University of Southern California Dissertations and Theses

Fabrication of silicon-based membranes via vapor-phase deposition and pyrolysis of organosilicon polymers

(USC Thesis Other)

Fabrication of silicon-based membranes via vapor-phase deposition and pyrolysis of organosilicon polymers

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content Copyright 2023 Bryan Nguyen

FABRICATION OF SILICON-BASED MEMBRANES VIA VAPOR-PHASE DEPOSITION AND
PYROLYSIS OF ORGANOSILICON POLYMERS

by

Bryan Hoang Nguyen

A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(CHEMICAL ENGINEERING)

May 2023
ii

Table of Contents
LIST OF TABLES....................................................................................................................................................... iii
LIST OF FIGURES ......................................................................................................................................................iv
ABSTRACT ............................................................................................................................................................... vii
1 CHAPTER ONE: INTRODUCTION ................................................................................................................... 1
1.1 CERAMIC MEMBRANES IN STEAM METHANE REFORMING APPLICATIONS .................................................... 1
1.2 INITIATED CHEMICAL VAPOR DEPOSITION ................................................................................................... 3
1.3 PLASMA ENHANCED CHEMICAL VAPOR DEPOSITION ................................................................................... 4
1.4 PYROLYSIS MODELING ................................................................................................................................. 5
2 CHAPTER TWO: FABRICATION OF HYDROGEN-SELECTIVE SILICA MEMBRANES VIA
PYROLYSIS OF VAPOR DEPOSITED POLYMER FILMS ...................................................................................... 8
2.1 ABSTRACT .................................................................................................................................................... 8
2.2 INTRODUCTION ............................................................................................................................................. 9
2.3 EXPERIMENTAL ........................................................................................................................................... 11
2.4 RESULTS AND DISCUSSION ......................................................................................................................... 17
2.5 CONCLUSION .............................................................................................................................................. 28
2.6 ACKNOWLEDGEMENTS ............................................................................................................................... 29
3 CHAPTER THREE: FABRICATION OF SIC-TYPE FILMS USING LOW-ENERGY PLASMA
ENHANCED CHEMICAL VAPOR DEPOSITION (PECVD) AND SUBSEQUENT PYROLYSIS ....................... 30
3.1 ABSTRACT .................................................................................................................................................. 30
3.2 INTRODUCTION ........................................................................................................................................... 30
3.3 EXPERIMENTAL ........................................................................................................................................... 34
3.4 RESULTS AND DISCUSSION ......................................................................................................................... 40
3.5 CONCLUSION .............................................................................................................................................. 69
3.6 ACKNOWLEDGEMENTS ............................................................................................................................... 70
4 CHAPTER FOUR: TIME-RESOLVED OPERANDO ANALYSIS OF THE PYROLYSIS OF A PECVD-
DEPOSITED SILOXANE POLYMER USING A COMBINED DRIFTS-MS SYSTEM ......................................... 71
4.1 ABSTRACT .................................................................................................................................................. 71
4.2 INTRODUCTION ........................................................................................................................................... 72
4.3 EXPERIMENTAL ........................................................................................................................................... 75
4.4 RESULTS AND DISCUSSION ......................................................................................................................... 79
4.5 CONCLUSION ............................................................................................................................................ 101
4.6 ACKNOWLEDGEMENTS ............................................................................................................................. 102
4.7 SUPPLEMENTARY MATERIALS SECTION .................................................................................................... 103
5 CHAPTER FIVE: CONCLUSION ................................................................................................................... 105
5.1 CLOSING REMARKS................................................................................................................................... 105
5.2 FUTURE RESEARCH DIRECTIONS ............................................................................................................... 106
5.2.1 Preparation of SiC Membranes via In-situ PECVD Deposition and Pyrolysis .................. 106
5.2.2 Testing of the Membranes in a Membrane Steam Reforming Reactor .............................. 107
6 REFERENCES ................................................................................................................................................. 109

iii

List of Tables

Table 2-1. XPS elemental analysis of the PV4D4 films pre- and post-pyrolysis ....................................... 23
Table 2-2. Peak areas of the silicon oxidation state from high resolution XPS analysis ........................... 26
Table 2-3. He and H2 permeances and ideal separation factors of He and H2 over Ar for membranes
resulting from the pyrolysis at 1100 ºC of PV4D4 films deposited on SiC supports. The testing
pressure was 241 kPa and the testing temperature was 473K. .................................................................... 28
Table 3-1. IR bands shown in Figures 3-8 ................................................................................................. 41
Table 3-2. Atomic Composition of the Deposited Polymer Films from EDS Analysis ............................. 43
Table 3-3. Atomic Composition of the Pyrolyzed Polymer Films from EDS Analysis ............................. 59
Table 3-4. Ceramic yield of various polymer films post-pyrolysis ............................................................ 65
Table 4-1. IR bands shown in Figures 4.2-4.4 ........................................................................................... 79
Table 4-2. At% of the pV4D4 films pyrolyzed at various temperatures .................................................... 94

iv

List of Figures

Figure 1-1. Schematic of the SMR reactor ................................................................................................... 1
Figure 1-2. (a) Schematic of the iCVD process. Monomer (M) and initiator (I-I) vapors enter the
reactor and react on the substrate surface to form a polymer film. (b) Schematic of a cylindrical
iCVD reactor ................................................................................................................................................. 4
Figure 2-1. Schematic of the iCVD reactor ............................................................................................... 13
Figure 2-2. a) Schematic of the iCVD polymerization of V4D4 to pV4D4. b) FTIR spectra of the
monomer and the resulting polymer. The dashed lines indicate the vinyl bond modes, which have a
noticeable decrease in the IR spectra of the polymer. ................................................................................. 19
Figure 2-3. Real-time DRIFTS data showing the IR absorbance spectra for pV4D4 films at different
pyrolysis temperatures. The dashed lines represent: a) symmetric -CH2 stretching, b) the stretching
of the C=C bond, c) the wagging mode of the vinyl bond, and d) the symmetric Si-CH3 stretching ........ 20
Figure 2-4. a) IR absorbance spectra showing the evolution of the PV4D4 film into a silica film. The
absorbance bands labelled above correspond to i) the hydrocarbon bonds from the polyethylene
bond, ii, iii, vi) vibration modes corresponding to vinyl bonds, iv) stretching of methyl bonds, v)
asymmetric stretching of Si-O-Si bonds, and vii) asymmetric rocking in the methyl group. b) An
expanded view of the IR region from 1250 to 1280 cm
-1
showing the methyl stretching band. ................ 23
Figure 2-5. High resolution scans of Si 2p for the a) pV4D4 polymer and for the polymer after
pyrolysis at b) 400 °C, c) 600 °C, d) 800 °C, and e) and 1100 °C. ............................................................. 26
Figure 2-6. a) Cross-sectional view of the asymmetric composite structure consisting of the iCVD
polymer layer, and the intermediate and the macroporous SiC layers, and b) top-down image of the
dense iCVD layer on top of the SiC support. .............................................................................................. 28
Figure 3-1. Schematic of the PECVD reactor. ........................................................................................... 35
Figure 3-2. The chemical structure of the VPDMS (left) and DVB (right) monomers. ............................ 36
Figure 3-3. FTIR spectra of PECVD-deposited polymer films deposited on Si wafers. The dashed
lines represent regions: (i) Si-C stretching and para- and meta-vibrations of DVB ring, (ii) vinyl bond
stretching, (iii) Si-C6H5 stretching, (iv) symmetric Si-CH3 stretch stretching, (v) aromatic C 6H 5 C-H
and C=C stretching bands, (vi) Si-H stretching, and (vii) C=C and C-H hydrocarbon bonds from
polyethylene bonds and aromatic rings. ...................................................................................................... 42
Figure 3-4. DRIFTS data during pyrolysis for the copolymer coated on KBr. The dashed lines
represent: (i) Si-C stretching, (iiD) Si-(CH2)n-Si stretching, (iii) Si-C5H6 stretching, (iv) symmetric
Si-CH3 stretching, (v) pDVB C-H and C=C stretching bands, (vi) Si-H stretching, and (vii) C-H
hydrocarbon bonds from polyethylene bonds and aromatic rings. ............................................................. 47
v

Figure 3-5. DRIFTS data for the pVPDMS-coated BaF2. The dashed lines represent: (i) Si-C
stretching, (iii) Si-C5H6 stretching, (iv) symmetric Si-CH3 stretching, (v) pDVB C-H and C=C
stretching bands, (vi) Si-H stretching, and (vii) C-H hydrocarbon bonds from polyethylene bond and
aromatic rings.............................................................................................................................................. 50
Figure 3-6. DRIFTS data for the 1:1 p(VPDMS-coDVB)-coated BaF2. The dashed lines represent:
(i) Si-C stretching, (iiD) Si-(CH2)n-Si stretching, (iii) Si-C5H6 stretching, (iv) symmetric Si-CH3
stretching, (v) pDVB C-H and C=C stretching bands, (vi) Si-H stretching, and (vii) C-H hydrocarbon
bonds from polyethylene bond and aromatic rings. .................................................................................... 51
Figure 3-7. DRIFTS data for the 2:1 p(VPDMS-coDVB)-coated BaF2. The dashed lines represent:
(i) Si-C stretching, (iiD) Si-(CH2)n-Si stretching, (iii) Si-C5H6 stretching, (iv) symmetric Si-CH3
stretching, (v) pDVB C-H and C=C stretching bands, (vi) Si-H stretching, and (vii) C-H hydrocarbon
bonds from polyethylene bond and aromatic rings. .................................................................................... 52
Figure 3-8. DRIFTS data for the 1:2 p(VPDMS-coDVB)-coated BaF2. The dashed lines represent:
(i) Si-C stretching, (iiD) Si-(CH2)n-Si stretching, (iii) Si-C5H6 stretching, (iv) symmetric Si-CH3
stretching, (v) pDVB C-H and C=C stretching bands, (vi) Si-H stretching, and (vii) C-H hydrocarbon
bonds from polyethylene bond and aromatic rings. .................................................................................... 53
Figure 3-9. Functional group peak area change as a function of pyrolysis temperature for different
polymer films: a) pVPDMS b) 1:1 p(VPDMS-co-DVB) c) 2:1 p(VPDMS-co-DVB) and d) 1:2
p(VPDMS-co-DVB). .................................................................................................................................. 58
Figure 3-10. DRIFTS IR spectra of polymer-coated BaF2 powders pyrolyzed at 1000 °C. .................... 59
Figure 3-11. TGA graph showing the fraction of decomposition of the polymer films deposited on
BaF2: a) pVPDMS b) 1:1 VPDMS:DVB p(VPDMS-co-DVB) c) 2:1 VPDMS:DVB p(VPDMS-co-
DVB) and d) 1:2 VPDMS:DVB p(VPDMS-co-DVB). .............................................................................. 61
Figure 3-12. Weight change loss during polymer pyrolysis under TGA experiments for polymer
films a) pVPDMS b) 1:1 VPDMS:DVB p(VPDMS-co-DVB) c) 2:1 VPDMS:DVB p(VPDMS-co-
DVB) and d) 1:2 VPDMS:DVB p(VPDMS-co-DVB). .............................................................................. 64
Figure 4-1. Schematic of the combined DRIFTS-RGA apparatus ............................................................. 77
Figure 4-2. Spectra of the V4D4 monomer and the pV4D4 polymer deposited via PECVD on a Si
wafer ........................................................................................................................................................... 81
Figure 4-3. a) Ex-situ IR absorbance spectra showing the transformation of the PECVD-deposited
pV4D4 polymer during pyrolysis; b) expanded view of the IR region from 2750 to 3100 cm
-1

showing the polyethylene stretching band; c) expanded view of the IR region from 1200 to 1300 cm
-1

showing the methyl stretching band............................................................................................................ 82
Figure 4-4. DRIFTS data showing the IR absorbance spectra for pV4D4 films at different pyrolysis
temperatures ................................................................................................................................................ 85
vi

