David_Valdesueiro_-_PhD_thesis.

David_Valdesueiro_-_PhD_thesis.
Gas-Phase Deposition of Thin Aluminium Oxide Films at Ambient Conditions
Gas-Phase Deposition of
Thin Aluminium Oxide Films
at Ambient Conditions
David Valdesueiro
David Valdesueiro
Gas-phase Deposition
of Thin Aluminium Oxide Films
at Ambient Conditions
Proefschrift
ter verkrijging van de graad van doctor
aan de Technische Universiteit Delft,
op gezag van Rector Magnificus prof. ir. K.C.A.M. Luyben,
voorzitter van het College van Promoties,
in het openbaar te verdedigen op
woensdag 4 november 2015
om 12:30 uur
door
David VALDESUEIRO
Ingeniero Químico
Universidad de Valladolid, Spanje
geboren te León, Spanje
This dissertation has been approved by the
promotor: Prof. dr. ir. M. T. Kreutzer
copromotor: Dr. ir. J. Ruud van Ommen
copromotor: Dr. ir. Gabrie M. H. Meesters
Composition of the doctoral committee:
Rector Magnificus
chairman
Prof. dr. ir. Michiel T. Kreutzer
Technische Universiteit Delft, promotor
Dr. ir. J. Ruud van Ommen
Technische Universiteit Delft, copromotor
Dr. ir. Gabrie M.H. Meesters
Technische Universiteit Delft, copromotor
Independent members:
Prof. dr. Andreas Schmidt-Ott,
Technische Universiteit Delft
Prof. dr. Christophe Detarvernier
Universiteit Gent
Prof. dr. Mato Knez
CIC nanoGUNE
Dr. Rob Geertman
DSM Fibre Intermediates
Prof. dr. Stephen J. Picken
Technische Universiteit Delft, reserve
The research reported in this thesis was conducted in the Product and Process
Engineering section of the Chemical Engineering department, Faculty of Applied
Sciences (TNW) of the Delft University of Technology and received funding from the
European Union Seventh Framework Program FP7/2007-2013 under grant agreement
no. 26722, and from Royal DSM.
ISBN: 978-94-6186-564-9
Copyright © 2015 by David Valdesueiro
Printed in The Netherlands by CPI Wöhrmann Print Service
All rights reserved. Parts of this thesis are published in scientific journals and
copyright is subjected to different terms and conditions.
To my parents
Contents
Summary
xi
Samenvatting
xv
1. Introduction
1
1.1 Research Field . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2
1.1.1 Particle technology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2
1.1.2 Coating techniques for particles . . . . . . . . . . . . . . . . . . . . . . . . . . .
2
1.1.3 Atomic Layer Deposition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3
1.1.4 Fluidization of particles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
5
1.2 Objectives of this study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6
1.3 Outline . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
7
2. Gas-phase deposition of ultrathin Al2O3 films on nanoparticles at ambient
13
conditions
2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
14
2.2 Experimental section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
15
2.2.1 Experimental setup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
15
2.2.2 Design of experiments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
16
2.2.3 Characterization of the coating . . . . . . . . . . . . . . . . . . . . . . . . . . . .
16
2.3 Results and discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
17
2.3.1 Base case coating experiments . . . . . . . . . . . . . . . . . . . . . . . . . . . .
18
2.3.2 Coating with excess of precursors at ambient conditions . . . . . . . .
19
2.3.3 Coating with excess of precursors at 1 bar and 170 °C . . . . . . . . .
21
2.3.4 Alumina film density . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
22
2.4 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
23
Supplementary information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
27
v
Contents
3. Tuning roughness and gloss of powder coating paint by encapsulating the
35
particles with Al2O3 films
3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
36
3.2 Experimental section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
37
3.2.1 Experimental setup and compounds . . . . . . . . . . . . . . . . . . . . . . . .
37
3.2.2 Coating experiments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
37
3.2.3 Characterization of the coated particles and panels . . . . . . . . . . . .
38
3.3 Results and discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
38
3.3.1 Characterization of the coated particles . . . . . . . . . . . . . . . . . . . . .
38
3.3.2 Characterization of the sprayed panels . . . . . . . . . . . . . . . . . . . . . .
45
3.4 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
51
Supplementary information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
55
4. The deposition mechanism of Al2O3 on QD films at atmospheric pressure and
69
room temperature
4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
70
4.2 Experimental . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
71
4.2.1 PbSe QD solution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
71
4.2.2 EDA in ethanol solution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
72
4.2.3 Quartz substrate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
72
4.2.4 Fabrication of QD films . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
72
4.2.5 Alumina coating reactor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
72
4.2.6 Alumina coating precursors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
73
4.2.7 Alumina coating experiments . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
73
4.2.8 Characterization of the deposition of Al2O3 . . . . . . . . . . . . . . . . . .
73
4.2.9 Characterization of the air stability of the coated QD films . . . . . .
74
4.2.10 Preparation of the dodecanethiol self-assembled monolayer . . . . .
74
4.3 Results and discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
74
4.3.1 Deposition of Al2O3 using water as precursor . . . . . . . . . . . . . . . .
75
4.3.2 Estimation of the growth per cycle of Al2O3 . . . . . . . . . . . . . . . . . .
77
4.3.3 Study of the deposition mechanism of Al2O3 . . . . . . . . . . . . . . . . .
77
4.3.4 Study of the reaction of TMA on a dodecanethiol SAM . . . . . . . . .
77
vi
Contents
4.3.5 XPS depth profile of an uncoated QD film . . . . . . . . . . . . . . . . . . .
78
4.3.6 Reaction between TMA and amine groups . . . . . . . . . . . . . . . . . . .
80
4.3.7 Study of the air stability of Al2O3-coated QD films . . . . . . . . . . . .
80
4.3.8 Study of the deposition of Al2O3 using synthetic air as oxidiser . .
81
4.4 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
82
Supplementary information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
87
5. Producing SiC tracer particles using gas-phase deposition of Al2O3 at ambient
91
conditions
5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
92
5.2 Experimental . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
93
5.3 Results and discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
95
5.4 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
99
Supplementary information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
104
6. Outlook
111
6.1 Open questions and complementary studies . . . . . . . . . . . . . . . . . . . . . . . .
112
6.1.1 Optimization of the results obtained in the previous chapters . . . .
112
6.1.2 Characterization of the strength of the alumina coating . . . . . . . . .
114
6.1.3 Necessity for accurate characterization techniques . . . . . . . . . . . . .
115
6.1.4 Industrial implementation of ALD on powders . . . . . . . . . . . . . . . .
116
6.1.5 Investigation of the interparticle forces . . . . . . . . . . . . . . . . . . . . . .
117
6.1.6 Study of different fluidization assistance methods . . . . . . . . . . . . .
117
6.2 Broadening the applicability of this work . . . . . . . . . . . . . . . . . . . . . . . . . .
118
6.2.1 Deposition of other compounds at ambient conditions . . . . . . . . .
118
6.2.2 Biocompatibility study of ultrathin films . . . . . . . . . . . . . . . . . . . .
118
6.3 Analysis of ALD for product development . . . . . . . . . . . . . . . . . . . . . . . . .
119
List of publications
125
Acknowledgements
127
Curriculum vitae
131
vii
Summary
Particles are widely used in the chemical industry as raw material and end product. In many
applications of particles, it is advantageous to give them a coating that can either protect the
particle from outside influences or give it an additional function. Technologies to provide
micron-sized particles with relatively thick coatings (in the order of µm) are readily available.
However, for many applications it would be attractive to apply coatings to fine particles (of
the order of 50 µm or smaller) and to make these coatings thinner (<< 1 µm). This opens
possibilities for new properties. In addition, a lower amount of the compound would be
needed to coat the material of interest. This thesis focuses on a technique for depositing
ultrathin and conformal films, with control down to nanometre scale, with high a utilisation of
the precursors: atomic layer deposition (ALD) applied to fine particles.
ALD is a gas-phase coating technique that relies on two self-terminating reactions, that form a
coating cycle, after which a fraction of a monolayer of compound is deposited on the
substrate. This cycle can be repeated a desired number of times to deposit thicker layers.
Amongst the wide variety of compounds that have been deposited with ALD, alumina (Al2O3)
has been the favourite one to study, due to the ideal layer-by-layer growth mode that is
observed when used trimethyl aluminium (TMA) and water as precursors, and its broad
applicability in different fields.
Thin Al2O3 ALD films have been used, for example, in the fabrication of semiconductor
devices, in the production of catalysts and the passivation of the cathode of the Li-ion
batteries. Besides these applications, other materials such as polymers or biological products
could benefit from the deposition of ultrathin alumina films. However, alumina has been
typically deposited at relatively high temperatures (~170 °C) and low pressures (< 1 mbar),
inhibiting its application to heat-sensitive materials. This is our main motivation to investigate
the deposition of alumina films at room temperature and atmospheric pressure. Working at
ambient conditions would, in first place, permit the coating of heat sensitive materials, and, in
second place, facilitate the coating process and improve the scaling-up prospects, since less
complex equipment would be required.
Atomic layer deposition provides accurate control over the film thickness based on the selfterminating nature of the ALD reactions. That means that deposition only takes place as long
as there are surface species available for reaction. The combination of the operating
temperature and pressure, and the purging of the reactor with nitrogen after each reaction,
ensure the removal of unreacted precursor molecules from the surface of the substrate.
However, the physisorption of these unreacted molecules is the main challenge that we face
when depositing alumina at 25 °C and 1 bar, since the normal boiling temperature of TMA
and water is 125 and 100 °C, respectively.
xi
Summary
In Al2O3 ALD processes at 170 °C and < 1 mbar, where no physisorption occurs, the
thickness of the deposited films depends solely on the number of coating cycles, since once
the active sites on the surface are depleted, no more reaction takes place, regardless of the
amount of precursors fed to the reactor in excess. In contrast, at ambient conditions, the
thickness of the alumina film will depend on the number of cycles and on the dosing time of
the precursors due to the accumulation of the unreacted precursor molecules that are fed in
excess.
In this work, we aimed at controlling the physisorption of the precursors at ambient conditions
by adjusting the dosing time of the precursors to the reactor. By feeding the amount of
precursor molecules needed to saturate the active sites present on the surface of the substrate
(TiO2 nanoparticles in our case), we obtained similar growths as in pure ALD, i.e., 0.1-0.2 nm
per cycle, even if the reactions are not self-limiting. Complementarily, having a faster growth
of the alumina film could be done by dosing both precursors in excess due to the
physisorption of unreacted molecules.
These findings were applied in the remainder of the thesis to deposit alumina, at ambient
conditions, on three different substrates: polymeric powder coating paint, lead-selenide
quatum dot (PbSe QD) films, and silicon carbide (SiC) particles. Two of these materials, the
powder coating and the QD films, are heat sensitive, thus working at room temperature was
essential.
Al2O3 films were used to tune the surface appearance of a powder coating (i.e., dry paint
without any solvent). The polymeric core particles were coated with alumina films of
different thicknesses. The alumina layers partially confined the core material when it softened
above the glass transition temperature. As the softened core did not flow during the curing of
the paint, it created roughness of the surface, and therefore, a matte surface appearance. The
mechanical properties of the matte powder coating were similar to the ones of the originally
glossy one. Thin alumina films were used to tune the final appearance of the powder coating
paint without the use of foreign particles that would produce the same effect.
Additionally, films of PbSe QD crystals, which are nanosized crystals with very interesting
properties for semiconductor applications, were passivated using alumina. QDs films have
very promising applications, e.g., as high-efficiency photovoltaic material. However, they are
air and heat sensitive and irreversibly oxidize after a short exposure to air, losing their good
properties. Previous work showed the passivation of the QDs with alumina at temperatures
between 25 and 90 °C, and pressures smaller than 1 mbar. In this work, alumina was
deposited at 25 °C and 1 bar to efficiently passivate the QDs. At ambient conditions, a fast
deposition of alumina took place as a consequence of different groups that can react with
TMA, such as hydroxyl and amine groups. Coating with alumina at ambient conditions can be
easily coupled with the fabrication of the QD films, which also is done at room temperature
and atmospheric pressure.
The final application we considered is the production of radioactive tracer particles. For
fluidized bed studies, we would like to label SiC particles with a minimum change to their
other physical properties. SiC particles have a rather inert surface towards radioactive ions, so
we used the Al2O3 coating to enhance the labelling efficiency of this material and produce a
tracer. Typically, a completely different material is used for the tracer particles, such as γAl2O3. However, this would produce a mismatch in properties such as size, shape and density
between the tracer and the material of study. In this work, alumina films were deposited on
Summary
SiC particle to mimic the affinity of the γ-Al2O3 particles towards the radioactive ions, and
produce a SiC-based tracer. Since the labelling efficiency depends on the alumina film
thickness, TMA and water were fed in large excess, achieving a growth per cycle of alumina
of about 10 nm. The resulting core-shell particles, with an alumina coating of about 400 nm,
had only a slightly lower relative activity as the γ-Al2O3 particle. Nevertheless, this SiC-based
tracer was used in a particle tracking experiment, having enough activity to be detected by the
sensors and reconstruct its trajectory during fluidization.
One might argue whether the studied deposition process at atmospheric pressure and room
temperature is still true ALD, since the reactions are not self-terminating and we deposit
several layers of alumina in each cycle. Nevertheless, it gives the possibility of depositing
alumina on heat sensitive materials, with good tunability and control over the films thickness,
at operating conditions that would simplify the scaled-up of this technology. The findings in
this thesis could be expanded to other applications, such as the coating of biological
compounds, that would require low-temperature processing, or the deposition of other
compounds at ambient conditions, i.e., zinc oxide (ZnO). This is further discussed in the
outlook of this thesis.
xiii
Samenvatting
Deeltjes worden op grote schaal toegepast in de chemische industrie als basismateriaal en als
eindproduct. Bij vele toepassingen van deeltjes is het aantrekkelijk om een coating aan te
brengen als een beschermende of functionele laag. Technieken om microdeeltjes te voorzien
van een relatief dikke laag (in de orde van micrometers) zijn reeds beschikbaar. Echter, voor
veel toepassingen is het aantrekkelijk een dunne laag (<< 1 µm) aan te brengen op fijne
deeltjes (50 µm of kleiner) om op deze manier nieuwe eigenschappen van de deeltjes
mogelijk maken. Bijkomend voordeel van de dunne laag is dat er minder materiaal nodig is
om de coating aan te brengen op het gewenste materiaal. Dit proefschrift richt zich op een
techniek voor de depositie van ultradunne en conforme lagen, met nanometer-precisie en met
zeer efficiënt gebruik van de precursors: atoomlaagdepositie (ALD) op kleine deeltjes.
ALD is een coatingtechniek in de gasfase, gebaseerd op twee zelf-terminerende reacties die
elke cyclus een fractie van een monolaag vormen op het substraat. De dikte van de laag wordt
bepaald door het aantal cycli. Van alle materialen die gebruikt worden voor ALD, is alumina
(Al2O3) een populair materiaal om te bestuderen, vanwege de ideale laag-per-laag-groei die
wordt waargenomen bij het gebruik van trimethylaluminium (TMA) en water als precursors,
en de vele mogelijke toepassingen in verschillende vakgebieden.
Dunne Al2O3-ALD-films worden onder andere gebruikt bij de fabricatie van halfgeleiders, in
de productie van katalysatoren en bij de passivering van de kathode van Lithium-ion-accu’s.
Daarnaast zou de depositie van ultradunne alumina films gebruikt kunnen worden voor andere
materialen zoals polymeren en biologische producten. Echter, alumina depositie gebeurt over
het algemeen bij hoge temperaturen (~170 °C) en lage drukken (< 1 mbar), waardoor deze
techniek niet toegepast kan worden op warmtegevoelige materialen. Dit is de belangrijkste
motivatie om de depositie van alumina bij kamertemperatuur en atmosferische druk te
onderzoeken. Op deze manier kan de ultradunne alumina laag aangebracht worden op
warmtegevoelige materialen; daarnaast wordt het proces vereenvoudigd en toepasbaar voor
grootschalig gebruik.
ALD biedt nauwkeurig controle over de filmdikte, gebaseerd op het zelf-terminerende
karakter van de ALD reacties. Dat betekent dat de depositie slechts plaats vindt waar
substraatoppervlak beschikbaar is voor de reactie. De combinatie van de temperatuur, de druk
en de zuivering van de reactor met stikstof na iedere reactie, zorgt voor de verwijdering van
het overschot aan precursors van het substraat oppervlak. Echter, de fysisorptie van deze nietgereageerde precursor moleculen is de voornaamste uitdaging die we ondervinden bij
alumina-depositie bij 25 °C en 1 bar, aangezien de kookpunten van TMA en water onder deze
condities respectievelijk 125 °C en 100 °C zijn.
xv
Samenvatting
Bij het ALD proces van Al2O3 bij 170 °C en < 1 mbar, waar geen fysisorptie plaatsvindt,
hangt de dikte van de aangebrachte laag enkel af van het aantal coating cycli, aangezien de
reactie stopt zodra alle reactieve plekken op het oppervlak bezet zijn, ondanks de hoeveelheid
aanwezige precursors. Bij standaardomstandigheden, kamertemperatuur en atmosferische
druk, daarentegen, hangt de dikte van de laag af van het aantal cycli en de doseringstijd van
de precursor, wegens accumulatie van de overmaat aan precursors in de reactor.
In dit onderzoek streven we naar controle over de fysisorptie bij standaardomstandigheden
door de doseringstijd van de precursor in de reactor aan te passen. Door de benodigde
hoeveelheid precursors toe te voegen om alle reactieve plekken op het substraat (in dit geval
TiO2 nanodeeltjes) te verzadigen, verkrijgen we vergelijkbare groei als in pure ALD (0.1-0.2
nm per cyclus), ondanks het ontbreken van de zelf-terminerende reactie. Daarnaast kan
snellere groei van de alumina film verkregen worden door beide precursors in overmaat te
doseren, dankzij de fysisorptie van de niet-gereageerde moleculen.
In het vervolg van dit proefschrift zijn bovenstaande bevindingen toegepast om, onder
standaardomstandigheden, een alumina-coating aan te brengen op drie verschillende
substraten: een polymere poedercoating-verf, lagen van lood-selenide quantum dots (PbSe
QD) en silicium carbide (SiC) deeltjes. Twee van deze materialen, de poedercoating verf en
de QD-film, zijn warmtegevoelig waardoor het essentieel is deze experimenten bij
kamertemperatuur uit te voeren.
Al2O3 films zijn gebruikt om de afwerking van een aangebrachte poedercoating-verflaag (verf
zonder oplosmiddel) te variëren. De polymere kerndeeltjes zijn gecoat met alumina lagen van
verschillende diktes. De alumina lagen bedekten het kernmateriaal gedeeltelijk wanneer deze
boven hun glas transitie temperatuur zachter werden. Doordat de zachtere kern niet vloeide
tijdens het harden van de verf, werd een ruw oppervlak gecreëerd met als resultaat een matte
afwerking. De mechanische eigenschappen van de matte poedercoating verf waren
vergelijkbaar met de originele glanzende verf. Dunne alumina lagen zijn gebruikt om de
uiteindelijke afwerking van de poedercoating verf te variëren zonder het gebruik van externe
materialen die hetzelfde effect creëren.
Hierna zijn films van PbSe QD-kristallen gepassiveerd met behulp van alumina. PbSe QD
kristallen zijn nanokristallen met zeer interessante eigenschappen voor toepassingen binnen
de halfgeleiders en fotovoltaïsche materialen. Deze laatste zijn echter lucht- en
warmtegevoelig en oxideren irreversibel na een korte blootstelling aan lucht, waarbij de goede
eigenschappen verloren gaan. Eerder werk heeft de passivering van quantum dots met
alumina laten zien bij temperaturen tussen de 25 en 90 °C en drukken lager dan 1 mbar. In
deze thesis is alumina aangebracht bij 25 °C en 1 bar om quantum dots efficiënt te passiveren.
Bij deze standaardomstandigheden vond een snelle depositie plaats van alumina als gevolg
van functionele groepen op de quantum dots, zoals hydroxyl en amine groepen, die kunnen
reageren met TMA. Het aanbrengen van een coating met alumina bij
standaardomstandigheden kan gemakkelijk gekoppeld worden aan het productieproces van
quantum dot films, diat ook uitgevoerd worden bij kamertemperatuur en atmosferische druk.
De laatste toepassing die we hebben bestudeerd, is de productie van radioactieve marker
deeltjes. Voor gefluïdiseerd-bed-studies wilden we SiC deeltjes labelen met een minieme
verandering aan hun fysische eigenschappen. SiC deeltjes hebben een vrijwel inert oppervlak
voor radioactieve ionen. We gebruikten de Al2O3 coating om de labeling-efficiëntie van dit
materiaal te vergroten en markers te produceren. Echter, dit zou leiden tot een discrepantie in
Samenvatting
de eigenschappen wat betreft de grootte, vorm en dichtheid tussen de marker en het
bestudeerde materiaal. In deze studie zijn alumina films aangebracht op SiC deeltjes om de
affiniteit van γ-Al2O3 deeltjes voor radioactieve ionen te imiteren en markers te produceren
gebaseerd op SiC. Aangezien de labeling-efficiëntie afhangt van de dikte van de alumina film,
zijn water en TMA in overmaat toegevoegd zodat een groei van ongeveer 10 nm per cyclus
plaatsvindt. De gevormde deeltjes, met een alumina schil van ongeveer 400 nm, blijken
slechts een iets lagere relatieve activiteit te hebben dan de γ-Al2O3 deeltjes. Deze SiC marker
is gebruikt in een deeltjes-tracking- experiment, aangezien deze marker genoeg activiteit bezit
om gedetecteerd te worden door de sensoren en een traject te reconstrueren tijdens het
fluïdiseren.
Men kan betwisten of het bestudeerde depositie-proces bij atmosferische druk en
kamertemperatuur nog steeds ALD is, aangezien de reacties niet zelf-terminerend zijn en we
iedere cyclus verschillende lagen alumina aanbrengen. Desondanks biedt deze methode de
mogelijkheid alumina aan te brengen op warmtegevoelige materialen, met een nauwkeurige
controle over de filmdikte en technieken die het proces op grote schaal zullen versimpelen. De
bevindingen van deze thesis kunnen uitgebreid worden tot andere toepassingen als coatings,
bijvoorbeeld zinkoxide (ZnO), voor biomaterialen bij standaardomstandigheden. Dit wordt
verder besproken in het laatste hoofdstuk van dit proefschrift.
xvii
Chapter 1
Introduction
1
Chapter 1
1.1 Research Field
1.1.1 Particle Technology
Powders play an important role in the process industry, as it is estimated that 75% of the raw
materials and 50% of the end products are found as particulate matter [1]. Thus, the
production and handling, the characterization of the physical and chemical properties, and the
development of applications of particles represent an important research field. The particle
size is one of the properties that define the behaviour of powders, and therefore, their
application. Particles in the millimetre range are typically found in food and pharmaceutical
products, detergents and building materials. Finer particles - in the micron and submicron
range - are used for catalysis, aerosols or electronics [2].
In the past two decades, nanoparticles (NPs) received increasing attention because of the
special properties associated to their size (dp < 100 nm), such as the large values of the
specific surface area, and the quantum confinement effects [3]. Therefore, NPs are widely
researched for the production of drugs, solar cells and catalysts amongst other fields [4-6]. In
several of those applications, the performance of nanoparticles can be enhanced by depositing
a compound that brings additional functionalities, such as protection or activation. For
example, a thin film of about 2 nm of aluminium oxide (Al2O3) was deposited on the cathode
material of Li-ion batteries (LiMnO2 nanoparticles) to prevent its dissolution in the
electrolyte, and thus, reduce the aging of the battery [7-9]. Another example is the activation
of a photocatalyst, i.e., TiO2 P25 nanoparticles, by depositing a small amount of platinum
clusters on its surface. The resulting material was five times more active than the bare
uncoated TiO2 nanoparticles [10, 11]. These two examples show that even small
modifications on the surface can have a significant influence on the performance of a
material.
In this thesis, we explore the field of particle coating and how we can, by depositing small
amounts of material, tailor the properties and add new functionalities to different type of
particles. We now introduce and briefly review relevant theory about this topic.
1.1.2 Coating techniques for particles
A wide range of techniques are used to coat particles, depending on the size of the solids and
the desired thickness and nature of the film (Fig. 1 left). Liquid-based techniques, such as
spray drying or fluid bed coating, are used in food and pharmaceutical industry to deposit
films from micrometre to millimetre scale. For example, the fluid bed coating technique was
used to spray a solution composed of a mixture of polymers to encapsulate micro-crystalline
cellulose (dp ~ 1200 µm) [12]. This process was performed at 70 °C, avoiding the degradation
of the core and shell materials. However, it requires a drying stage to evaporate the solvent.
This involves longer process times, the generation of larger amounts of waste, and a higher
energy consumption. Thinner coatings can be deposited with chemical vapour deposition
(CVD). This is a gas-phase technique based on the reaction of two precursors, that are fed to
the reactor simultaneously, on the substrate. The CVD process combines a relatively rapid
growth of layers with a precise control of the film thickness, down to hundreds of nanometres
[13]. As an example, 75 µm tungsten particles where coated, in a fluidized bed, with a film of
SiO2, with a thickness of few hundreds of nanometres. The coating process was carried out at
2
Introduction
650 °C, with a deposition rate of 1 – 16 nm/min. Around 1970’s, atomic layer deposition
(ALD), formerly known as atomic layer epitaxy or molecular layering [14, 15], was
developed based on the CVD process. ALD comprises the reaction of two gas precursors that
are fed in a consecutive manner to the reactor, as opposed to CVD. In this way, the two
reactions, one per precursor, only take place on the surface of the substrates. These two
reactions form one ALD cycle to deposit less than a full monolayer of compound, due to the
concentration of active sites on the surface, and the steric hindrance between precursor
molecules [16, 17]. This cycle is repeated a desired number of times to deposit thicker films
(Fig. 1 right). In this manner, ALD provides atomic control over the thickness of the films,
that can be as thin as few nanometres [16].
1.1.3 Atomic Layer Deposition
The high precision and controllability of the film growth of ALD is based on the selfterminating nature of the chemical reactions. That means that the precursors molecules only
react as long as there are unreacted active sites or functional groups available on the surface.
Once all the surface species have reacted, the so-called saturation regime, no more reaction
will take place independently of how much precursor is dosed in excess. To ensure the selftermination of the reactions, the operating temperature and pressure, and the purging step with
nitrogen, contribute to remove the unreacted molecules and prevent their absorption to the
substrate surface.
Fig. 1. (left) Map of several coating techniques for particles, based on the particle size and the
film thickness. (right) Reaction mechanism of ALD, based on two cyclic reactions.
The growth per cycle (GPC), which refers to the increase in the film thickness in each cycle,
is used to characterize the coating process. Besides the concentration of active sites and the
size of the precursor molecules, other parameters influence the GPC of a film, such as the
operating temperature and pressure, the flow of purging gas, and dosage of the precursors.
The removal of unreacted molecules is severely diminished at temperatures and pressures at
which the precursor cannot evaporate, such as room temperature and atmospheric pressure. At
these conditions, the unreacted molecules would accumulate on the surface of the substrate,
taking part in the upcoming coating reactions. After each reaction, nitrogen is flowed through
the reactor for purging, which is favoured at high temperatures and low pressures. At 25 °C
and 1 bar, it is possible that nitrogen is not sufficient to remove the unreacted molecules,
which would unavoidably accumulate on the surface [18, 19]. Consequently, the flow of
precursors would define the GPC, as the excess molecules after the surface has been saturated
3
Chapter 1
with precursor would accumulate. That results in the deposition of several layers of
compound in each cycle [20]. This process coating does not fit as ALD, since the reactions
are not self-limiting, nor as CVD, since the precursors are fed in separate steps. The term
parasitic CVD-growth is used to define this deposition [21, 22]. We refer to it simply as gasphase coating technique. In the bright side, working at ambient conditions allow the coating
of heat sensitive materials, and the hindered removal of unreacted precursor could be used to
deposit thicker films faster.
The versatility of ALD was shown by Miikkulainen et al. [23] as a periodic table with all the
compounds that had been deposited until 2013 (Fig. 2), including pure elements, oxides,
nitrides, etcetera. To carry out an ALD process, we just need two precursors that can be easily
taken into the gas phase, and which combined reaction results in the desired coating material.
Usually, the first precursor is the source of metal, often an organometallic compound, while
the second one would determine whether the deposited compound is a pure element, an oxide
or a nitride, by using as co-reactant water, oxygen or ammonia, respectively [23]. The
deposition of aluminium oxide using trimethylaluminium (TMA) and water is the most
studied ALD system, since (i) the deposition of alumina shows the typical layer-by-layer
growth mechanism of ALD, (ii) it can be carried out at a wide range of temperatures and
pressures given the high reactivity of TMA molecules, and (iii) it has been applied to a wide
range of applications [7, 24-29].
Fig. 2. Periodic table of ALD, adapted from Miikkulainen et al. [23]
ALD was initially developed and applied on flat substrates for semiconductor applications
[30]. Later, it expanded to substrates with irregular geometries [18, 31] and particles [32, 33].
Amongst the several reactor configurations available for ALD of particles [34], a fluidized
bed reactor (FBR) was used in the work described in this thesis [35].
4
Introduction
1.1.4 Fluidization of Particles
Fluidizing particles are characterized by a good gas-solid mixing that translates in efficient
mass and heat transfer [36]. Additionally, fluidized bed reactors can process large amounts of
particles, and are, in general, well scalable. Fluidized beds are widely used in industry for
gasification and combustion of biomass, and chemical looping combustion [36-38]. However,
the fluidization behaviour of powders depends on the size and density of the particles, as
described in the Geldart diagram (Fig. 3 left) [39]. The suspended state of the fluidizing
particles is achieved when the drag force exerted by the fluidizing gas equals the forces that
oppose the fluidization of the particles, i.e., gravity and interparticle forces (Fig. 3 right).
There are usually four contributions to the interparticle forces: van der Waals, capillary,
electrostatic forces and hydrogen bonding [40, 41], from which the capillary force can be
neglected if no liquid is present.
Fig. 3. (left) Geldart diagram of the fluidization properties of particles based on the particle
size and the difference between the particle and fluidizing gas densities. (right) Values of the
interparticle forces with the particle size.
In the fluidization of fine cohesive powder, with a particle diameter < 50 µm, the interparticle
forces predominate over the gravitational and drag forces, causing the agglomeration of the
particles (Fig. 3 right). The formation of the agglomerates results in a poor fluidization
behaviour, characterized by channels, dead zones and plugs that reduce the mixing, and
disturb the circulation patterns of the particles [42]. Nevertheless, uniform fluidization of fine
particles can be attained by using fluidization assistance methods. This consists of an external
supply of energy that contributes to the breakage of the agglomerates, and improves the
quality of the fluidization. Some of these assistance methods are mechanical vibration, both in
vertical and horizontal directions, stirring, microjets and ultrasounds amongst others [43-48].
The agglomeration of cohesive particles could be seen as detrimental for the coating process,
since the inner particles of the agglomerates would not be exposed to the precursor molecules.
However, nanoparticle agglomerates are open structures with a size of about 300 µm and a
void fraction of about 99% [49]. Additionally, the fluidization of nanoparticle agglomerates is
dynamic, meaning that the agglomerates break and recombine continuously as result of the
agglomerate-agglomerate and agglomerate-wall collisions [35]. Thus, the high porosity and
dynamic behaviour of these agglomerates ensure that the surface of the primary particles is
exposed to the coating precursors [50]. The good processing properties of fluidized bed
reactors, and the versatility and precise control over the film thickness of ALD reactions,
convert this technology in an interesting field, not only for research purposes, but also for
industrial production.
5
Chapter 1
1.2 Objectives of this Study
This PhD project was defined as a product-driven and process-development project, in
collaboration with DSM Coating Resins (Zwolle, The Netherlands). The objective of this
project is to develop the deposition of alumina at atmospheric pressure and room temperature
on particles, in a ALD-like process, and demonstrate its application to improve the
performance of several materials.
Alumina ALD films have been applied as passivating material, and in the production of
membranes and catalysts amongst others [26, 27, 51], in both flat substrates and particles.
Al2O3 has been deposited, using ALD, at different operating conditions, such as 170 °C and <
1mbar, at 170 °C and 1 bar, and at 33 °C and < 1 mbar [7, 18, 52]. However, it was never
investigated the deposition of alumina at atmospheric pressure and room temperature. In the
case of particles, the vast majority of the work was reported at 170 °C and < 1mbar. At these
conditions, (i) it will not be possible to coat heat sensitive materials, such as polymers and
biological compounds, and (ii), operating at low pressures would require certain equipment
and an exquisite process control at both lab and industrial scales. This was the reason to
investigate the deposition of alumina at ambient conditions.
The main challenge that we face at ambient conditions is the physisorption of unreacted
precursors molecules on the surface of the particles. As a consequence of the accumulation,
several layers of alumina would be deposited in each cycle, in a parasitic CVD-type of
process. Given that molecules accumulate at ambient conditions, we think that the dosage of
precursors will be crucial. Then, we could play with the dosing time of reactants to only
saturate the surface by feeding the same amount of precursor as the number of active sites, or
to overexpose the substrate to the precursors, and deposit several layers per cycle. This can be
used to deposit thicker films faster. Below, we list several sub-objectives in the form of
questions that were addressed during the completion of this research.
(i)
Are the coating reactions self-terminating at room temperature and atmospheric
pressure, as in ALD processes?
(ii)
Can we deposit ultrathin alumina films at ambient conditions? Are these alumina films
uniform and conformal? How good is the control over the film thickness at ambient
conditions?
(iii) What are the properties of the alumina films? Are there pores or unreacted molecules
entrapped in the alumina films?
(iv) Could we benefit from accumulation of molecules to purposely deposit thicker alumina
films faster?
(v)
Can we deposit Al2O3 films on heat sensitive materials without degrading them?
As discussed earlier, the agglomeration of fine particles represents a challenge for the coating
process [42]. Thus, we could investigate how to interfere on the agglomerate formation in two
different manners. First, by influencing directly the interparticle attractive forces and prevent
the agglomeration [41, 53, 54]. And second, by supplying external energy in the form of
fluidization assistance methods to induce the breakage of the agglomerates [43]. In this
6
Introduction
respect, there are still missing studies on the contribution of different interparticle forces to
the agglomeration, such as the electrostatic force and hydrogen bonding, and a comparison of
different assistance methods regarding energy consumption, influence on the fluidization
quality, etcetera. These studies would be very useful, but that obtained less emphasis in this
thesis. Nevertheless, we discuss some of these issues in the outlook of this thesis (chapter VI).
1.3 Outline
This thesis comprises four papers published in or submitted to peer-reviewed journals, and a
final chapter with the outlook of this work. A brief summary of the chapters is given below.
In chapter II, we studied the deposition of alumina films at atmospheric pressure and room
temperature on fluidizing TiO2 nanoparticles [20]. For that, we first fed enough precursor to
saturate the particles, and then we fed large excess of precursors, to compare the film
thickness and the GPC. We show that it is possible to deposit ultrathin films alumina at
ambient conditions, while maintaining accurate control over the film thickness.
In chapters III, IV and V we deposited alumina films on materials for three different
applications: to modify the surface appearance of a paint, to stabilize quantum dot (QD) films
for production of solar cells, and to enhance the labelling efficiency of a SiC-based (silicon
carbide) tracer. We now explain the main issues that are addressed in each of these chapters.
In chapter III, we deposited thin layers of aluminium oxide to alter notably the finish
properties of an industrial powder coating paint, and characterized the alumina-coated
particles and the powder coated panels, which is the final product. In agreement with
literature, we observed that alumina only formed complete films after several cycles [55, 56].
The partial confinement of the polymeric core particles by the alumina films above the glass
transition temperature reflected on the appearance of the final paint. We found that an
equivalent alumina film of 6 nm was able to tune the surface finish of the painted panel from
gloss to matte. This study aimed at depositing alumina films on a heat-sensitive industrial
material that would give a tailored property, such as converting a gloss paint into a matte one,
by using exactly the same material with a slight modification on the surface of the primary
solids.
In chapter IV, we deposited aluminium oxide on PbSe QD films, which are heat and air
sensitive. QD crystals oxidize and degrade irreversibly after short exposure to air. ALD of
alumina, performed at a range of temperatures of 27-100 °C and < 1 mbar, was used to
effectively passivated the QD films and prevent their degradation [57, 58]. The main research
question in this chapter was whether we could benefit from the physisorption of precursors at
ambient conditions to obtain full passivation of the QD films after a low number of cycles.
