С =
Suggestions for Effective Troubleshooting
Certain basic procedures make troubleshooting faster and
easier. Most important are close observation of operating
parameters and a good record keeping system (temperatures,
flow rates, chart speeds, column type, stationary phase type
and amount, solid support type and mesh size, etc.). Also of
primary importance is a reference standard containing known
concentrations of the components in your sample. The refer-
ence standard should contain no extraneous, unknown
components. Many hours can be wasted hunting problems
within an instrument or column, when the problem is in fact
the sample being analyzed. If your chromatographic system
separates the reference standard well and reproducibly, any
existing problem is most likely related to the sample.
Your troubleshooting will progress more smoothly if you
have on hand:
Chgcking Carrier
1. A duplicate analytical column, one you know will
provide acceptable separation under proper condi-
tions — Try this duplicate column in your malfunc-
tioning system. If it corrects your problem, the prob-
lem is related to the original column.
2. A new syringe to help isolate the source of ghost
peaks — Repeat the analysis with a new, clean syr-
inge. If your trouble disappears, the problem has been
isolated to a defective or dirty syringe used during the
original analysis.
3. Snoop® or other leak detection aids —Use these to
ensure that your entire system is free of leaks, manda-
tory for proper operation. (NOTE: Don't use these
liquids with capillary columns.)
4. Spare septa and high temperature septa — These
help to identify problems with reproducibility or
ghosts caused by leaking or bleeding septa. Simply
replace your septum with a new duplicate septum or
with a higher temperature septum. If the symptom
disappears, the trouble was a leaking or bleeding
Changing Column
and Septu
5. Detector cleaner — A dirty detector creates noisy
baselines. Flame ionization detectors (FIDs) may be
cleaned using either Freon® TF, an in-place cleaner, or
an ultrasonic bath filled with an immersion cleaner.
6. Thermometer — To verify the oven temperature,
Guide 792C ruling out a defective temperature control.
E 7. Spare ferrules — Often necessary to eliminate leaks
in connections.
Cy = U = E L с; O 8. Soap bubble meter — To check your gas flow.
A ROHM AND HAAS COM SUPPLIED BY 9. Spare recorder and electrometer cables — To elimi-
SUPELCO, INC. * Supelco Park » Bellefonte, PA 161 SUPELCHEM | nate the recording system as a source of trouble.
Phone [814] 359-3441 and [B14] 358-344 10. Instrument manual.
TWX 510 670-3600 + FAX 814-359-3044 SUPELCO CHROMATOGRAPHY
© Copyright 1983. Supelco, Inc., Belle = (0799) 513320/51 3288 (6 LINES)
FAX: (0799) 513283
There are five major sources of problems in gas chromatog-
raphy: (1) the operator, (2) the sample, (3) the column, (4) the
electronics, and (5) the gas flow system. Various problems can
stem from any one of these sources. Eliminate them one-by-one
to isolate the actual cause of the problem.
To define your problem, first refer to the Troubleshooting Table
on page 5. Using the symptoms index at the front of the table,
locate your trouble symptom. If there is more than one symp-
tom, note the possible causes for each. If one cause is common
to all symptoms, this is most likely the source of your problem.
(Note that while the troubleshooting table contains most of the
symptoms you will encounter, it cannot cover all potential prob-
lems.) When you cannot find a rapid solution using the
troubleshooting table, you must systematically isolate the trouble:
1. Carefully note the symptoms (e.g. broad peaks, unresolved
peaks, long retention times).
2. Rule out operator error by double checking all operating
parameters, such as temperature, carrier gas flow, column
description, etc.
3. Check for a sample problem by injecting a reference stan-
dard. If you get a good chromatogram, the problem is most
likely sample related. If not, the problem is probably col-
umn or instrument related.
4. Check for a column problem by replacing the column with
a duplicate column, one known to provide good results
under proper conditions. If good results are obtained, the
problem is related to the original column. If the symptom
persists, the problem is related to the instrument.
5. Isolate equipment related problems by listing the equipment
systems which can cause the observed symptoms (e.g. for
broad peaks with long retention times: (1) carrier gas sys-
tem, (2) column heating system, and (3) injection port heat-
ing system). Next, isolate the problem within the suspected
Dexsil — Dexsil Chemical Corporation
Freon — E.l. du Pont de Nemours & Co., Inc.
Glasrench — Supelco, Inc.
Graphlok — Scientific Glass Engineering Party Ltd. (Australia)
Pyrosep — Supelco, Inc.
Snoop — Nupro Company
Supeltex — Supelco, Inc.
Swagelok — Crawford Fitting Co.
Teflon — E.l. du Pont de Nemours & Co., Inc.
Thermogreen — Supelco, Inc.
VESPEL — E.l. du Pont de Nemours & Co., Inc.
For current product prices, and a complete explanation of Supelco’s limited
warranty, please see our most recent catalog.
6. Isolate possible electronic system malfunctions (detector,
electrometer, recorder, and associated wiring) by perform-
ing the following checks:
NOTE: If your instrument is equipped with
dual channels (detector, electrometer, record-
er, etc.), see paragraph (c).