Figure 4-5. Functional group peak area change as a function of pyrolysis temperature ............................ 86
Figure 4-6. Partial pressures of gaseous species as detected by RGA ....................................................... 88
Figure 4-7. Fraction of decomposition versus temperature for three different heating rates ..................... 92
Figure 4-8. Derivative of weight change during pV4D4 film pyrolysis .................................................... 93
Figure 4-9. Carbon content of the polymer strands as a function of pyrolysis temperature ...................... 97
Figure 4-10. Mass loss of the polymer strands as a function of pyrolysis temperature ............................. 98
Figure 4-11. Temperature evolution of the Si-C bonds during the pyrolysis of pV4D4 ........................... 99
Figure 4-12. Evolution of the compounds observed during the pyrolysis process for the a) 1 K/ps, b)
2 K/ps, and c) 3 K/ps heating rate simulations. ........................................................................................ 101
Figure 4-13. Comparison of the ReaxFF and DFT bond dissociation energies for the following
bonds: a) Si-O, b) Si-CH3, c) C-CH3, and d) Si-Vinyl. ........................................................................... 103
Figure 4-14. Comparison of the ReaxFF and DFT bond angle distortion energies for the following
bonds: a) Si-O-Si, b) O-Si-O, c) C-Si-C, and d) O-Si-C........................................................................... 104
Figure 5-1. Schematic of the modified iCVD reactor with in-situ heating apparatus .............................. 107

vii

Abstract

The efficient separation of H2 under steam reforming conditions is important for the
development of the clean energy industry and has helped drive inorganic membrane research for
several decades. Silicon-based ceramic materials have shown great potential as nanoporous
membranes for gas separations due to their high temperature resistance and excellent chemical and
mechanical stability. In our work, we demonstrate the ability to fabricate highly permeable silicon-
based membranes using chemical vapor deposition (CVD) and subsequent pyrolysis. Chapter 1 of
this Thesis introduces the initiated chemical vapor deposition (iCVD) and plasma-enhanced
chemical vapor deposition (PECVD) techniques, and the motivation for using polymer films as
precursors for the preparation of ceramic materials. Chapter 2 demonstrates the fabrication of a
silica membrane through the iCVD deposition of a polymer onto a ceramic substrate and its
subsequent pyrolysis. We demonstrate the ability to conformally deposit on a porous ceramic
substrate a cyclic siloxane polymer called pV4D4, and we pyrolyze the film in order to form a thin
silica membrane. Chapter 3 demonstrates the fabrication of a SiC-type film through the deposition
of an organosilicon polymer using PECVD and subsequent pyrolysis. We evaluate how changing
the monomer ratio in during the preparation of a precursor copolymer film affects the structure of
the resulting pyrolyzed ceramic. Chapter 4 discusses the deposition of pV4D4 via PECVD and its
subsequent pyrolysis using a unique DRIFTS-MS experimental system that allows for the in-situ
characterization and analysis of the resulting materials. We also perform complementary
computational simulations modeling the pyrolysis of the pV4D4 in order to gain further insight
into the reaction mechanism. Chapter 5 provides closing remarks on the pyrolysis studies
completed to date and discusses future research directions.
1

1 Chapter One: Introduction

1.1 Ceramic Membranes in Steam Methane Reforming Applications

The ability to efficiently separate hydrogen under high temperature and pressure steam
methane reforming (SMR) conditions is important for the further development of clean energy
sources.

Figure 1-1. Schematic of the SMR reactor

One way to improve the efficiency of SMR is through the use of high temperature
membrane reactors (MR), see Fig. 1-1.
1,2,3
A challenge with using such reactors is that the
membranes, in addition to having good H2 permselectivity, must also be able to function in the
extreme SMR environments.
4
,
56
Inorganic membranes have several inherent advantages for such
applications over their polymeric counterparts due to their superior chemical, thermal, and
structural stability.
1,7
Current materials used as inorganic membranes for in situ H2 separation in
MRs include Pd and its alloys
8,9
, zeolites
10
and carbon molecular sieves (CMS)
11,12,13
. Pd and Pd-
alloy membranes are the most commonly employed in SMR-MR studies. However, these
2

membranes are expensive and are susceptible to coking and embrittlement, both of which impart
inferior thermal and mechanical properties causing the membranes to eventually develop cracks
under stress
14
. In addition, poisoning from S-containing compounds degrades separation
performance. Zeolite membranes, while made of common materials, have poor H2 selectivity under
SMR conditions due to the presence of intercrystalline voids and also lack long-term stability in
the presence of steam and S-containing compounds.
10
CMS membranes (CMSM) are highly H2
selective, and our group is presently field-testing them in power generation applications
11
.
However, these membranes have an upper temperature limit of operation of ~325
o
C and, thus,
cannot be used in the SMR process which employs significantly higher temperatures.
Si-based membranes, with chemical structures ranging from amorphous SiO2 to pure SiC,
possess excellent H2 separation properties and do not have issues with embrittlement and S
poisoning
1,15
. Current methods for producing these membranes include solution-based techniques
such as spin-coating
16
and dip-coating
17, 18, 19
. However, films fabricated with such methods have
substrate compatibility issues and difficulties with solvent disposal
20
. Si-based membranes are also
produced using conventional chemical vapor deposition (CVD)
15,21,22
techniques. These
techniques often require high energy inputs that can damage the chemical functionality of the
precursors, which hinders their ability to control the structure of the final ceramic produced.
In this research, we plan to study a new synthesis method to create robust asymmetric,
nanoporous Si-based membranes using initiated chemical vapor deposition (iCVD) and
subsequent pyrolysis. Compared to other CVD processes, iCVD requires very low energy and mild
reactor conditions and employs a broad range of monomers to generate a variety of polymer
films.
23
The macroporous support used during membrane preparation for the SMR application
needs to be mechanically strong to withstand the high SMR temperatures and the high
3

transmembrane pressure gradients imposed during the MR operation; it must also be resistant to
the corrosive SMR atmosphere and be able to withstand thermal and pressure-cycling.
1,6
Here, we
will use SiC supports due to their superior chemical and thermal stability.
3,4

1.2 Initiated Chemical Vapor Deposition

The iCVD technique is a one-step process that utilizes a thermally-cleavable initiator, such
as tert-butyl peroxide, to initiate a nucleophilic free-radical polymerization reaction of monomers
with vinyl pendant groups.
23
Because the iCVD process uses a low energy input and mild reactor
conditions, the functionality of the monomer is retained in the resulting polymer and etching in the
film is avoided. The iCVD process, see Fig. 1-2, is a “non-line-of-site” process and has been used
to produce a variety of films on a variety of planar, curved, and porous surfaces. The reactor stage
is cooled to increase the adsorption of the monomer to the surface. The iCVD process is quite
flexible and can prepare a variety of polymer coatings whose functionality depends on the choice
of monomer. Though not commonly employed for that purpose, iCVD is ideally suited for the
deposition of membrane films. Our group has recently shown, for example, that we can create
porous polymer membranes by freezing the monomer prior to polymerization.
24,25,26,27,28,29,30
4

Figure 1-2. (a) Schematic of the iCVD process. Monomer (M) and initiator (I-I) vapors enter the
reactor and react on the substrate surface to form a polymer film. (b) Schematic of a cylindrical
iCVD reactor

In iCVD, the film deposition rate and the weight-averaged molecular weight, MWav, are
proportional to the ratio of PM to PSat, where PM is the partial pressure and PSat is the saturation
pressure of the monomer. The iCVD process is, typically, run at a PM/PSat ratio of 0.4 to 0.7, and
always below 1, in order to prevent condensation. P M can be increased by either increasing the
monomer molar feed flow-rate or by increasing the total reactor pressure, and PSat can be decreased
by decreasing the substrate temperature. Deposition rates vary from a few nm/min to as high as
500 nm/min, and polymers with average molecular weight (MWav) as high as 200,000 can be
produced with appropriate optimization of the reactor conditions
31
.

1.3 Plasma Enhanced Chemical Vapor Deposition

In this research, we also plan to demonstrate that SiC-type films can be fabricated through
plasma enhanced chemical vapor deposition (PECVD), and subsequent pyrolysis. Our study of
using iCVD showed that it is possible to deposit cross-linked organosilicon polymer films.
However, it also showed that siloxane and organosilicon monomers have a low propagation (kp)
rate
32
, relatively to other monomer classes. In addition, our work showed that using iCVD to
deposit organosilicon polymer incorporated oxygen into the chemical structure, reducing the purity
(a) (b)
5

of the pyrolyzed ceramic. The PECVD technique may be able to remedy this problem by
employing a low-power plasma to produce ions from an inert gas (such Argon) to initiate a
polymerization reaction without the use of an initiator. The plasma also activates the monomer,
increasing its reactivity and exponentially increasing the deposition rate.
32
In addition, the PECVD
technique is compatible with a wide variety of chemistries, including organosilicon monomers.
Our work has shown that one is able to deposit polymer films onto silicon wafer substrates over
ten times as fast as iCVD for certain siloxane polymers, and over one hundred times faster for
another organosilicon polymer. This increase in deposition rate has benefits in reducing reactor
operating time, therefore, reducing the costs of production for scale-up operations. To our
knowledge, we are the first group to attempt to produce SiC precursor films via either the iCVD
or the PECVD and subsequent pyrolysis technique. In this Thesis, we systematically investigate a
series of monomers with different moieties to determine which functionalities lead to the formation
of SiC-type membrane films with good separation properties.