The results showed a fast deposition of alumina from the first cycle as the combination of
several mechanism, i.e., presence of solvent molecules, reactivity of TMA with both amine
and hydroxyl groups, and the excess of precursors fed to the reactor. The alumina infilled and
overcoated the QD films, giving full passivation after a relatively low number of cycles. This
process could be combined with the fabrication process of the QD films by dip-coating, which
also is done at ambient conditions.
In chapter V, we deposited alumina films on silicon carbide (SiC) particles to increase its
affinity towards 18F radioactive ions, aiming at enhancing the labelling efficiency. SiC
7
Chapter 1
particles are fluidized to harvest solar energy in concentrated solar plants [59, 60].
Hydrodynamics studies of these systems are needed to assess the efficiency of this
technology, evaluated with the circulation frequency of the particles near the wall of the
column, where they get heated. Positron emission particle tracking (PEPT) is a tool to track a
radioactive tracer particle in a multiphase system, such as a fluidized bed, and reconstruct its
trajectory. Thus, the tracer must resemble the material of study. As the surface of the SiC
particles is inert, it has low affinity with the 18F ions and cannot be used as tracer. Therefore,
we deposited alumina films on SiC particles to improve the affinity and the labelling
efficiency. In this chapter we propose a general method to produce tracers for particle tracking
applications from originally inert particles by depositing films of aluminium oxide, which
shows a good affinity towards the radioactive ions of 18F.
This thesis was aimed as proof of principle on how thin films of alumina, deposited at
ambient conditions, improved the performance or properties of different materials. Thus, this
work could be further optimized, regarding the dosing of the precursors, number of cycles and
purging time, for instance. Additionally, there are studies that could be done to investigate the
fluidization of cohesive particles, or to expand the deposition at ambient conditions to other
compounds and substrates. These issues and other remaining open questions from this work
are discussed in the outlook of this thesis (chapter VI).
Bibliography
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
Kalman, H. and G.I. Tardos, Elements of Particle Technology in the Chemical
Industry. Particulate Science and Technology, 2005. 23(1): p. 1-19.
Stark, W.J., et al., Industrial applications of nanoparticles. Chemical Society
Reviews, 2015.
Burda, C., et al., Chemistry and properties of nanocrystals of different shapes.
Chemical Reviews, 2005. 105(4): p. 1025-1102.
Moshfegh, A.Z., Nanoparticle catalysts. Journal of Physics D: Applied Physics, 2009.
42(23).
Luther, J.M., et al., Schottky Solar Cells Based on Colloidal Nanocrystal Films. Nano
Letters, 2008. 8(10): p. 3488-3492.
Parveen, S., R. Misra, and S.K. Sahoo, Nanoparticles: a boon to drug delivery,
therapeutics, diagnostics and imaging. Nanomedicine: Nanotechnology, Biology and
Medicine, 2012. 8(2): p. 147-166.
Beetstra, R., et al., Atmospheric pressure process for coating particles using atomic
layer deposition. Chemical Vapor Deposition, 2009. 15(7-9): p. 227-233.
Beetstra, R., J. Nijenhuis, E.M. Kelder, J.R. van Ommen, Improved Li-Ion Battery
Performance by Coating Cathode Nano-Particles using Atomic Layer Deposition.
Conference proceedings of the 12th International Conference on Fluidization - New
Horizons in Fluidization Engineering, 2007.
Jung, Y.S., et al., Enhanced stability of LiCoO2 cathodes in lithium-ion batteries using
surface modification by atomic layer deposition. Journal of the Electrochemical
Society, 2010. 157(1): p. A75-A81.
Goulas, A. and J. Ruud van Ommen, Atomic layer deposition of platinum clusters on
titania nanoparticles at atmospheric pressure. Journal of Materials Chemistry A,
2013. 1(15): p. 4647-4650.
8
Introduction
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
Van Ommen, J.R., et al., Continuous production of nanostructured particles using
spatial atomic layer deposition. Journal of Vacuum Science and Technology A:
Vacuum, Surfaces and Films, 2015. 33(2).
Perfetti, G., et al., Attrition and abrasion resistance of particles coated with pre-mixed
polymer coating systems. Powder Technology, 2012. 230(0): p. 1-13.
Choy, K.L., Chemical vapour deposition of coatings. Progress in Materials Science,
2003. 48(2): p. 57-170.
Suntola, T., Atomic layer epitaxy. Materials Science Reports, 1989. 4(5): p. 261-312.
Puurunen, R.L., A Short History of Atomic Layer Deposition: Tuomo Suntola's Atomic
Layer Epitaxy. Chemical Vapor Deposition, 2014. 20(10-11-12): p. 332-344.
George, S.M., Atomic layer deposition: An overview. Chemical Reviews, 2010.
110(1): p. 111-131.
Puurunen, R.L., Surface chemistry of atomic layer deposition: A case study for the
trimethylaluminum/water process. Journal of Applied Physics, 2005. 97(12).
Groner, M.D., et al., Low-Temperature Al2O3 Atomic Layer Deposition. Chemistry of
Materials, 2004. 16(4): p. 639-645.
Jur, J.S. and G.N. Parsons, Atomic Layer Deposition of Al2O3 and ZnO at
Atmospheric Pressure in a Flow Tube Reactor. ACS Applied Materials & Interfaces,
2011. 3(2): p. 299-308.
Valdesueiro, D., et al., Gas-Phase Deposition of Ultrathin Aluminium Oxide Films on
Nanoparticles at Ambient Conditions. Materials, 2015. 8(3): p. 1249-1263.
Profijt, H.B., et al., Plasma-assisted atomic layer deposition: Basics, opportunities,
and challenges. Journal of Vacuum Science and Technology A: Vacuum, Surfaces and
Films, 2011. 29(5).
Potts, S.E. and W.M.M. Kessels, Energy-enhanced atomic layer deposition for more
process and precursor versatility. Coordination Chemistry Reviews, 2013. 257(23–
24): p. 3254-3270.
Miikkulainen, V., et al., Crystallinity of inorganic films grown by atomic layer
deposition: Overview and general trends. Journal of Applied Physics, 2013. 113(2).
Lichty, P., et al., Surface modification of graphite particles coated by atomic layer
deposition and advances in ceramic composites. International Journal of Applied
Ceramic Technology, 2013. 10(2): p. 257-265.
Liang, X., et al., Stabilization of supported metal nanoparticles using an ultrathin
porous shell. ACS Catalysis, 2011. 1(10): p. 1162-1165.
Moghtaderi, B., I. Shames, and E. Doroodchi, Combustion prevention of iron powders
by a novel coating method. Chemical Engineering and Technology, 2006. 29(1): p. 97103.
Hakim, L.F., et al., Synthesis of oxidation-resistant metal nanoparticles via atomic
layer deposition. Nanotechnology, 2007. 18(34).
King, D.M., et al., Passivation of pigment-grade TiO2 particles by nanothick atomic
layer deposited SiO2 films. Nanotechnology, 2008. 19(25).
Liang, X., et al., Biocompatible interface films deposited within porous polymers by
atomic layer deposition (ALD). ACS Applied Materials and Interfaces, 2009. 1(9): p.
1988-1995.
Elam, J.W., Z.A. Sechrist, and S.M. George, ZnO/Al2O3 nanolaminates fabricated by
atomic layer deposition: growth and surface roughness measurements. Thin Solid
Films, 2002. 414(1): p. 43-55.
Ritala, M., et al., Perfectly conformal TiN and Al2O3 films deposited by atomic layer
deposition. Chemical Vapor Deposition, 1999. 5(1): p. 7-9.
9
Chapter 1
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
Wank, J.R., S.M. George, and A.W. Weimer, Nanocoating individual cohesive boron
nitride particles in a fluidized bed by ALD. Powder Technology, 2004. 142(1): p. 5969.
Yakovlev, S.V., A. A. Malygin, S. I. Kol'tsov, V. B. Aleskovskii, Y. G. Chesnokov, I.
O. Protod'yakanov, Mathematical model of molecules layering with the aid of a
fluidized bed. Journal of Applied Chemistry - U.S.S.R., 52, 959, 1979.
Longrie, D., D. Deduytsche, and C. Detavernier, Reactor concepts for atomic layer
deposition on agitated particles: A review. Journal of Vacuum Science &amp;
Technology A, 2014. 32(1): p. -.
Hakim, L.F., et al. Fluidization behavior and conformal coating of nanoparticles in
fluidized beds by ALD. 2004. Austin, TX.
Kunii, D. and O. Levenspiel, Fluidization Engineering. Second ed. 1991: ButterworthHeinemann.
Lyngfelt, A., B. Leckner, and T. Mattisson, A fluidized-bed combustion process with
inherent CO2 separation; application of chemical-looping combustion. Chemical
Engineering Science, 2001. 56(10): p. 3101-3113.
Zhong, W., et al., Fluidization of Biomass Particles in a Gas−Solid Fluidized Bed.
Energy & Fuels, 2008. 22(6): p. 4170-4176.
Geldart, D., Types of gas fluidization. Powder Technology, 1973. 7(5): p. 285-292.
Seville, J.P.K., C.D. Willett, and P.C. Knight, Interparticle forces in fluidisation: A
review. Powder Technology, 2000. 113(3): p. 261-268.
Tahmasebpoor, M., et al., The role of the hydrogen bond in dense nanoparticle-gas
suspensions. Physical Chemistry Chemical Physics, 2013. 15(16): p. 5788-5793.
Sundaresan, S., Instabilities in fluidized beds. 2003. p. 63-88.
van Ommen, J.R., J.M. Valverde, and R. Pfeffer, Fluidization of nanopowders: a
review. Journal of Nanoparticle Research, 2012. 14(3): p. 1-29.
Mawatari, Y., et al., Comparison of three vibrational modes (twist, vertical and
horizontal) for fluidization of fine particles. Advanced Powder Technology, 2001.
12(2): p. 157-168.
Quevedo, J.A., A. Omosebi, and R. Pfeffer, Fluidization enhancement of
agglomerates of metal oxide nanopowders by microjets. AIChE Journal, 2010. 56(6):
p. 1456-1468.
Kuipers, N.J.M., E.J. Stamhuis, and A.A.C.M. Beenackers, Fluidization of potato
starch in a stirred vibrating fluidized bed. Chemical Engineering Science, 1996.
51(11): p. 2727-2732.
Yu, Q., et al., Enhanced fluidization of nanoparticles in an oscillating magnetic field.
AIChE Journal, 2005. 51(7): p. 1971-1979.
Chirone, R., L. Massimilla, and S. Russo, Bubble-free fluidization of a cohesive
powder in an acoustic field. Chemical Engineering Science, 1993. 48(1): p. 41-52.
de Martín, L., W.G. Bouwman, and J.R. van Ommen, Multidimensional Nature of
Fluidized Nanoparticle Agglomerates. Langmuir, 2014. 30(42): p. 12696-12702.
Grillo, F., M.T. Kreutzer, and J.R. van Ommen, Modeling the precursor utilization in
atomic layer deposition on nanostructured materials in fluidized bed reactors.
Chemical Engineering Journal, 2015. 268(0): p. 384-398.
Liang, X., et al., Barrier properties of polymer/alumina nanocomposite membranes
fabricated by atomic layer deposition. Journal of Membrane Science, 2008. 322(1): p.
105-112.
Hakim, L.F., et al., Surface modification of titania nanoparticles using ultrathin
ceramic films. Journal of the American Ceramic Society, 2006. 89(10): p. 3070-3075.
10
Introduction
53.
54.
55.
56.
57.
58.
59.
60.
Ciborowski, J. and A. Wlodarski, On electrostatic effects in fluidized beds. Chemical
Engineering Science, 1962. 17(1): p. 23-32.
Guardiola, J., V. Rojo, and G. Ramos, Influence of particle size, fluidization velocity
and relative humidity on fluidized bed electrostatics. Journal of Electrostatics, 1996.
37(1–2): p. 1-20.
Wilson, C.A., R.K. Grubbs, and S.M. George, Nucleation and growth during Al2O3
atomic layer deposition on polymers. Chemistry of Materials, 2005. 17(23): p. 56255634.
Liang, X., et al., Novel processing to produce polymer/ceramic nanocomposites by
atomic layer deposition. Journal of the American Ceramic Society, 2007. 90(1): p. 5763.
Liu, Y., et al., Robust, functional nanocrystal solids by infilling with atomic layer
deposition. Nano Letters, 2011. 11(12): p. 5349-5355.
Liu, Y., et al., PbSe quantum dot field-effect transistors with air-stable electron
mobilities above 7 cm2 V-1 s-1. Nano Letters, 2013. 13(4): p. 1578-1587.
Flamant, G., et al., Dense suspension of solid particles as a new heat transfer fluid for
concentrated solar thermal plants: On-sun proof of concept. Chemical Engineering
Science, 2013. 102: p. 567-576.
Flamant, G., et al., A New Heat Transfer Fluid for Concentrating Solar Systems:
Particle Flow in Tubes. Energy Procedia, 2014. 49(0): p. 617-626.
11
Chapter 2
Gas-phase deposition of ultrathin
Al2O3 films on nanoparticles
at ambient conditions
We have deposited aluminium oxide films by atomic layer deposition on titanium oxide
nanoparticles in a fluidized bed reactor at 27 ± 3 °C and atmospheric pressure. Working at
room temperature allows the coating of heat-sensitive materials, while working at atmospheric
pressure would simplify the scale-up of this process. We performed 4, 7 and 15 cycles by
dosing a predefined amount of precursors, i.e., trimethylaluminium and water. We obtained a
growth per cycle of 0.14–0.15 nm determined by transmission electron microscopy (TEM),
similar to atomic layer deposition (ALD) experiments at a few millibars and ~180 °C. We
also increased the amount of precursors dosed by a factor of 2, 4 and 6 compared to the base
case, maintaining the same purging time. The growth per cycle (GPC) increased, although not
linearly, with the dosing time. In addition, we performed an experiment at 170 °C and 1 bar
using the dosing times increased by factor 6, and obtained a growth per cycle of 0.16 nm.
These results were verified with elemental analysis, which showed a good agreement with the
results from TEM pictures. Thermal gravimetric analysis (TGA) showed a negligible amount
of unreacted molecules inside the alumina films. Overall, the dosage of the precursors is
crucial to control precisely the growth of the alumina films at atmospheric pressure and room
temperature. Dosing excess precursor induces a chemical vapour deposition type of growth due
to the physisorption of molecules on the particles, but this can be avoided by working at high
temperatures.
Published as: D. Valdesueiro, G.M.H. Meesters, M.T. Kreutzer and J.R. van Ommen. GasPhase Deposition of Ultrathin Aluminium Oxide Films on Nanoparticles at Ambient
Conditions. Materials, 2015. 8(3): p. 1249-1263, doi:10.3390/ma8031249.
13
Chapter 2
2.1 Introduction
The production of core-shell nanoparticles using atomic layer deposition (ALD) in a fluidized
bed reactor (FBR) is an attractive technology because of the good mixing between gas and
solids, and the possibility to process large amounts of solids up to industrial scale [1–4]. For
instance, aluminium oxide (Al2O3) coating obtained by ALD on particles in a FBR was applied
as a passivating barrier against oxidation, as a gas barrier on sensors, and in the production of
catalysts, amongst other applications [5–10]. Nanoparticles are widely used in diverse fields
such as catalysis, medicine, energy conversion and storage [11–14]. In addition to the large
specific surface area, surface modification can introduce additional functionalities to improve
these nanomaterials. Atomic layer deposition, which was developed for the semiconductor
industry [15,16], appears as an enabling coating technique to either protect or activate the surface
of substrates with a precise control over the amount of material deposited [11,17–21]. The ALD
layer-by-layer growth mechanism relies on two alternating gas phase reactions with a purging
step in between, using an inert gas to remove the unreacted molecules of precursor. These
reactions can be repeated a certain number of times to deposit conformal and pin-hole free
thin films with a precision down to atomic scale. In addition, the absence of solvent in ALD
processes reduces the waste generated when considering industrial scale production [3]. These
features establish the ALD in a FBR as a suitable technique to produce core-shell
nanoparticles with tailored functionalities.
The deposition of Al2O3 is typically done at low pressure (~1 mbar or lower) and elevated
temperatures (~180 °C) to enable the evaporation of the excess molecules of precursors, and in
this manner ensure atomic growth of the films by an efficient removal of the excess
precursors [22]. The physisorption of unreacted molecules at ambient conditions would result
in undesired parasitic CVD-type (chemical vapour deposition) of reactions [23,24]. ALD of
alumina has been studied at different reactor temperatures ranging from 33 °C to ~180 °C,
and pressures, from 10−5 bar to 1 bar, over a diverse variety of powders. To our knowledge,
alumina ALD at both low temperature and atmospheric pressure has not yet been investigated,
neither on flat substrates nor on particles. The growth per cycle (GPC) is used to characterize
the coating experiments [25,26]. ALD at pressures of a few millibars and ~180 °C provides
alumina layers with a growth between 0.1 and 0.2 nm per cycle [27–32]. At atmospheric
pressure and 160 °C Al2O3, ALD revealed the possibility of depositing few-nanometre films
with slightly larger growth per cycle [33]. Alumina ALD performed at 10−5 bar and 33 °C
gave a GPC of 0.3 nm [34,35]. With regards to flat substrates, more studies were reported
either at room temperature and low pressure [36–38], or at atmospheric pressure and a
temperature above 100 °C [39,40] showing an influence of these variables on the GPC, which
is higher than at lower pressure and higher temperature. These conditions can be improved in
two different ways. First, working at room temperature opens the possibility to apply ALD on
heat-sensitive materials [41]. Secondly, using atmospheric pressure would facilitate the scaleup towards the industrial production by easing the handling of the powder and reducing the
complexity of the equipment required. For these reasons, we study the deposition of alumina
on nanoparticles at room temperature and atmospheric pressure.
This experimental paper describes the aluminium oxide ALD in a FBR at 1 bar and 27 ± 3 °C,
using TiO2 P25 nanoparticles as the support. The objective of the paper is to understand
whether conformal alumina films can be deposited at ambient conditions, supressing the
physisorption of the excess precursor molecules by a careful dosage of the precursors. For that,
we performed three sets of experiments. In the first one, we examined whether we can achieve
a growth per cycle that is similar to that reported in the literature at lower pressure and higher
14
Gas-phase deposition of ultrathin Al2O3 films on nanoparticles at ambient conditions
temperature. In the second set, we determined if the precursor would physisorb on the surface
of the particles if the dosing of precursor is extended, at constant concentration, thus
increasing the excess of precursor relative to the amount of reactive sites inside the reactor. In
the third set, we investigated the GPC of the alumina films at 1 bar and 170 °C when both
precursors are fed in excess. In this work, we consider the dosed amount of the precursors as a
crucial factor to avoid the CVD-type growth expected at ambient conditions, particularly in
the case of water [39,42,43]. These molecules would react in the subsequent reaction,
resulting in thicker and non-uniform alumina films. We try to avoid the accumulation of
unreacted molecules with an accurate delivery of the precursors. That would allow the
deposition of ultrathin films at room conditions, while maintaining the control over the
properties of the films.
2.2 Experimental Section
2.2.1 Experimental Setup
Experiments are carried out in a fluidized bed reactor similar to the one described by Beetstra
et al. [33], which is composed by a vertical glass column of 26 mm internal diameter and 500
mm height, placed on a single motor Paja PTL 40/40-24 vertical vibration table to assist the
fluidization [2]. The vibration table is operated at 35 Hz, and provides a vibration amplitude
of 2 mm to the column. An infrared lamp placed parallel to the column, and a type-K
thermocouple inserted in the column, are used to control the bed temperature. The gas is
introduced to the column through a stainless steel SIKA-R 20 AX distributor plate of sintered
particles with a pore size of 37 μm, to achieve a homogenous flow of gas through the full
cross-section of the column. An identical distributor plate is placed on top of the column to
prevent nanoparticles from leaving the column. Although the pore size of the distributor
plates is several orders of magnitude larger than the particle size, the risk of losing particles is
small since they do not fluidize as individual particles, but as agglomerates of 200–300 μm
[2]. These agglomerates, which have a void fraction of 98%–99% [44], demonstrate a
dynamic behavior during fluidization [45]. This means that the agglomerates break and
recombine constantly as a consequence of the collisions between solids. The dynamic
behavior and large porosity of the agglomerates ensure that the precursor molecules reach the
whole surface area of the individual particles, even though they exist as agglomerates.
Aeroxide P25 titanium oxide (TiO2) particles from Evonik Industries (Hanau, Germany) are
used as the substrate for the coating experiments. This powder has a Sauter mean diameter
(d3,2) of 32.7 nm (Supplementary Information 2.A), a specific surface area of 52.4 m2/g, and a
surface concentration of hydroxyl groups of 5.0 OH/nm2 [46]. Semiconductor grade TMA
(trimethylaluminium) was provided by Akzo Nobel HPMO (Amersfoort, The Netherlands) in a
600 mL WW-600 stainless steel bubbler, which is kept at 30 °C during the coating
experiments. The second precursor, demineralized water, is kept in a similar bubbler.
Pressurized nitrogen grade 5.0 is provided to the column as the carrier gas; no pump is present
after the column. The column is always kept at atmospheric pressure. During
start-up of an experiment, we first used nitrogen to drive away the air before starting the
coating. The off-gas of the fluidized bed was led through a rack of five washing bubblers
filled with Kaydol oil to remove possible traces of unreacted precursors and the products of
the reactions.
15
Chapter 2
The precursor bubblers, the fluidized bed reactor and the washing bubblers are placed inside a
nitrogen-blanketing cabinet as a TMA safety measure. The cabinet is operated at an O2
concentration below 6%. For each experiment, 2.00 g of TiO2 powder were placed inside the
column. A flow of 0.4 L/min of nitrogen, which corresponds to a superficial gas velocity of 1.26
cm/s, was applied to fluidize the powder.
2.2.2 Design of the Experiments
To calculate the precursor dosing times, we estimated the total amount of active sites in the
bed of particles, i.e., hydroxyl groups. This amount is calculated with the surface area of the
TiO2 (52.4 m2/g), the mass of powder placed inside the column (2.00 g), and the surface
concentration of hydroxyl groups (5.0 OH/nm2) [46]. For 2.00 g of powder, we have an initial
amount of 8.7 × 10−4 mol of OH. To calculate the amount of TMA dosed to the reactor, we
assumed that at 30 °C, which is the TMA bubbler temperature, TMA forms dimers [47–50].
We estimated that the saturation of the nitrogen bubbles with TMA inside the bubbler, using
the model proposed by Mayer et al. [51], is about 50% when the bubbler is filled with TMA
up to the secure level. In addition, we assumed that the entire particle surface area is exposed
to TMA in the gas phase (i.e., no “dead zones”) because fluidized bed reactors provide intense
contact between precursors and solids [52,53]. Based on these assumptions,
and using the vapour pressure of TMA and the ideal gas law, we obtained a flow of TMA to
the reactor of 4.8 × 10−4 mol/min, which translates into an ideal dosing time of 1.7 min for
TMA. To account for the assumptions used in the theoretical estimation, and ensure that the
surface of the powder inside the column is fully saturated, we doubled the resulting dosage
time for TMA and water. As a result, a significant fraction of the TMA leaves the reactor
without reacting. This inefficient use of TMA is an acceptable penalty for the objective of this
paper, i.e., to find out how much excess of precursor can be tolerated close to physisorbing
conditions.
From the calculation above, we obtained a final dosing time of 3.5 min for TMA, and 2.5 min for
water. After dosing each precursor to the reactor, we purge the system with N2 for 10 min
(which corresponds to 13 residence times), establishing a feeding sequence of TMA–N2–
H2O–N2. With these dosing times (3.5–10–2.5–10 min), which we will consider as the base
case situation, we performed 4, 7, and 15 cycles. In the second set of experiments, we studied
the deposition of alumina when increasing the dosing times of precursors and keeping the
same purging time of nitrogen. For this study, we perform 7 cycles at three different dosing
times: 2 times (7–10–5–10 min), 4 times (14–10–10–10 min), and 6 times (21–10–15–10
min) larger than in the base case study. In addition, we performed 5 cycles at 1 bar and 170
°C with dosing times of 21–10–15–10 min, to investigate whether we can avoid the
physisorption of the molecules in excess and gain the self-limitation of the ALD reactions, by
increasing the reactor temperature.
2.2.3 Characterization of the Coating
To characterize the deposition process, we determined the thickness of the alumina film by
both TEM and elemental analysis, and divided it by the number of cycles to calculate the
growth per cycle. For these measurements, we used TEM (transmission electron microscopy)
with EDX (energy dispersive X-ray spectroscopy), done with a Tecnai TF20 (FEI, Hillsboro,
OR, USA), which provides direct observation of the thickness of the coating δTEM. Elemental
16
Gas-phase deposition of ultrathin Al2O3 films on nanoparticles at ambient conditions
analysis was carried out using ICP-OES (Induced Couple Plasma Optical Emission
Spectroscopy, performed with a PerkinElmer Optima 5300, PerkinElmer, Waltham, MA,
USA), which provides the mass fraction of aluminium in the coated sample xAl. To determine
xAl in the samples, we first destructed an amount of 50 mg of the coated powder in a solution
prepared with 1.5 mL 65% HNO3 + 4.5 mL 30% HCl + 1 mL 40% HF using the microwave.
After destruction, the samples were diluted to 50 mL with MQ (milli-Q) water. The samples
were then analyzed with ICP-OES to determine the mass fraction of aluminium in the
samples. With this value, we calculated, using Equation (2.1), the corresponding volume of
aluminium oxide in each particle
, assuming spherical TiO2 particles with a diameter
d3,2 and density
, using the molecular mass of alumina
and aluminium
,a
factor of ½ because there are 2 moles of Al in one mole of Al2O3, and a density of the film of
aluminium oxide
of 2500 kg/m3, equal to the value reported by Groner et al. [36] for
Al2O3 ALD at room temperature. We presume that the density of alumina deposited by ALD
on particles would be smaller than the one deposited on flat substrates, due to the
discontinuities between the facets of the particles.
(2.1)
Then, we used Equation (2) to calculate the thickness of the film
based on the volume of
alumina per particle determined with Equation (1). Assuming spherical particles, we
calculated the thickness of the alumina film as the difference of the volume between the coreshell particle, and the uncoated TiO2 P25 particle. Using Equations (1) and (2), we account for
the curvature of the particles to estimate the thickness of the alumina film
:
√
(2.2)
In the calculation of the film thickness, we used average values for particle size, mass fraction
of aluminium and density of the aluminium oxide; however, that implies an inaccuracy due to
the experimental error. To account for that, we performed the propagation of the uncertainty
(Supplementary Information 2.B), using a Sauter mean diameter
of 32.7 ± 3.3 nm
(Supplementary Information 2.A). The errors correspond to 10% of the measurement of the
diameter
, 3% of the mass fraction value from ICP-OES
, and 200 kg/m3 for the
density of alumina
, to obtain the interval of confidence for the film thickness
.
2.3 Results and Discussion
This section is divided in four parts: (i) experiments performed with the base case dosing
times of the precursors; (ii) experiments done to study the self-limitation of the reactions
when feeding excess of precursors at 1 bar and 27 °C; and (iii) at 1 bar and 170 °C, and (iv) a
study of density of the alumina films using BET (Brunauer-Emmett-Teller) N2-adsorption and
thermal gravimetric analysis (TGA) (Mettler Toledo TGA/SDTA 851 e, Greifensee,
Switzerland).
17
Chapter 2
2.3.1 Base Case Coating Experiments
We performed 4, 7, and 15 ALD cycles with a dosage sequence of 3.5–10–2.5–10 min, of TMA,
nitrogen, water, and nitrogen, respectively, using a flow of 0.4 L/min of N2. We measured the
film thickness of a significant number of points and particles (Supplementary Information 2.C
and 2.D) for a proper analysis of the coated samples. Measuring the film thickness after 4
cycles is complex since each cycle deposits about 30% of a monolayer as consequence of the
steric hindrance between the alkyl ligands, and the density of active sites on the surface of the
particles [54]. Such thin films of around 0.6 nm could be deceiving if the particles are not
properly focused in the TEM images. In addition, we calculated the growth per cycle after
dividing the thickness by the number of ALD cycles, plotted the histograms, and calculated
the mean value and the standard deviation of all the measurements. With this characterization
we evaluated the thickness of the film, the deposition rate, and conformality of the alumina
films.
Figure 2.1. TEM images of the alumina film on TiO2 nanoparticles for a dosing sequence of
3.5–10–2.5–10 min after (a) 4 ALD cycles; (b) 7 ALD cycles; and (c) 15 ALD cycles.
(d) Distribution of the growth per cycle from the TEM pictures after 4 ALD cycles;
(e) after 7 ALD cycles; and (f) after 15 ALD cycles. The mean growth per cycle and the
standard deviation values are given in each histogram.
Figure 2.1a–c shows TEM pictures, and the distribution of the growth per cycle (Figure 2.1d–
f) for the 4, 7, and 15 ALD-cycle experiments, respectively. The values of the growth per
cycle are similar to the results reported for ALD experiments at a few millibars and ~180 °C,
i.e., 0.1–0.2 nm. Although the histograms show some spread, from the TEM images we can
see that the alumina coating on each particle is rather uniform and conformal. Figure 2.2a
suggests a linear correlation between the number of cycles and the film thickness from TEM
with a GPC of about 0.14–0.15 nm (Figure 2.2a), although it would require more data points
to draw solid conclusions about linearity.
18
Gas-phase deposition of ultrathin Al2O3 films on nanoparticles at ambient conditions
To support the results from the TEM images, we performed ICP-OES (Supplementary
Information 2.E) on the samples to calculate the thickness of the alumina films with Equations
(2.1) and (2.2), using a density of alumina of 2500 kg/m3 (Figure 2.2b). First, we observe that
for the samples coated with 4 and 7 cycles (Supplementary Information 2.E) are similar,
while there is a clear difference in the film thickness observed from TEM pictures (Figure
2.1). We have not found a plausible explanation for the comparable values from ICP (Figure
2.2b). Besides these values for the 7-cycle experiment, which seem lower than expected, the
other two values are in good agreement with those from the TEM (Supplementary
Information 2.E). The error bars in Figure 2.2b represent the error in the measurements
obtained from the propagation of the uncertainty (Supplementary Information 2.B).
Figure 2.2. (a) Thickness of the alumina film, determined from TEM images, for the
different number of cycles for a dosing sequence of 3.5–10–2.5–10 min. The error bars
represent the 10% error assumed in the image analysis; (b) Comparison of the film thickness
from TEM images and ICP-OES measurements for a density of alumina of 2500 kg/m3. The
horizontal error bars represent the 10% error in the image analysis, while the vertical error
bars show the error made in the calculation of the thickness, which accounts for the size of
TiO2 particle, the experimental error of the ICP-OES device (3%) and the density of the
alumina. The red line represents ideal agreement between the results.
In this study, we used relatively long pulsing and purging times to provide precursor
molecules to the total surface area of powder, and to make sure that all non-absorbed reactants
were removed. Nevertheless, if one is not carrying out a research study, but rather an
industrial process, the initial load of powder inside the reactor may be increased and purging
times reduced to find a proper compromise between production rate, reactant removal and
operating costs.
2.3.2 Coating with Excess of Precursors at Ambient Conditions
After obtaining a reasonable value for the GPC in line with ALD literature, 0.14–0.15 nm,
we studied what occurs when we feed an excess of precursor by increasing the dosage by a
factor of 2, 4, and 6, compared to the base case dosing times (Figure 2.1b,e), and performed 7
cycles.
Figure 2.3 shows the TEM pictures of the coating experiments with a dosage of TMA of 7
min (Figure 2.3a), 14 min (Figure 2.3b), and 21 min (Figure 2.3c) per cycle. The pulsing time
19
Chapter 2
of water was also increased accordingly for each of the experiments. Figure 2.3d–f gives the
distribution of the growth per cycle for each of the experiments. The mean value and standard
deviation, calculated from over 125 measurements (Supplementary Information 2.D), is
shown in each histogram (Figure 2.3d–f). Although the films look uniform in the TEM pictures
we assessed (Figure 2.3a–c), we observed a noticeable spread in the histograms compared to
Figure 2.1. Nevertheless, if we calculate the normalized standard deviation (i.e., standard
deviation divided by the mean) for the growth per cycle, we observe similar relative spread of
the data. The larger deposition rates obtained from Figure 2.3 compared to the base case study
indicate the presence of a CVD type of growth, especially in the sample where TMA was fed
for 21 min (Figure 2.3c,f). This suggests that part of the excess molecules adsorbs to the
surface of the particles, and reacts during the subsequent reaction, which is a consequence of
operating below the boiling temperature of TMA (~128 °C) and water. Figure 2.4a shows the
increase of the growth per cycle with the dosing time of TMA.
Figure 2.3. TEM image of the alumina film on TiO2 nanoparticles after 7 cycles when the
dosing time of TMA per cycle is (a) 7 min; (b) 14 min; and (c) 21 min. The distribution of
the GPC obtained from TEM images for a dosing time of TMA of (d) 7 min; (e) 14 min; and
(f) 21 min. The mean and standard deviation values are shown in each histogram. The results
in this figure can be compared with the ones of 7 cycles, with a dosing time of 3.5 min for
TMA (Figure 2.1b,e), considered as the base case study.
We also calculated the thickness of the alumina films and the growth per cycle using the mass
fraction of aluminium
from the ICP measurements. We observed a discrepancy between
the measurements for the longest dosing time; this might be explained by the fact that CVD at
these conditions leads to a different alumina density than the assumed 2500 kg/m3.
Although there seems to be a notable CVD-component present at these long pulse times—so
we can no longer call the coating truly ALD—it is still possible to get conformal films of a
reasonably controlled thickness. Nevertheless, the CVD component that occurs at ambient
conditions can be avoided by increasing the reactor temperature.
20
Gas-phase deposition of ultrathin Al2O3 films on nanoparticles at ambient conditions
Figure 2.4. (a) GPC of the alumina film after 7 cycles with different dosing times of TMA
per cycle (3.5, 7, 14, and 21 min). The red open circle represents an experiment carried out at
170 °C with a dosing time of TMA of 21 min. The error bars represent the error made in the
image analysis of the TEM pictures, which is estimated to be around 10%. (b) Comparison of
the GPC from TEM measurements and calculation from ICP-OES measurements for a
density of alumina of 2500 kg/m3. The red open circle represents an experiment carried out at
170 °C with a dosing time of TMA of 21 min, for a density of the alumina of 3000 kg/m3.
The horizontal error bars represent the 10% error in the image analysis, while the vertical
error bars show the error made in the calculation of the thickness, which accounts for the size
of TiO2 particle, the experimental error of the ICP-OES device (3%), and the density of the
alumina. The red line represents the ideal agreement between the results.
2.3.3 Coating with Excess of Precursors at 1 bar and 170 °C
To evaluate the influence of the reactor temperature on the physisorption of the precursor
molecules in excess, we performed a coating experiment at higher temperature, i.e., 170 °C.
For that, we used the dosing times where we observed the strongest CVD component, i.e., 21–
10–15–10 min, and performed 5 cycles at 1 bar and 170 °C. We measured the thickness of the
alumina films of about 20 particles using TEM (Figure 2.5a and Supplementary Information
2.D), and plotted a histogram of these values (Figure 2.5b). We measured a mass fraction of
aluminum
in the sample of 0.044 with ICP-OES, which translated into a film thickness of
0.66 nm and a GPC of 0.13 nm (Figure 2.4a, red-open circular symbol). To calculate these
values we used an alumina density of 3000 kg/m3, typical for Al2O3 ALD at 177 °C [36]. The
results of film thickness and GPC obtained from elemental analysis agree with the values
from the TEM images (Figure 2.5), and in general, with the GPC reported for similar ALD
experiments at <1 mbar and ~180 °C, i.e., 0.1–0.2 nm. The decrease in the GPC from 0.51 nm
at room temperature to 0.16 nm at 170 °C (Figure 2.4a) indicates that the ALD reactions at
these conditions, i.e., 1 bar and 170 °C, are self-limiting even when both precursors are fed in
large excess.
This work, as a first attempt to describe the ALD coating at room conditions, opens up the
possibility for further research, such as that on the purging time. Being able to deposit
controlled thin films at these conditions widens the potential use of alumina ALD to other
heat sensitive materials, facilitating the coating process of particles at larger scales.