(a.) Check the recorder by setting the gas chromatograph's
attenuation to infinity. The recorder pen should go to
electronic zero. If the symptom (baseline drift, noise,
etc.) disappears, the recorder is not the problem. If the
symptom continues, refer to the recorder instruction
To isolate the detector (FID) as the source of trouble,
turn off the instrument and disconnect the cable (at the
detector end) from the detector to the electrometer.
(Note: To prevent inducing extraneous noise onto the
cable, it may be necessary to install a coaxial cap on
the free end.) If the symptom disappears when power
is on, the problem is the detector. If the symptom con-
tinues, disconnect the same cable at the electrometer.
If the symptom now disappears, the cable is defective.
Replace it. If the symptom still continues, refer to the
electrometer instrument manual.
(C.) If your chromatograph is equipped with dual detector
channels, you have a simple but effective alternate
means of identifying the problem source using the sec-
ond, or parallel, detector and electrometer. If the symp-
tom occurs in channel A, disconnect at the detectors
(shown below) the cables which connect channels A and
B detector outputs to channels A and B electrometer
inputs. Reconnect the cable from electrometer B input
to the output of detector A. This applies the signal from
the detector output of channel A to the electrometer in-
put and recorder of channel B. If the symptom does not
appear on recorder channel B after this cable change,
then either the electrometer, recorder, or cables of chan-
nel A are defective. If the symptom is not eliminated,
the detector of channel A is the problem source.
Alternate — Ju | |—— Alternate
Connection When Connection When
Problem 16 in Problem is in
Channel A | | Channel В
| u ~~] Input Input |
Output [CI Ll Output |
| Detector lectrometer Recorder
B B В |
Channel B
7. Check the carrier gas system for possible problems. Refer
to the following section, Checking the Carrier Gas System.
A typical carrier gas system is illustrated at the bottom of this
page. The most common problem within this system is insuffi-
cient carrier gas flow through the chromatographic column. This
is generally caused by (1) insufficient source pressure, (2) leaks,
and/or (3) an unusually large pressure drop across one of the
components in the system.
Verify the column carrier gas flow at the detector exit using
a soap bubble flow meter and a stopwatch. (We do not recom-
mend using rotameters for measuring gas flow because specific
rotameters are required for each type of gas used, and they pos-
sess a definite linear response range with pressure changes.)
To verify that the carrier gas flow rate is properly adjusted,
the time required for the soap bubble to travel a specified distance
at the desired flow rate should be calculated as follows:
Time (sec.)= Volume Observed (cc) x
60 sec./min.
Desired Flow (cc/min.)
Where: Time (Sec.) = time required for the bubble
to travel the observed
Volume Observed = volume indicated by the soap
bubble flow meter.
Desired Flow = the rate specified by the
method being used.
Example: Obtain a flow rate of 20 cc/min. using
a 10cc soap bubble meter.
Time (sec.)= 10cc x 60 sec./min.
Time (sec.)= 30 sec.
If the time required is not equal to the calculated time, adjust
the carrier gas flow rate. If sufficient flow cannot be obtained by
adjusting the flow control valve, the problem is probably due to
inadequate source pressure (observed at P1). Increasing this
pressure will usually provide adequate flow. Normally, a source
pressure of 60psig is sufficient for 6 to 12 foot columns. Remem-
ber, increasing the column length, oven temperature, and/or flow
rate will require raising the source pressure. The source pressure
ı is particularly important if you are using a temperature program,
since the pressure must be 10-15psig in excess of the column
pressure drop at the maximum temperature. This pressure dif-
ference allows the differential flow controller to function proper-
ly. If the correct pressure difference is not maintained, carrier
gas flow will drop drastically at elevated temperatures.
Typical Carrier Gas System
P1 Р2
Molecular Oxygen Hotameter Flow
Sieve Trap Control
Trap Valve
Other common causes of inadequate gas flow are leaks in the
system and large pressure drops across one or more of the sys-
tem components. The use of pressure gauges can save con-
siderable time when isolating these problems. Some common
leak points are: column connections, the septum, and connec-
tions for the various valves and gas purifiers.
A pressure gauge installed between the particle trap and the
injection port (P3 in the gas system illustration) indicates column
head pressure. A low reading at this point indicates a leak be-
tween P3 and the detector outlet (e.g. a defective column, sep-
tum, etc.) or a large pressure drop across a previous component
(e.g. a plugged oxygen trap). A high reading at this point indicates
an over-tightened septum, dirty detector, too-tightly packed col-
umn, etc. Similarly, a pressure gauge at P2 will reveal an ex-
hausted oxygen trap (low pressure readings indicate a larger than
normal pressure drop across the trap). Routine observation of
this pressure enables you to determine when the carrier gas
purifier should be changed.
NOTE: Many chromatographs have crimps in.the carrier
gas line between the flow controller and the injection port
or employ capillary tubing with a small internal diameter.