1.4 Pyrolysis Modeling

In order to design better membranes, it is important to have a fundamental understanding
of the processes involved during preparation that impact their final structure and properties. In
particular, a better understanding of the pyrolysis process and its effects on the chemical structure
of the resulting material would help optimize the pyrolysis temperature and, in the long-term,
reduce energy costs, especially in scaling-up the preparation method. Therefore, computational
simulations are an indispensable tool and a major part of this project. Their overarching goal is the
analysis of the mechanisms of film deposition and pyrolysis that determine the pore structure
characteristics and the preparation conditions leading to high-quality inorganic nanoporous
6

membranes optimal for application to SMR. For the pyrolysis step, we use process-based models,
recently developed by our group
33,34,35,36
, which offer predictive capabilities, as they use
information about the chemical structure of the pre-ceramic polymer precursor and the preparation
method of the membrane; such models enable progress towards the long-term goal of first-
principle molecular engineering and design of improved materials for adsorption, reaction, and
separation. The method requires the implementation of a reactive force field (FF) that models the
molecular interactions between complex molecules and atoms. A FF is a function and
accompanying set of parameters that describe the potential energy of a chemical system. These
FFs are generally derived from experimental data and knowledge derived from quantum chemistry
(QC). While FF methods, in general, do not simulate reactions, Goddard and his team at
Caltech
37,38
built a reactive force field (ReaxFF) that can simulate complex and reactive
interactions between molecules. ReaxFF uses interatomic distance in order to determine a bond
order, which can then be used to determine whether a bond should be built or broken. ReaxFF
helps bridge the gap between computational methods, based on QC, and empirical force-fields
(EFF) by simplifying the former. ReaxFF has been used to model hydrocarbons
39
, heterogeneous
catalysts
40,41
, atomic layer deposition (ALD)
42
, solid-gas interfaces
43
, electrodes
44
, and many more.
Known QC methods are, generally, applicable to all chemical systems, regardless of connectivity,
but their computational expense makes them inapplicable for large (typically, a few thousand
atoms or so) or complex systems (such as those that require reactions). In contrast, ReaxFF makes
possible reactive molecular dynamics (RMD) simulations of large-scale reactive chemical systems
with near QM accuracy at a fraction of the computational cost, thus, making it possible to examine
chemical processes such as polymer pyrolysis and degradation. In the past, our group expanded
and improved the ReaxFF for PDMS to adopt it for the pyrolysis of HPCS at high
7

temperatures
33,34,35,36
, as a key step leading towards the development of process-based models for
nanoporous materials, in general, and SiC materials, in particular. As part of this project, therefore,
we will extend our efforts to also simulate a variety of other Si-containing pre-ceramic polymers
prepared by the iCVD process.

8

2 Chapter Two: Fabrication of Hydrogen-Selective Silica Membranes
via Pyrolysis of Vapor Deposited Polymer Films
1

2.1 Abstract

Efficient separation of hydrogen under steam reforming conditions is important for the
development of clean energy sources. Although high-temperature and steam-stable membranes
with high fluxes and large separation factors would be valuable for such an application, their
fabrication remains a challenge. Silicon-based ceramic membranes are particularly promising due
to their high temperature resistance and excellent chemical stability. In this study, we propose a
new synthetic route for fabricating nanoporous, asymmetric membranes via the pyrolysis of
silicon-containing polymer films deposited by initiated chemical vapor deposition (iCVD) on
macroporous silicon carbide supports. Specifically, we systematically investigated the change in
the chemical structure of poly(2,4,6,8-tetravinyl-2,4,6,8-tetramethyl cyclotetrasiloxane) films at
different pyrolysis temperatures and found that the complete transition to a silica membrane
occurred at ~1100 °C. Three different supports composed of silicon carbide powders of varying
sizes were tested for membrane preparation. It was found that membranes formed with our process
were microporous with separation factors several times above the corresponding Knudsen factors.
Our synthetic route, therefore, offers a scalable and solventless method for producing silicon-based
ceramic membranes for high-temperature separation and sensor applications.

1
Published in Industrial & Engineering Chemistry Research: Nguyen, B.; Dabir, S.; Tsotsis, T.; Gupta, M. Fabrication
of Hydrogen-Selective Silica Membranes via Pyrolysis of Vapor Deposited Polymer Films. Ind. Eng. Chem. Res. 2019,
58 (33), 15190 –15198.
9

2.2 Introduction

As the global energy demand continues to rapidly grow, hydrogen has been increasingly
discussed as a viable alternative energy source for applications such as in hydrogen fuel cells.
Approximately 50% of commercial hydrogen is currently produced via steam methane reforming
(SMR), where methane reacts with high-pressure steam to form hydrogen and carbon monoxide,
and the hydrogen is then extracted using adsorption
45
or membrane separation.
46,47
In order to
reduce the cost of hydrogen for practical applications, it is important to optimize its production
through SMR.
47, 48
This highly endothermic reaction requires a large input of energy, with process
temperatures as high as 850 °C typically employed.
49,50
In order to improve the efficiency of the
reaction, membrane reactors have been studied as a promising alternative to the more conventional
packed-bed reactor processes.
51

One challenge in using such reactors is that the membranes, in addition to having good
hydrogen permselectivity, must also be able to function in these extreme environments.
48-50

Inorganic membranes have several inherent advantages over their polymeric counterparts due to
their superior chemical, thermal, and structural stability.
1,7
Current materials used as membranes
for hydrogen separation include palladium,
1,8,9
zeolites,
1,10,49
and carbon molecular sieves.
11-13

Palladium and palladium-alloy membranes are currently the most commonly employed in SMR
membrane reactor studies. However, these membranes are expensive and are susceptible to coking
and hydrogen embrittlement which impart weaker thermal and mechanical properties causing the
membranes to develop cracks under stress.
14
In addition, poisoning from sulfur-containing
compounds negatively impacts separation performance. Zeolite membranes, while made of
common materials, can have poor hydrogen selectivity under SMR conditions due to the presence
of intercrystalline voids and they also lack long-term stability in the presence of steam and sulfur-
10

containing compounds.
1,49
Carbon-based membranes are highly selective toward hydrogen
1
, and
we are presently field-testing them in power generation applications.
11
However, these membranes
have an upper temperature limit of operation of ~325
o
C and therefore they cannot be used in the
SMR process which employs significantly higher temperatures.
Silicon-based membranes, such as those with chemical structures ranging from amorphous
silica to pure SiC, have been shown to possess excellent hydrogen separation properties and do
not have issues with hydrogen embrittlement and sulfur poisoning.
7,15,49
Current methods for
producing these ceramic membrane films include solution-based techniques such as spin-coating
16
and dip-coating
17-19
. However, films fabricated with these solution-based methods have
disadvantages such as substrate compatibility issues and difficulties with disposal of solvents.
20

Silicon-based membranes have also been produced using chemical vapor deposition (CVD)
15,21,22

and plasma-enhanced chemical vapor deposition (PECVD)
52,53
techniques. These deposition
techniques often require high energy inputs that can damage the chemical functionality of the
precursors, which hinders the ability to control the structure of the final ceramic produced.
20
In
addition, PECVD reactors often employ processing conditions that induce a competition between
deposition and etching of the film, which presents a challenge for controlling the thickness of the
deposited coating.
20

In this study, we introduce a new synthetic method to create robust asymmetric,
nanoporous silica-based membranes using initiated chemical vapor deposition (iCVD) and
subsequent pyrolysis. The iCVD process is an one-step, solventless process that utilizes a
thermally-cleavable compound, such as tert-butyl peroxide, to initiate a free radical polymerization
process.
20,54
Compared to other CVD processes, the iCVD process requires very low energy and
mild reactor conditions and can employ a broad range of monomers to generate a variety of
11

polymer films.
20
It can be used to produce thin, conformal films over a wide variety of planar,
55

curved,
56
and porous surfaces.
57
In addition, the iCVD process can be scaled-up for roll-to-roll
operation.
58
In this paper, we deposited poly(2,4,6,8-tetravinyl-2,4,6,8-tetramethyl
cyclotetrasiloxane) (pV4D4) on a variety of silicon carbide supports fabricated from powders of
various sizes. The pyrolysis of silicon-based polymers has been shown previously to form a variety
of ceramics, ranging from silica to silicon oxycarbide and silicon carbide, for use in different
applications including membranes.
33,59,60,61,62

The macroporous support used during membrane preparation needs to be mechanically
strong to withstand the high temperatures typically utilized in SMR and the high transmembrane
pressure gradients imposed across during membrane reactor operation; it must also be resistant to
the corrosive SMR atmosphere and be able to withstand thermal and pressure cycling.
1,49,
We,
therefore, selected in our study silicon carbide as our support material due to its superior chemical
and thermal stability.
7
We deposited our pV4D4 film onto these supports via iCVD and
systematically studied the conversion of the polymer into a ceramic membrane at different
pyrolysis temperatures using Infrared Spectroscopy (IR) and X-ray Photoelectron Spectroscopy
(XPS). To our knowledge, we are the first group to report the fabrication of a microporous ceramic
membrane from the pyrolysis of polymer films deposited via iCVD.