21
Chapter 2
Figure 2.5. (a) TEM image of the alumina film on TiO2 nanoparticles after 5 ALD cycles for
a dosing sequence of 21–10–15–10 min at 1 bar and 170 °C; (b) The distribution of the GPC
obtained from TEM images for a dosing sequence of time of 21–10–15–10 min at 1 bar and
170 °C. The mean and standard deviation values are shown in the histogram.
2.3.4 Alumina Film Density
Measuring the density of thin films on nanoparticles is not trivial with flat-substrate
techniques such as ellipsometry. The low value of the density can indicate the presence of
micropores or unreacted molecules in the film, although the results from Figures 2b and 4b
show a good agreement with the value of 2500 kg/m3 [36]. We performed N2-adsorption
measurements using BET, and TGA as further analysis of the alumina films. We measured
the surface area of three samples: P25 TiO2, the sample coated with 15 cycles and base case
dosing times, and the sample coated with 7 cycles and a dosage of precursors six times larger
than the base case study. The three samples showed similar adsorption and desorption
isotherms (Supplementary Information 2.F), indicating that either there are no pores in the
films, or that the pores have a small volume since there is no significant hysteresis in the
isotherms. The surface area of the coated samples decreases compared to the uncoated
nanoparticles (Supplementary Information 2.F), although this is explained by (i) the increase
on the particle size due to the alumina coating, and (ii) by the decrease of the density of the
core-shell particle, since the alumina film has a lower density than the P25 TiO2. These two
factors do not provide evidence of pore formation in the alumina coating. Moreover, no pores
are visible in the TEM images (Figure 2.2, Figure 2.4 and Supplementary Information 2.C and
2.D), although that depends on the focus of the microscope. Both measurements suggest that
the obtained films are not porous.
As an alternative explanation, the presence of unreacted water molecules or methyl groups
inside the alumina coating could decrease the value of the film density. We performed TGA
on the uncoated TiO2 and the samples coated with 4, 7 and 15 cycles to measure the mass loss
when heating the samples from room temperature to 600 °C in air atmosphere. In this range of
temperatures, we would observe first the evaporation of water molecules—both physisorbed
and entrapped in the coating—and at higher temperatures the combustion of methyl groups.
We found a total weight loss of <3.5%, which was mostly produced below 200 °C
(Supplementary Information 2.G). This is attributed to water evaporation, part of it
physisorbed molecules assumed from the analysis of the uncoated TiO2 sample, and part of it
entrapped molecules within the coating. We compared the amount of water entrapped in the
22
Gas-phase deposition of ultrathin Al2O3 films on nanoparticles at ambient conditions
coating, with the amount of water dosed to the reactor, obtaining that the percentage of
entrapped water is <0.1% (Supplementary Information 2.G). This suggests that the
accumulation of unreacted water is negligible. The combustion of the unreacted methyl
groups, which occurs above 400 °C, would produce an exothermic peak in the heat flow
during the TGA and an increase in the SDTA (differential thermal analysis), which relates to
the temperature of the sample and the chamber (Supplementary Information 2.G). At
temperatures between 400 and 600 °C the mass loss recorded is <0.5% and the temperature
difference between the sample and chamber is <0.5 °C for all the cases. We calculated the
number of methyl groups of which combustion relates to this increase of temperature,
resulting in <0.001% of the total amount of methyl groups being fed to the reactor during the
whole experiment. This shows that there is not a substantial amount of unreacted water
molecules or methyl groups to alter the properties of the alumina films. The low density of
alumina we found—in agreement with the reported value for ALD at room temperature
[36]—seems to be caused neither by porosity nor by the presence of unreacted species.
2.4 Conclusions
We demonstrated that ultrathin films of aluminium oxide can be deposited on particles using
ALD at 27 °C and 1 bar in a well-mixed reactor, a fluidized bed. For that, we established that
controlling the amount of precursor dosed to the reactor is crucial. For a set of experiments of
4, 7 and 15 cycles, we obtained a GPC of 0.14 nm for the defined dosing times. This result is
comparable to the GPC of the experiments at ~170 °C and <1 mbar. We also calculated the
thickness of the alumina films based on the elemental composition of aluminium in the
samples, achieving a good agreement with the measured values from the TEM pictures. We
increased the amount of the precursors by a factor of 2, 4 and 6 at constant flow, feed
concentration and purging time, and performed 7 cycles. We observed that the thickness of
the alumina films increased with the dosage of the precursors. A tentative explanation is that
the unreacted molecules, which were fed in excess, accumulated on the surface of the
particles, resulting in parasitic CVD reactions and higher GPC. In addition, we performed a
coating experiment at 170 °C overdosing both precursors by a factor of 6 compared to the
base case study, and obtained a GPC of 0.16 nm. This result agrees with the range of values of
GPC reported for typical Al2O3 ALD reactions. We conclude that at room temperature and
atmospheric pressure, this ALD process ceases to be self-limiting, but allows precise
deposition of thin and uniform films by controlling the dosage of precursors. The obtained
films are non-porous and do not contain noticeable amounts of unreacted species.
Bibliography
1.
2.
3.
4.
King, D.M.; Liang, X.; Weimer, A.W. Functionalization of fine particles using atomic
and molecular layer deposition. Powder Technol. 2012, 221, 13–25.
Van Ommen, J.R.; Valverde, J.M.; Pfeffer, R. Fluidization of nanopowders: A review.
J. Nanopart. Res. 2012, 14, 1–29.
Goulas, A.; van Ommen, J.R. Scalable production of nanostructured particles using
atomic layer deposition. KONA Powder Part. J. 2014, 31, 234–246.
Longrie, D.; Deduytsche, D.; Detavernier, C. Reactor concepts for atomic layer
deposition on agitated particles: A review. J. Vac. Sci. Technol. A 2014, 32, 010802,
doi:10.1116/1.4851676.
23
Chapter 2
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
Moghtaderi, B.; Shames, I.; Doroodchi, E. Combustion prevention of iron powders by a
novel coating method. Chem. Eng. Technol. 2006, 29, 97–103.
Hakim, L.F.; Vaughn, C.L.; Dunsheath, H.J.; Carney, C.S.; Liang, X.; Li, P.; Weimer,
A.W. Synthesis of oxidation-resistant metal nanoparticles via atomic layer deposition.
Nanotechnology 2007, 18, 345603, doi:10.1088/0957-4484/18/34/345603.
King, D.M.; Liang, X.; Burton, B.B.; Kamal Akhtar, M.; Weimer, A.W. Passivation of
pigment-grade TiO2 particles by nanothick atomic layer deposited SiO2 films.
Nanotechnology 2008, 19, 255604, doi:10.1088/0957-4484/19/25/255604.
Liang, X.; King, D.M.; Groner, M.D.; Blackson, J.H.; Harris, J.D.; George, S.M.;
Weimer, A.W. Barrier properties of polymer/alumina nanocomposite membranes
fabricated by atomic layer deposition. J. Membr. Sci. 2008, 322, 105–112.
Liang, X.; Lynn, A.D.; King, D.M.; Bryant, S.J.; Weimer, A.W. Biocompatible interface
films deposited within porous polymers by atomic layer deposition (ALD). ACS Appl.
Mater. Interfaces 2009, 1, 1988–1995.
Lichty, P.; Liang, X.; Muhich, C.; Evanko, B.; Bingham, C.; Weimer, A.W. Atomic layer
deposited thin film metal oxides for fuel production in a solar cavity reactor. Int. J.
Hydrog. Energy 2012, 37, 16888–16894.
Li, J.; Liang, X.; King, D.M.; Jiang, Y.B.; Weimer, A.W. Highly dispersed Pt
nanoparticle catalyst prepared by atomic layer deposition. Appl. Catal. B Environ. 2010,
97, 220–226.
Hirsch, L.R.; Gobin, A.M.; Lowery, A.R.; Tam, F.; Drezek, R.A.; Halas, N.J.; West, J.L.
Metal nanoshells. Ann. Biomed. Eng. 2006, 34, 15–22.
Bawarski, W.E.; Chidlowsky, E.; Bharali, D.J.; Mousa, S.A. Emerging
nanopharmaceuticals. Nanomed. Nanotechnol. Biol. Med. 2008, 4, 273–282.
Elam, J.W.; Dasgupta, N.P.; Prinz, F.B. ALD for clean energy conversion, utilization,
and storage. MRS Bull. 2011, 36, 899–906.
Sveshnikova, G.V.; Kol’Tsov, S.I.; Aleskovskii, V.B. Formation of a silica layer of
predetermined thickness on silicon by the Molecular-Layering method. Zh Prikl Khim
1970, 43, 1150–1152.
Suntola, T. Atomic layer epitaxy. Thin Solid Films 1992, 216, 84–89.
George, S.M. Atomic layer deposition: An overview. Chem. Rev. 2010, 110, 111–131.
Miikkulainen, V.; Leskelä, M.; Ritala, M.; Puurunen, R.L. Crystallinity of inorganic
films grown by atomic layer deposition: Overview and general trends. J. Appl. Phys.
2013, 113, 021301, doi:10.1063/1.4757907.
Liang, X.; Zhou, Y.; Li, J.; Weimer, A. Reaction mechanism studies for platinum
nanoparticle growth by atomic layer deposition. J. Nanopart. Res. 2011, 13, 3781–3788.
Goulas, A.; van Ommen, J.R. Atomic layer deposition of platinum clusters on titania
nanoparticles at atmospheric pressure. J. Mater. Chem. A 2013, 1, 4647–4650.
Hakim, L.F.; King, D.M.; Zhou, Y.; Gump, C.J.; George, S.M.; Weimer, A.W.
Nanoparticle coating for advanced optical, mechanical and rheological properties. Adv.
Funct. Mater. 2007, 17, 3175–3181.
Puurunen, R.L. Growth per cycle in atomic layer deposition: A theoretical model. Chem.
Vap. Depos. 2003, 9, 249–257.
Profijt, H.B.; Potts, S.E.; van de Sanden, M.C.M.; Kessels, W.M.M. Plasma-assisted
atomic layer deposition: Basics, opportunities, and challenges. J. Vac. Sci. Technol. A
2011, 29, 050801, doi:10.1116/1.3609974.
Potts, S.E.; Kessels, W.M.M. Energy-enhanced atomic layer deposition for more process
and precursor versatility. Coord. Chem. Rev. 2013, 257, 3254–3270.
24
Gas-phase deposition of ultrathin Al2O3 films on nanoparticles at ambient conditions
25. Puurunen, R.L. Correlation between the growth-per-cycle and the surface hydroxyl group
concentration in the atomic layer deposition of aluminum oxide from trimethylaluminum
and water. Appl. Surf. Sci. 2005, 245, 6–10.
26. Philip, A.; Thomas, S.; Kumar, K.R. Calculation of growth per cycle (GPC) of atomic
layer deposited aluminium oxide nanolayers and dependence of GPC on surface OH
concentration. Pram. J. Phys. 2014, 82, 563–569.
27. Wank, J.R.; George, S.M.; Weimer, A.W. Coating fine nickel particles with Al2O3
utilizing an atomic layer deposition-fluidized bed reactor (ALD-FBR). J. Am. Ceram.
Soc. 2004, 87, 762–765.
28. Wank, J.R.; George, S.M.; Weimer, A.W. Nanocoating individual cohesive boron nitride
particles in a fluidized bed by ALD. Powder Technol. 2004, 142, 59–69.
29. Hakim, L.F.; George, S.M.; Weimer, A.W. Conformal nanocoating of zirconia
nanoparticles by atomic layer deposition in a fluidized bed reactor. Nanotechnology
2005, 16, S375–S381.
30. Hakim, L.F.; Blackson, J.; George, S.M.; Weimer, A.W. Nanocoating individual silica
nanoparticles by atomic layer deposition in a fluidized bed reactor. Chem. Vap. Depos.
2005, 11, 420–425.
31. Hakim, L.F.; McCormick, J.A.; Zhan, G.D.; Weimer, A.W.; Li, P.; George, S.M. Surface
modification of titania nanoparticles using ultrathin ceramic films. J. Am. Ceram. Soc.
2006, 89, 3070–3075.
32. Liang, X.; Zhan, G.D.; King, D.M.; McCormick, J.A.; Zhang, J.; George, S.M.; Weimer,
A.W. Alumina atomic layer deposition nanocoatings on primary diamond particles using
a fluidized bed reactor. Diam. Relat. Mater. 2008, 17, 185–189.
33. Beetstra, R.; Lafont, U.; Nijenhuis, J.; Kelder, E.M.; van Ommen, J.R. Atmospheric
pressure process for coating particles using atomic layer deposition. Chem. Vap. Depos.
2009, 15, 227–233.
34. Liang, X.; George, S.M.; Weimer, A.W.; Li, N.H.; Blackson, J.H.; Harris, J.D.; Li, P.
Synthesis of a novel porous polymer/ceramic composite material by low-temperature
atomic layer deposition. Chem. Mater. 2007, 19, 5388–5394.
35. Liang, X.; Hakim, L.F.; Zhan, G.D.; McCormick, J.A.; George, S.M.; Weimer, A.W.;
Spencer Ii, J.A.; Buechler, K.J.; Blackson, J.; Wood, C.J.; et al. Novel processing to
produce polymer/ceramic nanocomposites by atomic layer deposition. J. Am. Ceram.
Soc. 2007, 90, 57–63.
36. Groner, M.D.; Fabreguette, F.H.; Elam, J.W.; George, S.M. Low-temperature Al2O3
atomic layer deposition. Chem. Mater. 2004, 16, 639–645.
37. Potts, S.E.; Profijt, H.B.; Roelofs, R.; Kessels, W.M.M. Room-temperature ALD of
metal oxide thin films by energy-enhanced ALD. Chem. Vap. Depos. 2013, 19, 125–133.
38. Tang, X.; Francis, L.A.; Simonis, P.; Haslinger, M.; Delamare, R.; Deschaume, O.;
Flandre, D.; Defrance, P.; Jonas, A.M.; Vigneron, J.P.; et al. Room temperature atomic
layer deposition of Al2O3 and replication of butterfly wings for photovoltaic application.
J.
Vac.
Sci.
Technol.
A
2012,
30, 01A146, doi:10.1116/1.3669521.
39. Jur, J.S.; Parsons, G.N. Atomic layer deposition of Al2O3 and ZnO at atmospheric
pressure in a flow tube reactor. ACS Appl. Mater. Interfaces 2011, 3, 299–308.
40. Mousa, M.B.M.; Oldham, C.J.; Parsons, G.N. Atmospheric pressure atomic layer
deposition of Al2O3 using trimethyl aluminum and ozone. Langmuir 2014, 30, 3741–
3748.
41. Skoog, S.A.; Elam, J.W.; Narayan, R.J. Atomic layer deposition: Medical and biological
applications. Int. Mater. Rev. 2013, 58, 113–129.
25
Chapter 2
42. Salameh, S.; Schneider, J.; Laube, J.; Alessandrini, A.; Facci, P.; Seo, J.W.; Ciacchi, L.C.;
Mädler, L. Adhesion mechanisms of the contact interface of TiO2 nanoparticles in films
and aggregates. Langmuir 2012, 28, 11457–11464.
43. Mousa, M.B.M.; Oldham, C.J.; Jur, J.S.; Parsons, G.N. Effect of temperature and gas
velocity on growth per cycle during Al2O3 and ZnO atomic layer deposition at
atmospheric pressure. J. Vac. Sci. Technol. A. 2012, 30, 01A155, doi:10.1116/1.3670961.
44. De Martín, L.; Bouwman, W.G.; van Ommen, J.R. Multidimensional nature of fluidized
nanoparticle agglomerates. Langmuir 2014, 30, 12696–12702.
45. Hakim, L.F.; Portman, J.L.; Casper, M.D.; Weimer, A.W. Fluidization behavior and
conformal coating of nanoparticles in fluidized beds by ALD. In Proceedings of the 2004
AIChE Annual Meeting, Austin, TX, USA, 7–12 November 2004; pp. 2339–2359.
46. Mueller, R.; Kammler, H.K.; Wegner, K.; Pratsinis, S.E. OH surface density of SiO2 and
TiO2 by thermogravimetric analysis. Langmuir 2003, 19, 160–165.
47. Gow, T.R.; Lin, R.; Cadwell, L.A.; Lee, F.; Backman, A.L.; Masel, R.I. Decomposition
of trimethylaluminum on silicon(100). Chem. Mater. 1989, 1, 406–411.
48. Travis, C.; Adomaitis, R. Dynamic modeling for the design and cyclic operation of an
atomic layer deposition (ALD) reactor. Processes 2013, 1, 128–152.
49. Henrickson, C.H.; Eyman, D.P. Lewis acidity of alanes. Interactions of trimethylalane with
sulfides. Inorg. Chem. 1967, 6, 1461–1465.
50. Laubengayer, A.W.; Gilliam, W.F. The alkyls of the third group elements. I. Vapor phase
studies of the alkyls of aluminum, gallium and indium1. J. Am. Chem. Soc. 1941, 63,
477–479.
51. Mayer, B.; Collins, C.C.; Walton, M. Transient analysis of carrier gas saturation in liquid
source vapor generators. J. Vac. Sci. Technol. A 2001, 19, 329–344.
52. Kunii, D.; Levenspiel, O. Fluidization Engineering, 2nd ed.; Butterworth-Heinemann:
Oxford, UK, 1991.
53. Grillo, F.; Kreutzer, M.T.; van Ommen, J.R. Modeling the precursor utilization in atomic
layer deposition on nanostructured materials in fluidized bed reactors. Chem. Eng. J. 2015,
268, 384–398.
54. Puurunen, R.L. Surface chemistry of atomic layer deposition: A case study for the
trimethylaluminum/water
process.
J.
Appl.
Phys.
2005,
97,
121301,
doi:10.1063/1.1940727.
26
Gas-phase deposition of ultrathin Al2O3 films on nanoparticles at ambient conditions
2.A Particle Size Distribution of TiO2 from TEM Images
We obtained the PSD (particle size distribution) of the TiO2 P25 particles from the
measurements of the diameter of 25 TEM pictures. We determined the distribution for each
particle diameter, and plotted the PSD (Figures 2.A.1 and 2.A.2). From it, we calculated the
Sauter mean diameter (
) using the equation below. The Sauter diameter accounts for the
ration volume to surface area of the particles. The calculation of the alumina film thickness
(section 2.B) is based on the surface area of the particles, then we consider that the
is the
most accurate representation of the mean particle size in order to calculate the film thickness.
We also calculated the number-averaged diameter (
) from the transmission electron
microscopy (TEM) measurement, observing the difference with the
. We consider that we
do an error of 10% on the measurements of particle diameter from the TEM images (
=3
nm).
∑
∑
∑
Figure 2.A.1. PSD of the P25 TiO2 from TEM images. We measure the diameter in the TEM
pictures, and from them we calculated the
and plotted the distributions. We obtained an
average
= 32.7 nm.
Table 2.A.1. Determination of the PSD of TiO2 P25 from TEM images.
Sample
TiO2 P25
N TEM Pictures
25
N Particles
>300
27
N Points
533
[nm]
26.2
[nm]
32.7
Chapter 2
2.B Al2O3 Film Thickness Calculation
This calculation is based on the measurements of the mass fraction of aluminium ( ) on the
coated samples, obtain from the ICP-OES. From this value, we calculate the volume of
aluminium oxide per particle (
) related to the mass fraction of aluminium. Finally, we
calculate the thickness of the alumina film (
) considering spherical particles. To account
for the systematic error of the measurements of the particle size (
) and aluminium mass
fraction (
), we perform the propagation of the uncertainty to calculate the error originated
from the calculation of the film thickness (
).
2.B.1 Equation of the Volume of Alumina
[
]
[
]
[
]
[
[
[
[
]
[
]
]
]
]
Then the volume of aluminium oxide per particle can be calculated with this equation:
28
Gas-phase deposition of ultrathin Al2O3 films on nanoparticles at ambient conditions
2.B.2 Calculation of the Thickness of the Aluminium Oxide Film
(
)
[
]
where:
2.B.3 Error Propagation Equation
√(
)
(
)
29
(
)
Chapter 2
2.C TEM for Base Case Dosing Times after 4, 7, and 15 Cycles
(a)
(b)
(c)
Figure 2.C.1 TEM images of the base case experiments with (a) four; (b) seven; and (c) 15
cycles.
Table 2.C.1. Determination of alumina film thickness from TEM images, with the
standard deviation of the measurements.
Sample
4 cycles
7 cycles
15 cycles
N Pictures [-]
17
14
14
N Particles [-]
20
49
52
N Points [-]
73
148
163
30
[nm]
0.6 ± 0.1
1.0 ± 0.2
2.1 ± 0.6
GPC [nm]
0.15 ± 0.02
0.14 ± 0.02
0.14 ± 0.04
Gas-phase deposition of ultrathin Al2O3 films on nanoparticles at ambient conditions
2.D TEM after 7 Cycles, Feeding the Precursors in Excess
(a)
(b)
(c)
Figure 2.D.1. TEM images of the experiments where the precursors were dosing with
twice, four and six times the dosing times of the base case study. (a) 2× tbase; (b) 4× tbase; (c)
6× tbase.
Table 2.D.1. Determination of alumina film thickness from TEM images, with the tandard
deviation of the measurements. Temperature is 27 °C unless indicated otherwise.
Sample
t base case
2 × t base case
4 × t base case
6 × t base case
170 °C–6 × t base
N Pictures [-]
14
11
9
10
15
N Particles [-]
49
35
25
28
23
N Points [-]
148
131
126
113
96
[nm]
1.0 ± 0.2
1.5 ± 0.3
1.8 ± 0.4
3.6 ± 1.3
0.8 ± 0.1
GPC [nm]
0.14 ± 0.02
0.22 ± 0.04
0.26 ± 0.06
0.51 ± 0.19
0.16 ± 0.02
2.E ICP-OES Measurements
Table 2.E.1. Film thickness calculated from the ICP-OES measurements, and comparison
with the results from the TEM measurements. The interval of confidence represents the error
in the measurements as shown.
Sample
4 cycles
7 cycles
15 cycles
2 × t base case
4 × t base case
6 × t base case
170 °C–6 × t base
[-]
0.036 ± 0.001
0.035 ± 0.001
0.105 ± 0.003
0.086 ± 0.003
0.100 ± 0.003
0.131 ± 0.004
0.044 ± 0.001
[m3]
2.44 × 10−24
2.37 × 10−24
8.32 × 10−24
6.48 × 10−24
7.79 × 10−24
1.10 × 10−24
2.30 × 10−24
31
[nm]
0.64 ± 0.08
0.62 ± 0.08
2.02 ± 0.26
1.61 ± 0.21
1.90 ± 0.24
2.58 ± 0.33
0.66 ± 0.08
[nm]
0.59 ± 0.06
0.96 ± 0.10
2.10 ± 0.21
1.51 ± 0.15
1.84 ± 0.18
3.60 ± 0.36
0.78 ± 0.08
Chapter 2
2.F BET Measurements of the Porosity
Figure 2.F.1. BET isotherms for the uncoated TiO2 (left), sample coated with 15 cycles
(middle), and sample coated with a dosage of precursors six times larger than the base case
study (right).
Table 2.F.1. Comparison of the surface area measured with BET and calculated.
Samples
Un-coated TiO2
15 cycles
6 × tdosing
[m2/g]
52.4 ± 0.1
39.9 ± 0.1
38.2 ± 0.1
[nm]
2.1 ± 0.7
3.7 ± 1.5
[kg/m3]
3683 ± 141
3422 ± 213
[m2/g]
44.1 ± 4.6
43.7 ± 5.6
We use the values of the particle size measured (
= 33 ± 3 nm), density of TiO2
3
(
= 4200 kg/m ) and density of the alumina film (
= 2500 kg/m3) to compare the
measured surface area (
) with the calculated value (
). We calculated the interval
of confidence of the measurement by calculating the propagation of the uncertainty. These are
the equations used:
√(
√(
)
)
(
(
)
)
32
(
)
Gas-phase deposition of ultrathin Al2O3 films on nanoparticles at ambient conditions
2.G Thermo Gravimetric Analysis
Figure 2.G.1. (left) Mass loss profile for the uncoated TiO2 and the samples coated with four,
seven and 15 cycles; (right) Calculation of the percentage of the water molecules fed to the
system compared to the amount of water released during the TGA analysis.
Figure 2.G.2. (left) Mass loss profile for the uncoated TiO2 and the samples coated with an
excess of precursors of two and six times that of the base case study; (right) Calculation of the
percentage of the water molecules fed to the system compared to the amount of water released
during the TGA analysis.
Figure 2.G.3. Differential Thermal Analysis (SDTA) from the TGA, for the base case study
(left); and when the precursors where fed in excess (right).
33
Chapter 3
Tuning roughness and gloss of powder
coating paint by encapsulating
the particles with Al2O3 films
We report a method to change of the surface finish of a standard polyester-based powder
coating paint, from gloss to matt, by depositing ultrathin films of Al2O3 on the powder coating
particles, using trimethylaluminium and water as precursors. We performed this coating in a
fluidized bed reactor at 1bar and 27 °C, using a gas-phase coating process of alternating
exposure of the particles to the two precursors, similar to atomic layer deposition (ALD). We
varied the number of coating cycles (1, 2, 3, 5, 7 and 9 cycles) to obtain film thicknesses
ranging from 1 to 30 nm, determined by elemental analysis of the coated samples. The
average growth per cycle of the process is 3.5 nm, significantly larger than the growth per
cycle for pure self-limiting ALD process. When the average alumina shell was thicker than 6
nm, the shell prevented the flow of the core, that did soften above its glass transition
temperature. With the particles morphology intact, this resulted in a rough and matte finish
coating after curing. The surface roughness, with a value around 9 µm determined by surface
profilometry, is associated to the alumina coated particles as observed from SEM and EDX
analysis. In addition, the matte finish coating showed good mechanical resistance. Using this
technique, we avoid the use of a foreign particles to produce matte coatings with similar
properties as the gloss paints.
Submitted as: D. Valdesueiro, H. Hettinga, J.P. Drijfhout, P. Lips, G.M.H. Meesters, M.T.
Kreutzer and J.R. van Ommen. Tuning roughness and gloss of powder coating paint by
encapsulating the coating particles with thin Al2O3 films.
35
Chapter 3
3.1 Introduction
Powder coatings have important advantages over liquid-based paints: ease of application, high
utilization by electrostatic spraying, being environmentally friendly since they do not contain
solvents, and showing excellent performance once applied. These strong points are also
known as the Four E’s, standing for ecology, excellence of finish, economy and energy [1-3].
These properties allow powder coatings to be used in a wide variety of applications, i.e.
automotive, architectural, electronics and furniture amongst others [3]. A powder coating is
composed by a resin, a catalyst, a cross-linker, pigments and additives such as flow modifiers
and degassing agents, which define the properties of the final powder coating, including the
surface finish: gloss or matte.
Glossy coatings reflect all the incident light in a mirror-like fashion, whereas a matte finish
scatters part of the light, lessening the glossiness. The difference in gloss or matte appearance
relies on the surface texture, whether it is highly smooth or rough. Certain indoor applications
require gloss finish paints, while for industrial and agricultural purposes, a matte paint is
preferred to hide surface irregularities and damages. The transition from a gloss to a matte
finish is possible by partially blocking the activity of the catalyst [4], varying the size of the
powder coating, adding inert fillers, playing with the humidity of the environment or tuning
the conditions during the electro spraying [5]. In this work, we use a different approach to
change the surface finish of a powder coating paint, from gloss to matte, by modifying the
surface of the powder coating particles.
For that, we deposited ultrathin films of aluminium oxide (Al2O3) using a gas-phase technique
similar to atomic layer deposition (ALD) on the primary particles of a standard powder
coating, produced by DSM, which is characterized by a fast and low-temperature cure, good
flow and flexibility, and a gloss surface finish once it cures, ideal for architectural
applications [6]. By using the same particles as the original powder coating, but with modified
surface, we prevent issues such as lack of homogeneity or of adhesion induced by the
polydispersity in the size of the powder and the adhesion of a foreign material to the powder
coating paint. We investigated how aluminium oxide films will modify the flowing behaviour
of the powder coating particles above the glass transition temperature, and whether that would
induce a textured-surface paint.
ALD allows the deposition of inorganic films in a layer-by-layer growth manner based on two
consecutive self-terminating reactions, with a purging step using an inert gas after each
reaction [7, 8]. The self-terminating feature of the ALD reactions ensures that the precursor
molecules will only react where there is an active site available, preventing the growth of
several layers of compound in each cycle [9, 10]. Al2O3 ALD, using trimethylaluminium
(TMA) and water as precursors, is commonly carried out at ~170 °C and a few millibars of
pressure, achieving a growth per cycle (GPC) of 0.1-0.2 nm [11, 12]. The growth of the
alumina at these conditions is independent of the dosing time of the precursors due to the selfterminating property of the ALD reactions. Since our material cannot be heated to such
temperatures, we deposited alumina films using TMA and water at ambient conditions, i.e. 27
± 3 °C and 1 bar. Working at these conditions may involve the accumulation of the unreacted
species that are dosed in excess above the saturation regime, inducing a chemical vapour
deposition (CVD) type of reaction [13, 14]. That results in the deposition of multiple atomic
layers of alumina during each cycle, and therefore, a higher GPC than in a typical ALD
process, depending on the amount of precursors dosed to the reactor [15]. The experiments in
this work were carried out in a fluidized bed reactor, which provides good mixing between
36
Tuning roughness and gloss of powder coating paint by encapsulating the particles with Al2O3 films
gas and solids, and allows processing large amounts of particles [16-20]. In these
experiments, we fed an excess of both precursors to induce the accumulation of molecules on
the substrates, and in this manner grow thicker alumina films faster.
Alumina deposited with ALD was applied as passivating materials, in the production of
membranes and catalysts amongst others [21-28]. In this work, we altered the surface finish of
a powder coating paint by depositing small amounts of Al2O3 on the surface of the powder
coating particles. The alumina films act as physical barrier to confine the softened powder
coating paint above the glass transition temperature, and prevent it from mixing
homogeneously with other powder coating particles until reaching the curing temperature, at
which the resin cross-links and the softened material becomes a crosslinked paint. We
investigated the influence of the thickness of the Al2O3 films on the flowability of the coated
particles, and how this translated into different textures of the final paint. This experimental
paper is a proof of concept for the applicability of an emerging technology such as ALD in a
fluidized bed reactor to modify the surface appearance of an industrial product, i.e. standard
powder coating, while maintaining the same mechanical properties.
3.2 Experimental
3.2.1 Experimental setup and compounds
Al2O3 coating experiments were performed in a fluidized bed reactor similar to the one
described previously [15, 29], composed by a vertical glass column with a diameter of 26 mm
and 500 mm in length which was located in a vertical vibration table to assist the fluidization
[17]. Semiconductor grade TMA was provided by Akzo Nobel HPMO in a 600 mL WW-600
stainless steel bubbler, which is kept at 30 °C during operation. Water, the second precursor,
is kept in a similar bubbler, while nitrogen grade 5.0 is used as carrier and purging gas. In
each experiment we coated 110 g of white standard powder coating particles with glossy
finish, provided by DSM Coating Resins (Zwolle, The Netherlands). The powder coating
particles, with an Sauter mean diameter of 33 μm, is composed by five components: resin, i.e.
Uralac® P 3210, crosslinker, white pigment, i.e. titanium oxide, flow control agent and
degassing agent (more detail in [6] and Supplementary information 3.A). In order to fluidize
the particles, we used a flow of 0.4 L/min of nitrogen, which corresponds to a superficial gas
velocity of 1.26 cm/s.
3.2.2 Coating experiments
We used 8–10–4–10 min for the dosing sequence TMA–N2–H2O–N2. To estimate these
dosing times, we calculated the maximum amount of aluminium atoms and methyl groups
that can be allocated on the surface of a powder coating particle to obtain fully coverage.
These values are 4 and 5 species per nm2 of Al and CH3, respectively [30, 31]. Using the
Sauter mean diameter (d3,2) of 33 µm and a particle density of 1500 kg/m3 (Supplementary
information 3.A), and assuming that the particles are spherical, we calculated a specific
surface area of the powder of 0.12 m2/g, and a total surface area of inside the column of 13
m2, for the 110 g of powder used in each experiment. We used the vapour pressure of the
precursors inside the bubblers and the ideal gas law to calculate the amount of precursor
molecules fed to the reactor, assuming that the TMA is a dimer at 30 °C [32, 33], and that the
saturation of the nitrogen bubbles inside the TMA bubbler is about 50% [34]. We obtained
37
Chapter 3
theoretical dosing times of 0.25 and 0.24 min for TMA and water to saturate the surface of the
particles with precursor molecules. In order to obtain thicker alumina films, we overdosed
both precursors, fixing the dosing times in 8 and 4 min for TMA and water. At ambient
conditions, the molecules of water would physisorb onto the substrate [35-37], being involved
in the following reaction with TMA molecules, also dosed in excess. We used a purging time
of 10 min, approximately 13 times the residence time in the reactor. With these times, we
performed six deposition experiments with different number of cycles, i.e. 1, 2, 3, 5, 7 and 9.
3.2.3 Characterization of the coated particles and panels
To investigate the influence of the alumina film thickness on the surface finish of the cured
powder coating. To characterize this process, we first looked at individual powder coating
particles. We estimated the film thickness and the growth per cycle from the fraction of
aluminium on the samples measured by elemental analysis using ICP-OES (induced couple
plasma – optical emission spectroscopy) as shown elsewhere [15]. Further, we performed
DSC (differential scanning calorimetry) to study whether the alumina films influenced the
thermal properties of the coated particles, such as the glass transition temperature. Finally, we
studied the completeness of the alumina films and its barrier performance with a hot stage
microscope and the pill flow test, which compares the flowability of the coated powder down
an inclined panel while heating the samples in an oven from room to the curing temperature.
Subsequently, we studied the influence of the alumina films on the appearance of the paints
by spraying the coated samples onto panels, using a corona-discharge gun, and curing them in
an oven at 160 °C during 10 minutes. We found that the aluminium oxide films reduced the
charge acquired by the powder, which did not adhere properly to the panels during spraying.
To counteract the poor adhesion, we mixed the white Al2O3-coated powder (TiO2 pigment)
with brown uncoated powder (prepared using a mixture of pigments and BaSO4 as filler) in
order to improve the adhesion and give contrast to the painted panels, and facilitate the
characterization of the panels. The other compounds of both powders are the same. Before
spraying the panels, we prepared mixtures of the Al2O3-coated white and uncoated brown
powders using a 1:1 mass ratio.
We studied the topography of the panels by visual observation, and SEM (scanning electron
microscope) coupled with EDX (energy dispersive X-rays) to have a closer look at the surface
of the panels. In addition, we measured the roughness with surface pofilometry and the gloss
of the paints with a haze-gloss meter, and evaluated the mechanical resistance upon the
impact of a steel ball in the reverse side of the panels. More details of the characterization
techniques are given in Supplementary information 3.B.
3.3 Results and discussion
3.3.1 Characterization of the coated particles
We performed a series of six experiments to apply alumina film, with different thicknesses,
on uncoated white powder coating particles, and estimated the film thickness using the
fraction of aluminium obtained from elemental analysis with ICP. The uncoated powder
coating particles contained a mass fraction of aluminium of 0.0064, coming from the white
pigment (Supplementary information 3.C). From the fraction of aluminium on the coated
38
Tuning roughness and gloss of powder coating paint by encapsulating the particles with Al2O3 films
samples, we calculated the corresponding volume of aluminium oxide (
). From this
value we calculated an equivalent film thickness of alumina ( ) assuming that the particles
are spherical, and that the coatings are uniform. We refer to this value as “equivalent film
thickness” since the coating process requires a certain number of cycles to form a uniform and
conformal alumina film [31, 38], that is, there is not yet a complete film after few coating
cycles. With an increasing number of cycles, we found an increase in the content of
aluminium on the coated samples (i.e., in the deposited volume of aluminium oxide); see Fig.
3.1. We calculated the GPC from the slope of the linear fit shown by the red-dotted line. The
uncertainty in the calculation of the film thickness is represented by the error bars in Fig. 3.1
(more details can be found in Supplementary Information 3.C). We obtained a GPC of 3.5 nm,
much larger than the typically values for Al2O3 ALD, i.e. 0.1-0.2 nm. This is explained by the
accumulation of the overdosed precursor molecules at ambient conditions, which would react
in the subsequent reaction. Obtaining such high GPC benefited our process, since we were
able to deposit thick alumina films faster, although we can no longer refer to this coating
process as atomic layer deposition. Nevertheless, we maintain a rather good control over the
film thickness, as shown in Fig. 3.1.