Consequently, the pressure reading at point P3 will be dif-
ferent from the column head pressure reading taken through
the septum. These restrictions can also make it difficult to
obtain sufficient carrier gas flow, particularly when convert-
ing instruments for use with capillary columns.
The most common method of leak testing is to apply liquids
such as Snoop or HT-Leak Detector, and then watch for bub-
bles to appear. Exercise care when using these liquids since they
can contaminate your chromatographic system, causing unstable
baselines and ‘ghost peaks’ in subsequent chromatograms.
(These liquids can enter the GC system through an aspirator
Several devices are available for leak testing, such as ther-
mal conductivity leak detectors and pressure gauges. (The lat-
ter were discussed in the previous section.) When pressure
guage readings indicate a possible leak, take the appropriate
corrective action (such as tightening the fittings which would
cause this abnormal pressure reading) rather than verifying the
leak point with Snoop or HT-Leak Detector and possibly contam-
inating the chromatographic system. A specific technique for test-
ing small septum leaks while avoiding contamination employs
the Supelco leak tester, a plastic capillary tube with two conical
ends. One end is dipped into Snoop, and capillary action pulls
a small amount of Snoop into the tube. If a leak is present, bub-
bles appear at this end when the opposite end is pressed against
the septum nut.
P3 Septum
Particle Injection
Trap Port
When changing columns or septa, it is important that you first
turn off the chromatograph oven and allow the column to cool
for 10 to 15 minutes, then turn off the carrier gas. Cooling the
column and turning off the carrier gas protects your column in
two ways: allowing the column to cool prevents oxidation of the
column packing which can occur when a hot column is exposed
to oxygen in the air. Allowing the column pressure to drop to am-
bient pressure prevents the packing from popping or blowing out
of the column ends. A sudden change in pressure, when a col-
umn or a septum is removed with the carrier gas flowing, can
blow packing from the column.
NOTE: When storing columns, cap the ends with metal
Swagelok® caps to prevent diffusion of air (and
subsequent oxidation) into the column. Plastic caps
do not prevent diffusion of air into the column.
Improperly installed columns and septa are a frequent source
of leaks, and are the most common cause of glass column break-
age. Incorrectly tightened septum nuts present problems such
as excessive septum bleed, premature occurrence of septum
leaks, and low carrier gas flow rates.
Supelco offers two types of torque wrenches to help ensure
correct installation of columns and septa. The Glasrench, used
for installing columns, is available in two different torque settings
to provide the correct torque for the various types of ferrules.
This saves both analysis time and money by eliminating leaks,
over-tightening, and column breakage. Select and use the ap-
propriate Glasrench for the ferrule you are using.
The Supelco septum nut torque wrench ensures that the cor-
rect torque is consistently applied when installing septum nuts.
Improper sample injection can cause many problems in gas
chromatography. To ensure that your injections are accurate and
reproducible, the following general guidelines and procedures
are recommended:
A. Syringe Size. Always use a syringe large enough that the
desired sample volume does not fill it to capacity, and small
enough that the sample volume is not less than approx-
imately 10% of its capacity.
B. Injection Technique. Sample injection should be smooth
and rapid with quick removal of the syringe after injection,
in order to avoid peak broadening.
C. Sample Size Reproducibility. Many problems in chroma-
tography result from difficulties in reproducing the size of
a sample. Some techniques which will help ensure repro-
ducible samples are discussed below:
1) Automatic Injectors. These devices improve sample
reproducibility by virtue of consistent mechanical opera-
tion. Each step (syringe cleaning, sampling, and sam-
ple injection) is repeated precisely.
2) Sampling Valve Injection. Sample reproducibility is im-
proved because chances for variability are greatly re-
duced. Sample size is determined solely by sample loop
size, and injection is rapid and precise.
3) Solvent Flush Technique. The following technique is
recommended to reduce the problem of irreproducible
injection volumes when making syringe injections by
Air Plunger
Air Air Plunger
Sample Solvent
(a) Eliminate sample hang up in the needie by first
cleaning the syringe, then drawing in a small aliquot
of solvent.
(b) Remove the syringe from the solvent and draw in
a small amount of air.
(c) Draw in the desired amount of sample.
(d) Remove the syringe from the sample and draw in
a little more air.
(e) Verify the amount of sample in the syringe barrel.
This is only possible with syringes that do not have
plungers in the needle.
(1) Quickly and smoothly inject sample into the chro-
4) Syringes With Needle Plungers. Improve sample
reproducibility by using a syringe with a plunger in the
needle. This eliminates sample retention in the needle
dead volume. The solvent flush technique, (3) above,
may be useful, since a small amount of sample hang-
up can still occur.
In addition to information presented in this guide, more helpful
tips are available to save time and money in chromatography.
Some of this information is provided FREE, upon request, in the
following Supelco technical bulletins:
e Bulletin 739 — Reviews problems caused by using im-
pure carrier gas, and methods of extend-
ing column life.
e Bulletin 780 — Discusses the selection of a septum
which will resist leaks and exhibit low
bleed at high operating temperatures.