2.3 Experimental

Support Preparation
We prepared supports by cold-pressing and sintering SiC powders into flat disks. We
employed two different materials: a pure 0.6 μm β-SiC powder (Superior Graphite Co.) and a <80
nm β-SiC powder (US Research Nanomaterials). Three different supports were prepared: one
12

composed solely of the 0.6 μm powder, one composed solely of the <80 nm powder, and one
composed of a 50/50 blend of the 0.6 μm and <80 nm powders. Prior to being pressed into a flat
disk pellet, the SiC powder was mixed with boron carbide (0.1 wt.%) and phenolic resin (4 wt.%)
serving as sintering aids, employing acetone as the dispersing medium. For thorough mixing of
the SiC powder and the sintering aids, the resulting slurry was ultrasonicated for 20 min, then
manually mixed with a spatula, and then ultrasonicated again for another 20 min. The whole
procedure was repeated before the mixture was placed in a fume-hood to dry for 24 hr. The
resulting material was then ground into a powder using a pestle, and oleic acid and toluene as
pressing aids were then gradually added to the powder and mixed manually using a pestle and then
allowed to dry for 30 min. One gr of powder was then loaded into a disk-like mold and pressed
into a pellet using a hydraulic press for 2 min at a pressure of 86 MPa. The pellets were then placed
in a high-temperature graphite furnace (Thermal Technology, Inc., Model 1000-3060-FP20) where
they were heated (3
o
C/min) in flowing Ar to a sintering temperature of 1900
o
C, where they were
kept for 3 hr., and were then cooled down (6
o
C/min) to room temperature. Prior to deposition,
they were sonicated in acetone for 20 min and then dried.
5,63

A dense iCVD layer is necessary for separation of the molecules and, therefore, an
intermediate layer was first formed on the macroporous support. Without an intermediate layer,
the iCVD film would uniformly coat the pores of the macroporous support instead of forming a
dense layer since the pore size of the support is similar to the thickness of the iCVD layer. The
intermediate layer solution was prepared by mixing 10 wt.% of allyl-hydridopolycarbosilane
(AHPCS, Starfire Systems), 1.25 wt.% polystyrene (2500 MW, Sigma-Aldrich), and 5 wt.% fine
SiC powder (100-200 nm) in hexane. The fine SiC powder was prepared from the original 0.6 μm
SiC powder using a precipitation method described elsewhere.
64
The support was dip-coated by
13

immersing it in this solution for 12 sec before raising it out of the solution at a rate of 3 cm/min.
The disk was then heated to 250 °C (at a rate of 3 °C/min) in flowing Ar in a tube furnace
(Lindberg/Blue, Model STF55433C), where it was held for 1 hr. It was then heated to 400 °C,
where it was held for another hour, and then finally heated to 750 °C, where it was held for 2 hr.,
before being cooled down to room temperature at a rate of 3 °C/min.

iCVD Deposition and Subsequent Pyrolysis

Figure 2-1. Schematic of the iCVD reactor

The iCVD method was used to deposit poly(2,4,6,8-tetravinyl-2,4,6,8-tetramethyl
cyclotetrasiloxane) (pV4D4) onto the SiC support disks, as shown in Fig. 2-1. The custom-
designed vacuum reactor (250 mm in diameter, 48 mm in height, GVD Corporation) has a quartz
top that allows for in situ thickness measurements on a reference silicon wafer via a helium-neon
laser interferometer (Industrial Fiber Optics). The support disks were placed on the reactor stage
which was kept at a constant temperature of 40 °C using a recirculating chiller (Thermo Scientific
Haake A25). A piece of aluminum foil was tightly wrapped and taped around the support to ensure
that deposition only occurred on the top surface. A rotary-vane vacuum pump (Edwards E2M40)
was used to hold the reactor under vacuum and the pressure was regulated at 190 mTorr by a
14

throttle valve (MKS 153D) with active feedback control from a capacitance manometer (MKS
622C01TDE Baratron). The initiator and monomer liquids were loaded into stainless steel jars and
introduced into the reactor through heated lines. The monomer 2,4,6,8-tetravinyl-2,4,6,8-
tetramethyl cyclotetrasiloxane (V4D4) (Gelest, Inc.) was heated to 50 °C and the flow rate was
maintained at 0.7 sccm. The initiator tert-butyl peroxide (Aldrich, 98%) was kept at room
temperature and the flow rate was maintained at 1 sccm. A nichrome filament array (Omega
Engineering) was placed inside the reactor and heated to 250 °C to cleave the initiator.
Approximately 1 𝜇 m of pV4D4 was deposited on each support as measured via interferometry on
a reference silicon wafer.
After the deposition of the polymer, the disk was placed in a tube furnace (Lindberg/Blue,
Model STF55433C) and pyrolyzed at the desired temperature over flowing argon. Prior to this
pyrolysis step, flowing argon was introduced into the furnace chamber for 30 min to purge the
chamber from ambient air. The sample disk was heated at a rate of 3 °C/min to 250 °C, where it
was held for one hr., and then to 400 °C, where it was held for another hr., and finally to the desired
pyrolysis temperature, where it was held for two more hrs., before being cooled down to room
temperature at a rate of 3 °C/min.
Chemical Characterization
To determine the optimal film pyrolysis conditions, Diffuse Reflectance Infrared Fourier
Transform Spectroscopy (DRIFTS) (COLLECTOR II, Thermo Scientific) was utilized. DRIFTS
monitors the change in the IR spectra of a powder substrate in situ as a function of the temperature
of the sample.
65
Due to SiC having a high IR radiation absorption in the mid-IR region, DX-type
15

α-Al2O3 powder was chosen in this study for the DRIFTS experiments. To coat the alumina power
with pV4D4, it was first spread onto a silicon wafer using a razor blade and 1 𝜇 m of pV4D4 was
deposited on the alumina powder as measured on a reference wafer by interferometry. After the
deposition, the pV4D4-coated powder was mixed manually with barium fluoride (BaF2), a non-
absorbing matrix, at a ratio of 1:9 powder to BaF2. The powder was then placed into the sample
cup of the DRIFTS cell. Flowing Ar was introduced into the cell at a pressure of 69 kPa to ensure
an inert atmosphere. The temperature of the DRIFTS chamber was raised in increments of 10 °C,
and the IR absorbance spectra was measured every 100 °C. The DRIFTS absorbance spectra of
the uncoated alumina powder was used as the background spectra.
We also measured the IR absorbance spectra of pV4D4 films ex situ by deposition onto
clean silicon wafers and subsequent pyrolysis at preset temperatures. For the pyrolysis step, the
pV4D4-coated wafers were placed in a tube furnace which was then purged with Ar for 30 min to
ensure that no oxygen remains in the furnace. Throughout the pyrolysis process, Ar was allowed
to flow through the furnace in order to ensure an inert atmosphere. The wafers were first heated to
200 °C at a rate of 2 °C/min and then kept at this temperature for 1 hr. They were then heated to
400 °C at the same rate, where they were again kept for one hr., and then finally brought up to the
desired pyrolysis temperature, where they were kept for two additional hrs. The wafers were then
cooled back to room temperature at a rate of 3 °C/min. The slow heating and cooling rates are
known to prevent cracking in the film.
66
After the wafers were cooled down to room temperature,
the IR spectra were recorded using a FTIR spectrophotometer (Nicolet iS10, Thermo Scientific).
The DRIFTS and FTIR spectra were both analyzed using the freeware SpectraGryph 1.2. and the
deconvolution was performed using OriginPro Peak Analyzer.
16

X-ray Photoelectron Spectroscopy (XPS) analysis was used to determine the elemental
composition and bonding environments of the films before and after pyrolysis at various
temperatures. The Kratos Axis Ultra DLD XPS spectrometer equipped with a magnetic immersion
lens and charge neutralization system and a monochromator Al X-ray source was used to collect
the spectra. Survey scans were performed at a pass energy of 160 eV, while high-resolution scans
were captured at a pass energy of 40 eV. A charge correction value was applied to the C 1s
environment so that the peak value was shifted to 284.8 eV. The same correction factor was applied
to the Si 2p and O 1s orbitals, which resulted with the O 1s peak value being shifted to ~532 eV.
Since there were signals generated from both the Si 2p1/2 and Si 2p3/2 electron spin states, it was
necessary to deconvolute each silicon environment into two separate peaks. The peak area for the
Si 2p1/2 was set equal to half of the peak area for the Si 2p3/2, and the full-width half-maximum
(fwhm) of each spin state were set to be equal. In addition, the energy difference between the two
peaks was kept constant at 0.65 eV. The number of environments we fitted per peak was based on
the data obtained from our FTIR analysis. The XPS data were analyzed using the software
CasaXPS. A scanning electron microscope (SEM) (Nova NanoSEM 450 Field Emission Scanning
Electron Microscope) at a 15 kV accelerating voltage was also used to image the membranes.

Permeation Testing
The membranes were tested by measuring the single-gas permeance (mol/m
2
∙s∙Pa) of
helium, hydrogen, and argon. The ideal binary separation factors were calculated by dividing the
permeances of the two gasses. Prior to testing, the flat disks were glued to metal fender washer
17

supports using an epoxy (J-B Weld) and allowed to cure for 24 hr. To measure the permeance, the
metal supports were placed in between the two half-cells of the permeation system and sealed
using silicone O-rings to ensure no leaks. Compressed air was used to seal the chambers via a
hydraulic mechanism. The gas to be measured was introduced to the bottom half-cell at a set
pressure and allowed to permeate through the membrane and to exit in the top cell. The pressure
of the permeate side was kept at atmospheric condition and measured using a pressure gauge
(OMEGA Engineering, DPG1000B-15G). The pressure gradient in between the top and bottom
cells was also kept constant and measured using a differential pressure transducer. The pressure
on the feed side was controlled using the regulator attached to the gas cylinder and a needle valve
installed on the reject-side tube. The flow rate of the permeated gas was measured with bubble-
flow meters of different volumes (1 mL, 10 mL, 100 mL). The gas was allowed to fill the half-
cells and permeate through the membrane for at least 30 min before data were taken to assure that
steady state conditions prevail. For the permeation properties reported here, the temperature of the
cell was kept at 200
o
C and the transmembrane pressure difference at 241 kPA (for further details
about the permeation set-up, see
64
).