Fig. 3.1 Volume of aluminium oxide deposited on the particles (primary Y axis) and the
equivalent film thickness (secondary Y axis) assuming spherical particles, and uniform and
complete alumina films. The increase of the film thickness can be approximated with a linear
trend, where the slope represents a GPC of 3.5 nm. The error bars represent the uncertainty in
the calculation of the film thickness.
To study whether (i) the aluminium oxide films would act as an insulating layer, changing the
thermal properties of the powder such as the glass transition temperature ( ), or (ii) the
alumina would only act as a physical barrier, preventing the softened resin from flowing
freely, we performed DSC to determine the thermal properties and used a hot stage
microscope to observe the particles above their glass transition temperature. We combined the
results from both techniques to identify the influence of the alumina films on the powder
coating particles.
The DSC measurements in Fig. 3.2 show the heating phase, with positive normalized heat
flows, and the cooling phase, with negative normalized heat flows. During the heating we see
the glass transition as the step function at around 50 °C. The powder coating is amorphous, so
we refer as softening of the particles above the glass transition instead of melting, which only
applies for crystalline materials and would be seen as an endothermic peak with a constant
39
Chapter 3
baseline. To compare all the samples, we normalized the heat flow measured by the DSC
device with the amount of powder used in each measurement, which ranged between 10 and
20 mg. We used a multi-step program in the DSC, composed by: (i) equilibration of the
sample at 25 °C for 5 min, (ii) cooling down from 25 to 0 °C, (iii) isothermal period at 0 °C
for 5 min, heating ramp from 0 to 130 °C, (iv) isothermal period at 130 °C for 5 min, (v)
cooling down from 130 to 0 °C, (vi) isothermal period at 0 °C for 5 min, (vii) heating up from
0 to 170 °C, (viii) isothermal stage at 170 °C for 10 min, (ix) cooling down from 170 to 0 °C.
All the heating and cooling steps were carried out with a rate of 10 °C/min. Steps (i) and (ii)
were carried out so all the samples have the same thermal starting point. Steps (iii) to (vi)
were carried out to have “enthalpic relaxation” of the powder, which typically occurs to
amorphous resins during the glass transition [39]. Finally, steps (vii) to (ix) were done to
determine the glass transition of the particles. Within this approach, we heated the samples
above the glass transition temperature during step (iii), but below the curing temperature, then
we did not induce any irreversible change on the powder coating particles. Fig. 3.2 only
shows steps (vii), (viii) and (ix) of the full procedure. The full DSC profile is given in
Supplementary information 3.D.
With DSC, we first evaluated whether the fluidization process would induce any change in the
heat properties of the powder, by measuring the DSC profile of a sample that was fluidized
for 3 hours at the coating temperature, i.e. 27 °C, and compared it with the profile of the
unprocessed powder (Fig. 3.2a). We observed that both uncoated and fluidized samples have
the same glass transition temperature and heat flow profile. Thus we conclude that the
fluidization process does not alter the thermal properties of the powder. In Fig. 3.2b we
compared the normalized heat flow profiles from DSC for the uncoated and coated samples.
We can observe that the curves of all the samples show the step function at the same
temperature, meaning that the glass transition is not affected by the alumina films. Therefore,
we conclude that the alumina films did not act as thermal insulator for the powder, since the
glass transition of all the samples occurred at the same temperature. Then, we investigated the
effect of the alumina films as a physical barrier.
Fig. 3.2. (a) DSC profiles of the untreated and fluidized powder. (b) DSC profiles of the
powder of the untreated sample and the samples coated with 1, 2, 3, 5, 7 and 9 cycles.
A hot stage mounted on a light microscope was used to determine the confinement efficiency
of the alumina films by observing whether the softened powder, above its glass transition
temperature, would flow freely or remain enclosed. For that we placed a few particles of the
samples on a quartz plate which was placed on top the hot stage, with all the system on the
40
Tuning roughness and gloss of powder coating paint by encapsulating the particles with Al2O3 films
microscope stage. We heated the samples from 25 to 70 °C to study the behaviour of the
particles above the with a heating rate of 10 °C/min, similar to the one used in the DSC. In
Fig. 3.3 we compared the uncoated, 1-cycle, 2-cycle, 3-cycle and 9-cycle samples.
Fig. 3.3 Hot stage microscope images of the uncoated powder (a-c) and the samples coated
with 1 (d-f), 2 (g-i), 3 (j-l) and 9 (m-o) cycles, and three temperatures, i.e. 30, 50 and 70 °C.
By studying the uncoated sample under the hot stage microscope (Fig. 3.3a-c), we found that
the softening of the powder coating particles started at about 45 °C, slightly lower than the 50
°C observed in the DSC curves. In the hot stage microscope, there were loses of heat to the
surrounding air since the sample is in contact with the environment, while in DSC the samples
are placed inside an isothermal chamber. That could cause the difference in the measured
41
Chapter 3
softening temperatures in the DSC and the hot stage microscope. Despite that, we used the hot
stage microscope to compare all the samples, which would experience the same heating
process using the same equipment. In order to have a better view of the particles, we
illuminated the samples with a LED light source. This gave extra information by providing a
3D visualization of the particles if we compare them with the pictures taken using the built-in
light of the microscope (Supplementary information 3.E). The bluish colour seen in Fig. 3.3 is
produced by the LED light source. By using the light source from the microscope in
transmission mode, we counted the fraction of the particles that were contained above the
glass transition temperature (Supplementary information 3.E).
The images of the uncoated (Fig. 3.3a-c) and 1-cycle (Fig. 3.3d-f) powder coating particles,
showed a similar behaviour of the particles upon heating. In both samples, the softened
particles flowed at the same temperature. However, a fraction of the sample with 2 cycles
(Fig. 3.3g-i) was contained within the alumina film, similar to what was observed with the 3
cycles sample (Fig. 3.3j-l). The particles of the sample with 9 cycles preserved their shape,
indicating that the alumina coating was able to enclose the softened powder coating paint
(Fig. 3.3m-o). We took images with the transmission light source of the microscope
(Supplementary information 3.E), and found that all the particles of the uncoated and 1-cycle
samples observed under the microscope flowed out once softened, and eventually coalesced.
By investigating about 130 particles per sample, we determined that for the 2-cycle sample,
5% of the particles were retained, while for the 3-cycle sample it was 41%, and for the 5-cycle
sample it was 62% of the particles which remained encapsulated. All the particles from the
samples with 7 and 9 cycles were completely encapsulated, observing no deformation in any
of the particles. The combination of the DSC results, which showed that the glass transition
temperature is not altered by the alumina films, and the one from the hot stage microscope,
which showed that there is a key number of cycles above which the particles remained
encapsulated. We conclude that the alumina films do not act as a thermal insulating layer, but
as a physical barrier preventing the particles from flowing out of the alumina shell.
From the hot stage microscopy, we found that there is a critical number of cycles after which
the alumina film is sufficiently thick and completely closed to contain the softened particles.
We propose two hypotheses: (i) the alumina films are free of defects, and the stress caused
inside the shell by expansion upon softening is higher than the stress that the deposited
alumina shell can endure; and (ii) that a certain number of coating cycles are required to form
a complete alumina film, based on the nucleation of the alumina ALD during the initial cycles
when using polymeric particles as substrate [38, 40].
To calculate the resistance of the alumina films upon an increase of the internal pressure
(eq. 1), we modelled the core-shell particles as a “thin-walled spherical vessel” [41]. For this
model to be valid, the ratio of the film thickness and the particle diameter should be small (<
0.1), which is the case for our particles. By using this approach, we calculated the tensile
stress on the alumina coating (
) caused by the expansion of the core (Eq. 2), and
compared it to the critical tensile stress before cracks appear in the alumina coating (
),
calculated using (Eq. 3). A detailed explanation of these calculations can be found in
Supplementary information 3.F.
(eq. 3.1)
42
Tuning roughness and gloss of powder coating paint by encapsulating the particles with Al2O3 films
(eq. 3.2)
(eq. 3.3)
Here,
is the Sauter mean diameter of the particles,
is the volumetric coefficient of
thermal expansion of the core material,
is the linear coefficient of thermal expansion of the
alumina shell,
is the difference in temperature between the final
and initial state, i.e.
70 and 25 °C, and are the Young modulus and Poisson’s ratio of the aluminium oxide
film, is the compressibility factor of the core material (Table 1), and is the alumina film
thickness. The values of the coefficient of thermal expansion
and the compressibility
factor were taken from literature for a similar resin, since we could not determine these two
parameters for our material with a good accuracy. We performed a sensitivity analysis of
these two variables and observed that they do not have a strong impact on the value of
.
Table. 3.1. Values used in the calculation of the tensile stress
Variable
Units
[1/°C]
[1/°C]
1/MPa
GPa
-6
-6
-4
Value
100·10
5·10
9.2·10
170
Reference
[42]
[43]
[44]
[45]
.
0.24
[45]
To calculate the critical tensile stress of the alumina film
, we used (Eq. 3), where is
the Young modulus of the alumina, and
is the critical strain of alumina films
deposited by ALD under a tensile stress. The values of the critical strain before cracking
depends on the film thickness, as shown by Jen et al [46]. We used the values of the
equivalent film thickness determined by elemental analysis (Fig. 3.1) to calculate both tensile
stresses (
and
).
In Fig. 3.4 we can see that for the film thickness deposited in these coating experiments, the
aluminium oxide films would break under a much larger tensile stress than the one produced
by the expansion of the softened resin. The model predicts that the alumina films would resist
the internal pressure in all the cases, while experimental evidences show that the softened
powder flowed out the alumina shell for the samples with less than 7 coating cycles. Based on
modelling results of Fig. 3.4, we conclude that the onset of flowing is not caused by breaking
of the shell. Rather, samples with a small number of cycles start flowing because a closed
rigid film has not formed yet, allowing fluid to escape as soon as it softens.
This analysis agrees with the nucleation theory of alumina ALD on polymeric substrates [38,
40]. During the first coating cycles, nuclei of alumina penetrate through the surface of the
polymers before making a conformal coating (Fig. 3.5). Several factors promote the formation
of alumina nuclei below the substrate surface, such as the solubility of TMA on hydrophobic
surfaces as in the case of polymeric substrates, the rough surface of the resin particles, and the
free volume near the surface of these materials due to the lack of lattice structure and
crystallinity, which is seen as a porosity that can be accessed by the precursors. Although the
nucleation phenomena is qualitatively understood, there is no rule-of-thumb for the number of
cycles needed to deposit fully conformal films, since this depends on the polymer nature,
43
Chapter 3
preparation method, polymer history, etc. Based on the experiments, it appears that 7 coating
cycles are needed to obtain a conformal and complete aluminium oxide film for our material
(Fig. 3.3 and Supplementary information 3.E and 3.F).
Fig. 3.4. Tensile stress (
) exerted on the alumina-coated samples (blue symbols) and
critical tensile stress (
) of the alumina before breaking (open symbols) for the different
film thickness. The error bars represent the uncertainty in the calculation of the tensile
stresses.
Fig. 3.5. Nucleation mechanism of the deposited alumina. (a) Surface of the untreated
polymer particle. (b) Surface of the polymer particle during the first cycles, in which the
green spheres represent the alumina molecules.
In addition, we performed the pill flow test to study the degree of confinement of the coated
particles. With this test we analyse the flowability of a pill of powder coating paint down an
inclined panel upon heating between room temperature and the curing temperature of the
paint (Fig. 3.6a and b). We observe that even after 1 alumina coating cycle, there is a
reduction of the length of the trail compared to the uncoated sample, and kept decreasing
when increasing the number of cycles. Interestingly, we found that the sample coated with 9
cycles showed a displacement down the inclined panel, while it remained totally encapsulated
in the hot stage microscope. In the preparation of the experiment, a pressure of 5000 psi is
applied to the powder to compact the pill; it is possible that part of the alumina film broke
because of the compression, creating a way out for the softened core and letting it flow down
the panel.
44
Tuning roughness and gloss of powder coating paint by encapsulating the particles with Al2O3 films
The results from the pill flow test are in agreement with the DSC and hot stage microscopy,
showing that the alumina film acts as a physical barrier, preventing the softened particles from
flowing freely. We further investigated the impact of the alumina film in the appearance of
the final product, the powder coating paint.
Fig. 3.6. Pill Flow test panel before (a) and after (b) introducing it in the oven at 160 °C for
10min, which is the curing procedure for this powder. The pills correspond, from left to right,
to the 0, 1, 3, 5, 7 and 9-cycle samples. The pills have a diameter of about 10mm, and 5mm
height.
3.3.2 Characterization of the sprayed panels
To investigate the appearance of the cured powder coating, we initially sprayed the Al 2O3coated particles on the panels using a corona discharge gun. However, the powder did not
stick properly to the aluminium panel, presumably due to a reduction of the charge
transference between the spraying gun and the powder caused by the alumina coating. In
addition, the poor adhesion of the powder coating to the panel and the reduced flowability of
the coated particles resulted in a curing of the paint that lacked proper adhesion to the panel.
These observations suggested a detrimental effect of the alumina films on the curing of the
particles, as observed in the characterization of the coated particles. To overcome the
adhesion problems, we prepared mixtures, with a 1:1 mass ratio, of regular uncoated brown
powder coating particles and the Al2O3-coated white powder coating. We used brown powder
coating to facilitate the observation and characterization of the painted panels. In total, we
sprayed seven panels with the mixtures of brown and white particles, i.e. 0, 1, 2, 3, 5, 7 and 9
cycles, and cured them in the oven at 160 °C for 10 minutes. Visual inspection of the panels
(Fig. 3.7a) indicated a change in the appearance regarding both colour and roughness for the
sample coated with more than 2 cycles. The panels prepared with the uncoated white powder
and the 1-cycle powder (Fig. 3.7a, ⓪ and ①) showed a predominant white colour, while
brown colour dominated the panels prepared with the 2-, 3-, 5-, 7- and 9-cycle samples (Fig.
3.5a, ③, ⑤, ⑦ and ⑨). We illuminated the panels with LED light sources (Fig. 3.7b) in a
dark environment to visualize the reduction of the gloss of the paints due to increased
roughness. The panels ⓪ and ① are smooth, as can be seen by the reflection of the light on
the surface. Increasing number of cycles induced roughness of the surface, shown as a
diffused reflection of the LED lights for the ②, ③, ⑤, ⑦ and ⑨ panels.
45
Chapter 3
Fig. 3.7. (a) Piece of the sprayed panels after curing at 160 °C for 10min, prepared with the
mixture of white powder coated with 0, 1, 2, 3, 5, 7 and 9 cycles, and uncoated brown
powder coating paint (from left to right). The dimensions of the panels shown are 22 x 25
mm. The same panels were illuminated with LED lights in a light environment (b) and dark
environment (c) to show the gradual decrease in the glossiness of the panels with the
thickness of the alumina films that coated the primary particles used in the preparation of the
panels.
The measurements of the surface roughness with a profilometer (Fig. 3.8a) and the gloss of
the paints with a gloss meter (Fig. 3.8b) showed a similar transition to the one in the
glossiness measurements observed in the panel coated with the 2-cycle powder (Fig. 3.7c).
Both measurements agreed: the rougher the surface is, the more light it will scatter, resulting
in a less glossy paint. The value of the surface roughness increased with the number of cycles,
reaching a constant value of about 9 µm after 7 cycles, which corresponds to 1/3 of the
particle size. With the hot stage microscope (Fig. 3.3 and Supplementary Information 3.E) we
found that more particles were successfully contained when increasing the number of cycles.
We observed a similar trend in the surface roughness measurements, which reached a
levelled-off value. The alumina film deposited after 2, 3 and 5 cycles did not succeed to
encapsulate the full particles above the glass transition temperature, so that part of the core
flowed out the alumina shell. On the other hand, the particles coated with 7 and 9 cycles were
completely confined by the alumina film, increasing the surface roughness, where we
presume that the brown softened powder formed a continuous layer in which the coated white
powder is suspended. We observed a similar trend in the values of the gloss, which
significantly dropped for 2 or more cycles, in agreement with the observation of the surface
roughness. Values of the gloss above 60 correspond to a gloss surface, while below 35 the
surface is considered to have a matt finish, according to the internal standards of DSM
Powder Coating Resins. Both surface roughness and gloss measurements agree with other
46
Tuning roughness and gloss of powder coating paint by encapsulating the particles with Al2O3 films
results in which we found that the alumina film after 2 cycles altered the flowing of the single
particles, and the mixing with the brown particles.
Fig. 3.8. (a) Measurement of the roughness of the panels sprayed with a 1:1 mixture of
alumina-coated and uncoated Uralac powder. Error bars represent the standard deviation of
the measurements. (b) Gloss measurements of the panels sprayed with a 1:1 mixture of
alumina-coated and uncoated Uralac powder.
To study the transition in colour and roughness observed for the panels, and take a closer look
at the surface, we used an optical microscope (Fig. 3.9) and SEM-EDX (Fig. 3.10 and 3.11).
With the optical microscope, we observed a difference in the distribution of the cured white
and brown powder coated paint. The panels sprayed with uncoated (Fig. 3.9a) and 1-cycle
powder (Fig. 3.9b) coating particles showed an homogeneous distribution of the brown and
white colour, explained by the good flow of the white powders. The panel prepared with the
3-cycle sample (Fig. 3.9c) did not show a homogenous pattern of the colours, caused by the
partial confinement of the white powder and the consequent decrease of the flow and mixing
between both white and brown powder after softening. To evaluate this panel (Fig. 3.9c) we
modified the incident angle and intensity of the light source to overcome the reflection on the
panel cause by the roughness, creating a yellowish colour. It seems that in Fig. 3.9a and 3.9b
both white and brown particles softened, flowed and mixed homogeneously, creating a
uniform pattern. While Fig. 3.8c suggests that while the brown powder softened, the white
powder partially did, creating a kind of suspension of white grains in a brown uniform matrix.
Fig. 3.9. Images from the optical microscope. Panels sprayed with (a) uncoated powder, (b)
1-cycle coated powder, and (c) 3-cycle coated powder.
47
Chapter 3
Fig. 3.10. SEM pictures using topology mode (BET) (a, c, e) and composition mode (BEC)
(b, d, f) of the panels prepared with uncoated particles (a) and (b), 3-cycle particles (c) and
(d), and 9-cycle particles (e) and (f). The duplet of pictures for each sample represent the
same area of the panel.
To investigate the formation of the roughness, we evaluated the sprayed panels with the
combination of SEM (Fig. 3.10) and EDX (Fig. 3.11). We used two different SEM modes to
look at the surface: Back-scattering Topology BET, (Fig. 3.10a, 3.10c and 3.10e), and Backscattering Composition BEC, (Fig. 3.10b, 3.10d and 3.10f).
We observed that the panels sprayed with the uncoated particles (Fig. 3.10a) did not show
roughness in the BET mode, in agreement with the visual observation of the panels (Fig. 3.7).
Nevertheless, the BEC-mode picture revealed darker and brighter areas (Fig. 3.10b). As we
expected, BET-mode images of the 3-cycle (Fig. 3.10c) and the 9-cycle (Fig. 3.10e) panels
confirmed the presence of surface roughness, which are related with the location of the darker
areas observed in the corresponding BEC-mode pictures (Fig. 3.10d and 3.10f). We used
EDX analysis (Fig. 3.11) to examine the composition of the darker and brighter areas in the
BEC-mode images, and to determine how they relate with the roughness in the BET-mode
images.
We performed EDX analysis in the full area shown in the images of Fig. 3.11, and performed
spot analysis in the bright (①) and darker (②) areas. We found that the brighter areas (white
bars in Fig. 3.11) showed a higher concentration of Ba and S, while the darker areas (black
bars in Fig. 3.11) had a dominant concentration of Ti. That relates to the presence of BaSO 4,
which is the filler in the brown powder coatings, and white pigment TiO2 used in the
preparation of the white powder coatings. In agreement, the concentration of aluminium is
higher in the darker areas, which correspond to the Al2O3-coated white powder. Other
compounds identified by EDX analysis, such as C, O and N, are not included in Fig. 3.10
since they were common to all the panels. The combination of the SEM (Fig. 3.10), which
related the presence of darker and brighter areas with the roughness, and the EDX analysis
(Fig. 3.11), which determined the presence of white pigment and the filler used in the brown
powder in the darker and brighter areas respectively, evidences that the roughness is caused
by the Al2O3-coated particles, and thus, the reduction in the gloss of the paints.
48
Tuning roughness and gloss of powder coating paint by encapsulating the particles with Al2O3 films
Fig. 3.11. SEM images using BET mode for the samples sprayed with 0 cycles (a), 1 cycles
(d) and 3 cycles (g), and BEC mode for the samples with 0 cycles (b), 1 cycles (e) and 3
cycles (h). (c), (f) and (i) shows the EDX analysis of the full image (green bars), spot 1 (red
bars) and spot 2 (blue bars).
Finally, we evaluated the mechanical resistance of the paints prepared with the Al2O3-coated
powder coating particles using the reverse impact test (Supplementary information 3.G). After
the impact of the steel ball, we used a stereomicroscope to observe the cracks on the surface
of the panel (Fig. 3.12). The contrast and brightness of the pictures in Fig. 3.12 were modified
to have a better visualization of the formation of the cracks. In the centre of the images (Fig.
3.12) we can see the deformation of the panel after the impact, on which the cracks would
appear. We found that all the panels showed a good impact resistance.
From the results obtained, we conclude that powder coating particles coated with only 2
cycles of alumina induced a granular textured paint, characterized by a rough and matte
surface finish with enough surface resistance. Nevertheless, this product could be further
developed in the preparation of the alumina-coated powder and in the production of the
painted panels. For instance, the dosing and purging time, in combination with the number of
cycles, during the alumina coating could be further optimized towards a more time-efficient
process. Likewise, variables such as the particle size and the ratio of mixture of the coated-touncoated powder would influence the appearance, i.e. roughness, gloss and colour, and the
mechanical resistance of the paint. Nevertheless, this work did not aim at obtaining an
industrial solution for the production of a matte powder coating paint, but at proving the
applicability of gas-phase techniques on particles, such as ALD or CVD, to modify or
49
Chapter 3
improve the properties of industrial products, as in this case, where we induced roughness
while maintaining the good mechanical properties of a DSM standard powder coating paint.
Fig. 3.12. Reverse Impact test, performed over the painted panels with 0, 1, 3, 5, 7 and 9
cycles. The thickness of the paint was around 50 µm in all the samples. The contrast and
brightness of the pictures were modified for a better visualization of the cracks.
50
Tuning roughness and gloss of powder coating paint by encapsulating the particles with Al2O3 films
3.4 Conclusions
We showed that a thin coating of aluminium oxide deposited on particles of a standard dry
powder coating paint delayed or even completely suppressed fluid-like flow above the glass
transition temperature. As a result, the appearance of the final paint layer could be tuned
between a gloss finish and a matte one, depending on the pre-treatment of the paint particles.
We deposited these alumina films, with different thicknesses, on the powder coating particles
in a fluidized bed reactor at 27 °C and 1 bar, using trimethylaluminium and water as
precursors, in a process similar to ALD with a growth of about 3.5 nm of Al 2O3 per cycle. We
found that 2 coating cycles, which deposited an amount of alumina equivalent to a 6 nm
average film thickness, was sufficient to alter the flow of the particles and the appearance of
the paint. More cycles resulted in thicker alumina coatings and further modified the paint
behaviour, whereas more than 5 cycles had little additional effect. The suppression of flow
was not caused by delayed melting due to thermal insulation by the alumina films, as
determined by differential scanning calorimetry. Rather, the film acts as a hard physical shell
that prevents material release from the core, that softens irrespective of the coating around it.
The thickness of the film is not important as long as it covers the entire particle. We
calculated that the thinnest shell is strong enough to contain the core, even when the core
softens and expands due to heating, using an order-of-magnitude analysis in a thin-wall
spherical-vessel model. We found, after interpretation of the results from SEM and EDX, that
Al2O3 can readily alter the flowing behaviour and induce roughness in paints, while keeping
the mechanical resistance comparable to the reference paints.
Bibliography
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
Verlaak, J., Broader scope for powder coatings: Resin developments benefit
applicators, end users and the environment. Eur. Coat. J. 2009, (12), 86-89.
Misev, T. A.; van der Linde, R., Powder coatings technology: new developments at
the turn of the century. Progress in Organic Coatings 1998, 34, (1–4), 160-168.
Spyrou, E., Powder Coatings Chemistry and Technology. Vincentz Network: 2013.
Jacobs, W.; Foster, D.; Sansur, S.; Lees, R. G., Durable glossy, matte and wrinkle
finish powder coatings crosslinked with tetramethoxymethyl glycoluril. Progress in
Organic Coatings 1996, 29, (1–4), 127-138.
Biris, A. S.; Mazumder, M. K.; Yurteri, C. U.; Sims, R. A.; Snodgrass, J.; De, S.,
Gloss and texture control of powder coated films. Particulate Science and Technology
2001, 19, (3), 199-217.
DSM Powder Coating Resins, DSM Uralac P 3210 Technical Data Sheet [accessed on
2015 June 5]. www.dsmcoatingresins.com
Sveshnikova, G. V.; Kol'Tsov, S. I.; Aleskovskii, V. B., Formation of a silica layer of
predetermined thickness on silicon by the Molecular-Layering method. Zh Prikl Khim
1970, 43, (5), 1150-1152.
Suntola, T., Atomic layer epitaxy. Materials Science Reports 1989, 4, (5), 261-312.
George, S. M., Atomic layer deposition: An overview. Chem. Rev. 2010, 110, (1),
111-131.
Miikkulainen, V.; Leskelä, M.; Ritala, M.; Puurunen, R. L., Crystallinity of inorganic
films grown by atomic layer deposition: Overview and general trends. Journal of
Applied Physics 2013, 113, (2).
51
Chapter 3
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
Hakim, L. F.; George, S. M.; Weimer, A. W., Conformal nanocoating of zirconia
nanoparticles by atomic layer deposition in a fluidized bed reactor. Nanotechnology
2005, 16, (7), S375-S381.
Ferguson, J. D.; Weimer, A. W.; George, S. M., Atomic layer deposition of ultrathin
and conformal Al2O3 films on BN particles. Thin Solid Films 2000, 371, (1), 95-104.
Profijt, H. B.; Potts, S. E.; Van De Sanden, M. C. M.; Kessels, W. M. M., Plasmaassisted atomic layer deposition: Basics, opportunities, and challenges. J. Vac. Sci.
Technol. A Vac. Surf. Films 2011, 29, (5).
Potts, S. E.; Kessels, W. M. M., Energy-enhanced atomic layer deposition for more
process and precursor versatility. Coordination Chemistry Reviews 2013, 257, (23–
24), 3254-3270.
Valdesueiro, D.; Meesters, G.; Kreutzer, M.; van Ommen, J., Gas-Phase Deposition of
Ultrathin Aluminium Oxide Films on Nanoparticles at Ambient Conditions. Materials
2015, 8, (3), 1249-1263.
Kunii, D.; Levenspiel, O., Fluidization Engineering. Second ed.; ButterworthHeinemann: 1991.
van Ommen, J. R.; Valverde, J. M.; Pfeffer, R., Fluidization of nanopowders: a
review. Journal of Nanoparticle Research 2012, 14, (3), 1-29.
Hakim, L. F.; Blackson, J.; George, S. M.; Weimer, A. W., Nanocoating individual
silica nanoparticles by atomic layer deposition in a fluidized bed reactor. Chemical
Vapor Deposition 2005, 11, (10), 420-425.
Goulas, A.; van Ommen, J. R., Scalable production of nanostructured particles using
atomic layer deposition. KONA Powder and Particle Journal 2014, 31, (1), 234-246.
Longrie, D.; Deduytsche, D.; Detavernier, C., Reactor concepts for atomic layer
deposition on agitated particles: A review. Journal of Vacuum Science &amp;
Technology A 2014, 32, (1), -.
Moghtaderi, B.; Shames, I.; Doroodchi, E., Combustion prevention of iron powders by
a novel coating method. Chemical Engineering and Technology 2006, 29, (1), 97-103.
Hakim, L. F.; Vaughn, C. L.; Dunsheath, H. J.; Carney, C. S.; Liang, X.; Li, P.;
Weimer, A. W., Synthesis of oxidation-resistant metal nanoparticles via atomic layer
deposition. Nanotechnology 2007, 18, (34).
King, D. M.; Liang, X.; Burton, B. B.; Kamal Akhtar, M.; Weimer, A. W., Passivation
of pigment-grade TiO2 particles by nanothick atomic layer deposited SiO2 films.
Nanotechnology 2008, 19, (25).
Liang, X.; King, D. M.; Groner, M. D.; Blackson, J. H.; Harris, J. D.; George, S. M.;
Weimer, A. W., Barrier properties of polymer/alumina nanocomposite membranes
fabricated by atomic layer deposition. Journal of Membrane Science 2008, 322, (1),
105-112.
Liang, X.; Lynn, A. D.; King, D. M.; Bryant, S. J.; Weimer, A. W., Biocompatible
interface films deposited within porous polymers by atomic layer deposition (ALD).
ACS Applied Materials and Interfaces 2009, 1, (9), 1988-1995.
Lichty, P.; Liang, X.; Muhich, C.; Evanko, B.; Bingham, C.; Weimer, A. W., Atomic
layer deposited thin film metal oxides for fuel production in a solar cavity reactor.
International Journal of Hydrogen Energy 2012, 37, (22), 16888-16894.
Lichty, P.; Wirz, M.; Kreider, P.; Kilbury, O.; Dinair, D.; King, D.; Steinfeld, A.;
Weimer, A. W., Surface modification of graphite particles coated by atomic layer
deposition and advances in ceramic composites. International Journal of Applied
Ceramic Technology 2013, 10, (2), 257-265.
Valdesueiro, D., P. Garcia-Trinanes, G.M.H. Meesters, M.T. Kreutzer, J. Gargiuli, T.
Leadbeater, D.J. Parker, J. Seville, J.R. van Ommen, Enhancing the activation of
52
Tuning roughness and gloss of powder coating paint by encapsulating the particles with Al2O3 films
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
silicon carbide tracer particles for PEPT applications using gas-phase deposition of
alumina at room temperature and atmospheric pressure. (submitted for publication)
2015.
Beetstra, R.; Lafont, U.; Nijenhuis, J.; Kelder, E. M.; Van Ommen, J. R., Atmospheric
pressure process for coating particles using atomic layer deposition. Chemical Vapor
Deposition 2009, 15, (7-9), 227-233.
Puurunen, R. L., Correlation between the growth-per-cycle and the surface hydroxyl
group concentration in the atomic layer deposition of aluminum oxide from
trimethylaluminum and water. Applied Surface Science 2005, 245, (1–4), 6-10.
Puurunen, R. L., Growth Per Cycle in Atomic Layer Deposition: A Theoretical Model.
Chemical Vapor Deposition 2003, 9, (5), 249-257.
Laubengayer, A. W.; Gilliam, W. F., The Alkyls of the Third Group Elements. I.
Vapor Phase Studies of the Alkyls of Aluminum, Gallium and Indium1. Journal of the
American Chemical Society 1941, 63, (2), 477-479.
Henrickson, C. H.; Eyman, D. P., Lewis acidity of alanes. Interactions of
trimethylalane with sulfides. Inorganic Chemistry 1967, 6, (8), 1461-1465.
Mayer, B.; Collins, C. C.; Walton, M., Transient analysis of carrier gas saturation in
liquid source vapor generators. Journal of Vacuum Science & Technology A 2001, 19,
(1), 329-344.
Jur, J. S.; Parsons, G. N., Atomic Layer Deposition of Al2O3 and ZnO at Atmospheric
Pressure in a Flow Tube Reactor. ACS Applied Materials & Interfaces 2011, 3, (2),
299-308.
Salameh, S.; Schneider, J.; Laube, J.; Alessandrini, A.; Facci, P.; Seo, J. W.; Ciacchi,
L. C.; Mädler, L., Adhesion mechanisms of the contact interface of TiO 2
nanoparticles in films and aggregates. Langmuir 2012, 28, (31), 11457-11464.
Mousa, M. B. M.; Oldham, C. J.; Jur, J. S.; Parsons, G. N., Effect of temperature and
gas velocity on growth per cycle during Al 2O 3 and ZnO atomic layer deposition at
atmospheric pressure. J. Vac. Sci. Technol. A Vac. Surf. Films 2012, 30, (1), 01A155.
Wilson, C. A.; Grubbs, R. K.; George, S. M., Nucleation and growth during Al2O3
atomic layer deposition on polymers. Chemistry of Materials 2005, 17, (23), 56255634.
Bailey, N. A.; Hay, J. N.; Price, D. M., A study of enthalpic relaxation of
poly(ethylene terephthalate) by conventional and modulated temperature DSC.
Thermochimica Acta 2001, 367–368, (0), 425-431.
Liang, X.; Hakim, L. F.; Zhan, G. D.; McCormick, J. A.; George, S. M.; Weimer, A.
W.; Spencer Ii, J. A.; Buechler, K. J.; Blackson, J.; Wood, C. J.; Dorgan, J. R., Novel
processing to produce polymer/ceramic nanocomposites by atomic layer deposition.
Journal of the American Ceramic Society 2007, 90, (1), 57-63.
Fenner, R. T.; Reddy, J. N., Mechanics of Solids and Structures, Second Edition. CRC
Press: 2012.
Zarrelli, M.; Skordos, A. A.; Partridge, I. K., Investigation of cure induced shrinkage
in unreinforced epoxy resin. Plastics, Rubber and Composites 2002, 31, (9), 377-384.
Miller, D. C.; Foster, R. R.; Jen, S.-H.; Bertrand, J. A.; Cunningham, S. J.; Morris, A.
S.; Lee, Y.-C.; George, S. M.; Dunn, M. L., Thermo-mechanical properties of alumina
films created using the atomic layer deposition technique. Sensors and Actuators A:
Physical 2010, 164, (1–2), 58-67.
Hatzikiriakos, S. G.; Dealy, J. M., Start-up pressure transients in a capillary rheometer.
Polymer Engineering and Science 1994, 34, (6), 493-499.
Tripp, M. K.; Stampfer, C.; Miller, D. C.; Helbling, T.; Herrmann, C. F.; Hierold, C.;
Gall, K.; George, S. M.; Bright, V. M., The mechanical properties of atomic layer
53
Chapter 3
46.
47.
deposited alumina for use in micro- and nano-electromechanical systems. Sensors and
Actuators A: Physical 2006, 130–131, (0), 419-429.
Jen, S.-H.; Bertrand, J. A.; George, S. M., Critical tensile and compressive strains for
cracking of Al2O3 films grown by atomic layer deposition. Journal of Applied Physics
2011, 109, (8), 084305.
Groner, M. D.; Fabreguette, F. H.; Elam, J. W.; George, S. M., Low-Temperature
Al2O3 Atomic Layer Deposition. Chemistry of Materials 2004, 16, (4), 639-645.
54
Tuning roughness and gloss of powder coating paint by encapsulating the particles with Al2O3 films
3.A Powder coating paint characterization
3.A.1 Particle Size Distribution
To determine the particle size, which is used later in the calculation of the alumina-film
thickness, we used a Beckman Coulter LS 230 Diffraction Particle Size Analyser. We
obtained a value for the Sauter mean diameter (
) of 33 µm (Fig. 3.A.1).
Fig. 3.A.1. PSD of DSM Uralac ® powder coating paint used in the ALD experiments
3.A.2 Composition
This is the composition of the powder coating used in the Al2O3-coating experiment [6].
Table 3.A.1. Composition of the powder coating paint used in the coating experiments.
Component
Mass percentage [%]
DSM Uralac® P 3210 resin
62.4
3.3
HAA ( -Hydroxyalkylamide) cross-linker
TiO2 white pigment
33
Flow control agent
1
Degassing agent
0.3
3.A.3 Density calculation
We measured the density of the powder using a 25 mL volumetric flask, and ethanol with
known density. We placed a weighted mass of powder, and filled the flask with ethanol. By
subtracting the added ethanol volume to the flask volume, we determined the volume
occupied by the amount of powder placed inside the flask. We performed error propagation to
determine the uncertainty of this measurement, obtaining a density for the powder of 1500 ±
100 kg/m3. The details of the calculation are shown in Table 3.A.2.
value
error
Table 3.A.2. Calculation of the density of the powder.