This bulletin provides test results
demonstrating the quality of Ther-
mogreen™ LB-1 septa.
e Bulletin 783 — Provides information pertaining to dirty
flame ionization detectors (FIDs), and
offers hints to help prevent contam-
ination, as well as instructions for
s Bulletin 741 — Reviews various types of ferrules and
methods of preventing problems when
connecting glass columns.
s Bulletin B14 — Discusses use of wide bore capillary
columns in packed column gas chro-
ECD — Electron Capture Detector
FID — Flame lonization Detector
FPD — Flame Photometric Detector
GC — Gas Chromatography
NPD — Nitrogen Phosphorous Detector
TCD — Thermal Conductivity Detector
Symptom Symptom Symptom
Symptom No. Symptom No. Symptom No.
Baseline — Detector response, low 3,4 Peaks —
changing 23 missing 2
cycling 9 Ghost peaks — negative 12
dip 25, 26 random 13 no peaks 1
drift 7 sample memory solvent, broad 27
drop 24 peaks 14 unresolved 29
noise 8
off scale Peak shapes, incorrect — Quantitation,
(zeroing) 6 cigar top 20 irreproducible 5
rise 22 clipped 21 a
round top 19 Retention time,
Carrier gas, low skewed too long or short 28
flow rate 32 (leading edge) 16 |
split 17 Spikes —
Column life, short 33 square top 18 irregular 11
tailing 15 regular 10
Column packing —
compacted 31
gaps in 30
Possible Cause
Symptom No. 1
No Peaks
1. Detector or electrometer power
off or fuse blown
2. Sample injected in wrong col-
umn in multiple-column chromato-
3. Flame ionization detector (FID)
not lit
4. No carrier gas flow
5. Defective syringe
1. Check detector, electrometer set-
ting, and fuses.
2. Re-injectsample in proper column.
3. Check FID with mirror over ex-
haust. If lit: H, O condensates on
mirror. If not lit, light flame.
Check H, and air flows.
4, Measure carrier gas at detector or
column exit. If no carrier gas flow
is present, check for leaks or ob-
structions at column connection
and septum,
Check for adequate pressure at
tank and ensure septum is not
over-tightened (flow cut off by
septum obstructing the column).
5. Replace syringe
nents from solvent
Symptom Possible Cause Remedy
(Symptom No. 1, cont'd.) 6. Column or septum leak 6. Replace the septum.
Check column connections.
7. Injection port temperature too 7. Increase injection port tempera-
low, sample not vaporized when ture, (but not in excess of liquid
analyzing high boilers phase temperature limit), or inject
sample directly onto column
8. Defective recorder 8. Check recorder connections.
Check recorder zero.
Troubleshoot recorder according
to recorder manual.
9, Defective detector, electrometer, 9. Check collector voltage and con-
or cables nections per instrument manual.
10. Bad connection between FID col- | 10. Check “collector spring clip’ con-
lector and voltage source nection.
Symptom No. 2
Solvent peak only, or missing 1. Sample too dilute 1. Check system by injecting with
peak known standard, If okay, increase
sensitivity or inject larger or more
concentrated sample.
2. Column or septum leak 2. Check for leaks (see page 4).
Tighten connections.
Replace septum.
3. Incorrect temperatures 3,
| | | (a.) Injection port or column (a.) Ensure that the selected col-
temperature too low (sam- umn temperature setting Is
Normal ple not vaporized) correct for column being
used and the sample being
analyzed. Then verify that
the oven is operating at the
selected temperature. In-
crease the temperature as re-
(b.) Injection port temperature (b.) Decrease injection port tem-
IL too high for thermally labile perature.
(c.) Column temperature too (c.) Decrease column tempera-
high, sample eluting in sol- ture,
vent peak
4. Flow rate incorrect 4. Measure column flow rate and ad-
just if necessary (see page 3).
5. Sample adsorption by column 5. Inject standard on known “good”
or glass wool column. If okay, original column is
№ =
Use properly treated glass wool
Problem (i.e., Hy PO, for free acid analysis,
silicone-treated for other com-
If sample has never been analyzed
and is active chemically, you may
need a special column.
6. Column cannot separate compo- 6. Change column or solvent.
Possible Cause
Symptom No. 3
Low response (all) — correct 1. Poor injection technique 1. Use correct syringe size and sol-
retention time vent flush techniques (see pages 4
and 5).
2. Wrong sensitivity setting or sam- 2. Check and correct if necessary.
ple size too small Inject standard sample for com-
3. Defective syringe 3. Use new syringe.
4. Septum leak 4. Replace septum,
| В 5. Injection port temperature too 5. Increase injection port tempera-
Normal low for sample ture.
6. FID only — Hs or air flow incor- 6. Measure and correct if necessary.
7. FID only — low O; level in com- 7. Replace air tank,
pressed air
8. FID only — faulty connection be- 8. Clean "collector spring clip” with
tween FID collector and voltage emory paper.