2.4 Results and Discussion
Fig. 2-2a shows the iCVD polymerization of V4D4 to pV4D4. The monomer consists of a
Si-O-Si cyclical structure with four vinyl bonds. Fig. 2-2b displays the IR spectra for both the
monomer and the polymer formed via iCVD. The peaks at 960, 1400, and 1600 cm
-1
in the
monomer spectra represent the vinyl bonds. The reduction in the size of these peaks in the polymer
18

spectra along with the increase of the symmetric stretching of CH2 at 2920 cm
-1
confirm that there
is polymerization taking place through the vinyl bonds. In addition, there is a shift of the broad Si-
O-Si bond from 1080 cm
-1
to 1065 cm
-1
, which is associated with the polymer.
67,68
The resulting
P4V4 polymer is highly crosslinked, which has been shown to be beneficial to retain silicon groups
during pyrolysis. For example, Carlsson et al.
69
found that crosslinking PDMS films, which were
cast from toluene onto silicon wafers, minimized chain-scission reactions during pyrolysis which
in turn prevented loss of silane groups.
To study the transition of the polymer film into a ceramic membrane at elevated
temperatures, the pV4D4 film was deposited onto alumina particles and pyrolyzed in flowing Ar.
The changes in the IR absorbance spectra were recorded in situ in a DRIFTS cell as its temperature
was raised up to the maximum rated temperature of the cell (~900 °C). As shown in Fig. 2-3,
several of the distinct peaks of the pV4D4 film prior to pyrolysis were detected such as the
symmetric -CH2 stretching related to the polyethylene backbone centered at 2920 cm
-1
(3a), the
stretching of the C=C bond at 1600 cm
-1
(3b), the wagging mode of the vinyl bond at 1415 cm
-1

(3c), and the symmetric Si-CH3 stretching at ~1260 cm
-1
(3d). There was no significant change in
the spectra when heating to 300 °C, indicating that the chemical structure of the polymer film stays
relatively intact in this range of temperatures. There were significant changes at 400 °C such as
the disappearance of the peaks corresponding to the stretching of the C=C bond and the wagging
mode of the vinyl CH2, indicating that there were no vinyl bonds left in the material. At 600 °C,
there was a significant reduction in the size of the peaks corresponding to the polyethylene bonds
and the symmetric Si-CH3 bond at 1260 cm
-1
indicating that these bonds were likely cleaved. The
polyethylene peak continued to be present in small quantities until ~800 °C. The results of the
DRIFTS experiments demonstrate that 600 °C is a critical temperature during which most of the
19

hydrocarbon bonds are removed, signaling the transition to a pure silica film. Our results agree
with those of Narisawa
70
who reported that the decomposition of cross-linked polysiloxane resins
with methyl side groups begins first with the removal of methyl groups at ~600 °C, and the
formation of silica begins beyond 1000 °C even in an oxidizing environment.

Figure 2-2. a) Schematic of the iCVD polymerization of V4D4 to pV4D4. b) FTIR spectra of the
monomer and the resulting polymer. The dashed lines indicate the vinyl bond modes, which have
a noticeable decrease in the IR spectra of the polymer.

20

Figure 2-3. Real-time DRIFTS data showing the IR absorbance spectra for pV4D4 films at
different pyrolysis temperatures. The dashed lines represent: a) symmetric -CH2 stretching, b)
the stretching of the C=C bond, c) the wagging mode of the vinyl bond, and d) the symmetric Si-
CH3 stretching
Since pV4D4 does not completely convert into the ceramic until 1000
o
C and the maximum
operating temperature of the DRIFTS cell is ~900
o
C, we also carried out complimentary ex situ
FTIR measurements to study pyrolysis at higher temperatures. For these experiments, pV4D4 was
deposited onto silicon wafers using the iCVD process, pyrolyzed in a tube furnace in flowing Ar
under select pyrolysis conditions, and then the IR spectra of the resulting films were measured
(Fig. 2-4). We monitored the change in the oxidation state of the film by monitoring the following
four silicon environments: ‘M’ representing Si bonded to a single oxygen atom (SiO1R3), ‘D’
representing Si bonded to two oxygen atoms such as in the network polymer film (SiO 2R2), ‘T’
representing Si bonded to three oxygen atoms such as in the silsesquioxane cage structure (SiO3R),
and ‘Q’ representing Si bonded to four oxygen atoms such as in pure silica environment (SiO 4).
There is no significant change between the polymer before pyrolysis (4a) and the polymer after
21

pyrolysis at 300 °C (Fig. 2-4b), which is consistent with our DRIFTS experiment. For the sample
pyrolyzed at 400 °C (Fig. 2-4c), the absorbance band at 2850-2950 cm
-1
corresponding to the
polyethylene carbon chain decreased slightly, indicating the cleavage of some of the Si-C
polyethylene bonds. In addition, the bands at 960 cm
-1
, 1415 cm
-1
, and 1600 cm
-1
corresponding
to the vinyl bonds completely disappeared, in agreement with our results from the DRIFTS
experiment. Previously, Trujillo et al.
68
studied the annealing of pV4D4 films in air up to a
temperature of 410 °C, and they found that the films oxidized to form silsesquioxane-type
structures. Our FTIR results show that although there are bands at 1028 cm
-1
,

corresponding to a
small bond angle Si-O-Si, and at 1120 cm
-1
,

corresponding to a silsesquioxane cage (‘T’), the band
for the intact network at 1065 cm
-1
(‘D’) still dominates the structure. This is likely because we
pyrolyzed our samples under inert conditions and, therefore, there were fewer oxygen atoms
readily available to allow the network rings to form the silsesquioxane cages. The Si-CH3
stretching band shifted from 1260 cm
-1
to a double peak at 1262 cm
-1
and 1275 cm
-1
and the Si-
CH3 asymmetric rocking band at 800 cm
-1
transitioned to a double peak at 780 cm
-1
and 800 cm
-1
,
which provides further evidence of the oxidation of some of the network groups.
68,71
For the sample
pyrolyzed at 600 °C (Fig. 2-4d), there was an overall shift in the Si-O-Si peak from 1028 cm
-1
to
1047 cm
-1
. The polyethylene band centered at 2920 cm
-1
and the bands at 1270 cm
-1
and 2965 cm
-
1
mostly disappeared at this pyrolysis temperature. These observations are consistent with the
findings of Mantz et al.
72
who reported that cylic dimethylsiloxane oligomers started decomposing
at ~400 °C and that the silsesquioxane ring lost its structure in the temperature range from 450 °C
- 650 ºC due to methyl abstraction. The deconvolution of the IR region between 1250 to 1270 cm
-
1
can be used to determine the ratio of M, D, and T groups in the bulk film at each pyrolysis
temperature.
73,74
Figure 2-4b shows an expanded view of this region. It can be seen that ‘D’ groups
22

dominated the deposited polymer film prior to pyrolysis. At pyrolysis temperatures of 300 °C and
400 °C, ‘D’ groups still dominate the structure, accounting for ~81 % and ~73 %, respectively.
However, at a pyrolysis temperature of 500 °C, there was a noticeable transition to a silsesquioxane
structure, with ‘T’ groups accounting for 48% of the structure. At a pyrolysis temperature of 600
°C, the methyl peak decreased and ‘T’ groups accounted for 94% of the remaining methyl bonds.
Beyond 600 °C, the methyl stretching band was enveloped in the larger Si-O-Si asymmetric
stretching band and it was not possible to deconvolute it. At a pyrolysis temperature of 800 °C
(Fig. 2-4e), the Si-O-Si band continued transitioning to a silica structure with a peak shift to 1065
cm
-1
and the silsesquioxane peak, the methyl bond, and the polyethylene absorption bands were
greatly reduced.
68,71
Fig. 2-4f shows that pyrolysis at 1100 °C led to the formation of silica thin
films as indicated by the presence of the following characteristic peaks of Si-O-Si: a rocking mode
at ~450 cm
-1
, a weak bending mode near 800 cm
-1
, and a tall stretching vibration at ~1080 cm
-1
.
75,76,77
The shifts to these peaks, signaling the change to a pure silica ‘Q’ environment, were not
apparent in our spectra at pyrolysis temperatures lower than 1100 °C. We also pyrolyzed pV4D4
films at 1200 °C and 1300 °C and the FTIR spectra showed no further change in the structure of
the ceramic. Burns et al.
61
and Babonneau et al.
78
reported that the pyrolysis of siloxanes resins
beyond 1300 °C-1500 °C resulted in the scission of Si-O bonds, with oxygen loss of up to 80% at
1600 °C. Therefore, we did not test pyrolysis at those temperatures. The consistency among the
FTIR results, for which the pV4D4 film was deposited on a silicon wafer, and the DRIFTS results,
for which the pV4D4 film was deposited on alumina powder, indicate that the processes that take
place during the conversion of the polymer into the ceramic are substrate independent.
23

Table 2-1. XPS elemental analysis of the pV4D4 films pre- and post-pyrolysis
Element % Silicon % Oxygen % Carbon
pV4D4, theoretical 20 20 60
pV4D4, experimental 23 28 49
pV4D4, pyrolyzed at 400 °C 22 25 53
pV4D4, pyrolyzed at 600 °C 37 58 5
pV4D4, pyrolyzed at 800 °C 37 59 4
pV4D4, pyrolyzed at 1100 °C 38 60 2

The pyrolyzed pV4D4 films on silicon wafers were also analyzed by XPS to determine
their chemical structure and composition. Survey scans were performed on each of the pyrolyzed
a)
b)
Figure 2-4. a) IR absorbance spectra showing the evolution of the PV4D4 film into a silica
film. The absorbance bands labelled above correspond to i) the hydrocarbon bonds from the
polyethylene bond, ii, iii, vi) vibration modes corresponding to vinyl bonds, iv) stretching of
methyl bonds, v) asymmetric stretching of Si-O-Si bonds, and vii) asymmetric rocking in the
methyl group. b) An expanded view of the IR region from 1250 to 1280 cm
-1
showing the
methyl stretching band.