V flask [mL]
mass powder [g]
V EtOH [mL]
25
19.2
12.2
0.5
0.5
0.5
55
[kg/m3]
1497
92
Chapter 3
3.B Characterization techniques
To calculate the Al2O3 film thickness, we measured the mass fraction of aluminium on the
coated samples ( ) with a PerkinElmer Optima 5300 Induce Coupled Plasma – Optical
Emission Spectroscope. We performed DSC measurements on the coated samples using a TA
Instruments DSC Q-20 device. To evaluate the confinement of the Al2O3-coated particles, we
placed several particles on a Metler ToledoFP82 HT Hot Stage to heat up the sample above
the glass transition temperature of the powder. The heating procedure, which is controlled
with a Metler Toledo FP90 Central Processor, consisted of a ramp-up of 10 °C/min from
room temperature to 70 °C, followed by an isothermal state at 70 °C for 3 minutes before the
samples were brought back to room temperature. The hot stage was placed under a Bresser
MPO-401 light microscope using a 4x magnification objective to observe the particles. We
performed a Pill Flow test as a measurement of the curability of the powder. For that, we
applied a pressure of 5000 psi to prepare pills with the uncoated and Al2O3-coated coated
samples, and placed them on an aluminium Q-panel. This panel was introduced in the oven
with an inclination of 45°, and heated up to 170 °C, which is the curing temperature. As the
temperature increased above the glass transition, the particles started to flow down the panel,
until curing. The comparison was based on the measurement of the trail of the pills when
flowing down the panel, which indicated whether the alumina coating influenced the curing of
the powder.
The sprayed panels with the mixture of uncoated-coated Uralac were evaluated under a Leica
EZ4 optical microscope, and a JEOL JSM 6010LA SEM microscope equipped with EDX
detector to determine the surface topology and its composition. Furthermore, we measured the
roughness of the panels using a VECCO DEKTAK8 Stylus Profilometer with a scanning force
of 1.0mg, and a scan speed of 4mm/s. We used a Byk-Gardner haze-gloss meter to measure
the gloss at 20 and 60° according to both standards ASTM-D-523/70 and ISO 2813. Finally
we studied the resistance of the panels under the impact of a 0.45 kg ball in the reverse of the
painted panel falling from 1.5m height. The resistance and flexibility of the powder coating
paint, and how well it is cured, were assessed by the number of cracks that appeared on the
panels for two different paint thickness: 50 µm and 75 µm. We evaluated the area of the
impact on the panels with a Zeiss Discovery V8 stereomicroscope. The appearance of crack
qualified whether the powder coating paint produced with the alumina-coated primary
particles showed the same resistance as the standard Uralac powder coating paint.
56
Tuning roughness and gloss of powder coating paint by encapsulating the particles with Al2O3 films
3.C Calculation of the alumina film thickness
This calculation is based on the measurements of the mass fraction of aluminium ( ) on the
coated samples, obtained from the ICP-OES. From this value, we calculated the volume of
aluminium oxide per particle (
) related to the mass fraction of aluminium.
Subsequently, we calculated the thickness of the alumina film
considering spherical
particles. To account for uncertainty in the film thickness
, we determined the error
considering the error in the particle size
, the aluminium mass fraction
, the density
of the powder
and the density of the aluminium oxide
. We used the
equations below to calculate the film thickness , and its error
. The error for the mass
fraction
is 3% of the value measured from ICP-OES, for the particle size
is 15 μm,
3
for the density of the powder
is 100 kg/m , and for the density of the alumina is 200
kg/m3. The density of the alumina
used to calculate the film thickness is 2500 kg/m3
[47]. More details of the calculation are given in [15].
(eq. 3.C.1)
(eq. 3.C.2)
√(
)
(
Table 3.C.1. Results
characterization
# ALD %Al ICP
cycles
[%w]
0
0.64 ± 0.02
0.67 ± 0.03
1
0.73 ± 0.09
2
0.85 ± 0.21
3
0.84 ± 0.20
5
1.13 ± 0.49
7
1.10 ± 0.46
9
)
from
(
)
(
)
(eq. 3.C.3)
ICP-OES
δAl2O3
[nm]
0
1.9 ± 0.9
5.7 ± 2.6
13.3 ± 6.1
12.7 ± 5.8
31.1 ± 14.4
29.2 ± 13.5
Fig. 3.C.1. Calculation of the thickness of
Al2O3 for different number of cycles. The
error bars indicate the uncertainty of the
measurement.
57
Chapter 3
3.D DSC measurements
We performed the DSC measurements using a multiple-step heating/cooling process to
prevent that the entalphic relaxation of the powder could influence the interpretation of the
results for the glass transition temperature. Fig. 3.D.1 (left) shows the full heating/cooling
process, while Fig. 3.D.1 (right) only shows the last heating and cooling cycle. The figures
shown in the manuscript represent the last heating cooling cycle, similar to the one plotted in
Fig. 3.D.1 (right).
Fig. 3.D.1. DSC profiles for the uncoated samples, showing the full heating/cooling process
(left), and the reduced heating/cooling (right).
58
Tuning roughness and gloss of powder coating paint by encapsulating the particles with Al2O3 films
3.E Hot Stage Microscopy images.
59
Chapter 3
Fig. 3.E.1. Hot stage pictures before (25 °C) and after heating, for the samples
with 0, 1, 2, 3, 5 and cycles. We observe a difference for the samples with 2 and
less cycles, which melted and coalesced, and the samples with 3 and more
cycles, where most of the particles showed resistance to temperature.
Table 3.E.1. Summary of the hot stage analysis of the samples.
# cycles
0
1
2
3
5
7
N particles analysed [-]
136 130 120 140 137 156
T max [°C]
70
70
70
70
70
70
Contained particles [%] 0
0
5
41
62
97
60
9
136
70
100
Tuning roughness and gloss of powder coating paint by encapsulating the particles with Al2O3 films
3.F Calculation of the tensile stress of the alumina film
3.F.1 Equations
We base this calculation in the book Mechanics of Solids and Structures, by Roger T. Fenner
and J.N. Reddy [41], by modelling a core-shell particle as a thin-walled spherical vessel.
Fig. 3.F.1. Model proposed with the force balance to calculate the tensile stress of the film.
∫
∫
(eq. 3.F.1)
(eq. 3.F.2)
Developing the equation above, and assuming that
, we obtain:
(eq. 3.F.3)
(eq. 3.F.4)
For the strain of the shell of a thin-walled spherical vessel, we have two different
contributions: radial direction ( ) and circumferential direction ( ). Similarly, the stress has
the same two contributions ( and ).
[
]
(eq. 3.F.5)
[
]
(eq. 3.F.6)
The strain in radial direction
only provides information about how much the thickness
diminishes under the radial stress . However, the
value is in the order of the inner
61
Chapter 3
pressure, and therefore, very small compared to , which is proportional to the ratio
According to this, we rewrite the circumferential strain as:
.
(eq. 3.F.7)
We assume that the strain of the polymeric core
will equal the strain of the alumina
shell
. The strain of the core have two components: the thermal expansion and the
compressibility of the liquid. On the other hand, the alumina shell will have a component
related to the deformation induced by the stress based on the pressure
which depends
on the film material, and a component that shows the thermal expansion of the film, knowing
that due to geometrical approximation
.
(eq. 3.F.8)
(eq. 3.F.9)
(eq. 3.F.10)
(eq. 3.F.11)
(eq. 3.F.12)
Where
is the difference in temperature between the final
and initial
state, is the
particle diameter, is the increase in pressure between the inside and outside of the coreshell, is the alumina wall thickness,
is the volumetric coefficient of thermal expansion of
the core material,
is the linear coefficient of thermal expansion of the alumina shell, and
are the Young’s modulus and Poisson’s ratio of the aluminium oxide film, and is the
compressibility factor of the core material.
Once we have calculated the stress that results from the expansion of the softened resin, we
can compare the values with the ones from the critical tensile stress before breaking. We can
calculate these values using the values of the critical strain of alumina films
with
different thickness and the Young modulus of the ALD alumina films.
(eq. 3.F.13)
62
Tuning roughness and gloss of powder coating paint by encapsulating the particles with Al2O3 films
F.2
Error propagation
We estimate the uncertainty of the previous calculations.
(eq. 3.F.14)
√
(eq. 3.F.15)
(eq. 3.F.16)
[
]
[
[
]
]
(eq. 3.F.17)
[
]
(eq. 3.F.18)
Critical tensile stress that the alumina films can withstand before breaking. The values of the
critical strain (figure below) were reported by Jen et al. 2011 [46].
(eq. 3.F.19)
√
(eq. 3.F.20)
√
(eq. 3.F.21)
√
√
63
(eq. 3.F.22)
Chapter 3
To calculate whether the alumina shell would break or not, we use the following equation for
the breakage probability
:
(eq. 3.F.23)
(eq. 3.F.24)
√
(eq. 3.F.25)
(eq. 3.F.26)
Table 3.F.1. Calculated values of the tensile stresses
# cycles
[nm]
[bar]
[bar]
[%] [46]
[bar]
[-]
Does it break?
1
1.9
0.7
3150 ± 60
3.0
50800 ± 5250
0.06 ± 0.01
no
2
5.7
2.2
3130 ± 70
2.3
38300 ± 5250
0.08 ± 0.01
no
3
13.3
5.0
3080 ± 90
1.7
28700 ± 5250
0.11 ± 0.02
no
64
5
12.7
4.7
3080 ± 80
1.7
29250 ± 5250
0.11 ± 0.02
no
7
31.1
11.2
2960 ± 140
1.1
19000 ± 5250
0.16 ± 0.04
no
9
29.2
10.5
2970 ± 130
1.2
19700 ± 5250
0.15 ± 0.04
no
Tuning roughness and gloss of powder coating paint by encapsulating the particles with Al2O3 films
3.G Results of several techniques
3.G.1 Pill Flow Test
Table 3.G.1. Length of the “tongue” of
the pill flow test for the different
number of cycles.
Sample Flow
length
[mm]
0 cycles
1 cycle
3 cycles
5 cycles
7 cycles
9 cycles
99
83
71
72
51
53
Fig. 3.G.1. Pill flow test results.
3.G.2 Surface profilometry
Fig. 3.G.2. Representation of the areas of the samples that were measured with the
profilometre.
Table 3.G.2. Roughness from the profilometre
ALD
N areas
Roughness
cycles
[-]
[µm]
0
4
0.3 ± 0.1
1
7
0.7 ± 0.2
3
30
4.1 ± 1.2
5
10
5.4 ± 0.8
7
17
10.1 ± 2.7
9
17
8.8 ± 2.8
65
Fig. 3.G.3. Surface roughness of the
samples
Chapter 3
3.G.3 Reverse impact test
Table 3.G.3. Number of cracks appeared after the impact test.
# cycles
thickness [µm]
number cracks
52
0
0 cycles
64
0
50
2
1 cycles
68
>25
60
2
3 cycles
82
>25
56
0
5 cycles
82
0
60
0
7 cycles
76
>25
48
0
9 cycles
80
<25
66
Chapter 4
The deposition mechanism of Al2O3
on QD films at atmospheric pressure and
room temperature
Stability of quantum dot (QD) films is an issue of concern for applications in devices such as
solar cells, LEDs and transistors. This paper analyzes and optimizes the passivation of such
QD films using atomic layer deposition (ALD), resulting in enhanced optical stability.
Crucially, we deposited alumina at economically attractive conditions, room temperature and
atmospheric pressure, on PbSe EDA (1,2-ethanediamine) capped QD films using
trimethylaluminum (TMA) and water as precursors. We performed coating experiments from
1 to 25 cycles on the QD films, finding that alumina formed from the first exposure of TMA,
as determined by X-ray photoelectron spectroscopy (XPS) depth profiling measurements.
XPS points to the presence of oxygen-rich compounds on the bare QD films, most likely from
entrapped solvent molecules during the dip-coating assembly of the QD films. These
oxygenated compounds and the amine groups of the organic ligands act as active sites to react
with TMA in the first cycle, and this results in a fast growth of alumina. Using 10 cycles of
ALD resulted in infilling (between the QDs) and overcoating (a layer on top of the QDs), to
produce a product that was optically stable for at least 27 days. We also investigated the
deposition of alumina with synthetic air as an alternative for the use of water as precursors.
We found that the absence of physisorption of precursor molecules yielded a thinner coating,
and the QD film oxidized during the first exposure to synthetic air. Depositing alumina on the
QD films at ambient conditions is preferred, since the deposition of alumina occurs faster, and
the production of the QD films is also carried out at room temperature and atmospheric
pressure, allowing an interesting combination of both processes in a single go.
Submitted as: D. Valdesueiro, M. Krishna Prabhu, C. Guerra-Nuñez, C.S. Suchand Sandeep,
Sachin Kinge, L.D.A. Siebbeles, L.C.P.M. de Smet, G.M.H. Meesters, M.T. Kreutzer, A.J.
Houtepen, J.R. van Ommen. The deposition mechanism of aluminium oxide on quantum do
films at atmospheric pressure and room temperature.
69
Chapter 4
4.1 Introduction
This paper describes, analyzes and optimizes a passivation process for assembled films of
colloidal quantum dots (QD). Such assembled QD films have come to the fore in potential
applications for photovoltaic devices [1-7] due to their size-tunable band gap and ease of
processing [4, 8]. In particular, colloidal QDs exhibit carrier multiplication [8-18], a process
in which a high energy photon creates more than one charge carrier pair. That boosts the
efficiency in applications such as photovoltaic devices, LEDs and transistors.
In optoelectronic devices, QDs are immobilized as films and separated by organic ligands that
are bound to the QD surface. These ligands prevent sintering of the crystals and ensure
electronic contact between the QDs [19, 20], but they also make the assembled structure more
open, allowing oxygen to diffuse inside the film easily and degrade the crystals by oxidation.
To begin with the positive aspects, the tunable physical separation between crystals impacts
the absorption spectrum to allow the so-called band-tuning, while the electronic contact
improves the charge-carrier mobilities. Good results have recently been obtained by using
films of QDs capped with EDA (1,2-ethanediamine) and EDT (1,2-ethanedithiol) for
applications in optoelectronic materials [21-24]. Returning to the adverse effect of reduced
stability: the organic ligands act as separators, inducing voids inside the QD films that create a
pathway for air to diffuse through and degrade the crystals irreversibly. In fact, the
susceptibility to air oxidation [21], Ostwald ripening from solid-state diffusion [25], nonpassivated surface states and the significant tunneling barriers [19, 20] found in these films
have been the major challenges that have frustrated the use in working devices.
The open space created by the separating ligands also allows other molecules to diffuse into
the film, and this feature is used in this paper: precursor molecules for coating the QDs can
penetrate easily, such one can coat deep layers of QDs in the assembled film without diffusion
limitations just as well as the outer layers of QDs close to the surface [26, 27]. Recently,
atomic layer deposition (ALD) of aluminium oxide (a.k.a. alumina; i.e., Al2O3), carried out at
25–75 °C and ~0.15 Torr, was used to infill and overcoat PbSe QD films and stabilize them
against air degradation [15, 28]. ALD of alumina involves cyclic, self-terminating, gas-phase
reactions between trimethylaluminium (TMA) and an oxidizer, such as oxygen or water, to
deposit thin and conformal films [29-31].
Crucially, we modified this ALD process to work at economically much more attractive
conditions, room temperature and atmospheric pressure [32, 33]. Typically, Al2O3 ALD is
carried out at near vacuum to ensure the removal of unreacted species from the substrate [3436]. Obviously, avoiding vacuum would be attractive, and a room-temperature process would
be completely compatible with the fabrication of the QD films, also done at ambient
conditions [4]. The potential problem that requires analysis and optimization is the
physisorption of unreacted species on the substrate, inducing a less precise chemical vapor
deposition (CVD) type of growth characterized by the deposition of several layers of alumina
in each cycle [33, 34, 37]. We show in this paper that indeed occurs, but that this not reduce
the capacity of the alumina in enhancing stability. Replacing the liquid precursors, i.e., water,
by gaseous ones, i.e., air, would reduce the accumulation of unreacted species on the QD
films, providing a more precise control over the deposition of ultrathin films of alumina.
In previous work by Law and co-workers [15, 22, 28, 38], the deposition of alumina on QD
films successfully prevented oxidation and photo-thermal degradation, allowing direct use in
open air. According to reference [28], the alumina also acted as a matrix to prevent solid-state
70
The deposition mechanism of Al2O3 on QD films at atmospheric pressure and room temperature
diffusion of atoms, that reduced the tunneling barrier for charge transfer between the quantum
dots and passivated the surface charge traps, thereby enhancing the charge mobility by one
order of magnitude. However, despite the excellent performance of Al2O3-ALD on QD films
as passivating material, there is a lack of studies on the deposition mechanism of alumina
inside the films. Recent work done in the field has triggered interest into understanding the
process of alumina deposition in the pores of quantum dot films [28]. Parameters such as the
combined effect of temperature and pressure on the deposition of alumina, and the minimum
amount of alumina that is required to fully protect the QD films from degradation, have not
been investigated yet. Also, little is known about the growth of alumina in the pores of the QD
film during the ALD process.
Here we demonstrate the deposition of alumina inside EDA capped PbSe QD films at room
temperature and atmospheric pressure. The first cycles lead dominantly to infilling, i.e., the
alumina grows between the QDs, and when the interstitial space between the QDs is filled,
additional cycles lead to overcoating of the films, i.e., the formation of a layer of alumina on
top of the QD film. For this aim, we investigated (i) the amount of alumina deposited at
ambient conditions after different number of cycles, (ii) the deposition mechanism of alumina
inside the films, (iii) the air-stability of QD films coated with different number of cycles, and
(iv) the deposition of alumina using a diluted flow of dry air after 25 cycles and its stability
under air. We observed the infilling of alumina from the first dosage of TMA and explain this
by the reactivity of TMA molecules with hydroxyl and amine groups present as remaining of
the solvent used during the sip-coating of the QD film and the organic ligands, respectively.
Furthermore, we also demonstrate that efficient infilling with Al2O3 ALD at ambient pressure
and room temperature is possible, significantly enhancing the speed and ease of the process.
This could be combined with the fabrication process of the QD films, also done at ambient
conditions, to develop a continuous process in which the QD films would be produced and
passivated in a one-go process, using a similar technology as the spatial ALD reactors [3941].
4.2 Experimental
4.2.1 PbSe QD solution
To prepare the QD solution, we used the following reactants: PbO (99.999%), oleic acid
(technical grade, 90%), 1-octadecene (technical grade, 90%), diphenylphosphine (98%),
anhydrous 1-butanol (98%), methanol (99.8%), n-hexane (95%), acetonitrile (99.8%) and
tetrachloroethylene (TCE, >99%), which were purchased from Sigma-Aldrich,
trioctylphosphine (TOP, >90%) and 1,2-ethanediamine (EDA, >99.5%), purchased from
Fluka, and selenium powder (200 mesh, 99.999%), purchased from Alfa Aesar. In the
preparation, 0.66 g of PbO were dissolved in a mixture of 30 mL of 1-octadecene and 2.2 mL
of oleic acid, which were degassed at 100 °C for one hour under vacuum. 10.8 mL of that
solution were injected into a solution 1 M of Se in TOP. Then, 84 μL of DPP
(diphenylphosphine) were injected into 28.4 mL of the prepared Pb(oleate) at 180 °C. This
mixture was allowed to react at 160 °C for 2 min to form ~6.4 nm QDs, determined from the
first absorption peak, after which it was quickly cooled by a water bath. The QDs were
precipitated with butanol/methanol mixture and centrifuged at 5000 rpm for 5 minutes. The
precipitate was redissolved in hexane and washed with butanol/methanol mixture once again.
The QDs were then dispersed in TCE for optical absorption measurements and in hexane for
film preparation. All the synthesis was done in a Schlenk line under nitrogen atmosphere.
71
Chapter 4
4.2.2 EDA in methanol solution
A 1 M solution of EDA in methanol was used to replace the long ligands, i.e. C18 chains, by
short ligands, i.e. EDA, to couple the QDs electronically.
4.2.3 Quartz substrate
The dimensions of the quartz substrates used as support are 25 mm x 10 mm x 1 mm. The
substrates were cleaned by sonication for 5 minutes in 5% Triton solution followed by 5
minutes sonication in water and finally 5 minutes sonication in ethanol, to ensure good
adhesion of the QDs to the substrate. The quartz substrates were dried in a flow of nitrogen
and clamped to the mechanical arm of the dip coater.
4.2.4 Fabrication of QD films
We deposited the PbSe QDs on quartz substrates via layer-by-layer dip-coating in three
solutions inside a N2-purged glovebox: (i) PbSe QDs in hexane for 30 s, (ii) 1 M solution of
EDA in methanol for 30 s, and (iii) rinsing solution of methanol for 30 s. The dip-coating of
the substrates was repeated 20 times. The mechanical dip-coater used in the fabrication of the
films was a DC Multi-8 from Nima Technology. The synthesized QD films were stored in a
glovebox to avoid contact with air. The fabricated films had a thickness of approximately 35
nm, determined by a depth profile measurement by a stylus profilometer.
4.2.5 Alumina coating reactor
The alumina coating experiments were performed in an air-tight purpose-made reactor
operated at atmospheric pressure and 27 °C. The operating temperature was controlled by an
external IR lamp and measured with a thermocouple inserted in the reactor. The reactor
consisted of a glass column, 42 mm of diameter and 140 mm in length, with a Teflon holder
inside in which the QD films were placed facing downwards, so that they were exposed to the
precursors during the coating. The holder has 10 slits, separated vertically by 1 cm, allowing
the coating of multiple samples simultaneously. The glass column and the Teflon holder are
fixed to the inlet and outlet connections with a ring, that sealed the reactor. The reactor was
provided with two valves to maintain an oxygen-free environment. The QD films are loaded
into the reactor inside a glovebox, to avoid that the QD are exposed to air, and then
transported to the coating setup, in which we connected the reactor to the inlet of the
precursors. A sintered stainless steel SIKA-R 20 AX distributor plate with a pore size of 37
μm was used to homogeneously distribute the flow of precursors inside the reactor. The offgas from the reactor is taken to a series of bubblers filled with Kaydol oil, to trap unreacted
molecules of the precursors and the by-products of the coating reaction. The precursor
bubblers, the glass column and the washing bubblers are placed inside a nitrogen-blanketing
cabinet as a TMA safety measure. The cabinet is operated at an O2 concentration below 6%.
72
The deposition mechanism of Al2O3 on QD films at atmospheric pressure and room temperature
4.2.6 Alumina coating precursors
Semiconductor grade trimethylaluminium (TMA), purchased from Akzo Nobel, distilled
water and synthetic air were used as precursors, while pressurized N2 grade 5.0 was used to
transport the precursors to the reactor, and to purge it after each reaction. TMA and water
were placed in a 600 mL WW-600 stainless steel bubbler and kept at 30 °C during the coating
experiments. To have a comparable concentration of oxygen to the equivalent experiments
with water, we mixed a flow of 0.5 L/min of N2 with 0.1 L/min of dry air. Some of the
experiments were done using H2O as second precursor, and few others using synthetic air. A
fixed gas flow of 0.6 L/min was chosen for all the experiments performed.
4.2.7 Alumina coating experiments
We used dosing times per cycle of 15 s – 5 min – 15 s – 5 min for the dosing sequence TMA –
N2 – H2O/synthetic air – N2 to coat two QD films, which were placed at the same height of the
holder in all the experiments. The same dosing times were used in all the coating experiments. We
performed coating experiments at 27 °C and 1 bar with different number of cycles, i.e., 1, 3, 5,
8, and 25 cycles, using three different substrates: PbSe QD films prepared with EDAmethanol solution, PbSe QD films prepared using EDA-acetonitrile solution, and a
dodecanethiol self-assembled monolayer on a gold-coated substrate. To study the air-stability,
we prepared other Al2O3-coated films with 1, 10 and 25 cycles, using the same dosing and
purging times, i.e. 15 s – 5 min – 15 s – 5 min. We also performed a 90-cycle experiment on a
PbSe QD film using shorter dosing times, i.e., 0.5 s – 5 min – 0.5 s – 5 min. The total amount
of both precursors dosed during the 90 shorter cycles would be equivalent to the amount one
dosed during 3 cycles using the same dosing times as in the rest of the samples, i.e., 15 s – 5
min – 15 s – 5 min. For this reason, we refer to this 90-shorter-cycle sample as the sample coated
with “3 equivalent cycles” (Figure 4.8b), in order to compare the air-stability with the rest of
the coated QD films.
4.2.8 Characterization of the deposition of Al2O3
To study the deposition of alumina, we measured the concentration depth profile of the
different elements with a Thermo Scientific X-ray photoelectron spectrometer (XPS) KAlpha, equipped with a monochromatic Al Kα radiation source and a pass energy of 100 eV
for the survey scan, and ion-beam etching unit. The XPS device, which operates at ultrahigh
vacuum, is equipped with an etching unit, that uses Ar+ ions with energy of 1000 eV and a
raster size of 2 mm, to remove layers of the QD films with an etching rate of 0.5 nm/s
(Supplementary Information SI I). We used the combination of the XPS and etching unit to
measure the concentration depth profile of the coated substrates. During the XPS analysis, the
spectra of the elements was charge-corrected with the adventitious carbon peak at 284.8 eV.
We used the software Thermo Avantage 5.913 and Gaussian curves to deconvolute the peaks.
During the XPS measurements, we used the flood gun to compensate for the positive charge.
We used a microtome to prepare a slice of one alumina-coated QD film and examine it under
a Tecnai F20 TEM microscope. We measured the film thickness of the alumina overcoat in
130 points from nine TEM images to determine the mean value of the alumina film and the
standard deviation.
73
Chapter 4
4.2.9 Characterization of the air stability of the coated QD films
The alumina-coated and uncoated QD films were exposed to air at 80 °C and UV light, and
their optical absorption spectra were measured using a Perkin-Elmer Lambda 900
spectrophotometer, equipped with an integrating sphere. The position of the maximum peak
of the band gap indicates the stability of the QD crystals, since the band gap absorption peak
blue shifts and broadens in the case of oxidation of the PbSe QD films. We measured the
absorption spectrum of all the uncoated QD films before the coating experiments. The coated
QD films were exposed to air, and the optical absorption was measured periodically, from 1 to
63 days.
4.2.10 Preparation of the dodecanethiol self-assembled monolayer
We used the following solutions to prepare the self-assembled monolayer (SAM):
dodecanethiol (>99.5%), ethanol (>99%) and nitric acid (63%), purchased from Sigma
Aldrich. As substrate, we used 10 mm x 10 mm x 1 mm glass plates, sputter-coated with 200
nm thick gold layer (purchased from SSens BV, The Netherlands). The gold-coated substrates
were washed in a 2% nitric acid solution for 5 min and then rinsed with ethanol and dried
under nitrogen. The plates were then immersed in a 1 M solution of dodecanethiol in ethanol
for 18 h. After that, the plates were taken out, washed with pure ethanol and dried under a
nitrogen atmosphere.
4.3 Results and discussion
In this work, we infilled and overcoated QD films with alumina at atmospheric pressure and
room temperature. We used PbSe QD films with a thickness of 35 nm, measured with a stepprofilometer, consisting of 6 nm QDs separated by EDA organic ligands. In each experiment,
two films were placed in the ALD reactor holder at the same height. Alumina ALD is
typically deposited using TMA and water as precursors according to the general reaction
mechanism consisting of two subsequent reactions, in which || denotes the surface species
[29].
Reaction A: ||– Al – OH + Al(CH3)3 → ||– Al – O – Al(CH3)2 + CH4
Reaction B: ||– Al – O – Al(CH3) 2 + 2 H2O → ||– Al – O – Al – (OH)2 + 2 CH4
In reaction A of the first cycle, functional groups are needed to react with the precursor
molecules and initiate the deposition. These active sites can be any group that readily reacts
with TMA molecules, such as hydroxyl and amine groups [29, 42]. During the first cycles,
both precursors can diffuse through the voids of the QD films, depositing alumina inside the
film, i.e., the infilling stage. By increasing the number of cycles, the voids would get fully
infilled and the precursors react on the surface of the QD film, creating an overcoating (Fig.
4.1).
74
The deposition mechanism of Al2O3 on QD films at atmospheric pressure and room temperature
Fig. 4.1. Cartoon showing (from left to right) an uncoated, an infilled and overcoated QD film
(red spheres) with alumina (blue matrix).
4.3.1 Deposition of Al2O3 using water as precursor
We performed 1, 3, 5, 8 (Fig. 4.2) and 25 cycles (Fig. 4.3), at 27 °C and 1 bar, to study the
growth of alumina inside the PbSe QD films. In the concentration depth profile, we observed
the evolution of the intensity of the different peaks of Pb4f and Se3d belonging to the QDs,
Al2p coming from the deposited alumina, C1s from the EDA ligands, Si2p from the quartz
substrate and O1s from both deposited alumina and quartz substrate.
Fig. 4.2. XPS concentration-depth profile of the PbSe QD films coated with alumina after (a)
1 cycle, (b) 3 cycles, (c) 5 cycles and (d) 8 cycles of ALD. The depth profiles were obtained
by etching the QD films with Ar+ ions.
During the initial ALD cycles, i.e., 1 and 3 cycles (Fig. 4.2a and 4.2b), both TMA and water
penetrated all the way to the wafer through the pore network between the QDs and deposited
as alumina in the entire film. This is clearly seen from the atomic percentages of Pb, Se and
75
Chapter 4
Al in Fig. 4.2a, that remain more or less constant with the increased etching of the assembled
film, all the way down to the quartz substrate. Similarly, the depth profiles of the films coated
with 5 and 8 cycles (Fig. 4.2c and 4.2d) show that alumina is deposited down to the quartz
substrate. In Fig. 4.2c, the amount of Al slightly decreases with increasing etch depth. This
indicates that diffusion limitations start to occur as the overcoated layer becomes thicker. Also
Figs. 4.2b-c show that he layer grows, because the atomic percentage of aluminium increases
with increasing number of cycles.
Fig. 4.3a shows that overcoating of alumina has occurred in the sample coated with 25 cycles,
since only the signals of Al and O are present during the first tenths of nanometers in the
concentration depth profile (Fig. 4.3a). The Al:O ratio is around 1.5 in the overcoating as
expected for Al2O3 molecule. The overcoating can also be clearly seen in the cross sectional
TEM image of a 25-cycle sample (Fig. 4.3b). The XPS profile and TEM image in Fig. 4.3 do
not correspond to the same sample, but to two different batches. We can see that by the
thickness of the PbSe QD film, which is larger than in the XPS depth profile.
Fig. 4.3. (a) XPS depth profile of the QD sample coated with 25 cycles of alumina. (b)
Cross-section TEM image of the 25-cycle QD film, in which the alumina overcoating is
visible.
Carbon was also detected on the surface of the alumina overcoating in the 25-cycle QD film
(Fig. 4.3a). This carbon signal probably originated from unreacted methyl groups from the
TMA molecules, although it quickly decreased to an atomic percentage close to zero after
etching the upper layers of the coated sample. This shows that the amount of unreacted
methyl groups is negligible within the alumina overcoat. Once the alumina overcoat is etched
away and the QD film measured, the atomic percentage of C increased again, most likely
from the EDA ligands. In the XPS spectra of the films coated with 1, 3, 5 and 8 cycles (Fig.
4.2), C1s was also detected in a similar atomic percentage. This indicates that the alumina
does not replace the organic ligands, but would deposit around them and on the surface of the
QD crystals. We conclude that during the first coating cycles, alumina is deposited inside the
QD film, and after a several cycles, the alumina formed an overcoating, although we do not
have a clear evidence about the transition between the infilling and overcoating regimes (Fig.
4.2 and 4.3).
To gain more insight into the deposition process, we first calculated a growth per cycle (GPC)
of alumina during the infilling, considering that the alumina only deposits on the surface of
the QD crystals. And secondly, we studied the deposition of alumina on different substrates to
study the reaction of the precursors in the QD films and define a reaction mechanism. For
that, we dosed TMA to different substrates, i.e., a QD film prepared with methanol as solvent,
76
The deposition mechanism of Al2O3 on QD films at atmospheric pressure and room temperature
a QD film prepared using acetonitrile as solvent, and a substrate covered with a dodecanethiol
self-assembled monolayer.
4.3.2 Estimation of the growth per cycle of Al2O3
To calculate the amount of alumina deposited in each cycle, we modelled the PbSe QD film
as an array of monodispersed spheres, with a particle diameter ( ) of 6.4 nm, separated by
1,2-ethanediamine organic ligands. We assumed that alumina was only deposited on the
surface of the QD crystals. The atomic ratio between Pb and Se atoms in a QD is, in average,
1.9:1 according to the XPS results (Fig. 4.2). An excess of Pb is common for PbSe QDs [43,
44], although we note that the ratio observed here is higher than usually reported. We used the
atomic ratio of Pb and Al in the film coated with 1 cycle (Fig. 4.2a), to calculate the number
of atoms of aluminium in one QD. Using the molecular mass of Al2O3 and its density [35], we
calculated that 0.6 nm of alumina were deposited in the first cycle. Considering an average
value of the GPC for alumina ALD on nanoparticles of 0.15 nm [45-49], and that the
theoretical size of an Al2O3 monolayer is about 0.38 nm [34], we estimate that 1.5 monolayers
of alumina were deposited per cycle during the infilling of the QDs films. However, this value
of the GPC could be lower if we consider that alumina could deposit also around the organic
ligands. More detail of the calculation can be found in the Supplementary Information 4.B.
4.3.3 Study of the deposition mechanism of Al2O3
The GPC of alumina during infilling and the relatively large overcoating-film thickness
measured from TEM (Fig. 4.3b) indicates that the deposition of alumina is faster than the one
expected for ALD experiments, i.e., 0.1-0.2 nm per cycle. Working at ambient conditions can
involve a CVD-type of growth due to accumulation of unreacted precursor molecules on the
substrate [33]. However, we found a high deposition of alumina during the first cycle, where
no unreacted precursors were present. To explain this, we investigated three possible
mechanisms. First, we studied whether TMA molecules could physisorb on the ligands by
exposing a dodecanethiol self-assembled monolayer (SAM) film to TMA. Secondly, we
investigated if there are any remaining solvent molecules entrapped in the QD film that may
act as active sites. This was realized by measuring the concentration depth profile of an
uncoated QD film prepared using different solvents while dip-coating. And third, we
considered whether the amine groups from the EDA organic ligands could act as active sites
and react with TMA molecules.
4.3.4 Study of the reaction of TMA on a dodecanethiol SAM
We used a self-assembled monolayer of dodecanethiol on a gold substrate to study whether
TMA molecules would physisorb on the organic ligands. If true, it would eventually result in
a faster deposition of alumina. We prepared the dodecanethiol SAM , and loaded it into the
coating reactor inside a glovebox to prevent contact with air and humidity, and carried out a
half-coating cycle, i.e., we just exposed the film to TMA. We measured the atomic
percentages on the surface of the exposed SAM (Fig. 4.4), finding a small amount of
aluminium. In case TMA molecules would physisorb on carbon groups, we would have found
a larger amount of aluminium with XPS. This result suggests that TMA does not physisorb on
carbon groups, thus it is unlikely that it would on the EDA organic ligands with similar
77
Chapter 4
methyl groups. We calculated that a full monolayer of physisorbed TMA would produce an
atomic percentage of about 3.5%, while in Fig. 4.4 it is only ~ 0.3% (Supplementary
information 4.C). This is an order of magnitude difference; of course, a SAM is different from
the 3D complex structure of ligands in the assembled QD film, but the results obtained here
are convincing enough to, as a first approximation, rule out physisorption of carbon groups.
Fig. 4.4. Atomic concentrations from XPS on the surface of the SAM after being exposed to
TMA.
4.3.5 XPS depth profile of an uncoated QD film
We measured the concentration depth profile of the uncoated QD film to study the presence of
solvent molecules (Fig. 4.5). As expected, no Al traces were detected; however, O1s signals
were observed. This may point to the presence of oxygenated compounds that could possibly
react with TMA molecules during the first cycle causing a fast deposition of alumina.
Fig. 4.5. XPS concentration depth profile of an uncoated QD film.
To determine the nature of the source of oxygen, we studied the spectra of the elements
detected by XPS after etching 7 nm of the PbSe QD film (Fig. 4.6). At this depth, N1s signals
were observed, which can be attributed to the NH bonds present in the EDA organic ligands,
and a Se Auger peak, both in the same range of binding energies (Fig. 4.6a). The intense peak
at 284.8 eV (Fig. 4.6b) indicates the presence of C-C bonds [50-53]. The peaks at 286.1 and
288.7 eV are attributed to C atoms bound to electronegative atoms [53-55], such as –COOH
or –COH (Fig. 4.6b). These results are consistent with the presence of a peak at 533 eV,
which is typical of –OH groups (Fig. 4.6c). The acid groups may originate from the oleic acid
used to prepare the colloidal solution of PbSe QD crystals, while the observed alcohol
features could come from the methanol used during the fabrication of the QD films via dip78
The deposition mechanism of Al2O3 on QD films at atmospheric pressure and room temperature
coating. To conclude this part, the XPS spectra gives strong evidence of the presence of
oxygen-rich compounds and –NH bonds, which would readily react with TMA molecules
during the first cycle, causing a fast deposition of alumina.