9. Dirty electron capture detector 9, Clean per instrument manual.
Problem (ECD)
10. (a.) Thermal conductivity detec- 10. (a.) Measure and adjust carrier
tor (TCD) carrier gas flow gas flow if necessary.
rate is incorrect
(b.) Cell voltage is incorrect (b.) Refer to instrument manual.
11. Adsorption of sample by column, 11. Use deactivated column materials.
glass wool, tubing, etc.
12. Flame photometric detector (FPD) 12. Check with known standard (hy-
only — hydrocarbon eluting with drocarbon free) and change to col-
sample, causing diminished re- umn that will separate hydrocar-
sponse due to quenching effect bons from components of interest.
Symptom No. 4
Low response — long retention 1. Low carrier gas flow rate 1. Measure and adjust carrier gas flow
time rate if necessary (see page 3).
2. Carrier gas leak at septum or col- 2. Check for leaks and correct if nec-
umn connections essary (see page 4).
3. Column temperature too low 3. Increase column temperature.
| 4. Column worn out or conditioned 4. Verify column temperature and
at too high a temperature stationary phase temperature li-
== - mits.
Normal Check sample with known good
Repack first 6" of column or re-
place column.
Symptom No. 5
Quantitation not reproducible 1. Sample incorrect 1. Verify using known standard.
a. Retention time okay. Compo-
nents with longest retention
time show low value when
using “normalization” tech-
a. 1. Incomplete sample injection
2. Injection port or column
temperature too low
3. Incorrect slope sensitivity
with electronic integrator
a. 1. Use solvent flush techniques.
2. Increase temperature (see
page 4).
3. Adjust slope sensitivity.
Possible Cause
Remedy |
(Symptom No. 5, cont'd.)
b. Retention time okay. Different
components not yielding simi-
lar peak areas for same amount
c. Quantitation varies for one
component eluting over wide
time span, even using internal
standard technique
d. Inconsistent quantitation for
same sample on successive
e. Low values for minor com-
f. Increased peak response with
successive injections
1. Differing detector response
for different components
2. Adsorption of components
by packing, glass wool, tub-
ing, or transfer lines
: 1. Internal standard not com-
pensating for all components
in sample
2. Slope sensitivity of integra-
tion not high enough for late
_ Insufficient resolution of peaks,
or peak tailing
‚ Sample size too small for ac-
curate counting by integrator
_ Adsorption of components and
saturation of active sites with
sample (priming the column)
_ 1. Determine correction fac-
tors and/or use internal
standards techniques.
2. Use deactivated system.
‚ 1. Use multiple internal stan-
2 Use multiple internal stan-
. Modify operating parameters
or replace column to improve
resolution and eliminate tailing.
Increase sample size or elec-
trometer range setting.
Use deactivated system.
Symptom No. 6
Baseline off-scale,
cannot zero
Y ‘|
0 n=
Column not conditioned properly,
contaminated column, or temper-
ature too high
. Recorder problem
. Septum leak
. Wrong gases (e.g., argon/methane)
with flame
. Contamination
. Too much or too little gas flow
Imbalance in column flow if using
_ Contaminated detector, such as
nitrogen phosphorous detector
(NPD) with Snoop, or ECD with
chlorinated solvents
. Electrometer or detector problem
. Verify by reducing column tem-
perature to ambient. If baseline is
normal, check system with good
column. If okay, recondition
“bad” column.
Get attenuation to infinity. If re-
corder goes to electrical zero, it's
okay. If it does not go to zero,
troubleshoot recorder per recor-
der manual.
Check septum for leaks and cor-
rect leaks if required (see page 4).
_ Verify gases are correct for instru-
ment and detector as specified in
_ Turn off injection port heat. If
zeroing capability returns, clean
injection port liners, etc.
. Check and adjust to within manu-
al specifications.
. Check and adjust as required.
~ Avoid sources of contamination.
(i.e. Don't use Snoop with NPD or
chlorinated solvents with ECD.)
_ Troubleshoot instrument per man-
Possible Cause
Symptom No. 7
Baseline drift
. Carrier gas flow changing with
temperature during temperature
. Septum or column leaks
. Septum bleed or septum fragments
in column
. Column bleed or contamination
. Gas flows not within minimum/
maximum limits (including H,
and air on FID) or poorly regu-
lated flow
. Insufficient instrument warm up
time or temperature equilibration
. Defective electrometer or detector
. Contaminated detector or injec-
tion port
. Increase carrier gas source pres-
sure to 15psig above column head
. Check and correct as required (see
page 4).
. Replace septum with higher tem-
perature type, and repack inlet
side of column,
Always use sharp syringe needle
and inject in same place. Use of a
needle guide is recommended.
. Replace column with known good
column. If okay, column requires
. Measure flows and verify against
manual specifications,
. Allow time for instrument to
equilibrate when changing oper-
ating temperature or installing
another column.
. Troubleshoot as in Isolation of
Problem Source section.
. Clean as recommended in instru-
ment manual.