24

films to determine the atomic fraction of carbon, oxygen, and silicon present pre- and post-
pyrolysis, and the results are shown in Table 2-1. In agreement with the IR data, these results
indicate that organic groups are removed from the pV4D4 film in increasing amounts as the
temperature of pyrolysis increases. The deposited pV4D4 films had atomic compositions that were
comparable to previously reported pV4D4 films deposited via iCVD.
68
There was no noteworthy
shift in the elemental composition at a pyrolysis temperature of 400 °C, but there was a significant
decrease in carbon content when the temperature was increased to 600 °C. This observation is
consistent with the IR studies that show the disappearance of the polyethylene backbone and
methyl groups, thus causing the composition of the film to shift toward the pure silica structure.
At a pyrolysis temperature of 1100 °C, the silicon fraction was higher than expected from a film
composed purely of ‘Q’ groups which would have a composition of 33.33% silicon and 66.66%
oxygen.
79
The likely reason for this is the penetration of the X-ray photons through the pyrolyzed
film to the underlying silicon substrate, since these films are porous due to the void space left by
the organic groups that cleave during pyrolysis. The small amount of carbon is likely indicative of
surface contamination or the presence of trace amounts of carbon embedded within the structure.
In order to confirm the bonding environments in the film during the pyrolysis process, high
resolution XPS scans of the silicon environment were collected. Alexander et al.
80
have shown
previously that in siloxanes, the oxidation state is the biggest contributing factor to the shifts in the
binding energy of silicon, with reported binding energies of Si(-O1)=101.5 eV, Si(-O2)=102.1 eV,
Si(-O3)=102.8 eV, and Si(-O4)=103.4 eV. O’Hare et al.
81
identified similar binding energies for
these components of a siloxane film: ‘M’=101.63 eV, ‘D’=101.99 eV, ‘T’=102.67 eV, and
‘Q’=103.47 eV. The high resolution scans of the pV4D4 films pyrolyzed at different temperatures
are shown in Fig. 2-5. The areas corresponding to each oxidation state are shown in Table 2-2. As
25

was previously seen in the IR data, there was a small shift from the network structure and small
bond angle Si-O-Si ‘D’ groups to silsesquioxane ‘T’ groups at a pyrolysis temperature of 400 °C.
At 600 °C, there was a dramatic shift toward a mixed oxidation state, with the ‘Q’ groups
dominating. This observation is consistent with the IR data that show that the removal of the
polyethylene backbone and the methyl groups also occurs primarily between 400 °C and 600 °C.
However, as with the FTIR data, there is evidence in the XPS data of leftover hydrocarbon bonds,
with a small number of the ‘D’ groups remaining. Overall, we found that there were similarities in
the ratios of the silicon environments in the bulk and the surface as determined by FTIR and XPS,
respectively, up to a pyrolysis temperature of 400 °C. At a pyrolysis temperature of 600 °C, ‘T’
groups dominated the bulk whereas ‘D’ groups dominated the surface. This is consistent with the
observations of Corriu et al.
82
, who investigated the pyrolysis of siloxane gels and found that after
a pyrolysis temperature of 600 °C, ‘T’ siloxane materials transitioned to both ‘D’ and ‘Q’
structures due to restructuring of Si-C and Si-O bonds. Since methyl groups are cleaved and diffuse
out of the film, there is likely a higher level of recombination at the surface compared to the bulk.
At ~800 °C, these groups mostly transitioned to ‘Q’ groups, however residual ‘D’ groups were
still present. The peak shifted to 103.35 eV at a pyrolysis temperature of 1100 °C, indicating the
transformation of the film into silica. In summary, based on the FTIR/DRIFTS and XPS analysis
results, the formation of silica film occurs at a pyrolysis temperature of 1100 °C.

26

Figure 2-5. High resolution scans of Si 2p for the a) pV4D4 polymer and for the polymer after
pyrolysis at b) 400 °C, c) 600 °C, d) 800 °C, and e) and 1100 °C.

Table 2-2. Peak areas of the silicon oxidation state from high resolution XPS analysis
Oxidation levels SiO 1 SiO 2 SiO 3 SiO 4
Binding energy - 101.8 102.6 103.4
pV4D4 0 100 0 0
400°C 0 72 28 0
600°C 0 15 0 85
800°C 0 6 0 94
1100°C 0 0 0 100
27

Table 2-3 shows the permeance data for three distinct asymmetric membranes made with
different supports: SiC supports composed of 50% of particles with an average diameter less than
80 nm and 50% of particles with average diameter of 0.6 μm, supports made of pure 0.6 μm SiC
powders, and supports made with pure <80 nm particles. Table 2-3 shows the average measured
permeances of He, H2, and their ideal separation factors (defined here as the ratio of permeances)
over Ar for each different type of membrane. It is important to note that the permeances and ideal
separation factors for each membrane were calculated from three distinct samples prepared under
the same conditions. As can be seen in the Table, all silica films are microporous with ideal
separation factors which are several times higher than the corresponding Knudsen separation factor
𝛼 𝑖 𝑗 =
√
𝑚 𝑗 𝑚 𝑖 , where mj is the molecular weight of Ar and mi the molecular weight of He or H 2. The
mixed powder supports (0.6 μm/80 nm) resulted in an asymmetric membrane with a higher
separation factor than the pure powder supports (either 0.6 μm or 80 nm). This is consistent with
prior observations by our group
64
, which show that utilizing larger particle sizes for supports results
in bigger pores and, therefore, higher permeances and lower separation factors. However, using a
mixture of powders with distinct particles sizes allows for better packing and, therefore, better
separation factors with only a small sacrifice in permeance. It is important to note that there have
been other silica membranes produced by both solution-phase
7,12
and conventional CVD
15,21
methods with separation factors several times larger than ours. Our future work will involve
optimizing the thickness, composition, and density of the iCVD polymer film prior to its pyrolysis
to improve the separation factor and permeance of the resulting membranes.
28

Figure 2-6. a) Cross-sectional view of the asymmetric composite structure consisting of the
iCVD polymer layer, and the intermediate and the macroporous SiC layers, and b) top-down
image of the dense iCVD layer on top of the SiC support.

Table 2-3. He and H2 permeances and ideal separation factors of He and H2 over Ar for
membranes resulting from the pyrolysis at 1100 ºC of pV4D4 films deposited on SiC supports.
The testing pressure was 241 kPa and the testing temperature was 473K.
Composition of the
support
He permeance
(mol/m
2
•Pa•s)
H 2 permeance
(mol/m
2
•Pa•s)
He/Ar
separation factor
H 2/Ar
separation factor
100% <80 nm 4.19E-08 2.25E-08 20.7 11.0
100% 0.6μm 7.45E-07 4.81E-07 16.4 10.6
50%(<80nm)/50%(0.6μm) 2.74E-07 1.89E-07 25.1 17.4

2.5 Conclusion

In this study, we fabricated microporous silica membrane films capable of selectively
separating lighter gases (e.g., H2 and He) from bulkier gases (e.g., Ar) through the deposition and
29

subsequent pyrolysis of pV4D4 films. We used in situ DRIFTS and complimentary ex situ FTIR
studies to investigate the chemical structure of pV4D4 at different pyrolysis temperatures and
found that the network and silsesquioxane-dominated bonding environment shifted to a largely
silica ‘Q’ bonding environment at 600 °C. It was not until the pyrolysis temperature reached 1100
°C, however, that we saw a complete transformation into a silica structure. In addition to the
DRIFTS/FTIR studies, we also used XPS to analyze the films after pyrolysis. The XPS results
confirm that organic groups were removed from the film structure as the temperature of pyrolysis
increased. The high resolution XPS analysis allowed us to identify the oxidation state of each film
and to establish that 1100 °C was the pyrolysis temperature at which a pure silica film was formed.
Therefore, we used this pyrolysis temperature to prepare silica membrane films on SiC support
disks to measure their permeation properties. The resulting membrane films were shown to be
microporous and capable of separating light gases. The nature of the supports also has an influence
on the properties of the resulting membrane films with mixed-powder supports exhibiting higher
separation factors.

2.6 Acknowledgements

The authors acknowledge the support of the ACS Petroleum Research Fund (PRF). XPS
data and SEM images were acquired at the Core Center of Excellence in Nano Imaging at the
University of Southern California.

30

3 Chapter Three: Fabrication of SiC-Type Films Using Low-Energy
Plasma Enhanced Chemical Vapor Deposition (PECVD) and
Subsequent Pyrolysis

3.1 Abstract

Silicon carbide (SiC) is a promising material for a variety of applications in the biomedical,
aerospace, and energy industries. Solution-phase techniques have long been used to deposit
precursor films prior to pyrolysis into SiC, but they tend to face difficulties with substrate
compatibility and the use of toxic solvents. In this study, we introduce a solventless synthesis route
for fabricating SiC-type films by depositing an organosilicon copolymer
poly(vinylphenyldimethylsilane-co-divinylbenzene) (p(VPDMS-co-DVB)) film using low-energy
plasma chemical vapor deposition (PECVD) followed by subsequent pyrolysis. The chemical
structure of the film was systematically studied in-situ during pyrolysis as a function of
temperature using Diffuse Reflection Infrared Fourier Transform Spectroscopy (DRIFTS). The
majority of the functional groups were found to have disappeared by a temperature of 800 °C, with
most of the mass loss occurring between 350 °C and 520 °C. Thermogravimetric Analysis (TGA)
was used to measure the loss of mass as the pyrolysis temperature was increased, and the observed
pyrolysis rates were compared to estimates of such rates from the DRIFTS analysis. Our proposed
synthesis route provides a scalable and solventless method of producing SiC-type ceramic films
for such applications as high-temperature sensors and membranes.

3.2 Introduction

31

Silicon carbide (SiC) has been shown to be a valuable material due to its high thermal
conductivity, chemical resistance in corrosive (including acidic) environments, high mechanical
stability, and low thermal expansion coefficient.
83
In addition, SiC has excellent dielectric
properties and biocompatibility. Because of its versatility, SiC films have recently been finding
many uses including, among others, a variety of biomedical
84
and electronic
85
applications. For
example, in the aerospace field, SiC has been used in the manufacturing of nanoelectronic devices
and integrated circuits as well as in the preparation of high-temperature sensors and actuators.
84
In
addition to these applications, SiC/metal composites have been synthesized as a replacement for
pure metals in order to reduce weight while retaining durable strength properties.
86
SiC membranes
have also been used as both macroporous and microporous filters, for applications such as
wastewater treatment
87
and hydrogen gas separations
3
.
SiC thin films can be fabricated using solution-phase
65
and gas-phase
88
techniques, both of
which have been studied by our group. Processes such as dip-coating
89
and polymer infiltration
have been popular methods for depositing precursor SiC films on underlying supports prior to
pyrolysis
90
. Solution-phase techniques, however, usually have substrate compatibility issues, and
also pose safety and environmental concerns due to the use of toxic solvents.
20
As a result, in
addition to the solution-phase processes, a variety of chemical vapor deposition (CVD) techniques,
including hot-wall chemical vapor deposition (HWCVD)
91, 92
and plasma-enhanced chemical
vapor deposition (PECVD),
93, 94, 95
have also been used to prepare SiC films with sub-micron
thicknesses using organosilicon precursors or a combination of silane and gaseous hydrocarbons
as a source.
32