Fig. 4.6. XPS spectra of the different elements found in an uncoated PbSe QD film, at an
etch-depth of approximately 7 nm. (a) Elements corresponding to the QD film, such as
nitrogen and selenium. (b) C1s region spectrum, showing different contributions of carbon,
including alcohol and acid bonds, and (c) hydroxyl groups.
To investigate a possible role of the solvent used during dip-coating, we exposed three QD
films, prepared in different manners, to a single pulse of TMA without completing the cycle
with water. These three samples were: (i) QD film prepared using a solution of EDA in
methanol (Fig. 4.7a), exactly the same as the QD films that were coated with alumina; (ii) QD
film prepared exactly the same way as the previous one, but drying it after the fabrication for
24h in a vacuum chamber at room temperature (Fig. 4.7b); (iii) QD film prepared using a
solution of the EDA ligands in acetonitrile instead of methanol (Fig. 4.7c), aiming to reduce
the presence of oxygenated compounds inside the QD films and study the reaction between
TMA molecules and nitrogen-based functional groups, such as –NH2 from the EDA ligands
and –CN from the acetonitrile. In the sample that was not dried (Fig. 4.7a), hardly any
nitrogen was detected. We found, however, that by decreasing the exposure time to the beam
N2 could be detected (Fig. 4.7b and 4.7c). The rest of elements showed similar profiles
between the samples with no treatment and drying post-treatment (Fig. 4.7a and 4.7b).
Oxygen was detected in the sample prepared with acetonitrile (Fig. 4.7c). One would expect
this sample to contain much less oxygen than the samples prepared with methanol. This
suggests that the oxygen might come from the remaining oleate on the QDs surface, as it is
known that protic solvents such as methanol are more efficient in removing oleate [56], which
also explains why the QD film prepared with acetonitrile is thinner than the ones prepared
with methanol.
Fig. 4.7. XPS depth-profiles after dosing one pulse of TMA to (a) a QD film prepared with a
solution of EDA ligands in methanol, (b) a QD film prepared with a solution of EDA ligands
in methanol and dried for 24h under vacuum, and (c) a QD film prepared with a solution of
EDA ligands in acetonitrile.
79
Chapter 4
4.3.6 Reaction between TMA and amine groups
As the third possibility, we could explain the high deposition of alumina by the reaction
between TMA molecules and the amine groups of the EDA organic ligands. According to the
general scheme of Al2O3-ALD, TMA would react with hydroxyl groups to form alumina.
However, it does not necessarily have to begin with OH groups. Al2O3 has been deposited on
boron nitride (BN) hexagonal platelet-like particles via ALD [46]. BN surfaces contain both
hydroxyl and amine functional groups [57-59], which would react with TMA molecules. In
addition, ammonia has been used as second precursor in ALD, to deposit aluminium nitride
(AlN) [30, 42], proving the reactivity between TMA and amine groups. Therefore, the amine
groups from the EDA ligands could act as active sites where TMA can chemisorb during the
first cycle. In the second cycle, after the first exposure of water, only hydroxyl groups would
be available for reaction, since amine groups cannot be regenerated by using water as second
precursor.
We conclude that the fast deposition of alumina during the first cycle cannot be explained by
one single cause, but rather as the contribution of the hydroxyl and amine active sites present
in the uncoated film and the large excess of precursor molecules, which resulted in a CVDtype of deposition. We note that, while unexpected, this allows very easy and rapid infilling of
these QD films by a single exposure to TMA at room temperature and atmospheric pressure.
4.3.7 Study of the air stability of Al2O3-coated QD films
We investigated the passivation efficiency of the alumina coatings with different film
thickness. We evaluated four samples: uncoated QD film, film coated at 27 °C with an
amount of alumina equivalent to 3 cycles (see section 4.2.7), film coated with 10 cycles at 70
°C, and film coated with 25 cycles at 27 °C. In this study, we used a solution of EDA in
methanol to prepare the QD films. The three coated films and one uncoated QD film were
placed in an oven and exposed to air at 80 °C, and separately to UV light, similarly to the
accelerated test for photovoltaic devices [60]. We measured the optical absorption spectra
periodically, and compared them with the spectra of each film before the coating experiment,
shown as the 0 days curves. We assessed the stability of the QD films by measuring the
position of the maximum of the peak in the optical absorption spectra, which corresponds to
the band gap of the QD film. The peak exhibits a blue shift and broadens due to the oxidation
of the QDs, which results from an effective reduction of the QD size, as seen for the uncoated
QD film (Fig. 4.8a). The quick oxidation of the QDs agrees with previous studies [21]. We
found that 3 cycles does not deposit enough amount of alumina to fully passivate the QD film
(Fig. 4.8b). However, the film coated with 10 cycles was stable upon air exposure at least for
27 days (Fig. 4.8c). This film was coated at 70 °C rather than 27 °C as was done the other
films. We think that at 70 °C, less amount of alumina would be deposited since unreacted
molecules could not physisorb as easily as at room temperature. That means that we could
safely assume that 10 cycles at 27 °C would stabilize the QD film also, and probably with
lower number of cycles. As expected, 25 cycles at 27 °C provided full protection of the QD
film (Fig. 4.8d), at least for 63 days. We found that the optimum for the stabilization of these
PbSe QD films was between 3 and 10 cycles. Therefore, going to higher temperatures would
not be required to protect the QD films from oxidation after a relatively small number of
cycles.
80
The deposition mechanism of Al2O3 on QD films at atmospheric pressure and room temperature
Fig. 4.8. Optical absorption spectra of (a) uncoated QD film, (b) QD coated at 27 °C with an
amount of alumina equivalent to 3 cycles, (c) QD coated with 10 cycles at 70 °C, and (d) QD
coated with 25 cycles at 27 °C, measured periodically after exposure to air at 80 °C and UV
light. The coating experiments were carried out at 27 °C.
4.3.8 Study of the deposition of Al2O3 using synthetic air as oxidizer
To complete this study, we used an alternative second precursor, a diluted flow of air instead
of water, to investigate whether we could reduce the deposition rate of alumina and get closer
to the ALD growth, since it is easier to remove the excess of oxygen molecules. We
performed 25 cycles, and measured the concentration profile with XPS (Fig. 4.9a). It seemed
that both infilling and overcoating occurred, although in this case the overcoating thickness is
thinner than the one with water (Fig. 4.3a). Water has a strong tendency to adsorb on the
surface, specially at room temperature and atmospheric pressure [61]. Contrary to that, when
using a diluted flow of air as in our case, no accumulation would be produced as the unreacted
molecules of oxygen would be expected to leave the reactor. That results in lower deposition
rates, and therefore, a thinner overcoating layer of alumina. However, we found that the QD
film showed signs of oxidation during the coating with oxygen (Fig. 4.9b); this was not
observed when using water as co-reactant. We performed an additional coating experiment
with 1 cycle, using synthetic air as second precursor, to study the oxidation of the QDs. We
measured the optical absorption of the films before and after coating them with 1 and 25
cycles, and found a blue shift of the band gap. The shift of the band gap is similar for both 1and 25-cycle QD films, indicating that the crystals mainly oxidized during the first cycle. The
band gap shifted to a wavelength of 1800 nm, very similar to the shift of the uncoated QD
film after air exposure (Fig. 4.9b), which shifted to 1750 nm approximately. That indicates
that using a synthetic stream of air as second precursor, even with a low concentration of
81
Chapter 4
oxygen, fully degraded the QD crystals, thus not being an adequate oxidizer for the deposition
of Al2O3 on QD films.
Fig. 4.9. (a) XPS depth profile of PbSe QD film coated with 25 cycles of alumina, using
synthetic air as oxidizer. (b) Optical absorption spectra of QD films coated with 1 and 25
cycles of alumina when using synthetic air as oxidizer.
4.4 Conclusions
We deposited alumina on PbSe QD films at room temperature and atmospheric pressure,
using an ALD-like approach. During the first cycles, the alumina infilled the voids of the QD
films, and after several cycles, there was a transition at which an alumina overcoating was
formed. The combination of infilling and overcoating for a number of cycles between 3 and
10 effectively passivated the QD film, which did not show any oxidation after an exposure to
air at 80 °C and UV light for a maximum time tested of 27 days. We calculated that during the
first cycle, we deposited about 1.5 monolayers of alumina on the QD crystals. We think that
this fast growth is the combination of the dosage of excess of precursors, and the presence of
hydroxyl and amine groups inside the film, which acted as active sites. Having a higher
deposition rate of alumina at atmospheric pressure and 27 °C, was beneficial to infill and
overcoat the QD films faster to stabilize the samples against oxidation. Finally, we deposited
alumina using synthetic air as second precursor, instead of water, to reduce the physisorption
of water molecules that produced the rapid growth of alumina, obtaining a thinner overcoating
than the one deposited with water. However, the PbSe crystals degraded even during the first
exposure to synthetic air. This work shows that depositing alumina on QD films at
atmospheric pressure and room temperature gives results that are similar to those obtained at
lower pressures. Working at ambient conditions facilitates the processing of QD films, since
less equipment is required, and can be coupled directly to the fabrication of QD films, which
is also done at atmospheric pressure and room temperature.
82
The deposition mechanism of Al2O3 on QD films at atmospheric pressure and room temperature
Bibliography
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
Ip, A. H.; Thon, S. M.; Hoogland, S.; Voznyy, O.; Zhitomirsky, D.; Debnath, R.;
Levina, L.; Rollny, L. R.; Carey, G. H.; Fischer, A.; Kemp, K. W.; Kramer, I. J.; Ning,
Z.; Labelle, A. J.; Chou, K. W.; Amassian, A.; Sargent, E. H., Hybrid passivated
colloidal quantum dot solids. Nat Nano 2012, 7, (9), 577-582.
Nozik, A. J.; Beard, M. C.; Luther, J. M.; Law, M.; Ellingson, R. J.; Johnson, J. C.,
Semiconductor quantum dots and quantum dot arrays and applications of multiple
exciton generation to third-generation photovoltaic solar cells. Chem. Rev. 2010, 110,
(11), 6873-6890.
Hillhouse, H. W.; Beard, M. C., Solar cells from colloidal nanocrystals: Fundamentals,
materials, devices, and economics. Current Opinion in Colloid & Interface Science
2009, 14, (4), 245-259.
Tang, J.; Kemp, K. W.; Hoogland, S.; Jeong, K. S.; Liu, H.; Levina, L.; Furukawa, M.;
Wang, X.; Debnath, R.; Cha, D.; Chou, K. W.; Fischer, A.; Amassian, A.; Asbury, J.
B.; Sargent, E. H., Colloidal-quantum-dot photovoltaics using atomic-ligand
passivation. Nat Mater 2011, 10, (10), 765-771.
Lee, H. J.; Yum, J. H.; Leventis, H. C.; Zakeeruddin, S. M.; Haque, S. A.; Chen, P.;
Seok, S. I.; Grätzel, M.; Nazeeruddin, M. K., CdSe quantum dot-sensitized solar cells
exceeding efficiency 1% at full-sun intensity. J. Phys. Chem. C 2008, 112, (30),
11600-11608.
Sandeep, C. S. S.; Cate, S. T.; Schins, J. M.; Savenije, T. J.; Liu, Y.; Law, M.; Kinge,
S.; Houtepen, A. J.; Siebbeles, L. D. A., High charge-carrier mobility enables
exploitation of carrier multiplication in quantum-dot films. Nature Communications
2013, 4.
Luther, J. M.; Law, M.; Beard, M. C.; Song, Q.; Reese, M. O.; Ellingson, R. J.; Nozik,
A. J., Schottky Solar Cells Based on Colloidal Nanocrystal Films. Nano Lett. 2008, 8,
(10), 3488-3492.
Nordell, K. J.; Boatman, E. M.; Lisensky, G. C., A Safer, Easier, Faster Synthesis for
CdSe Quantum Dot Nanocrystals. Journal of Chemical Education 2005, 82, (11),
1697.
Beard, M. C., Multiple Exciton Generation in Semiconductor Quantum Dots. J. Phys.
Chem. Lett. 2011, 2, 1282-1288.
Beard, M. C.; Luther, J. M.; Semonin, O. E.; Nozik, A. J., Third generation
photovoltaics based on multiple exciton generation in quantum confined
semiconductors. Accounts of chemical research 2013, 46, (6), 1252-60.
Nair, G. C., L. Y.; Geyer, S. M.; Bawendi, M. G. , Perspective on the Prospects of a
Carrier Multiplication Nanocrystal Solar Cell. Nano Lett. 2011, (11), 2145-2151.
Beard, M. C. E., R. J. , Multiple Exciton Generation in Semiconductor Nanocrystals:
Toward Efficient Solar Energy Conversion. Laser Photonics Rev. 2008, (2), 377-399.
McGuire, J. A. J., J.; Pietryga, J. M.; Schaller, R. D.; Klimov, V. I. , New Aspects of
Carrier Multiplication in Semiconductor Nanocrystals. . Acc. Chem. Res. 2008, (41),
1810-1819.
Nozik, A. J., Multiple Exciton Generation in Semiconductor Quantum Dots. Chem.
Phys. Lett. 2008, (457), 3-11.
Ten Cate, S.; Liu, Y.; Suchand Sandeep, C. S.; Kinge, S.; Houtepen, A. J.; Savenije,
T. J.; Schins, J. M.; Law, M.; Siebbeles, L. D. A., Activating carrier multiplication in
PbSe quantum dot solids by infilling with atomic layer deposition. Journal of Physical
Chemistry Letters 2013, 4, (11), 1766-1770.
83
Chapter 4
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
Aerts, M.; Suchand Sandeep, C. S.; Gao, Y.; Savenije, T. J.; Schins, J. M.; Houtepen,
A. J.; Kinge, S.; Siebbeles, L. D. A., Free charges produced by carrier multiplication
in strongly coupled pbse quantum dot films. Nano Lett. 2011, 11, (10), 4485-4489.
Gao, Y.; Talgorn, E.; Aerts, M.; Trinh, M. T.; Schins, J. M.; Houtepen, A. J.;
Siebbeles, L. D. A., Enhanced hot-carrier cooling and ultrafast spectral diffusion in
strongly coupled PbSe quantum-dot solids. Nano Lett. 2011, 11, (12), 5471-5476.
Ten Cate, S.; Sandeep, C. S. S.; Liu, Y.; Law, M.; Kinge, S.; Houtepen, A. J.; Schins,
J. M.; Siebbeles, L. D. A., Generating free charges by carrier multiplication in
quantum dots for highly efficient photovoltaics. Accounts of chemical research 2015,
48, (2), 174-181.
Gao, Y.; Aerts, M.; Sandeep, C. S. S.; Talgorn, E.; Savenije, T. J.; Kinge, S.;
Siebbeles, L. D. A.; Houtepen, A. J., Photoconductivity of PbSe quantum-dot solids:
Dependence on ligand anchor group and length. ACS Nano 2012, 6, (11), 9606-9614.
Liu, Y. G., M.; Puthussery, J.; Gaik, S.; Ihly, R.; Hillhouse, H. W.; Law, M. ,
Dependence of Carrier Mobility on Nanocrystal Size and Ligand Length in PbSe
Nanocrystal Solids. Nano Lett. 2010, (10), 1960-1969.
Ihly, R. T., J.; Liu, Y.; Gibbs, M.; Law, M. , The Photothermal Stability of PbS
Quantum Dot Solids. ACS Nano 2011, (5), 8175−8186.
Liu, Y.; Gibbs, M.; Perkins, C. L.; Tolentino, J.; Zarghami, M. H.; Bustamante, J.;
Law, M., Robust, functional nanocrystal solids by infilling with atomic layer
deposition. Nano Lett. 2011, 11, (12), 5349-5355.
Gai, Y.; Peng, H.; Li, J., Electronic properties of nonstoichiometric PbSe quantum
dots from first principles. J. Phys. Chem. C 2009, 113, (52), 21506-21511.
Etgar, L.; Lifshitz, E.; Tannenbaum, R., Hierarchical conjugate structure of γ-Fe2O3
nanoparticles and PbSe quantum dots for biological applications. J. Phys. Chem. C
2007, 111, (17), 6238-6244.
Min, B. D.; Kim, Y.; Kim, E. K.; Min, S. K.; Park, M. J., Suppression of Ostwald
ripening in In0.5Ga0.5As quantum dots on a vicinal (100) substrate. Phys. Rev. B
Condens. Matter Mater. Phys. 1998, 57, (19), 11879-11882.
Dameron, A. A.; Davidson, S. D.; Burton, B. B.; Carcia, P. F.; Scott McLean, R.;
George, S. M., Gas diffusion barriers on polymers using multilayers fabricated by Al
2O3 and rapid SiO2 atomic layer deposition. J. Phys. Chem. C 2008, 112, (12), 45734580.
Elam, J. W.; Libera, J. A.; Huynh, T. H.; Feng, H.; Pellin, M. J., Atomic layer
deposition of aluminum oxide in mesoporous silica gel. J. Phys. Chem. C 2010, 114,
(41), 17286-17292.
Liu, Y.; Tolentino, J.; Gibbs, M.; Ihly, R.; Perkins, C. L.; Crawford, N.; Hemminger,
J. C.; Law, M., PbSe quantum dot field-effect transistors with air-stable electron
mobilities above 7 cm2 V-1 s-1. Nano Lett. 2013, 13, (4), 1578-1587.
Puurunen, R. L., Surface chemistry of atomic layer deposition: A case study for the
trimethylaluminum/water process. Journal of Applied Physics 2005, 97, (12).
Miikkulainen, V.; Leskelä, M.; Ritala, M.; Puurunen, R. L., Crystallinity of inorganic
films grown by atomic layer deposition: Overview and general trends. Journal of
Applied Physics 2013, 113, (2).
Suntola, T., Atomic layer epitaxy. Thin Solid Films 1992, 216, (1), 84-89.
Beetstra, R.; Lafont, U.; Nijenhuis, J.; Kelder, E. M.; Van Ommen, J. R., Atmospheric
pressure process for coating particles using atomic layer deposition. Chemical Vapor
Deposition 2009, 15, (7-9), 227-233.
84
The deposition mechanism of Al2O3 on QD films at atmospheric pressure and room temperature
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
Valdesueiro, D.; Meesters, G.; Kreutzer, M.; van Ommen, J., Gas-Phase Deposition of
Ultrathin Aluminium Oxide Films on Nanoparticles at Ambient Conditions. Materials
2015, 8, (3), 1249-1263.
George, S. M., Atomic layer deposition: An overview. Chem. Rev. 2010, 110, (1),
111-131.
Groner, M. D.; Fabreguette, F. H.; Elam, J. W.; George, S. M., Low-Temperature
Al2O3 Atomic Layer Deposition. Chemistry of Materials 2004, 16, (4), 639-645.
Tripp, M. K.; Stampfer, C.; Miller, D. C.; Helbling, T.; Herrmann, C. F.; Hierold, C.;
Gall, K.; George, S. M.; Bright, V. M., The mechanical properties of atomic layer
deposited alumina for use in micro- and nano-electromechanical systems. Sensors and
Actuators A: Physical 2006, 130–131, (0), 419-429.
Lu, P.; Demirkan, K.; Opila, R. L.; Walker, A. V., Room-temperature Chemical Vapor
Deposition of Aluminum and Aluminum Oxides on Alkanethiolate Self-Assembled
Monolayers. The Journal of Physical Chemistry C 2008, 112, (6), 2091-2098.
Zhang, J.; Tolentino, J.; Smith, E. R.; Beard, M. C.; Nozik, A. J.; Law, M.; Johnson, J.
C., Carrier transport in PbS and PbSe QD films measured by photoluminescence
quenching. J. Phys. Chem. C 2014, 118, (29), 16228-16235.
Poodt, P.; Lankhorst, A.; Roozeboom, F.; Spee, K.; Maas, D.; Vermeer, A., Highspeed spatial atomic-layer deposition of aluminum oxide layers for solar cell
passivation. Advanced Materials 2010, 22, (32), 3564-3567.
Poodt, P.; Cameron, D. C.; Dickey, E.; George, S. M.; Kuznetsov, V.; Parsons, G. N.;
Roozeboom, F.; Sundaram, G.; Vermeer, A., Spatial atomic layer deposition: A route
towards further industrialization of atomic layer deposition. Journal of Vacuum
Science &amp; Technology A 2012, 30, (1), 010802.
Werner, F.; Stals, W.; Görtzen, R.; Veith, B.; Brendel, R.; Schmidt, J. In High-rate
atomic layer deposition of Al2O3 for the surface passivation of Si solar cells, 1st
International Conference on Crystalline Silicon Photovoltaics, SiliconPV 2011,
Freiburg, 2011; Freiburg, 2011; pp 301-306.
Puurunen, R. L.; Lindblad, M.; Rootc, A.; Krausea, A. O. I., Successive reactions of
gaseous trimethylaluminium and ammonia on porous alumina. Physical Chemistry
Chemical Physics 2001, 3, (6), 1093-1102.
Moreels, I.; Lambert, K.; De Muynck, D.; Vanhaecke, F.; Poelman, D.; Martins, J. C.;
Allan, G.; Hens, Z., Composition and Size-Dependent Extinction Coefficient of
Colloidal PbSe Quantum Dots. Chemistry of Materials 2007, 19, (25), 6101-6106.
Sandeep, C. S. S.; Azpiroz, J. M.; Evers, W. H.; Boehme, S. C.; Moreels, I.; Kinge, S.;
Siebbeles, L. D. A.; Infante, I.; Houtepen, A. J., Epitaxially connected PbSe quantumdot films: Controlled neck formation and optoelectronic properties. ACS Nano 2014,
8, (11), 11499-11511.
Wank, J. R.; George, S. M.; Weimer, A. W., Coating fine nickel particles with Al2O3
utilizing an atomic layer deposition-fluidized bed reactor (ALD-FBR). Journal of the
American Ceramic Society 2004, 87, (4), 762-765.
Wank, J. R.; George, S. M.; Weimer, A. W., Nanocoating individual cohesive boron
nitride particles in a fluidized bed by ALD. Powder Technology 2004, 142, (1), 59-69.
Hakim, L. F.; George, S. M.; Weimer, A. W., Conformal nanocoating of zirconia
nanoparticles by atomic layer deposition in a fluidized bed reactor. Nanotechnology
2005, 16, (7), S375-S381.
Hakim, L. F.; Blackson, J.; George, S. M.; Weimer, A. W., Nanocoating individual
silica nanoparticles by atomic layer deposition in a fluidized bed reactor. Chemical
Vapor Deposition 2005, 11, (10), 420-425.
85
Chapter 4
49.
50.
51.
52.
53.
54.
55.
56.
57.
58.
59.
60.
61.
Hakim, L. F.; McCormick, J. A.; Zhan, G. D.; Weimer, A. W.; Li, P.; George, S. M.,
Surface modification of titania nanoparticles using ultrathin ceramic films. Journal of
the American Ceramic Society 2006, 89, (10), 3070-3075.
Barr, T. L.; Seal, S., Nature of the use of adventitious carbon as a binding energy
standard. Journal of Vacuum Science &amp; Technology A 1995, 13, (3), 1239-1246.
Smith, K. L.; Black, K. M., Characterization of the treated surfaces of silicon alloyed
pyrolytic carbon and SiC. Journal of Vacuum Science &amp; Technology A 1984, 2,
(2), 744-747.
Bastl, Z., X-Ray Photoelectron Spectroscopic Studies of Palladium Dispersed on
Carbon Surfaces Modified by Ion Beams and Plasmatic Oxidation. Collection of
Czechoslovak Chemical Communications 1995, (60), 383.
Gelius, U.; Hedén, P. F.; Hedman, J.; Lindberg, B. J.; Manne, R.; Nordberg, R.;
Nordling, C.; Siegbahn, K., Molecular Spectroscopy by Means of ESCA III. Carbon
compounds. Physica Scripta 1970, 2, (1-2), 70.
Chen, J. J.; Jiang, Z. C.; Zhou, Y.; Chakraborty, B. R.; Winograd, N., Spectroscopic
studies of methanol decomposition on Pd[lcub]111[rcub]. Surface Science 1995, 328,
(3), 248-262.
Solymosi, F.; Berkó, A.; Tóth, Z., Adsorption and dissociation of CH3OH on clean
and K-promoted Pd(100) surfaces. Surface Science 1993, 285, (3), 197-208.
Hassinen, A.; Moreels, I.; De Nolf, K.; Smet, P. F.; Martins, J. C.; Hens, Z., ShortChain Alcohols Strip X-Type Ligands and Quench the Luminescence of PbSe and
CdSe Quantum Dots, Acetonitrile Does Not. Journal of the American Chemical
Society 2012, 134, (51), 20705-20712.
Huang, M. T.; Ishida, H., Surface study of hexagonal boron nitride powder by diffuse
reflectance Fourier transform infrared spectroscopy. Surface and Interface Analysis
2005, 37, (7), 621-627.
Ikuno, T.; Sainsbury, T.; Okawa, D.; Fréchet, J. M. J.; Zettl, A., Amine-functionalized
boron nitride nanotubes. Solid State Communications 2007, 142, (11), 643-646.
Li, J.; Xiao, X.; Xu, X.; Lin, J.; Huang, Y.; Xue, Y.; Jin, P.; Zou, J.; Tang, C.,
Activated boron nitride as an effective adsorbent for metal ions and organic pollutants.
Sci. Rep. 2013, 3.
Igari, S.; Takahisa, K., Accelerated irradiance and temperature cycle test for
amorphous silicon photovoltaic devices. Prog Photovoltaics Res Appl 2014, 22, (6),
690-696.
Salameh, S.; Schneider, J.; Laube, J.; Alessandrini, A.; Facci, P.; Seo, J. W.; Ciacchi,
L. C.; Mädler, L., Adhesion mechanisms of the contact interface of TiO 2
nanoparticles in films and aggregates. Langmuir 2012, 28, (31), 11457-11464.
86
The deposition mechanism of Al2O3 on QD films at atmospheric pressure and room temperature
4.A Validity of XPS measurements to determine film thickness
Determining the thickness of the alumina overcoating with XPS is not accurate, since the
beam of X-rays has a penetration depth in which the concentration of the elements is
measured. We fixed an etching rate of 0.5 nm/s, although it depends on the material etched; a
harder material would etch slower than a softer one. Despite the etching depth measured from
XPS cannot be considered as the conclusive film thickness, we can still use these values to
compare the depth profiles between the different coated QD samples.
4.B Estimation of the growth per cycle of Al2O3
To calculate the amount of alumina deposited in each cycle, we modelled the PbSe QD film
as an array of monodispersed spheres, with a particle diameter ( ) of 6.4 nm, separated by
1,2-ethanediamine organic ligands. We assumed that alumina was only deposited on the
surface of the QD crystals. We used a density of the QD of 8.3 g/cm3, and a calculated
molecular mass for the QD of 162.3 g/mol, based on the atomic masses of Pb and Se, and the
Pb:Se atomic ratio. Using the size, molecular mass and density, we calculated the volume,
1.4·10-19 cm3, and the number of moles, 7·10-21 mol, of one QD. Therefore, one QD consists
of 4.6·10-21 mol of Pb and 2.4·10-21 mol of Se. In the XPS profile of the 1-cycle coated film
(Fig. 4.2a), we measured the same atomic percentages of Pb and Al. Hence, there were also
4.6·10-21 mol of aluminium in each QD, that corresponds with 2.3·10-21 mol of aluminium
oxide, since there are two moles of aluminium in one mole of aluminium oxide. Using the
density of the aluminium oxide of 2.5 g/cm3 [35] and molecular weight, 102 g/mol, we
calculated a volume of alumina (
) in each crystal of 9.3·10-23 cm3. We used (Eq. 1) to
calculate a film thickness of alumina (
) of 0.6 nm in each QD crystal. This calculation
considers that the volume of the Al2O3-QD core-shell is the sum of the volume of the QD
crystal and the volume of the alumina.
[(
)
]
(Eq. 4.B.1)
By using an average value of the GPC for alumina ALD on nanoparticles of 0.15 nm [45-49],
we estimated that 4 monolayers of alumina were deposited per cycle during the infilling of the
QD dots.
87
Chapter 4
4.C Concentration of a monolayer of TMA physisorbed on the
dodecanethiol SAM
The atomic percentages measured with XPS after one exposure of TMA are shown in the
table below.
Element
Atomic percentage [%]
C
47.07
Au
48.00
O
0.56
Al
0.34
S
4.03
Fig. 4.C.1. Atomic content of the dodecanethiol after exposure to TMA.
As expected for dodecanethiol, the ratio C/S is about 12. The physisorption of a full
monolayer of TMA on the SAM would mean that one TMA molecule would physisorb in
each of the dodecanethiol chains, therefore, we would have the same atomic percentage of S
than of Al. In addition, the number of carbon atoms would increase as well due to the three
methyl groups in a TMA molecule. Therefore, recalculating the atomic percentages, we could
expect 3.5% of sulfur and aluminium, 51% of carbon, 41.5% of gold and 0.5% of oxygen.
88
Chapter 5
Producing SiC tracer particles
using gas-phase deposition of Al2O3
at ambient conditions
We have enhanced the radio-activation efficiency of SiC (silicon carbide) particles,
which by nature have a poor affinity towards 18F ions, to be employed as tracers in
studies using PEPT (Positron Emission Particle Tracking). The resulting SiC-Al2O3
core-shell structure shows a good labelling efficiency, comparable to γ-Al2O3 tracer
particles, which are commonly used in PEPT. The coating of the SiC particles was
carried at 27 ± 3 °C and 1 bar in a fluidized bed reactor, using trimethylaluminium and
water as precursors, by a gas phase technique similar to atomic layer deposition. The
thickness of the alumina films, which ranged from 5 to 500 nm, was measured by
elemental analysis and confirmed with FIB-TEM (focus ion beam – transmission
electron microscope), obtaining consistent results from both techniques. By depositing
such a thin film of alumina, properties that influence the hydrodynamic behaviour of
the SiC particles, such as size, shape and density, are hardly altered, ensuring that the
tracer particle shows the same flow behaviour as the other particles. The paper
describes a general method to improve the activation efficiency of materials, which can
be applied for the production of tracer particles for many other applications too.
Submitted as: D. Valdesueiro, P. Garcia-Trinanes, G.M.H. Meesters, M.T. Kreutzer, J.
Gargiuli, T. Leadbeater, D.J. Parker, J. Seville and J.R. van Ommen. Enhancing the
activation of silicon carbide tracer particles for PEPT applications using gas-phase
deposition of alumina at room temperature and atmospheric pressure.
91
Chapter 5
5.1 Introduction
Positron Emission Particle Tracking (PEPT) is a powerful non-invasive technique to
follow the motion of individual particles in industrial processes [1, 2], which are
opaque to other tracking methods [3-5]. The positron emitter most commonly used in
such studies is 18F, which has a half-life of 110 minutes. The level of radioactivity of
the tracer will define the performance of the PEPT measurement, which depends on the
intensity of the signal in the “positron camera” detectors to reconstruct the trajectory of
the tracer in the three dimensions [6]. Using a tracer with low emission intensity results
in poor resolution of the spatial location of the tracer [7-9].
In previous PEPT studies [10-12], tracers have been produced either by direct
irradiation of the sample in a suitable cyclotron, converting oxygen in the sample
directly to 18F, or by irradiation of water, which is then exchanged with, or attached to,
molecules on the surface of the tracer. The trajectory of the tracer is understood to be
representative of the motion of all the particles in the system, which is only the case if
the emitting particle is identical, from a granular-matter point of view, to the particles
of interest. This can be readily achieved if the particles can adsorb the emitter.
However, in some cases this does not occur. This work deals with the problem that
appears when the particles do not adsorb the emitter. Then, one can take a different
particle to be used as tracer, accepting the mismatch in some properties, or one can
develop a particle that emits sufficiently and remains practically identical to the other
particles. This paper explores this last option for silicon carbide (SiC) particles.
SiC particles are used as a heat transfer medium in fluidized beds to harvest solar
energy in concentrated solar thermal plants [13, 14]. The advantages of SiC in this
application include high heat capacity, high sintering temperature, good availability
and low cost. The favourable properties of fluidized beds regarding mixing and
processability at large scale make them attractive in energy applications such as
gasification and combustion of biomass, and chemical looping combustion [15-17]. In
all these applications, ensuring and quantifying the good circulation of the particles is
essential, and for this PEPT is a uniquely powerful technique. Unfortunately, SiC
hardly adsorbs 18F.
We demonstrate the production of a core-shell structure to be used as a PEPT tracer
particle that better adsorbs the radioactive ions than the core itself. We used SiC as
core material and deposited films of Al2O3 using a gas-phase coating technique, similar
to atomic layer deposition (ALD), using trimethylaluminium (TMA) and water as
precursors, at atmospheric pressure and room temperature [18]. Providing the native
SiC particles with a thin coating that can be made radioactive is an attractive
alternative to enhance the labelling efficiency of these particles , defined as the ratio of
radioactivity absorbed by the SiC particles to the radioactivity of the water solution
[19]. Historically aluminium oxide has proved to be a very successful material used for
PEPT tracers due to its high affinity for 18F ions [20].
The SiC particles used here, with an average particle size (
) of 68 µm and density
3
of 3210 kg/m (Supplementary information 5.A), have the required thermal properties
and fluidization behaviour (Geldart A type [21]). However, the inert surface of SiC
particles causes poor adsorption of 18F. Other particles with higher labelling
performance, such as ion exchange resins or γ-Al2O3 particles [22], are effective as
92
Producing SiC tracer particles using gas-phase deposition of Al2O3 at ambient conditions
emitters, but their different density, size and shape make them poor tracers in this
application because their trajectory is different from the SiC particles in the fluidized
bed.
ALD is used to deposit inorganic compounds with accurate control based on a set of
two reactions repeated a certain number of times [23, 24]. ALD has been applied either
to functionalize [25, 26] or protect [27, 28] the surface of flat substrates or powders.
We used ALD to reproduce the affinity between 18F ions and the surface of γ-Al2O3
particles [20, 22]. Normally, ALD of alumina is performed at about 170 °C and
absolute pressures of about 1 mbar to ensure the removal of the excess of precursors
from the reactor, and obtain atomic growth of the films, i.e. between 0.1 and 0.2 nm
per cycle [29-31]. We carried out the coating of SiC particles in a fluidized bed
reactor30, 31 at 1 bar and 27 °C. At these conditions, the removal of the excess of
precursors is diminished, and this excess can physisorb on the surface of the SiC
particles [32], resulting in a CVD (chemical vapour deposition) type of reaction and
thus, higher growth per cycle (GPC) of alumina [18, 33].
To further accelerate the growth of the alumina films, we treated two samples with
oxygen plasma prior to coating to increase the initial surface density of hydroxyl
groups, crucial to initiate the deposition of alumina (reaction A). The surface of SiC is
formed by carbon- and silicon-terminated groups. While carbon-terminated groups are
stable and rather inert, silicon-terminated groups are prone to oxidation [34, 35],
providing the SiC surface with hydroxyl groups. By using stronger oxidising media,
such as oxygen plasma, we can increase the initial density of hydroxyl groups on the
surface [36, 37], enhancing the deposition of alumina during the first cycles, achieving
higher GPC. Nevertheless, having relatively high GPC is a good compromise between
depositing thick films of Al2O3 in a fast way, improving the activation with 18F, and
preserving the particle properties relevant for the hydrodynamic behaviour of the
particles, i.e. size, shape and density.
This experimental paper describes a generic method –using gas phase deposition of
alumina– for making tracer particles that closely resemble the original particles. The
resulting alumina films are very thin compared to the size of the original particles, and
therefore have a negligible influence on properties such as size, shape and density. We
demonstrate this with the specific example of the improved labelling efficiency of SiC
particles towards 18F ions. For that, we coated five samples of SiC with different
thicknesses of Al2O3 films, using two deposition temperatures, and surface prefunctionalizing treatment in two of the experiments. After the coating, we radioactivated the five samples with 18F ions, and compared the activity with the uncoated
SiC and γ-Al2O3 particles. We demonstrate the use of an activated SiC-Al2O3 coreshell particle as a tracer in a PEPT experiment on fluidizing SiC particles. Using this
novel tracer, more accurate studies on the hydrodynamic behaviour of SiC particles can
be performed, increasing knowledge of their behaviour in industrial applications such
as direct solar harvesting.