Symptom No. 8
Irregular or unstable baseline
Laser sca
. Column bleed or contamination
. Contaminated detector or injec-
tion port
. Carrier gas leak
. Poor carrier gas regulation
. Gas impurities or contaminated
gas line
. Gas flows not within minimum/
maximum limits (including Hz;
and air on FID) or poorly regu-
lated flow
. Defective electrometer, detector,
or cable
. FID only — collector incorrectly
. ECD only — heater wire too close
to detector wire, causes AC noise
. Check out with good column; if
okay, recondition “bad” column.
. Clean detector and/or injection
. Check for and correct septum or
column leaks (see page 4).
. Check carrier gas supply for suf-
ficient pressure. Replace tank if it
is near empty.
. Change gas tank, use gas purifier
and clean metal tubing.
. Measure flows and verify against
manual specifications.
_ Troubleshoot as in Isolation of
Problem Source section.
. Realign as required.
. Reposition heater wire.
Possible Cause
Symptom No. 9
Cycling baseline drift . Poor instrument location (drafts, 1. Relocate instrument, close win-
changes in ambient temperature, dows, etc.
. Defective detector or oven tem- 2. Replace temperature sensing
perature controller probe.
| . Defective column oven tempera- 3. Replace temperature sensing
ture controller probe.
. Carrier gas flow irregular: insuf- 4. Replace gas tank.
ficient supply pressure
. Defective carrier gas regulator 5. Replace regulator.
LLL — . Defective carrier gas flow con- 6. Replace flow controller.
Problem troller
. Sensitivity too high if using 7. Reduce detector sensitivity or de-
pumped gases, such as from hy- crease output pressure from gener-
drogen generator ator.
Symptom No. 10
Spikes (regular) . Condensate or dust particles in 1. Clean detector and check ends of
FID column to ensure glass wool is in
Normal . Contaminated gases 2. Replace gases or insert liquid ni-
trogen trap in gas line.
ТЕТ! | . Defective electronics or detector 3. Troubleshoot electronics and de-
Problem tector as in Isolation of Problem
Source section. Closely check the
1 Li 0 A recorder cables and detector/elec-
For LONA trometer cables,
Symptom No. 11
Spikes (irregular or erratic) . Defective cable, intermittent short- 1. Replace cable.
L : ECD heater wires and detector 2. Check wire position and relocate
Norma wire too close, or loose if necessary.
. FID — insufficient H, flow 3. Increase H, flow.
a | | | | | | . Electronic interference from ex- 4. Relocate instrument, and deter-
Problem ternal source mine possible interference sources
(e.g. nearby transmitter site, etc.).
Symptom No. 12
Negative Peaks . Recorder improperly connected, 1. Reverse recorder connections or
polarity reversed, or sample in- polarity switch.
jected into wrong column
. TCD only — impurity in carrier 2. Install or replace carrier gas puri-
gas fier.
Possible Cause
Symptom No. 13
Ghost peaks (peaks similar to pre-
vious sample appear when solvent
alone is injected)
Previous Sample
(solvent injected after sample)
(solvent injected after sample)
. Dirty syringe
. Column adsorption and subse-
quent desorption of sample (par-
ticularly in temperature program)
. Adsorption in transfer line
1. Try new syringe with clean sol-
vent, If ghost peaks disappear,
clean syringes more thoroughly.
2. Try more inert column materials
(i.e. tubing, packing and glass
3. Use glass-lined stainless steel for
transfer lines.
Symptom No. 14
“Extra” peaks (unlike sample
. Septum bleed, particularly in tem-
perature program 2
. Peaks from previous runs, particu-
larly if very broad with short re-
tention time
. Impurities from sample, solvent,
sample container (e.g. plasticizer
from cap liners or contaminated
glassware), labware and reagents
used in sample preparation, parti-
cularly when excess reagents are
concentrated in work-up
. Condensed carrier gas impurities
eluting during temperature pro-
1. Turn off injector heater. If ghost
peaks disappear, operate at lower
injector temperatures or use high
temperature septum.
2. Let analysis run longer and then
3. Run solvent blank with clean sy-
ringe. If extra peaks appear, change
solvent; if no extra peaks appear,
run solvent blank through entire
sample work-up without sample.
If no extra peaks appear, impuri-
ties are from sample. If extra
peaks appear, repeat solvent blank
from each step of work-up to iso-
late source.
4. Install or replace carrier gas puri-
5. Verify by analyzing lab environ-
"ram . Trace impurities in lab atmosphere ment. Take corrective action as
La necessary.
Cola кА. |. Cosi leds) Arno ELEC 7 as Cora. e NA
So stom
Fon a Ce 5
Coad sen a ox в
RR i ites —
Possible Cause
(Symptom No. 14 contd.)
6. Air peaks or H, O peaks
7. Multiple (or incomplete) deriva-
tives formed in sample work-up
8. Sample decomposition
6. This is not a trouble symptom,
but normal with TCD, using
syringe injection or an aqueous
7. Re-evaluate derivatization proce-
8. Try lower temperature and/or dif-
ferent column.