Previously, Pagès et al. deposited, via high-energy PECVD, SiC membrane films on porous
alumina tubes using diethylsilane as a precursor, Ar as the ionizing gas, and employing a substrate
temperature of 300 °C.
96
Huran et. al deposited SiC films on silicon wafers via PECVD using SiH4
and CH4 with flow rates of 10 sccm and 40 sccm, respectively, and a deposition temperature of
450 °C.
97
Similarly, Wei et al. deposited amorphous SiC films on silicon wafers using SiH4, CH4,
and Ar with flow rates of 70, 500, and 700 sccm, respectively, a deposition temperature of 400 °C,
and plasma power of 600 Watts.
98
Despite its success in preparing SiC films, high-energy PECVD
has a number of shortcomings. Because it requires high energy inputs and substrate temperatures,
the technique is not appropriate for applications requiring temperature-sensitive substrates like
those, for example, employing sacrificial porous polymer templates such as polyurethanes with
low degradation temperatures.
99
The aim to directly, in one step, produce SiC, in addition to
dictating the use of high temperatures and power densities, also limits the availability of
appropriate gaseous precursors to mostly silanes with or without low molecular weight
hydrocarbon (e.g., CH4) co-reactants. In addition, the ceramic yield of these reactors is, typically,
quite low, thus dictating the use of high precursor flow rates. For the production of nanoporous
SiC membrane and sensor films deposited on underlying macroporous supports, which is the focus
of our own research, it is difficult to use the high-energy PECVD technique and still be able to
prevent infiltration of the precursors which leads to SiC formation deep into the structure of the
underlying support, which then adversely impacts the permeability of the resulting membrane
materials.
In this paper we propose, instead, the use of low-energy PECVD to fabricate SiC films on
underlying supports. In contrast to directly forming the SiC film on the support, like in the case of
high-energy PECVD, our technique, similarly to solution-phase methods, first prepares a precursor
33

pre-ceramic polymer film, and then converts it, in a subsequent step, via pyrolysis into a SiC film.
However, our technique avoids the common shortcomings of the solution-phase methods noted
above.
We applied previously a similar two-step approach to prepare silica microporous
membrane films
100
, during which we deposited a cross-linked siloxane polymer film on
macroporous supports using, however, a different technique called initiated chemical vapor
deposition (iCVD). We subsequently pyrolyzed these pre-ceramic films to form membranes
capable of separating helium from its binary mixtures with larger molecules (e.g., argon) through
a molecular sieving mechanism. Unlike PECVD, iCVD uses an initiator, typically, a peroxide
compound, to initiate free-radical polymerization of monomers.
55
For the fabrication of SiC
precursor polymer films, it is important to limit the amount of oxygen embedded in the backbone
of the polymer to avoid it from being incorporated into the structure of the resulting ceramic film,
which makes it challenging to prepare such films using iCVD. It is possible to convert siloxane
moieties directly into SiC, but it typically requires the use of much higher temperatures or
employing an additional carbon source.
61
Low-energy PECVD, in contrast to iCVD, does not
require the use of an initiator molecule. In low-energy PECVD, a low-energy plasma is used to
initiate the reactions between chemical species to prepare a variety of organic films while
employing low precursor flow rates and relatively low substrate temperatures (<100 °C).
101, 102, 103

The technique has been used to deposit homogeneous films on large area substrates, and such films
tend to be more cross-linked and to have stable chemical and physical features.
104
Increased cross-
linking in the structure of the precursor polymer has been shown to help improve the ceramic yield
during film pyrolysis.
70
The technique is amenable to “roll-to-roll” processing,
105
to produce large-
34

area, uniform precursor polymer films that can then pyrolyzed to produce large-area inorganic
SiC-type films with correspondingly large areas.
In this project, we study the deposition of the pre-ceramic polymer films on porous
substrates using low-energy PECVD, and subsequently pyrolyze these films to produce SiC-type
ceramic films. We systematically study the deposition of the polymer at various monomer flow
rate ratios. We also analyze the structural changes of the polymer as it undergoes pyrolysis to
eventually form a SiC-type ceramic and examine the kinetics of the pyrolysis process. To our
knowledge, we are the first group to study the pyrolysis of PECVD-deposited polymer films to
form inorganic SiC-type films.

3.3 Experimental

PECVD Deposition
A schematic of the lab-scale low-energy PECVD reactor used in this study to deposit the
various polymer films is shown in Figure 3-1. It is a cylindrical reactor chamber, 250 mm in
diameter and 48 mm in height, purchased from the GVD Corporation. The reactor operates under
vacuum conditions via the use of a rotary-vane vacuum pump (Edwards E2M40). For the
experiments in this paper, the reactor pressure was maintained at 180 mTorr using a feedback-
type, throttle-valve controller (MKS 153D) with the aid of a capacitance manometer (MKS
622C01TDE Baratron). The temperature of the deposition stage inside reactor was kept constant
at 50 °C using a recirculating bath chiller (Thermo Scientific NESLAB RTE 7).
35

Figure 3-1. Schematic of the PECVD reactor.
Prior to initiating film deposition, we identified a number of organosilicon candidate
monomers, all of which had physical properties amenable to low-energy PECVD deposition and
chemical structures pre-disposed to form SiC-type films upon pyrolysis; they varied in the type of
organic moieties they contained. Among these candidate monomers, we selected for further study
vinylphenylmethylsilane (VPDMS) because it forms polymers with phenyl pendant groups which
tend to have good SiC yields post-pyrolysis.
69
We also selected divinylbenzene (DVB) as a cross-
linker, since it has been shown that cross-linked pre-ceramic polymers tend to have higher ceramic
yields.
69
Figure 3-2 shows the chemical structures of VPDMS and DVB.
36

Figure 3-2. The chemical structure of the VPDMS (left) and DVB (right) monomers.

The VPDMS (Gelest, 95%) and DVB (Aldrich, 80%) monomers were used as received
without further purification. They were loaded into stainless steel jars that were then mounted onto
the reactor; the temperature of both jars was kept constant at 25 °C. The monomers were introduced
into the reactor via heated lines maintained at a temperature of 40 °C. The VPDMS and DVB flow
rates were varied between 1.8 and 4.0 sccm, depending on the desired molar flow ratios. Argon
gas (99.999% pure) was also fed into the reactor as the ionizing gas, and its flow rate was
maintained at 40 sccm using a mass flow controller (MKS Type 1479A). Prior to the deposition,
the chamber of the reactor was purged six times in order to reduce the oxygen content of the
ambient environment. This was done by slowly increasing the reactor pressure from vacuum to a
positive pressure using argon gas. In addition, all of the lines were held under vacuum prior to the
deposition in order to further reduce the oxygen content. We studied the deposition of the pre-
ceramic polymer film onto two different powder substrates, potassium bromide (KBr, Alfa Aesar)
and barium fluoride (BaF2, Aldrich, 99.95%). The deposition rate on the powder substrates was
37

determined by measuring in situ via interferometry with a He-Ne laser (Industrial Fiber Optics,
633 nm) the deposition rate on a reference silicon wafer (Wafer World 119) placed nearby on the
reactor stage.
We used an external radio frequency (RF) plasma generator (Diener; 13.56 MHz, 100 W)
with a manual matchbox (Diener) which was connected to a nichrome filament array (Omega
Engineering, 80%/20% Ni/Cr) positioned 4.6 cm above the reactor stage. The RF plasma generator
was operated at 35 W.
Chemical Characterization
To analyze the chemical structure of the polymer films, we measured their IR absorbance
spectra ex situ using a FTIR spectrophotometer (Nicolet iS10, Thermo Scientific). For such
measurements, the films (~1 µm thick, as measured via interferometry) were deposited onto clean
silicon (Si) wafers. The clean bare Si wafer was also used as the background for the spectra. To
acquire the monomer IR spectra, a drop of liquid monomer was sandwiched in between two Si
wafers and the IR spectrum taken, with two Si wafers being used as the background.
To analyze the change in the polymer structure as a function of temperature during
pyrolysis, we used Diffuse Reflectance Infrared Fourier Transform Spectroscopy (DRIFTS,
COLLECTOR II, Thermo Scientific). The DRIFTS technique allows for monitoring the change in
the IR spectra of a powder substrate in situ as the sample is heated to the desired temperature.

To
prepare the polymer-coated powder samples for the DRIFTS experiment, the KBr and BaF2
powders were spread manually with the aid of a razor blade onto a Si wafer. The polymer was then
38

deposited onto the powder-covered wafers. The film deposition was allowed to proceed until a
nearby bare Si substrate was coated with a ~ 2 μm thick polymer film. The polymer-coated
powders were then inserted into the DRIFTS sample cup and leveled with a razor blade, as
instructed in the instrument manual.
65
The DRIFTS apparatus was then closed and heated at a
temperature of 100 °C, under flowing ultra-high purity (UHP) Ar at a pressure of 170 kPa for 2 hr
in order to eliminate any moisture effects and to ensure an inert atmosphere during pyrolysis. For
the DRIFTS measurements, the cell was heated at a rate of 1 °C/min from a temperature of 100 °C
to 800 °C, which is the maximum operating temperature of the high temperature and pressure
DRIFTS cell used in this experiment. The IR absorbance spectra were recorded at 20 °C
temperature intervals, or approximately every 20 min. After every temperature rise interval of 100
°C, the increase in temperature was stopped and the samples were allowed to stay (“soak”) at that
temperature for 1 hr. During this pyrolysis period under isothermal conditions, the IR absorbance
spectra were again recorded every 10 min. Uncoated KBr and BaF2 powders were used as the
background spectra for the polymer-coated KBr and BaF2 powders.
The polymer-coated BaF2 powder was also pyrolyzed up to a higher temperature (1000 °C)
outside the DRIFTS cell. For that, the powder was placed inside a graphite furnace, evacuated
first, and then purged with UHP Ar for 1 hr to eliminate potential oxygen contamination. After six
cycles of evacuation and purging, the powder was heated at a rate of 1 °C/min in flowing Ar up to
a temperature of 1000 °C. As with the DRIFTS experiments, for every 100
o
C interval, the ramping
of temperature was stopped, and the sample was kept at that temperature for 1 hr. After the
temperature reached 1000 °C, the powder was kept at that temperature for 3 hr, after which it was
cooled back to room temperature at a rate of 3 °C/min. The slow heating and cooling rates were
chosen because they have been shown previously to prevent cracking during the preparation of
39

ceramic materials.
66
After being cooled down to room temperature, the powder was placed in the
DRIFTS sample cup, and the chamber was purged with Ar for 30 min before IR absorbance spectra
were taken. The DRIFTS and FTIR spectra were both analyzed using the freeware SpectraGryph
1.2. and the deconvolution was performed using OriginPro Peak Analyzer.
For the atomic composition and surface analysis of the polymers, a plasma focused ion
beam (PFIB)/scanning electron microscope (SEM) system was utilized (Thermo Scientific, Helios
G4 PFIB UXe DualBeam FIB/SEM). The microscope is equipped with an x-ray energy dispersive
spectroscopy (EDS) system for elemental analysis equipped with an Oxford UltimMax 170 Silicon
Drift Detector. Data were collected with a voltage of 10 kV and a 1.6 nA beam current. The
working distance of the apparatus was set to 5.6 mm and for each spectrum, 1,000,000 counts was
chosen as the setting.
Thermogravimetric Analysis
Thermogravimetric analysis (TGA, Cahn Instruments, 12130-01) was used to measure the
mass loss of the polymer during pyrolysis as a function of temperature. For the analysis,
approximately 100 mg of the polymer-coated BaF2 powder was placed inside the cup of the TGA
apparatus. The TGA chamber was then purged with flowing Ar (flow rate of 30 sccm) for 20 min.
The powder was heated (ramp rate of 3 °C/min) to 100 °C and allowed to rest there for 2 hr under
flowing Ar in order to eliminate any residual sample moisture. Then, the sample was heated from
100 °C to 300 °C at a ramp rate of 2 °C/min, and subsequently in increments of 100 °C from 300
°C to 900 °C at a ramp rate of 5 °C/min, with soaking times of 1 hr at the end-point of each 100
°C heating interval.
40

Film Thickness Measurement
The change in the thickness of the films deposited on bare silicon wafers before and after
polymer pyrolysis was measured using a stylus profilometer (AMBiOS XP-2). The scan length
was set at 5 mm.