5.2 Experimental
Al2O3 films were deposited in a purpose-built fluidized bed reactor consisting of a
glass column of 26 mm in diameter and 500 mm in length. Two stainless-steel
93
Chapter 5
distributor plates with pore size of 37 µm, placed at the bottom and top of the column,
are used to obtain a homogeneous distribution of the gas inside the column and to
prevent particles from leaving the reactor. The reactor and the rest of the setup have
been described in detail previously [18, 38]. We use TMA and water as precursors to
deposit Al2O3 films according to the reaction mechanism (A) and (B). In an ALD
process, the surface species in reactions (A) and (B), respectively OH* and CH3*,
determine the completion of the reactions, and once they are depleted, the reactions
end. That confers the self-terminating feature to the ALD process, which ensures
atomic growth of the aluminium oxide film. N2 is pulsed into the reactor in between
the reactions for purging purposes. This cycle of reactions can be repeated to grow
thicker coatings of aluminium oxide.
(A) SiC-OH* + Al(CH3)3  SiC-O-Al(CH3)2* + CH4
(B) SiC-O-Al(CH3)2* + 2 H2O  SiC-O-Al-(OH)2* + 2 CH4
We calculated the dosing time for TMA based on the maximum amount of TMA
molecules that can be accommodated on the surface of the particles [23], considering
the steric hindrance between the methyl groups of the TMA, and 0.12 nm as the ligand
radius for a TMA molecule [39]. We measured with BET (Brunauer-Emmett-Teller) a
specific surface area of 0.12 m2/g for the SiC particles, and calculated the total particle
surface area inside the column for the 8.00 g of powder loaded in the column in each
experiment. A N2 flow of 0.8 L/min, which corresponds to a superficial gas velocity of
2.5 cm/s, was applied to fluidize the powder. To calculate the amount of TMA we dose
to the column, we assumed that at the bubbler temperature of 30 °C, TMA is found as
dimers [40, 41] and that the components follow the ideal gas law. We estimated, using
the model proposed by Mayer et al. [42], a saturation of the N2 bubbles with TMA of
about 50%. With these assumptions, we calculated a dosing time of 2.6 seconds for
TMA and 2.0 seconds for water to obtain saturation of the particles with the precursors
(Supplementary information 5.B). In order to ensure a faster growth of the alumina
films, and therefore, have a higher improvement in the radio-activation of the SiC
particles, we overdosed both precursors to the reactor by a factor of about 120 more
than the calculated times. With this, we established the dosing times for the sequence
of TMA–N2–H2O–N2, in 5–10–4–10 minutes in all the coating experiments.
We considered that at ambient conditions, the large amount of excess molecules of
precursor introduced in each cycle will accumulate on the surface of the particles,
resulting in a parasitic-CVD type of growth and thicker films [18]. In addition, we prefunctionalised two samples with oxygen plasma before the coating to obtain a higher
GPC. For that, SiC was uniformly spread over a glass Petri dish and introduced into a
Harrick Plasma PDC-002 plasma cleaner device for 1 minute; the pressure of the
chamber was kept at 6 mbar. Immediately after exposing the SiC to the oxygen plasma,
the powder was introduced inside the column to start with the coating experiments.
In total, we performed experiments at five different settings, modifying the number of
cycles, operating temperature and pre-functionalization of the SiC particles (Table 5.1).
At 100 °C, we performed two experiments with 5 and 7 cycles. To increase the layer
thickness, we lowered the temperature to 27 °C and performed 20 cycles. We carried
out the oxygen plasma functionalization, and performed 20 and 40 cycles. In all the
experiments we kept constant the initial mass of SiC powder loaded inside the reactor
94
Producing SiC tracer particles using gas-phase deposition of Al2O3 at ambient conditions
(8.00 g), the flow of nitrogen (0.8 L/min), and dosing times of the precursors and
purging N2 (5–10–4–10 minutes).
To characterize the coating process, we calculated the thickness of the alumina films
(
) from the elemental analysis of the samples carried out with a Perkin Elmer
Optima 500 ICP-OES (Induced Coupled Plasma – Optical Emission Spectroscopy).
We obtain the mass fraction of aluminium ( ) from ICP, from which we calculate the
thickness of the alumina coating [18]. For this calculation, we used a density for the
SiC particles of 3210 kg/m3, and for alumina of 2500 kg/m3 [43]. We measured a
fraction of aluminium of 0.0007 in the uncoated SiC particles, and used this value to
correct the fraction of aluminium in the coated samples, to consider only the
aluminium deposited as Al2O3 in the calculation of the film thickness. To compare the
thickness calculated from elemental analysis, we measured the thickness directly from
a TEM (Transmission Electron Microscope) image for the sample with 40 cycles. For
that, we produced a lamella in the nanometre range of the coated SiC using FIB
(Focused Ion Beam) [44] combined with TEM.
Extending the earlier discussion, radio-activation of the tracer can be achieved
according to three different techniques: direct activation, ion exchange and surface
modification [7, 45]. In this work, we used a procedure similar to the ion exchange
method, using 18F as radioactive source [6]. To activate the particles we prepared an
extremely dilute solution of 18F ions in pure water produced by bombardment with an
energetic 3He beam from the Birmingham MC40 cyclotron. Oxygen atoms within the
solution are converted into 18F ions in the two competing reactions described by Fan et
al [7]. A sample of our coated particles was immersed into the solution for around one
hour allowing contact between 18F ions and the Al2O3 surface layer of the coated SiC
particles. The liquid was evaporated and the particles dried, thus allowing their
recovery and subsequent measurement. The activity of the samples was measured with
a CRC-15R Capintenc Inc. radioisotope calibrator [45]. To compare the activation of
the different SiC samples and the γ-Al2O3, we calculated the relative activity as the
ratio of the activity of the particles, with a value in the µCi range, and the activity of
the radioactive solution, in the mCi range. Finally, to confirm the applicability of the
tracer particle produced, we used one radio-activated particle of the SiC sample coated
for 40 cycles with aluminium oxide to perform a PEPT experiment in a fluidized bed.
We fluidized SiC for 30minutes in a column of 90mm in diameter and 500mm in
length, and reconstructed the trajectory of the tracer particle based on the triplets
( ̅ ̅ ̅ measured by the ADAC “positron camera” [2].
5.3 Results and discussion
Table 5.1 shows the experiments carried out with different numbers of coating cycles
(5, 7, 20 and 40 cycles), different reaction temperatures (100 and 27 °C) and prefunctionalization of the powder with O2 plasma. As we expected, the mass fraction of
aluminium ( ) and, therefore, the thickness of the alumina films (
) increased
with: (i) an increasing number of cycles, (ii) a decrease in the reaction temperature, and
(iii) the pre-functionalization of the SiC particles (Table 5.1). We obtained a GPC of 12 nm for the experiments at 100 °C, which is calculated after dividing the thickness of
the alumina film over the number of cycles. When reducing the temperature to 27 °C
we achieved a GPC of 9 nm after 20 cycles. This shows the influence of the
95
Chapter 5
temperature on the accumulation of unreacted precursor molecules. In addition, we
observed an increase in the GPC after the functionalization of the SiC particles with O2
plasma for the experiment with 20 cycles. That can be explained by an increase of the
surface density of hydroxyl groups during the plasma treatment, which will promote a
larger deposition during the initial cycles of the experiment. We obtained a GPC from
1 to 12 nm in the different experiments, which is much larger than the characteristic
GPC for ALD, typically 0.1-0.2 nm [18, 46-49]. This is due to the long dosing times of
both precursors, far beyond the dosage to fully saturate the surface of the particles. At
27 °C, the molecules of the precursors in excess accumulate on the surface, losing the
self-terminating feature of the ALD reactions, and inducing higher GPC [18]. Based on
the GPC, these experiments cannot be considered as ALD. However, neither do the
GPC values in this work indicate a typical CVD mechanism, where the precursors are
dosed simultaneously to the reactor, and films grow with rates around 0.1 µm/min for
similar metal oxides [50], nor as Rapid ALD process [51], where the precursors acts as
catalyst to deposit layers of about 2 nm per cycle.
Table 5.1. Mass fraction of aluminium, determined by ICP, and the thickness of the alumina
film, calculated from the results from ICP, for the different experiments.
# cycles
O2 Plasma T [°C]
GPC [-]
[-]
[nm]
5
100
0.02
5.5 ± 0.6
1.1 ± 0.1
7
100
0.06
16.5 ± 1.7
2.4 ± 0.2
20
27
0.67
183.2 ± 19.4
9.2 ± 1.0
20
●
27
0.84
229.4 ± 24.3
11.5 ± 1.2
40
●
27
1.79
484.2 ± 52.3
12.1 ± 1.3
To validate the thickness calculated from the elemental analysis, we measured the
thickness of the Al2O3 film of the sample coated for 40 cycles using FIB combined
with TEM and EDX (Fig. 5.1 and Supplementary information 5.C and 5.D). To
prepare a lamella with FIB, we first deposited a protective layer of platinum of about
100 nm to avoid damaging the alumina film during the bombardment of ions
(Supplementary information 5.C). We placed the lamella under the TEM, and
measured a film thickness of about 400 nm (Fig. 5.1) after measuring the thickness at
50 points taken from two TEM images. This result is comparable to the one calculated
from elemental analysis, 484 ± 52 nm (Table 5.1).
Fig. 5.1. FIB-TEM image of the SiC sample coated for 40 cycles of Al2O3. The film
thickness is about 400nm.
96
Producing SiC tracer particles using gas-phase deposition of Al2O3 at ambient conditions
In Fig. 5.1, lighter areas in the alumina film are visible. These might be air pockets or
pores in the film. We verified the porosity of the alumina film of the sample coated for
40 cycles of Al2O3 with BET measurements (Supplementary information 5.E). The
specific surface area measured for the uncoated SiC particles, 0.12 m2/g, is near the
detection limit. The larger diameter of the coated sample produces a decrease in the
BET surface area, falling below the detection limit of the measuring device
(Supplementary information 5.E). That can be seen by the shape of the BET isotherms
for the coated sample, which is similar to that for an empty measuring probe.
Nevertheless, the pores seen in Fig. 5.1 would have produced a strong increase of the
surface area, especially since we are working near the detection limit. It might be that
these non-homogeneities found in the alumina film (Fig. 5.1) are either inaccessible to
the nitrogen during the BET measurement, or produced during the sample preparation
with FIB. In any case, these cannot be measured with nitrogen adsorption.
Fig. 5.2a shows the activity of the samples relative to the activity of the radioactive
water used in the activation, expressed as a percentage. Between the uncoated sample
(0.009%) and the γ-Al2O3 (0.143%), we observed an increase in the relative activity
with the thickness of the alumina film. The samples with the thin coating (5 and 7
cycles) showed a slight improvement with regard to the uncoated SiC. The experiment
with 20 cycles of Al2O3 at 27 °C showed a strong improvement of the activity, which
was further increased when the SiC particles were pre-functionalized with the oxygen
plasma. The sample with 40 cycles at 27 °C, and O2 plasma pre-treatment showed the
best relative activity, 0.108%. As we expected, the efficiency of the radio-activation
increased with the thickness of the alumina films (Fig. 5.2b). The sample with 40
cycles shows a relative activity comparable to the one of γ-Al2O3, often used as tracer.
Figure 5.2. (a) Relative activity, in percentage, of the uncoated SiC, coated samples
and the γ-Al2O3 sample. (b) Relative activity of the coated samples with respect to the
film thickness. In both pictures, the vertical error bars represent the standard deviation
of the activity measurements over the square root of the number of measurements. The
horizontal error bars in Fig. 5.2b represent the error introduced in the calculation of
the film thickness, based on the error of the ICP-OES equipment and the density of the
alumina film.
To evaluate the properties of the SiC sample coated with 40 cycles, we calculated an
equivalent density of the core-shell particle using (Eq. 5.1) and compared the value to
the uncoated SiC (Table 5.2). For that, we defined the core-shell density (
)
97
Chapter 5
as a weighted average considering the volume fractions of SiC (
coating (
).
(
(
) and the alumina
(Eq. 5.1)
(
We calculated the density of the core-shell for the 40-cycle sample, since it showed the
thickest alumina film and the highest relative activity with 18F. For this sample, an
alumina film of 484nm corresponds to a volume fraction of alumina (
) of 0.03.
Considering the density of SiC of 3210 kg/m3, and the density of Al2O3 deposited with
ALD at room temperature of 2500 kg/m3 [43], we obtained a density of the core-shell
of 3180 kg/m3. This represents a density difference of 0.8% with respect to the
uncoated SiC (Table 5.2), which is a negligible difference as far as the particle
dynamic behaviour is concerned. Moreover, we consider that neither the size nor the
shape of the SiC particles substantially changed with the alumina film, since the
thickness of the alumina film is much smaller than the particle diameter. Therefore, we
can conclude that the deposited alumina film does not alter the density, shape and size
of the SiC particles, which are the most relevant properties for the hydrodynamic
behaviour of fluidizing particles.
Table 5.2. Density comparison of the uncoated SiC
and the sample coated with 40 cycles.
SiC
Al2O3-SiC
68
68.8
[µm]
0.4
[µm]
1
0.97
[-]
0
0.03
[-]
3210
3185
[kg/m3]
We used a radio-activated particle of the SiC sample coated with Al 2O3 during 40
cycles to perform a PEPT experiment. During the 30 minutes that the experiment
lasted, we obtained over 17000 x,y,z locations in time (Fig. 5.3), which translates to an
average sampling frequency of about 10 Hz. The size of the tracer will affect the
activity, which influences the sampling data, and eventually may affect the precision of
the PEPT reconstruction; i.e. smaller particles result in lower activities which result in
lower spatial resolution [19].
In this work, we used as tracer the same 68 µm particles as in the rest of the bed,
coated with 40 cycles of aluminium oxide. The corresponding average activity was 22
µCi which allowed accurate measurement of the position of the tracer at a frequency of
10 Hz. The bare SiC particles, with an activity of 2 µCi, would have allowed only
measurements at a lower frequency (~1 Hz). The most relevant dynamics in fluidized
beds take place in the frequency range 0-6 Hz based on pressure fluctuations [52], or
even the lower part of that range based on solids motion. In the case of the experiment
with the 40-cycle SiC, we found that the power spectrum has little powder above 2 Hz,
such as sampling at 10 Hz is sufficient to capture all the dynamics, while sampling at 1
Hz would certainly miss relevant dynamics.
98
Producing SiC tracer particles using gas-phase deposition of Al2O3 at ambient conditions
Figure 5.3. (a) 3D representation of the trajectory of the SiC-Al2O3 tracer during the
10 first seconds of fluidization in the PEPT experiment. (b) Representation of all the
data points during the 30 minutes of the PEPT experiment. The projections of the data
in each of the planes are shown with red, green and blue symbols. (c) Mobility of the
tracer in each of the axes during the first 200 seconds of the PEPT experiments.
Within this 200 second period, in the first 20 seconds there is no fluidization, so that
the tracer remained stationary near the bottom of the bed of particles.
In Fig. 5.3a we see the trajectory of the tracer during the first 10 seconds of
fluidization. We observed that the intensity of the SiC sample coated with 40 cycles
provides sufficient data points for the PEPT algorithm to be applied and for description
of the location and movement of the tracer. In Fig. 5.3b, we show all the data points
recorded during the 30 minutes of the experiment. The projections of the data points on
the different planes give an idea of the uniform spread of the measured locations inside
the fluidized bed, illustrating that tracking was possible throughout the measurement
volume. Fig. 5.3c shows the mobility of the tracer in each of the three axes over a
duration of 200 seconds. During the first 20 seconds of the measurement (Fig. 5.3c),
there was no fluidization and the tracer rested almost at the bottom of the bed of
particles, and once the fluidization began, the tracer started moving inside the bed.
Despite the lower activity and sampling frequency for the alumina-coated SiC tracer
when compared to other tracers, the coating method used here allowed us to track and
reconstruct the 3D trajectory of a SiC particle, which would have been impossible
without the alumina coating. This generic approach could be extended to other types of
applications where a tracer particle is needed, such as PEPT or RPT (Radioactive
Particle Tracking).
5.4 Conclusions
We demonstrated that initially inert particles, such as SiC, can be activated with 18F
ions by modifying the surface of the primary SiC particles. For that, we deposited
aluminium oxide films on the SiC particles in a fluidized reactor using a gas-phase
coating technique similar to atomic layer deposition. Contrary to conventional ALD,
we carried out coating at atmospheric pressure and room temperature. At these
conditions, we fed the precursors in large excess to ensure a fast growth of the alumina
films. On the sample coated for 40 cycles, which was pre-functionalized with O2
plasma, we deposited a film of about 500 nm, resulting in a GPC of about 12 nm. This
40-cycles coated sample presented a labelling efficiency with 18F similar to that for the
γ-Al2O3, which is often used as a tracer. We conclude that the layer is thick enough to
enable sufficient activity, yet thin enough to make the changes in density, size and
shape of the particles negligible.
99
Chapter 5
We showed that the activated core-shell structure formed by the SiC particle coated
with an Al2O3 film of about 500 nm can be used as a tracer particle in a typical PEPT
experiment. The emission intensity of this tracer was sufficient to reconstruct its
trajectory inside the bed of particles albeit at suboptimal performance (i.e. low location
rate and corresponding lower precision than generally quoted for PEPT). That proves
that the deposition of alumina films can be used to produce tracers, mainly consisting
of the same material as the bulk. This will enable researchers to obtain more accurate
information about the flow patterns in systems with moving particles.
Bibliography
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
Parker, D. J.; Broadbent, C. J.; Fowles, P.; Hawkesworth, M. R.; McNeil, P.,
Positron emission particle tracking - a technique for studying flow within
engineering equipment. Nuclear Instruments and Methods in Physics Research
Section A: Accelerators, Spectrometers, Detectors and Associated Equipment
1993, 326, (3), 592-607.
Parker, D. J.; Forster, R. N.; Fowles, P.; Takhar, P. S., Positron emission
particle tracking using the new Birmingham positron camera. Nuclear
Instruments and Methods in Physics Research Section A: Accelerators,
Spectrometers, Detectors and Associated Equipment 2002, 477, (1–3), 540-545.
Seville, J. P. K.; Ingram, A.; Parker, D. J., Probing processes using positrons.
Chemical Engineering Research and Design 2005, 83, (7 A), 788-793.
Hoffmann, A. C.; Dechsiri, C.; Van De Wiel, F.; Dehling, H. G., PET
investigation of a fluidized particle: Spatial and temporal resolution and short
term motion. Measurement Science and Technology 2005, 16, (3), 851-858.
Chaouki, J.; Larachi, F.; Duduković, M. P., Noninvasive Tomographic and
Velocimetric Monitoring of Multiphase Flows. Industrial and Engineering
Chemistry Research 1997, 36, (11), 4476-4503.
Parker, D. J.; Fan, X., Positron emission particle tracking—Application and
labelling techniques. Particuology 2008, 6, (1), 16-23.
Fan, X.; Parker, D. J.; Smith, M. D., Labelling a single particle for positron
emission particle tracking using direct activation and ion-exchange techniques.
Nuclear Instruments and Methods in Physics Research Section A: Accelerators,
Spectrometers, Detectors and Associated Equipment 2006, 562, (1), 345-350.
Volkwyn, T. S.; Buffler, A.; Govender, I.; Franzidis, J. P.; Morrison, A. J.; Odo,
A.; van der Meulen, N. P.; Vermeulen, C., Studies of the effect of tracer activity
on time-averaged positron emission particle tracking measurements on tumbling
mills at PEPT Cape Town. Minerals Engineering 2011, 24, (3–4), 261-266.
Bickell, M.; Buffler, A.; Govender, I.; Parker, D. J., A new line density tracking
algorithm for PEPT and its application to multiple tracers. Nuclear Instruments
and Methods in Physics Research Section A: Accelerators, Spectrometers,
Detectors and Associated Equipment 2012, 682, (0), 36-41.
Boucher, D.; Deng, Z.; Leadbeater, T.; Langlois, R.; Renaud, M.; Waters, K. E.,
PEPT studies of heavy particle flow within a spiral concentrator. Minerals
Engineering 2014, 62, (0), 120-128.
Chan, C. W.; Brems, A.; Mahmoudi, S.; Baeyens, J.; Seville, J.; Parker, D.;
Leadbeater, T.; Gargiuli, J., PEPT study of particle motion for different riser
exit geometries. Particuology 2010, 8, (6), 623-630.
Chan, C. W.; Seville, J.; Fan, X.; Baeyens, J., Solid particle motion in a
standpipe as observed by Positron Emission Particle Tracking. Powder
Technology 2009, 194, (1-2), 58-66.
Flamant, G.; Gauthier, D.; Benoit, H.; Sans, J. L.; Garcia, R.; Boissière, B.;
Ansart, R.; Hemati, M., Dense suspension of solid particles as a new heat
100
Producing SiC tracer particles using gas-phase deposition of Al2O3 at ambient conditions
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
transfer fluid for concentrated solar thermal plants: On-sun proof of concept.
Chem. Eng. Sci. 2013, 102, 567-576.
Flamant, G.; Gauthier, D.; Benoit, H.; Sans, J. L.; Boissière, B.; Ansart, R.;
Hemati, M., A New Heat Transfer Fluid for Concentrating Solar Systems:
Particle Flow in Tubes. Energy Procedia 2014, 49, (0), 617-626.
Kunii, D.; Levenspiel, O., Fluidization Engineering. Second ed.; ButterworthHeinemann: 1991.
Lyngfelt, A.; Leckner, B.; Mattisson, T., A fluidized-bed combustion process
with inherent CO2 separation; application of chemical-looping combustion.
Chem. Eng. Sci. 2001, 56, (10), 3101-3113.
Zhong, W.; Jin, B.; Zhang, Y.; Wang, X.; Xiao, R., Fluidization of Biomass
Particles in a Gas−Solid Fluidized Bed. Energy & Fuels 2008, 22, (6), 41704176.
Valdesueiro, D.; Meesters, G.; Kreutzer, M.; van Ommen, J., Gas-Phase
Deposition of Ultrathin Aluminium Oxide Films on Nanoparticles at Ambient
Conditions. Materials 2015, 8, (3), 1249-1263.
Cole, K. E.; Buffler, A.; van der Meulen, N. P.; Cilliers, J. J.; Franzidis, J. P.;
Govender, I.; Liu, C.; van Heerden, M. R., Positron emission particle tracking
measurements with 50 micron tracers. Chem. Eng. Sci. 2012, 75, 235-242.
Kasprzyk-Hordern, B., Chemistry of alumina, reactions in aqueous solution and
its application in water treatment. Advances in Colloid and Interface Science
2004, 110, (1–2), 19-48.
Geldart, D., Types of gas fluidization. Powder Technology 1973, 7, (5), 285292.
Fan, X.; Parker, D. J.; Smith, M. D., Adsorption kinetics of fluoride on low cost
materials. Water Research 2003, 37, (20), 4929-4937.
Puurunen, R. L., Growth Per Cycle in Atomic Layer Deposition: A Theoretical
Model. Chemical Vapor Deposition 2003, 9, (5), 249-257.
George, S. M., Atomic layer deposition: An overview. Chem. Rev. 2010, 110,
(1), 111-131.
Goulas, A.; Ruud van Ommen, J., Atomic layer deposition of platinum clusters
on titania nanoparticles at atmospheric pressure. Journal of Materials Chemistry
A 2013, 1, (15), 4647-4650.
Elliott, S. D., Mechanism, products, and growth rate of atomic layer deposition
of noble metals. Langmuir 2010, 26, (12), 9179-9182.
Moghtaderi, B.; Shames, I.; Doroodchi, E., Combustion prevention of iron
powders by a novel coating method. Chemical Engineering and Technology
2006, 29, (1), 97-103.
King, D. M.; Liang, X.; Burton, B. B.; Kamal Akhtar, M.; Weimer, A. W.,
Passivation of pigment-grade TiO2 particles by nanothick atomic layer
deposited SiO2 films. Nanotechnology 2008, 19, (25).
Wank, J. R.; George, S. M.; Weimer, A. W., Coating fine nickel particles with
Al2O3 utilizing an atomic layer deposition-fluidized bed reactor (ALD-FBR).
Journal of the American Ceramic Society 2004, 87, (4), 762-765.
Hakim, L. F.; Blackson, J.; George, S. M.; Weimer, A. W., Nanocoating
individual silica nanoparticles by atomic layer deposition in a fluidized bed
reactor. Chemical Vapor Deposition 2005, 11, (10), 420-425.
King, D. M.; Spencer Ii, J. A.; Liang, X.; Hakim, L. F.; Weimer, A. W., Atomic
layer deposition on particles using a fluidized bed reactor with in situ mass
spectrometry. Surface and Coatings Technology 2007, 201, (22-23 SPEC. ISS.),
9163-9171.
Salameh, S.; Schneider, J.; Laube, J.; Alessandrini, A.; Facci, P.; Seo, J. W.;
Ciacchi, L. C.; Mädler, L., Adhesion mechanisms of the contact interface of
TiO 2 nanoparticles in films and aggregates. Langmuir 2012, 28, (31), 1145711464.
George, S. M.; Ott, A. W.; Klaus, J. W., Surface chemistry for atomic layer
growth. Journal of Physical Chemistry 1996, 100, (31), 13121-13131.
101
Chapter 5
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
Socha, R. P.; Laajalehto, K.; Nowak, P., Influence of the surface properties of
silicon carbide on the process of SiC particles codeposition with nickel.
Colloids and Surfaces A: Physicochemical and Engineering Aspects 2002, 208,
(1–3), 267-275.
Cicero, G.; Catellani, A.; Galli, G., Atomic Control of Water Interaction with
Biocompatible Surfaces: The Case of SiC(001). Physical Review Letters 2004,
93, (1), 016102.
Hollahan, J. R.; Carlson, G. L., Hydroxylation of polymethylsiloxane surfaces
by oxidizing plasmas. Journal of Applied Polymer Science 1970, 14, (10), 24992508.
Morra, M.; Occhiello, E.; Garbassi, F., Contact angle hysteresis in oxygen
plasma treated poly(tetrafluoroethylene). Langmuir 1989, 5, (3), 872-876.
Beetstra, R.; Lafont, U.; Nijenhuis, J.; Kelder, E. M.; Van Ommen, J. R.,
Atmospheric pressure process for coating particles using atomic layer
deposition. Chemical Vapor Deposition 2009, 15, (7-9), 227-233.
Dwivedi, V.; Adomaitis, R. A. In Multiscale simulation and optimization of an
atomic layer deposition process in a nanoporous material, 17th International
Chemical Vapor Deposition Symposium (CVD-XVII) - 216th Meeting of the
Electrochemical Society, Vienna, 2009; Vienna, 2009; pp 115-122.
Henrickson, C. H.; Eyman, D. P., Lewis acidity of alanes. Interactions of
trimethylalane with sulfides. Inorganic Chemistry 1967, 6, (8), 1461-1465.
Laubengayer, A. W.; Gilliam, W. F., The Alkyls of the Third Group Elements.
I. Vapor Phase Studies of the Alkyls of Aluminum, Gallium and Indium1.
Journal of the American Chemical Society 1941, 63, (2), 477-479.
Mayer, B.; Collins, C. C.; Walton, M., Transient analysis of carrier gas
saturation in liquid source vapor generators. Journal of Vacuum Science &
Technology A 2001, 19, (1), 329-344.
Groner, M. D.; Fabreguette, F. H.; Elam, J. W.; George, S. M., LowTemperature Al2O3 Atomic Layer Deposition. Chemistry of Materials 2004,
16, (4), 639-645.
Wirth, R., Focused Ion Beam (FIB) combined with SEM and TEM: Advanced
analytical tools for studies of chemical composition, microstructure and crystal
structure in geomaterials on a nanometre scale. Chemical Geology 2009, 261,
(3–4), 217-229.
Fan, X.; Parker, D. J.; Smith, M. D., Enhancing 18F uptake in a single particle
for positron emission particle tracking through modification of solid surface
chemistry. Nuclear Instruments and Methods in Physics Research, Section A:
Accelerators, Spectrometers, Detectors and Associated Equipment 2006, 558,
(2), 542-546.
Hakim, L. F.; Portman, J. L.; Casper, M. D.; Weimer, A. W. In Fluidization
behavior and conformal coating of nanoparticles in fluidized beds by ALD,
Austin, TX, 2004; Austin, TX, 2004; pp 2339-2359.
Wank, J. R.; Hakim, L. F.; George, S. M.; Weimer, A. W., Coating fine iron
particles with an oxidation-resistant alumina monolayer using ALD in a
fluidized bed reactor Fluidization XI - Present and Future of Fluidization
Engineering, ECI International (Brooklyn, NY). U. Arena, R. Chirone, Ml.
Miccio, and P. Salatino, editors. p. 603-610 2004, 8.
Ferguson, J. D.; Weimer, A. W.; George, S. M., Atomic layer deposition of
ultrathin and conformal Al2O3 films on BN particles. Thin Solid Films 2000,
371, (1), 95-104.
Hakim, L. F.; McCormick, J. A.; Zhan, G. D.; Weimer, A. W.; Li, P.; George,
S. M., Surface modification of titania nanoparticles using ultrathin ceramic
films. Journal of the American Ceramic Society 2006, 89, (10), 3070-3075.
Fujino, K.; Nishimoto, Y.; Tokumasu, N.; Maeda, K., Silicon Dioxide
Deposition by Atmospheric Pressure and Low‐Temperature CVD Using TEOS
and Ozone. Journal of the Electrochemical Society 1990, 137, (9), 2883-2887.
Hausmann, D.; Becker, J.; Wang, S.; Gordon, R. G., Rapid vapor deposition of
highly conformal silica nanolaminates. Science 2002, 298, (5592), 402-406.
102
Producing SiC tracer particles using gas-phase deposition of Al2O3 at ambient conditions
52.
53.
van Ommen, J. R.; de Korte, R.-J.; van den Bleek, C. M., Rapid detection of
defluidization using the standard deviation of pressure fluctuations. Chemical
Engineering and Processing: Process Intensification 2004, 43, (10), 1329-1335.
Travis, C.; Adomaitis, R., Modeling alumina atomic layer deposition reaction
kinetics during the trimethylaluminum exposure. Theor Chem Acc 2013, 133,
(1), 1-11.
103
Chapter 5
5.A Properties of SiC powder and γ-Al2O3 used as tracer
Fig. 5.A.1. TEM picture and PSD of the F220 SiC used in this work.
Table 5.A.1. Physical properties of SiC F220
and γ- Al2O3 used in this work. Information
provided by the SiC F220 provider
SiC F220 γ- Al2O3
3
[kg/m ]
3210
3950
3
[kg/m ]
1550
1220
[µm]
68
48
3
[kJ/m ·K]
2150
[°C]
1620
[W/m·K]
18
[mm/s]
5.5
[mm/s]
8
-
104
Producing SiC tracer particles using gas-phase deposition of Al2O3 at ambient conditions
5.B Calculation of the dosing time the precursors
Table 5.B.1. Properties of SiC F220 particles and fluidization experiment.
BET SiC [m2/g]
0.12
mass SiC [g]
8.00
68
[µm]
Total Surf Area [m2]
0.96
Flow N2 [L/min]
0.8
Table 5.B.2. Parameters for TMA [53] used in the steric hindrance model[23]
[nm]
0.125
[m2]
5.4 E-20
[molec]
1.8 E+19
[mol]
2.9 E-05
[mol]
4.4 E-05
Table 5.B.3. Calculation of the dosage times based on the ideal gas law.
0.021
[atm]
0.042
[atm]
[°C]
30
[mol/L]
8.6 E-4
[mol/min] 6.9 E-4
[s]
2.6
[mol/L]
1.7 E-3
[mol/min] 1.4 E-3
[s]
2.0
105
Chapter 5
5.C Preparation of the sample lamella with FIB-TEM
Fig. 5.C.1. Steps in the preparation of the lamella with FIB-TEM.
106
Producing SiC tracer particles using gas-phase deposition of Al2O3 at ambient conditions
5.D FIB-TEM and EDX spectra
Fig. 5.D.1. TEM image of the lamella of the coated
sample, and the EDX spectra for the different
components.
107
Chapter 5
5.E Study of the porosity of the alumina films
Fig. 5.E.1. Absorption and desorption
isotherms for the uncoated SiC F220 powder
(a), and the sample coated with 40 cycles (b).
108
Chapter 6
Outlook
The purpose of this chapter is to identify the questions that remained open and to provide a
broader perspective on the applicability of our work. We describe how our approach can be
further refined by optimizing certain process variables, such as the dosing time of the
precursors and the purging time. We also propose a detailed study on how to improve the
fluidization of fine cohesive particles. In addition, we discuss the deposition of other
compounds at ambient conditions, and the necessity to study the toxicology of the ultrathin
coatings before being applied to biological materials. Finally, we present an evaluation of
ALD for product development.
111
Chapter 6
6.1 OPEN QUESTIONS AND COMPLEMENTARY STUDIES
These are the following six key questions that are addressed in this section, including
suggestions for further investigation.
(i)
Optimization of the results obtained in the previous chapters
(ii)
Characterization of the strength of the alumina films
(iii) Necessity for accurate characterization techniques
(iv) Industrial implementation of ALD on powders
(v)
Investigation of the interparticle forces
(vi) Study of different fluidization assistance methods
6.1.1 Optimization of the results obtained in the previous chapters
In chapters II, III, IV and V, we showed how thin alumina films, deposited at ambient
conditions, improved the performance of three materials: a powder coating paint, PbSe
quantum dot (QD) films, and a silicon carbide (SiC) tracer. Despite that, further research
could be conducted to refine the production of these materials towards industrial
implementation. In this section we discuss the variables that have an influence in the coating
process, and how we can modify them to optimize the end product.
In chapter II, alumina films were deposited at ambient conditions with the same growth per
cycle (GPC) as in pure atomic layer deposition (ALD) processes, i.e., 0.1-0.2 nm, by
controlling the dosing time of the precursors. Also, an increase in the dosing time of the
precursors resulted in thicker alumina films, and therefore, higher GPC. At ambient
conditions, the purging flow of nitrogen was not sufficient to remove the unreacted precursor
molecules from the surface of the substrates. It would be worthwhile to investigate whether
we could achieve an efficient removal of the accumulated molecules with longer purging
times. And complementarily, we could study whether using shorter purging time would have
an influence on the film thickness. Unravelling the influence of the purging time on the
alumina film thickness would be useful to either eliminate the parasitic CVD growth of
alumina, or to reduce the process time by shortening the purging step. For example, a purging
reduction of 50%, from 10 to 5 minutes, would be reflected in a cycle time decrease of about
15%, on average, for the experiments shown in chapters II, III, IV and V.
Reduction of the cycle time can be also achieved by playing with the temperature of the
precursor bubblers. An increase in this temperature results in an increase in the concentration
of precursor in the nitrogen flow to the reactor. Therefore, less time would be required to
transport a certain amount of precursor molecules to the reactor. As an example, the vapour
pressure of trimethylaluminium (TMA) at 30 °C is 21 mbar, while at 80 °C is 200 mbar. This
increase in the temperature translates in a ten times reduction of the dosing time of TMA.
Nevertheless, we did not consider this possibility in our experiments since we wanted to keep
the reactor at room temperature during the whole coating process because of the thermal
sensitivity of the materials we worked with, i.e., polymeric particles and quantum dot crystals.
112
Outlook
In chapter III, we showed the production of a matte powder coating paint by encapsulating the
particles of a glossy powder coating with alumina. We found that 2 cycles influenced the final
surface properties of the powder coating paint without altering its mechanical properties.
However, there are a number of variables in this process that can be further optimized to
increase the tunability and production rate of the matte powder coating product.
1. We should investigate different ways to feed faster the same amount of precursors to the
reactor by increasing the dosing time and reducing the number of cycles simultaneously. We
could benefit from the accumulation of precursor molecules on the substrate to deposit the
alumina films faster. A hot stage microscope could be used to quickly determine if the
alumina coating with longer dosing times shows similar barrier performance as the one shown
in chapter III.
2. The particle size of the powder coating could have also an influence, since the gloss or
matte surface appearance depends on the surface roughness. Smooth surface would reflect the
incident light, while irregular surface would scatter it, producing a glossy and matte surface,
respectively. In our opinion, smaller particle sizes would reduce the surface roughness while
keeping the matte finish, since it is likely that the encapsulated particles could still scatter the
incident light. This would provide more options to finely tune the appearance of the powder
coating paint.
3. It would be interesting to investigate a reduction of the coated-to-uncoated mass ratio of
powders used in the preparation of the mixed paints. In chapter III, a ratio 1:1 was used.