Symptom No. 15
Tailing Peaks
= =
1. Column or injection port temper-
ature too low
2. Column deteriorating
3. Active sample adsorbing on injec-
tion port, transfer lines, column,
or glass wool
4. Two compounds eluting
5. Needle hitting and breaking pack-
ing in column inlet
1. Increase temperature (do not ex-
ceed maximum temperature for
2. If peaks tail on old column but re-
tention times have not changed
from what they were when col-
umn was new, replacing first 6"
of packing or replacing the pre-
column may help. If retention
times have changed, column prob-
ably should be replaced.
3. Use more inert system: “‘all glass”
or Teflon” specially designed pack-
ing, on-column injection, and
proper glass wool type.
4, Increase sensitivity and reduce
sample size. Reduce temperature
approximately 20°C and look for
partial separation,
5. Remove some packing so needle
won't penetrate it.
Symptom No. 16
Leading Peaks
— —
1. Column overload
2. Two components eluting
3. Sample condensation
4. Sample decomposition
1. Decrease sample size.
2. Decrease sample size, increase sen-
sitivity, and decrease temperature
by 200C. Look for partial separa-
3. Check injection port and column
temperatures, and increase if
4, Use inert system and deactivated
E 1
Possible Cause
Symptom No. 17
Split Peaks
1. Gross detector overload
2. Sample flashing prior to injec-
tion — simulates two injections
1. Reduce sample size.
2. Use syringe with solvent flush
technique so that sample is con-
tained in barrel, not in needle.
Use less volatile solvent,
Symptom No. 18
Square or flat top peaks
1. Electrometer saturated (normal
for solvent)
2. Defective recorder
1. Reduce sample size.
2. Troubleshoot recorder per re-
corder manual.
Symptom No. 19
Round top peaks
1. Detector overload — FID
2. Recorder gain too low
1. Decrease sample size,
2. Adjust gain control.
Possible Cause
Symptom No. 20
Cigar top peak
Detector overload — ECD
Reduce sample size,
Symptom No. 21
Clipped peaks — column
efficiency exceptionally high
Recorder or instrument zero be-
low minimum moveable range of
recorder pen
Shunt recorder leads and set re-
corder baseline adjustment zero to
approximately 5% of full scale.
Symptom No. 22
Baseline rise before
or after peak
Problem Problem
Sample decomposition
Use inert columns and packings.
Possible Cause
Symptom No. 23
Baseline change after
large peak
1. Contamination — water or large
component is stripping contami-
nants from column
. Column not conditioned properly
— liquid phase being stripped
. Pressure imbalance when gas sam-
pling valve activated
1. Repack first 6” of column or re-
place pre-column,
2. Recondition column.
3. Correct pressure imbalance.
Symptom No. 24
Baseline drop after peak —
FID only — flame extinguished
1. Sample too large
. Incorrect gas flows
. Flame tip plugged
. Collector and tip not located
properly: whistling or humming
noise often heard
. Decrease sample size.
. Check and adjust carrier gas, Hs,
and air.
. Clean or replace flame tip.
. Adjust collector position.
Possible Cause
Symptom No. 25
Negative dips after peaks
1. Only after large peak such as sol-
vent: sample too large
2. After all peaks with ECD: dirty
detector cell
1. Decrease sample size.
2. Clean detector.
Symptom No. 26
Negative dip before peak
1. Pressure imbalance when gas sam-
pling valve activated
1. Correct pressure imbalance.
Symptom No. 27
Broad solvent peaks
1. Dead volume in injection port due
to poor column installation
2. Normal with very dilute sample as
in trace analysis
3. Poor injection technique
4. Injection port temperature too
5. Sample solvent interacts with the
6. Sample solvent retained by col-
umn (e.g. methanol by active
1. Use on-column injection. Ensure
proper column connections, par-
ticularly when changing from one
column diameter to another.
2. —
3. Use smooth, rapid injection (see
page 4).
4, Increase injection port tempera-
5. Change sample solvent.
6. Change sample solvent.
Possible Cause
Symptom No. 28
Retention time longer
(or shorter) on same column
. Column temperature too high
(too low)
. Carrier gas flow rate too low (too
. Septum or column leak
. Column contamination or deterio-
. Recorder problem
. Sample overload
. Check temperature with indepen-
dent thermometer, and adjust
temperature as required.
. Measure flow rate with soap bub-
ble flow meter at column exit
and adjust as required.
. Check and correct as required.
. Repack first 6" of column, or re-
place column,
. Check recorder chart speed.
. Reduce sample size,
Retention time on new column is
longer (or shorter) than on previ-
ous column of the same composi-
. Column too long (too short)
. More (less) packing in column
due to:
a. Support density greater
(less) than previously used
b. Column packed more tight-
ly (loosely)
c. Column inside diameter
greater (smaller)
. Percent coating of stationary
phase on support is too high (low)
due either to different manufactur-
ing procedures or errors
. Different stationary phase used,
or slight variation in phase com-
position (particularly commercial
chemicals not manufactured for
. Different support type
This problem can be remedied by
using pretested, standardized col-
umns and packing materials.
Possible Cause
Symptom No. 29
Unresolved peaks, on column
which previously produced
good results
1. Wrong column temperature
2. Wrong carrier gas flow rate
3. Sample problem
a. Sample size too large
b. Sample concentration different
than previous analysis, causing
“swamping” of minor peak by
major peak
4, Poor injection techniques (slow)
5. Column contaminated or deteri-
. Check and adjust temperature.
. Check and adjust flow rate,
. Use smooth, rapid injection.
. Repack first 6 of column, or re-
a, Reduce sample size.
b. Reduce sample size.
place column,
0 №
Unresolved peaks, after previous
column of the same composition
produced good results
1. Column too long (too short)
2. More (less) packing in column due
a. change in support density
b. packed more tightly (loose-
c. column inner diameter dif-
3. Percent coating increased (de-
creased) either due to different
manufacturing procedures or er-
4. Different stationary phase used,
or slight variation in phase compo-
sition (particularly commercial
chemicals not manufactured for
5. Different support type
Each of these problem sources
can be remedied by using pre-
tested, standardized columns and
packing materials.
Possible Cause
Symptom No. 30
Large gaps appear in
packing in glass column
Improperly packed column
Add enough packing to fill in voids,
then gently vibrate until smooth. If
this does not solve the problem, col-
umn should be repacked.
Symptom No. 31
Packing compacting or
shrinking after conditioning
1. Slight compacting may occur
when packings are exposed to
2. Large shift (3" or greater)
1. You can use this column. Shifts
are not normally a problem.
2. Column may not be properly
packed. Add more packing.
Symptom No. 32
Low carrier gas flow and
high pressure drop
1. Overtightened septum
2. Insufficient carrier gas source pres-
3. Insufficient source pressure for
temperature program
4. Plugged injection port, carrier gas
line, or gas purifiers
5. Column over-packed or glass wool
too tight
1. Loosen septum.
2. Increase gauge pressure by 10psig.
3. Flow control must have 10-15psig
higher than maximum pressure
(reached at maximum tempera-
ture) to function properly.
4. Replace tubing or gas purifiers as
5. Increase carrier gas pressure.
(Note: not all packings have same
pressure drop.) If flow is still in-
sufficient, install another column.
If flow is okay, column was the
problem. If not, check plumbing
system for flow restrictions (e.g.
plugged detector, plugged gas fil-
ters, etc.).
Possible Cause
Symptom No. 33
Column deteriorates too soon
after installation (peaks tail, are
poorly resolved, etc.)
1. Operating column near or above
maximum temperature limit of
2. Water or O, in carrier gas is con-
taminating column
3. O, contamination due to column
4. Aqueous samples, serum, plasma
and other complex samples
1. a. Use higher temperature phase.
b. Use shorter column and lower
temperature if possible.
c. Reduce temperature when col-
umn is not in use.
d. Remove column when another
column is used at higher tem-
2. Use carrier gas purifier and appro-
priate grades of gases. Replace
tanks before pressure becomes too
low (300psig).
3. Check for leaks prior to Use. Al-
low columns to cool before re-
moving from GC to prevent col-
umn exposure to air while hot.
4. These samples can (1) strip phase
from support, (2) chemically react
with phase, (3) build up on col-
umn and possibly destroy the col-
umn, injection side first. Use of
pre-column, or repacking the inlet
will extend column life.
List of Bulletins by Subject
(Analyses are by GC, unless otherwise indicated)
FREE Technical Literature to Help You Select
the Right Column for Your Analysis
Consistent and accurate chromatographic analyses are dependent
not only on proper system operation, but also on packing materials
and analysis techniques.
Examine the following
list of Supelco
Bulletins. Select the right one (or more) to
help you perform your particular analysis.
Dibasic (Dicarboxylic) Acid
Methyl Esters (up to C10) 762
Elaidate/Oleate — separation
of cis/trans isomers 752
Fatty Acids, Free (C14-C20) 727
Fatty Acid Methyl Esters (C1 4-C24) 746
Fatty Acid Methyl Esters
(postional cis/trans isomers) 822
Fatty Acids, Volatile (C2-C5) —
at ppm levels in water 751
Fatty Acids, Volatile and Nonvolatile —
esterification and separation 721
Acylation 721
Alcoholic Beverage Analyses 790
Blood Alcohols (GC/HPLC) - 810
Acylation 721
Aliphatic and Aromatic Amines —
acylation, columns, separations 737
Nitrosamine Priority Pollutants 775
N-Nitrosamines — contaminants
in foods or wastewater 791
Phenylenediamines (GC/HPLC) 805
Homologs and Polymers — separation
by size exclusion HPLC (HPLC) 796
Proteins (HPLC) 795
Capillary Column Butt Connector 801
Glass GC Columns of
Consistent Dimensions 804
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