However, by reducing this ratio we could produce larger amounts of matte powder coating
paint with the same amount of alumina-coated powder. It has to be investigated the properties
of the final product, such as roughness, gloss and mechanical resistance, with different mass
ratios.
In chapter IV, PbSe QD films, which are prone to oxidation, were stabilized by depositing
aluminium oxide at ambient conditions. According to theoretical estimation, we deposited
around four monolayers of alumina in each cycle, finding that after 25 coating cycles the QD
films were successfully passivated for at least 60 days in an air environment. Despite these
good results, we still do not know the minimum amount of alumina required to passivate the
PbSe crystals. For that, we would need to perform experiments with a number of cycles
between 10 and 25, to determine the minimum number needed. Additionally, we should
increase the dosing times of the precursors to deposit the same amount alumina with lower
number of cycles. The concentration depth profile of the different elements, obtained with
XPS (X-ray photoelectron spectroscopy), and the passivation efficiency of the alumina
coating, assessed with the measurement of the optical absorption spectra, would be used to
compare different samples. Furthermore, other coating materials, such as zinc oxide (ZnO),
could be used to passivate the QD films. In principle, the deposition of ZnO, using diethyl
zinc (DEZ) and water as precursors, could be carried out at room conditions since DEZ, also a
metal alkyl, is as reactive as TMA. Additionally, ZnO would increase the electron mobility
inside the QD film [1]. This study would provide a promising alternative to the production of
QD films with enhanced properties, and could improve the efficiency of the solar cell
technology.
In chapter V, we deposited alumina films on SiC particles to improve the labelling efficiency
of the core materials towards 18F radioactive ions. The best result suggested that even if the
113
Chapter 6
SiC particles coated with 400 nm of aluminium oxide had a slightly lower relative activity
than the γ-Al2O3 particles, they had sufficient activity to be detected by the positron emission
particle tracking (PEPT) sensors. However, a thicker film of alumina could be deposited to
investigate the influence of the film thickness on the activity of the core-shell particle. This
could be tested by increasing the dosing times of the precursors, increasing the number of
cycles, or both. A higher activity of the tracer would result in a higher sampling frequency,
and therefore in a more accurate reconstruction of the trajectory of the tracer. Nevertheless,
the sampling frequency not only depends on the activity of the tracer, but also on other
parameters such as the diameter of the column, the nature of the material of study and the
velocity of the tracer.
In these three applications, we have benefited from the accumulation of unreacted precursor
molecules at atmospheric pressure and room temperature to deposit several layers of alumina
per cycle. Typical values of the GPC of alumina are 0.1-0.2 nm, while in our experiments we
had film growths ranging from 0.14 to 13 nm of alumina per cycle. Otherwise, we would have
had to carry out a much larger number of cycles in each experiment to obtain similar results.
Therefore, it would be interesting, as stated above, to investigate a further increase in the
dosing times to further reduce the process time, and thus, increase the production rate. This
study could be combined with the investigation of the alumina film properties, such as
density, porosity or presence of unreacted species. These measurements can be done using
TGA (thermo gravimetric analysis), BET N2 adsorption (Brunauer-Emmett-Teller) and ICPOES (induced coupled plasma – optical emission spectroscopy) combined with TEM
(transmission electron microscopy). Reducing the dosing and purging times could result in a
shortened deposition of alumina, while maintaining good control over the film thickness.
From all these parameters, we consider that the most interesting to investigate are the ones
that lead to a reduction in the cycle time, i.e., the dosing and purging times, since they could
be directly applied in the work shown in this thesis. Additionally, obtaining a similar alumina
coating in a shorter time would be interesting for an industrial implementation.
6.1.2 Characterization of the strength of the alumina coating
During the coating process in a fluidized bed, particles are constantly colliding with other
particles and the walls of the reactor. Therefore, investigating the physical properties of the
coated particles, and in particular the mechanical resistance of the deposited alumina films,
would be important for the production and final application of the core-shell structures.
Techniques such as nano-indentation and attrition tests could be used to determine the
strength, Young modulus, Poisson’s ration, coefficient of restitution and attrition resistance of
the alumina films. The nano-indenter provides force-displacement curves of a tip penetrating
the coating. Nevertheless, this measurement is challenging for particles, since the particles can
have different facets, and should remain immobile once the indenter is applied on their
surface. The force-displacement forces are used to determine the energy dissipated during the
indentation, giving a measurement of the elasticity of the alumina coatings. On the other hand,
the attrition test could be done with a vortex shaker [2]. In this technique, a mass of particles
is introduced in a glass cylinder which is placed in the vortex shaker with a rotational speed
from 1500 to 2500 rpm. The glass cylinder has a input for a gas, and an output that is
connected to a particle sizer device to measure the distribution of the fines formed. A sample
of the fines generated could be analysed with EDX, to determine whether the fines are coming
114
Outlook
from the core or the shell materials. This measurement provides an indication of the strength
of the particles and the coating upon collision.
6.1.3 Necessity for accurate characterization techniques
Good analytical characterization of the core-shell particles is crucial to interpret the
experimental results correctly. Different techniques rely on different properties to characterize
the material. For example, transmission electron microscope (TEM) can be used to clearly
measure a film thickness of few nanometres of alumina deposited on TiO2 nanoparticles, as in
chapter II, since both have a similar length scale. However, measuring the thickness of a few
nanometre film deposited on a 30 µm polymer particle, as in chapter III, is not
straightforward. In this section we discuss the capabilities and limitations of different
analytical techniques to characterise ultrathin films in both nanoparticles and micron-size
particles.
The core-shell particles produced in this thesis could be classified in two groups: /
~
0.1 and /
> 500, where
is the particle diameter and
is the film thickness. The
first group corresponds to the core-shells in which film thickness is approximately one order
of magnitude smaller than the particle size, as in the coating of TiO2 nanoparticles (chapter
II). This material offered a straightforward characterization, since the size and volume of the
core and coating are in the same order of magnitude. This allows direct measurement of the
composition with ICP-OES, and easy visualization with TEM. In addition, the large values of
the specific surface area of nanoparticles, usually > 10 m2/g, enable the use of surface
techniques such as BET N2 adsorption, XRD (X-ray diffraction spectroscopy), FTIR (Fourier
transform infrared spectroscopy) and TGA, amongst others. These techniques are used to
measure the surface area and porosity of a sample (BET), crystal structure and composition of
the coating (XRD and FTIR), or the humidity content and the presence of any other volatile
components (TGA). Therefore, we can accurately determine the chemical composition and
infer other properties of the coating such as density. On the other hand, some techniques are
not suitable for nanoparticles or nanoparticle agglomerates, such as nano-indentation. This
technique can provide information on the mechanical properties of the alumina coatings.
However, there are some challenges in its application. Namely, the size of the indenter tip is
similar to the size of the nanoparticle, and the non-uniform geometry of the agglomerates
complicates the interpretation of the force-displacement measurements.
Larger core-shell particles, in which the size of the core is in the micron range and the film in
the nanometre range, disallow the use of some of these techniques, as the measured values
would approach the detection limit of the equipment. For instance, ICP-OES would measure
mainly the composition of the core material, since the volume of the coating is very small
compared to the volume of the particle. In the case of the SiC tracer (chapter V), the volume
of the core material (68 µm SiC) represents 97% of the total volume of the core-shell (Al2O3
film thickness of 400 nm). Therefore, the compounds of the coating would be found in very
low concentrations. Additionally, using TEM is not straightforward for two reasons. First, the
microscope has to use large magnification on the edge of the particle, which may induce
artefacts on the images. And secondly, the micron sized particles would absorb a large part of
the beam of electrons causing low contrast between the particles and the coating. Surface
techniques, such as FTIR, BET or XRD, are not practical due to the low values of the specific
surface area of the non-porous micron sized particles, which is about 0.1 m2/g. For example,
BET requires a total surface area of about 5 m2 to measure accurately the specific surface
115
Chapter 6
area. However, the testing tube can only allocate a few grams of powder. Some techniques
such as FIB-TEM (focused ion beam transmission electron microscope) can be used to
observe the film thickness of the larger core-shell particles. A thin lamellae of the core-shell,
that is prepared with FIB by ion bombardment, can be introduced into the TEM to measure
the film thickness and perform elemental analysis with EDX (energy dispersive X-ray
spectroscopy). However, the sample preparation is time consuming and requires specific
equipment. Hence, FIB-TEM is not an optimal technique to measure the film thickness in a
systematic manner. Finally, other techniques to study the macroscopic behaviour of the coreshell, such as DSC (differential scanning calorimetry) and HSM (hot state microscopy), are
favoured with larger particle sizes.
In general, there is a large interest in the coating of micron and nano-size particles to be used
in catalysis, food, pharmaceutical and other fields. Depositing ultrathin films of a wide range
of materials by ALD for tailored functionalization is of especial interest [2]. For micron-size
particles, characterization is considerably more challenging, while particle processing is more
straightforward, especially in fluidized beds. On other hand, handling and coating of
nanoparticles is not an easy task, one reason being the strict operation protocols to avoid
dispersion of nanoparticles onto the environment. Nonetheless, the nano scale offers more
techniques for simple and accurate characterization of the coated nanoparticles.
6.1.4 Industrial implementation of ALD on powders
ALD has been used to produce tailored materials with improved performance in a variety of
fields, such as catalysis, solar cells, Li-ion batteries and many others [3]. However, these
experiments were done, in most of the cases, at lab scale [4, 5]. Based on the enhanced
performance of the materials after depositing a few nanometre films, the next step would be
the implementation at an industrial scale. In this section, we briefly delineate some general
ideas on how to attempt an scaled up production of engineered materials using ALD.
Reactors for coating of powder can have different designs [5]. Ongoing work at the Product
and Process Engineering (PPE) group (Chemical Engineering Department, Faculty of Applied
Sciences, Delft University of Technology) focuses on studying two reactor configurations: a
fluidized bed reactor operating in batch mode, and a pneumatic transport reactor operating in
a continuous mode [6, 7]. Some advantages and disadvantages of both reactors are discussed
below.
The fluidization technology is well-known and widely expanded in industrial activities such
as food and pharmaceutical. Fluidized bed reactors (FBR) allow the processing of a large
amount of solids of a variety of particle sizes. Additionally, FBRs can operate optimally with
different capacities of solids, since it only requires the adaptation of the dosing times to the
amount of solids inside the reactor. A down side of large fluidized beds would be the mixing
between the precursors and solids, since the hydrodynamic behaviour of the solid can
deteriorate when scaling up the diameter of the fluidized beds [8]. The second reactor
configuration is a pneumatic transport reactor [6], that consists of a tube through which the
particles are transported in a diluted gas stream of nitrogen, with the precursors being injected
at different locations. This type of reactor provides a continuous production of nano-structure
materials [7]. The current prototype can perform one cycle of ALD; nevertheless, the scaling
up of this reactor can be done by connecting several one-cycle reactors in series, in a modular
manner, to perform more than one cycle. This reactor has a length of about 37 m, giving
116
Outlook
enough time for the precursors to mix and react at the particle surface. Scaling up of these two
reactor configurations is a worthwhile research challenge, providing a promising future for
industrial ALD production.
6.1.5 Investigation of the interparticle forces
The fluidization of cohesive particles, as described in the introduction of this thesis, is
hindered by the tendency of the particles to form agglomerates. This results in the formation
of preferential paths and channels, dead zones, poor gas-solid mixing, and non-suspended
particles [9, 10]. Three interparticle forces are considered to play a major role in the particle
agglomeration in fluidized beds: van der Waals, electrostatics and hydrogen bonding [11-13],
while capillary force would be only relevant at high humidity level, usually above 70% [14].
However, the role of each force is not well defined yet.
Determining the contribution of each force is crucial to reduce or prevent the influence of the
agglomeration on the fluidization. The van der Waals force is an inherent force based on the
properties of the particles, thus, it would be difficult to influence without changing the
material of study. Electrostatics can be prevented by grounding the equipment, or dispersing a
solvent in the fluidization gas to increase its conductivity and dissipate the charges inside the
reactor [15]. Similarly, dispersing isopropyl alcohol in the gas stream was reported to reduce
the hydrogen bonding between hydrophilic particles [12]. This research would investigate the
influence of dispersing a liquid in the fluidizing gas on the electrostatic forces and the
hydrogen bond formation. For this study, 15 µm SiO2 particles were fluidized in a pseudo-2D
column, in which several induction probes we placed in contact with the external wall of the
column. The induction probes consisted of an isolated co-axial cable connected to a copper
plate on one side, which was in contact with the column, and to an amplifier on the other side.
The probe could detect the change in the charge related to the agglomerates fluidizing in front
of the probe [16-18]. Additionally, we could measure the bed height for a certain fluidization
gas velocity, and the pressure drop and pressure fluctuations at different points of the bed, to
characterize the quality of fluidization. This study would be done using two types of particles,
i.e., hydrophilic and hydrophobic SiO2, and three different fluidizing gases, i.e., dry nitrogen,
nitrogen with a relative humidity of 20% and nitrogen with dispersed isopropyl alcohol. These
six experiments would provide an evidence on the influence of the solvent used in any of the
interparticle forces considered, namely, electrostatics and hydrogen bonding.
This study would contribute in the scaling up of a fluidized bed system. Nevertheless, care
must be taken when using this approach in the coating process. Altering the composition of
the fluidizing gas by dispersing a solvent will, most likely, induce undesired reactions with
the coating precursors.
6.1.6 Study of different fluidization assistance methods
Despite the agglomeration of fine cohesive particles during processing in FBR, a good
fluidization can be achieved by supplying external energy to the system in the form
fluidization assistance methods. This supply could be as mechanical and acoustic vibration,
mechanical stirring, secondary gas injection using a microjet, pulsed gas flow, centrifugal and
electric field, amongst others [10, 19-22]. Although many of them have been investigated,
there is not a detailed comparison. It would be useful to define an optimal fluidization aid,
117
Chapter 6
based on the amount of energy added to the system and its influence on the fluidization
quality. In order to find the optimum, one could use a model to approximate the microscopic
properties of the agglomerates, such as size and porosity, by measuring the macroscopic
properties of the system, i.e., bed height [23, 24]. This model is a modified version for
nanoparticle agglomerates of Richardson-Zaki equation [25], which describes the settling
mechanism of a particle surrounded by other particles. By a simple measurement of the
settling velocity of the bed once the fluidization gas is switched off, we can obtain
information about the structure of the fluidized agglomerates. The validation of this model
can be done using a high speed camera connected to a borescope which is inserted in the
fluidized bed [26], allowing the measurement of the size and settling velocity of the fluidized
agglomerates. Additionally, the energy balances of the fluidizing system would be a
complementary tool to study the different assistance methods.
6.2 BROADENING THE APPLICABILITY OF THIS WORK
6.2.1 Deposition of other compounds at ambient conditions
In chapter II, we demonstrated the deposition of ultrathin films of alumina, using TMA and
water as precursors, at 27 °C and 1 bar. But, is it possible to deposit other coating materials at
room temperature and atmospheric pressure? Alumina can be deposited at ambient conditions
due to the high reactivity of the TMA molecules, which readily react in the presence of a
small concentration of oxygenated compounds, such as air or water, even at room
temperature. This high reactivity is characteristic for metal alkyls, which are pyrophoric, so
they can catch fire when in contact with an oxidizer, such as TMA. Other metal alkyls that
could be used for ALD reactions at moderate temperature are zinc, boron and magnesium
alkyls [27-29], since supposedly they are also highly reactive.
6.2.2 Biocompatibility study of ultrathin films
The deposition of alumina at atmospheric pressure and room temperature opens the door to
the application of nanometre coatings to heat sensitive materials such as polymers, food and
pharmaceutical products. However, the toxicology of the alumina films should be
investigated, especially if we use biological compounds as core material. Some studies have
been done on the toxicity of metal nanoparticles on E-coli bacteria and zebrafish embryos
[30-35]. Thus, can we use a similar approach to study the toxicology of nanometre-thick
alumina coatings? For that, we can describe some experiments to study their toxicity. We
would select a non-toxic powder as material of study, such as a food ingredient, and we would
deposit alumina films with different thicknesses at ambient conditions. To investigate the
toxicology, we prepare cultures of our cell of study, e.g., E-coli, and assess their reproduction
capability. Under the same conditions, we prepare identical cultures and add different
amounts of the coated powders and same amounts of powder with different film thickness, to
investigate the effect of the coated powder amount and alumina film thickness on the E-coli.
In parallel, we would prepare a control experiment consisting of a culture with the uncoated
powder. In this way, we can compare whether the alumina films have a toxic effect on the
bacteria, and whether this toxicity level would depend on the amount of powder and/or
thickness of the alumina film. A follow up study could use a living system that resembles
more the human cells. Applying nanocoatings to biological materials, i.e., food and
118
Outlook
pharmaceutical compounds, could be of great interest. However, toxicology studies such as
the one described in this section are necessary to determine the effects of these ultrathin
coatings on living systems.
6.3 ANALYSIS OF ALD FOR PRODUCT DEVELOPMENT
ALD is a powerful coating technique since it provides atomic control over the amount of
compound deposited on a substrate, i.e., flat supports, irregular geometries, and porous and
non-porous powders [36]. In this section, we describe why ALD is an interesting technique to
produce nanostructure materials, from the process and product point of view.
Looking from a process perspective, there is a need to move towards sustainable production
methods to reduce the impact on the environment in general. In this respect, ALD scores
highly as a sustainable technology. As being a gas-phase technique, the mass of waste per
mass of product, the so-called E-factor [37], is lower than in liquid-based coating techniques.
Such techniques usually require an energy-intensive drying step to evaporate the solvent used,
which is absent in ALD. Although several ALD processes, such as the ones to deposit noble
metals, require relatively high temperatures (200-300 °C), we have shown that with a highly
reactive precursor, such as trimethylaluminium, coating at room temperature is possible. ALD
– when expressed in film growth per unit of time – is a relatively slow process; its use is
doubtful when films of >> 10 nm are needed. The precursors are typically expensive, which
can be an addition economic driver to keep the film thickness as low as possible.
From a product point of view, ALD is also highly interesting, since depositing small amounts
of compounds can readily enhance the properties of materials, such as in photocatalysis and
Li-ion batteries [7, 38]. ALD allows controlling the amount of precursor deposited based on
the two self-terminating reactions, which only take place as long there are surface sites
available, ensuring a highly utilization of the precursors. That becomes of vital importance to
deposit scarce and costly materials, such as Pt [39], used in the production of catalysts. ALD
is a versatile technique to produce catalysts for several reasons [40]. It can be used to deposit
one [7, 41] or two active phases [42], and to deposit a protective overcoating to prevent the
active phase from sintering at high temperatures [43]. Additionally, the size and dispersion of
the catalytically-active clusters is also important in the production of nanostructured catalysts,
since the surface area determines their activity. Therefore, having small and dispersed clusters
of atoms increase the surface area, and thus, the activity of the catalyst. The cyclic reactions
of a ALD process can nicely control the size and distribution of the catalytic nanoparticles
deposited with the number of coating cycles, as can be seen in the production of TiO2
photocatalyst decorated with Pt clusters [4].
Finally, just as an overview, ALD has been used in a wide variety of applications, such as in
the production of semiconductor devices [44], in the fabrication of supercapacitors [45], in the
production of Li-ion batteries [46], in synthesis of catalysts [40], in the fabrication of
photovoltaic devices [47], and in the enhancement of phosphor materials with
photoluminescence properties [48]. It is to be expected that ALD, as an important part of the
nanotechnology toolbox, will be used for even more products in the near future.
119
Chapter 6
Bibliography
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
Janotti, A. and C.G. Van De Walle, Fundamentals of zinc oxide as a semiconductor.
Reports on Progress in Physics, 2009. 72(12).
Morgeneyer, M., et al., Experimental study of the aerosolization of fine alumina
particles from bulk by a vortex shaker. Powder Technology, 2013. 246(0): p. 583-589.
Kim, H., H.B.R. Lee, and W.J. Maeng, Applications of atomic layer deposition to
nanofabrication and emerging nanodevices. Thin Solid Films, 2009. 517(8): p. 25632580.
Goulas, A. and J.R. van Ommen, Scalable production of nanostructured particles
using atomic layer deposition. KONA Powder and Particle Journal, 2014. 31(1): p.
234-246.
Longrie, D., D. Deduytsche, and C. Detavernier, Reactor concepts for atomic layer
deposition on agitated particles: A review. Journal of Vacuum Science &amp;
Technology A, 2014. 32(1): p. -.
van Ommen, J.R., Apparatus and process for atomic or molecular layer deposition
onto particles during pneumatic transport. 2010, Google Patents.
Van Ommen, J.R., et al., Continuous production of nanostructured particles using
spatial atomic layer deposition. Journal of Vacuum Science and Technology A:
Vacuum, Surfaces and Films, 2015. 33(2).
Knowlton, T.M., 10 - Fluidized bed reactor design and scale-up, in Fluidized Bed
Technologies for Near-Zero Emission Combustion and Gasification, F. Scala, Editor.
2013, Woodhead Publishing. p. 481-523.
Sundaresan, S., Instabilities in fluidized beds. 2003. p. 63-88.
van Ommen, J.R., J.M. Valverde, and R. Pfeffer, Fluidization of nanopowders: a
review. Journal of Nanoparticle Research, 2012. 14(3): p. 1-29.
Seville, J.P.K., C.D. Willett, and P.C. Knight, Interparticle forces in fluidisation: A
review. Powder Technology, 2000. 113(3): p. 261-268.
Tahmasebpoor, M., et al., The role of the hydrogen bond in dense nanoparticle-gas
suspensions. Physical Chemistry Chemical Physics, 2013. 15(16): p. 5788-5793.
Salameh, S., et al., Adhesion mechanisms of the contact interface of TiO 2
nanoparticles in films and aggregates. Langmuir, 2012. 28(31): p. 11457-11464.
Guardiola, J., V. Rojo, and G. Ramos, Influence of particle size, fluidization velocity
and relative humidity on fluidized bed electrostatics. Journal of Electrostatics, 1996.
37(1–2): p. 1-20.
Pfeffer, R.S., AZ, US), Quevedo, Jose A. (Brick, NJ, US), Systems And Methods For
Reducing Electrostatic Charge In A Fluidized Bed. 2009, NEW JERSEY INSTITUTE
OF TECHNOLOGY (Newark, NJ, US): United States.
Boland, D. and D. Geldart, Electrostatic charging in gas fluidised beds. Powder
Technology, 1972. 5(5): p. 289-297.
Chen, A., et al., Measurement of charge distribution around a rising bubble in a 2-D
fluidized bed. AIChE Journal, 2006. 52(1): p. 174-184.
Demirbas, B., et al., Towards monitoring electrostatics in gas-solid fluidized beds.
Canadian Journal of Chemical Engineering, 2008. 86(3): p. 493-505.
Morooka, S., et al., FLUIDIZATION STATE OF ULTRAFINE POWDERS. Journal of
Chemical Engineering of Japan, 1988. 21(1): p. 41-46.
Xu, C. and J. Zhu, Experimental and theoretical study on the agglomeration arising
from fluidization of cohesive particles - Effects of mechanical vibration. Chemical
Engineering Science, 2005. 60(23): p. 6529-6541.
120
Outlook
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
Zhou, L., et al., Model of estimating nano-particle agglomerate sizes in a vibrofluidized bed. Advanced Powder Technology, 2013. 24(1): p. 311-316.
Moreno-Atanasio, R. and M. Ghadiri, Mechanistic analysis and computer simulation
of impact breakage of agglomerates: Effect of surface energy. Chemical Engineering
Science, 2006. 61(8): p. 2476-2481.
Castellanos, A., J.M. Valverde, and M.A.S. Quintanilla, Aggregation and
sedimentation in gas-fluidized beds of cohesive powders. Physical Review E, 2001.
64(4): p. 041304.
Nam, C.H., et al., Aerated vibrofluidization of silica nanoparticles. AIChE Journal,
2004. 50(8): p. 1776-1785.
Richardson, J.F. and W.N. Zaki, The sedimentation of a suspension of uniform spheres
under conditions of viscous flow. Chemical Engineering Science, 1954. 3(2): p. 65-73.
De Martín, L., et al., A settling tube to determine the terminal velocity and size
distribution of fluidized nanoparticle agglomerates. Journal of Nanoparticle Research,
2014. 16(1).
Elers, K.E., et al., Production of elemental thin films using a boron-containing
reducing agent. 2004, Google Patents.
Tynell, T. and M. Karppinen, Atomic layer deposition of ZnO: A review.
Semiconductor Science and Technology, 2014. 29(4).
Burton, B.B., D.N. Goldstein, and S.M. George, Atomic layer deposition of MgO
using bis(ethylcyclopentadienyl)magnesium and H2O. Journal of Physical Chemistry
C, 2009. 113(5): p. 1939-1946.
Lin, S., et al., Zebrafish high-throughput screening to study the impact of dissolvable
metal oxide nanoparticles on the hatching enzyme, ZHE1. Small, 2013. 9(9-10): p.
1776-1785.
Kaweeteerawat, C., et al., Toxicity of metal oxide nanoparticles in Escherichia coli
correlates with conduction band and hydration energies. Environmental Science and
Technology, 2015. 49(2): p. 1105-1112.
Nel, A., et al., Toxic potential of materials at the nanolevel. Science, 2006. 311(5761):
p. 622-627.
Nel, A., Y. Zhao, and L. Mädler, Environmental health and safety considerations for
nanotechnology. Acc Chem Res, 2013. 46(3): p. 605-606.
Nel, A.E., et al., Understanding biophysicochemical interactions at the nano-bio
interface. Nature Materials, 2009. 8(7): p. 543-557.
Xia, T., et al., Comparison of the mechanism of toxicity of zinc oxide and cerium oxide
nanoparticles based on dissolution and oxidative stress properties. ACS Nano, 2008.
2(10): p. 2121-2134.
George, S.M., Atomic layer deposition: An overview. Chemical Reviews, 2010.
110(1): p. 111-131.
Eckelman, M.J., J.B. Zimmerman, and P.T. Anastas, Toward Green Nano. Journal of
Industrial Ecology, 2008. 12(3): p. 316-328.
Beetstra, R., et al., Atmospheric pressure process for coating particles using atomic
layer deposition. Chemical Vapor Deposition, 2009. 15(7-9): p. 227-233.
Alonso, E., et al. Strategies to address risks of platinum scarcity for supply chain
downstream firms. in 2009 IEEE International Symposium on Sustainable Systems
and Technology, ISSST '09 in Cooperation with 2009 IEEE International Symposium
on Technology and Society, ISTAS. 2009. Tempe, AZ.
O’Neill, B.J., et al., Catalyst Design with Atomic Layer Deposition. ACS Catalysis,
2015. 5(3): p. 1804-1825.
121
Chapter 6
41.
42.
43.
44.
45.
46.
47.
48.
Goulas, A. and J. Ruud van Ommen, Atomic layer deposition of platinum clusters on
titania nanoparticles at atmospheric pressure. Journal of Materials Chemistry A,
2013. 1(15): p. 4647-4650.
Lu, J., et al., Toward atomically-precise synthesis of supported bimetallic
nanoparticles using atomic layer deposition. Nat Commun, 2014. 5.
Lu, J., et al., Porous alumina protective coatings on palladium nanoparticles by selfpoisoned atomic layer deposition. Chemistry of Materials, 2012. 24(11): p. 20472055.
Kim, H., Atomic layer deposition of metal and nitride thin films: Current research
efforts and applications for semiconductor device processing. Journal of Vacuum
Science and Technology B: Microelectronics and Nanometer Structures, 2003. 21(6):
p. 2231-2261.
Guan, C., et al., Atomic-Layer-Deposition-Assisted Formation of Carbon Nanoflakes
on Metal Oxides and Energy Storage Application. Small, 2014. 10(2): p. 300-307.
Meng, X., X.-Q. Yang, and X. Sun, Emerging Applications of Atomic Layer
Deposition for Lithium-Ion Battery Studies. Advanced Materials, 2012. 24(27): p.
3589-3615.
Liu, Y., et al., PbSe quantum dot field-effect transistors with air-stable electron
mobilities above 7 cm2 V-1 s-1. Nano Letters, 2013. 13(4): p. 1578-1587.
Kim, H.J., et al., Atomic layer deposition of aluminum oxide thin film on BaMgAl
10O17:Eu2+ phosphor. Molecular Crystals and Liquid Crystals, 2006. 459(1): p.
239/[519]-245/[525].
122
List of Publications
Papers
1.
D. Valdesueiro, G.M.H. Meesters, M.T. Kreutzer, and J.R. van Ommen. Gas-phase
deposition of ultrathin aluminium oxide films on nanoparticles at ambient conditions.
Materials, 2015. 8(3): p. 1249-1263.
2.
D. Valdesueiro, P. Garcia-Triñanes, G.M.H. Meesters, M.T. Kreutzer, J. Gargiuli, T.
Leadbeater, D.J. Parker, J. Seville, and J.R. van Ommen. Enhancing the activation of
silicon carbide tracer particles for PEPT applications using gas-phase deposition of
alumina at room temperature and atmospheric pressure (submitted for publication).
3.
D. Valdesueiro, H. Hettinga, J.P. Drijfhout, P. Lips, G.M.H. Meesters, M.T. Kreutzer,
and J.R. van Ommen. Tuning roughness and gloss of powder coating paint by
encapsulating glassy particles with thin Al2O3 films (submitted for publication).
4.
D. Valdesueiro, M. Krishna Prabhu, C. Guerra-Nuñez, C.S. Suchand Sandeep, S.
Kinge, L.D.A. Siebbeles, L.C.P.M. de Smet, G.M.H. Meesters, M.T. Kreutzer, A.J.
Houtepen, and J.R. van Ommen. The deposition mechanism of aluminium oxide on
quantum dot films at atmospheric pressure and room temperature (submitted for
publication).
Proceedings
1.
D. Valdesueiro, G.M.H. Meesters, and J.R. van Ommen. Coating Fine Particles with
an Ultrathin Film in a Fluidized Bed. In Proceedings of International Conference of
Particle Technology ParTec2013, Nuremberg, 23-25 April 2013.
2.
D. Valdesueiro, G.M.H. Meesters, and J.R. van Ommen. Microscopic characterization
of mechanically assisted fluidized beds. In Proceedings of Fluidization XIV: From
Fundamentals to Products, Noordwijkerhout, 26-31 May 2013.
3.
G.M.H. Meesters, D. Valdesueiro, and J.R. van Ommen. Coating Fine Particles with
Ultrathin Films using Atomic Layer Deposition. In Proceedings of 2013 AIChE
Annual Meeting: Global Challenges for engineering a Sustainable Future, San
Francisco, 3-8 November 2013.
125
Acknowledgements
This thesis is the result of four years of research at the Product and Process Engineering (PPE)
group. The Marie Curie Actions, under the program PowTech Initial Training Network, and
DSM Coating Resins in Zwolle, funded this project. In this chapter, I take the opportunity to
thank to many people that contributed to this thesis, and without whom, the previous chapters
would not have been written.
First of all, I would like to thank to my promotor, Michiel Kreutzer, and co-promotors, Ruud
van Ommen and Gabrie Meesters. It has been great to work under your guidance and
supervision. Michiel, your analytical thinking and critical feedback have made this work
better. Always identifying the loose ends and challenging to find explanations to all the
details. Thanks, it was fun working with you! Ruud, I am really lucky to have had you as
supervisor. Your support has been essential to make it all the way until here. Thank you for
your meticulous revisions and dedicated supervision, for all the quick talks in the early
mornings, and for having always a constructive feed-back. Your contribution has been
priceless for me from the beginning of the project. Thank you very very much! Gabrie, you
added to the project your industrial expertise and large network, and above all, a young spirit
and a lot of stories and anecdotes during the progress meetings. I enjoyed every non-scientific
discussion along these years. Thank you for bringing to the table your industrial point of view.
It is logical to think that without supervisors, it is not possible to complete a PhD. I would like
to add to it that, without the technical support at the lab, would not be possible either. Mojgan
Talebi, Mark de Niet and Peter Gouweleeuw, you have done a tremendous contribution to this
work. Thank you very much! Doing experiments without your help would just have been
inconceivable. Mojgan and Mark, we spent so many hours in front of the ALD setup, and not
only coating particles, but placing and replacing valves, lines and bubblers, troubleshooting
and a couple of times even observing the pyrophoricity of TMA… Your constant positive
attitude converted these tedious work into enjoyable tasks. It was great having your support at
the lab! And Peter, until you left for your retirement, you have been the watchman of the
Proeffabriek. You fixed everything that we messed in the lab, quickly and precisely, making
possible that this project never was never stopped due to technical difficulties. In addition, I
would like to thank the administrative work of Elly Hilkhuijsen, Caroline Monna, Els
Arkesteijn, Astrid Barrow and Marian de Bruijn. You make working at PPE much easier for
us. I would also like to thank a lot of people at the ChemE and other departments for the
technical support in the analysis of samples: Marcel Bus, Bart Boshuizen, Duco Bosma, Piet
Droppert, Ben Norder, Ruben Abellon, Baukje Terpstra, Meng Yue Wu, Bart van der Linden
and Willy Rook. Thanks for your help!
Special thanks to all the people I had the pleasure to collaborate with. The people at DSM in
Zwolle: Jan Pieter Drijfhout, Cindy van Tent, Hans Hettinga, Priscilla Lips and Paul
Vercoulen. Working with you was an important part of the project, and each time I went to
127
Acknowledgements
Zwolle to do experiments I felt like coming back home. Thank you! Likewise, I would like to
thank to all the collaborators in different projects. Arjan Houtepen, Suchand Sandeep, Ruud
Westerwaal, Zhi Zhou, Bert Hintzen and Pablo García-Triñanes. Depositing alumina films on
your special materials was an important addition to this thesis. Thank you for your effort and
compromise. I enjoyed working with you! In addition, I would like to thank also to the LO,
BSc and MSc students that I supervised during the PhD. Tim, Jacob, Jochem, Sam, Minouk,
Joost, Hayati, Niek, Louis, Sophie, Gaurav, Mahesh, Shiva, Marybeth, Ricardo and Carlos,
thanks a lot for your dedication and contribution to this work!
I would like to thank also to all the people from the Marie Curie network, and in particular, to
the fellow ESR’s. There were many stories and highlights in each Powtech meeting… The
drinks in Paris and Cork, the trains and the Oktoberfest party with Mr. Tambourine in Zurich,
the bowling in Budapest, the visits to the hospital and the Poco Loco factory in Gotteborg, or
the road trip in China. The Powtech program was really fun, and a great complement to this
PhD!
I have special words for all the colleagues and office mates at PPE and the Proeffabriek.
Michiel, Ruud, Pouyan, Volkert, Peter, Henk, Mojgan, Wim, David, Nasim, Mahsa, Duong,
Barbara, Michiel, Rajat, Bhaskar, Amer, Venky, Hrushi, Piotr, Josette, Daoyin, Jilian, Zhi
Zhou, Piotr, and especially, Hao, Jesús, Javi, Wenjie, Alberto, Samir, Durgesh, Floris,
Dayinta, Yogesh, Lilian, Aris, Andrea, Fabio. All the VriMiBo’s, after work drinks and
brewing attempts, betting events and shooting competitions, social activities and road trips,
Fun Fridays… It has been a wonderful experience to share these four years with you, and I
can certainly say that doing a PhD at PPE is, amongst many good things, a fun experience. I
also want to remember the friends with whom I spent a great time in Delft, eating fish on
Saturdays, playing board games, poker nights, basketball... You all contributed greatly to have
a relaxing and fun time in Delft.
Finally, I would like to thank to the most important part for me: my parents and family.
Gracias por todo lo que me habéis dado, vuestro apoyo incondicional y ser un ejemplo.
Siempre os tengo presentes y os siento cerca. Esto es en gran parte vuestro también.
128
Curriculum Vitae
David was born on 31 January 1984 in León, Spain. In 2009 he graduated in Chemical
Engineering from the University of Valladolid, Spain. David conducted his engineering thesis
between September 2008 and June 2009 at the University of Maribor, Slovenia. In 2010
David worked in the production factory of INERGY Automotive Systems in Arévalo, Spain.
In August 2010, he moved to Zwolle, The Netherlands, to work for the technology department
at DSM. In July 2011, David started his PhD in the Department of Chemical Engineering
within the Faculty of Applied Sciences. His promotor was Prof. dr. Michiel T. Kreutzer, and
his co-promotors were Dr. ir. J. Ruud van Ommen and Dr. ir. Gabrie M.H. Meesters. Since
September 2015, he works in Delft IMP as process engineer.
131
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

advertising