3.4 Results and Discussion

Polymer Film Analysis Using FTIR
PECVD was used to deposit five different polymer films: poly(vinylphenyldimethylsilane)
(pVPDMS), poly(divinylbenzene) (pDVB), and three poly(vinylphenyldimethylsilane-co-
divinylbenzene) (p(VPDMS-co-DVB)) copolymers employing VPDMS:DVB feed flow rate ratios
of 1:1, 2:1, and 1:2. Figure 3-3 shows the transmission FTIR spectra of the pVPDMS, the pDVB,
and the p(VPDMS-co-DVB) copolymers prepared by the PECVD technique on Si wafers as well
as of the monomers themselves (for a reference to the various IR bands in this and subsequent
Figures, please see Table 3-1). The monomers contain vinyl stretching bands (region iia and iib)
at approximately 903 cm
-1
and 960 cm
-1
. For the pVPDMS and pDVB polymers as well as the
copolymers, the vinyl bands are significantly reduced, something which has also been reported
previously by other investigators.
106
The reduction of the vinyl stretching bands area (region iia
and iib) is accompanied by an increase in the ethylene backbone C-H stretching bands (region viia)
at approximately 2940 cm
-1
formed during the polymerization reaction.
41

Table 3-1. IR bands shown in Figures 3-3 to 3-8
Regions IR Bands Corresponding to Bonds
i Si-C stretching
ii vinyl bond stretching
iiD Si-(CH2)n-Si stretching
iii Si-C6H5 stretching
iv symmetric Si-CH3 stretching
v aromatic C-H and C=C stretching bands
vi Si-H stretching
vii C=C and C-H hydrocarbon bonds from polyethylene bonds and aromatic rings

The increase in these bands is particularly noticeable in the pDVB and the 1:1 and 1:2
VPDMS:DVB copolymer films, indicating that the increased incorporation of the methylene
bridges result from the DVB polymerization.
107
The decrease of the DVB monomer bands between
700 and 850 cm
-1
(region ia and ib) are indicative of decrease of the meta and para vibrations of
the benzene ring, as has previously been seen by iCVD polymerization of DVB.
107
The bands
between 3000 and 3100 cm
-1
(region viib) are attributed to the various v(=C-H) modes of the
aromatic rings. Their continued presence in the spectra, as the polymer is being deposited, is
indicative of the fact that the polymerization reaction occurs primarily through the vinyl bonds.
The methyl (Si-CH3) stretching bands at 1250 cm
-1
(region iv) which are prominent in the VPDMS
spectra also remain in the pVPDMS polymer and in the p(VPDMS-co-DVB) copolymer spectra,
indicating that these bonds have remained relatively intact during polymerization.
108
The aromatic
(C6H5) C-H and C=C (region v) stretching bands between 1300 and 1600 cm
-1
that are prominent
in both the VPDMS and DVB spectra are still present in the pVPDMS and in the p(VPDMS-co-
DVB) copolymer spectra, again indicating that these bonds remained relatively intact during
polymerization. In addition, the Si-C stretching band (region ib) between 700 and 900 cm
-1
also
42

remains intact in both the pVPDMS and the p(VPDMS-co-DVB) copolymers. This peak size is
noticeable in the pVPDMS, 1:1, and in the 2:1 VPDMS:DVB copolymer film spectra, but is
reduced in the 1:2 VPDMS:DVB spectra, indicating that there is much less VPDMS incorporated
in the polymer prepared using this volumetric flow rate ratio. There is also a small band around
2100 cm
-1
(region vi) corresponding to the Si-H stretching vibration in all of the deposited
polymers films (except pDVB), which is, potentially, indicative of a small degree of polymer (Si-
C bond) cleaving.
109

Figure 3-3. FTIR spectra of PECVD-deposited polymer films deposited on Si wafers. The dashed
lines represent regions: (i) Si-C stretching and para- and meta-vibrations of DVB ring, (ii) vinyl
bond stretching, (iii) Si-C6H5 stretching, (iv) symmetric Si-CH3 stretch stretching, (v) aromatic
C6H5 C-H and C=C stretching bands, (vi) Si-H stretching, and (vii) C=C and C-H hydrocarbon
bonds from polyethylene bonds and aromatic rings.
In both the pDVB and the copolymer spectra, the majority of the vinyl bands are reduced,
indicating that most of the vinyl bonds have reacted and the polymer is, therefore, thoroughly
43

cross-linked. To confirm that the copolymer was, indeed, cross-linked, we investigated the
solubility of such films deposited on Si wafers in acetone at room temperature. After performing
a number of solubility tests, we found that the pDVB film was insoluble in acetone, while the
pVPDMS film completely dissolved in it. The copolymer films, similarly to the pDVB film, were
also not soluble in acetone, which indicates that they are, likely, also cross-linked.
The atomic carbon to silicon (C/Si) ratio in the precursor film plays an important role in
determining the composition of the final ceramic film, in particular its carbon content. To measure
this ratio, we carried out EDS measurements in order to determine the atomic composition of the
films. EDS has been shown previously to be an effective method for determining the elemental
composition of silicon-based polymer films.
110,111
Table 3-2 shows the atomic composition of each
film based on the EDS testing.

Table 3-2. Atomic Composition of the Deposited Polymer Films from EDS Analysis
Sample Spot
Atomic Composition
C:Si Ratio Theoretical C:Si Ratio Si:O Ratio
C Si O Other
pVPDMS
1
89.2 8.9 1.7 0.1 10.0
10
5.2
2
89.1 8.9 1.8 0.1 10.0 4.9
3
89.2 8.9 1.8 0.2 10.0 4.9
pDVB
1
93.1 0.1 6.8 0 -
-
-
2
93 0.1 6.9 0 - -
3
93 0.1 6.9 0 - -
1:1
1
94.7 3.8 1.4 0.1 24.9
20
2.7
2
94.7 3.8 1.4 0.1 24.9 2.7
3
94.8 3.8 1.4 0.1 24.9 2.7
2:1
1
94 4.4 1.6 0 21.4
15
2.8
2
94 4.4 1.6 0 21.4 2.8
3
94 4.4 1.6 0 21.4 2.8
1:2
1
96.1 2.5 1.4 0 38.4
30
1.8
2
96.1 2.5 1.4 0 38.4 1.8
3
96 2.5 1.5 0 38.4 1.7

44

We analyzed three different spots on each sample, with at least a distance of 1 mm apart
from each other; for each spot, we carried out three different measurements, with the average
composition value for each spot shown in Table 3-2. As Table 3-2 indicates, the composition of
the different spots is quite similar, which signifies spatially uniform deposition kinetics. All films
were shown to be carbon rich, as expected from the original (C/Si) ratios of both the VPDMS and
DVB monomers. Previous studies
83
have shown that increasing the plasma energy during PECVD,
beyond 100 Watts, generally, resulted in decreased carbon content of the resulting polymer; this
is, likely, due to the cleavage of various functional monomer moieties taking place during the
polymerization process. Because we utilize relatively low power (<40 Watts) during our PECVD
studies, the retention of carbon was very high, as can be seen from the (C/Si) ratio of the pVPDMS
film that matches that of the original monomer. For the copolymer films, the (C/Si) ratio is higher
(24 -25%) than the predicted value based on the feed volumetric flow rate ratios (on the assumption
that they incorporate into the structure of the growing polymer at rates which are proportional to
their flow rates). This indicates that during the formation of these copolymers, the DVB monomer
incorporates at a faster rate into the growing polymer chain than the VPDMS, thereby resulting in
a greater carbon content in its structure.
In all polymer films, we find a small oxygen content (the silicon to oxygen (Si/O) ratio is
also included in the Table). Since neither monomer contains oxygen, its presence in the EDS
spectrum of the final polymer may signify either oxygen incorporation during the preparation of
the films, as a result of trace oxygen contaminants being present in the PECVD chamber, or due
to exposure to the laboratory air during the transfer step from the PECVD chamber to the
SEM/EDS laboratory. It should be also noted, that the DVB monomer contains an inhibitor (4-
tert-butylpyrocatechol, 1-5 %) which has oxygen in its structure, so it is conceivable that this
45

inhibitor may have been incorporated into the polymer structure during the deposition (this may
also explain the lower Si/O ratio of the copolymer films relative to pVPDMS). There are no peaks
at the Si-O-Si position of the FTIR spectra, so the oxygen in the polymer films is, likely, either
adsorbed oxygen (or hydroxyls) as a result of exposure to laboratory air, or due to C=O bonds
being formed during the polymer deposition process.

Study of the Pyrolysis Process Using DRIFTS
The transformation of the polymer film into a ceramic during pyrolysis was studied using
the DRIFTS technique. For that, the polymer was deposited, via PECVD, onto two different
powder substrates, made of KBr and BaF2, in order to investigate whether the substrate itself
influences the composition and/or structure of the ceramic produced. We have previously used the
DRIFTS technique to study the scission of functional groups during the pyrolysis of a polymer
film and its conversion into a ceramic.
100
Other groups
112,113
have also used the DRIFTS technique
to study the heat treatment of organosilicon materials to produce siloxane films.
Figure 3-4 shows the DRIFTS data generated during the pyrolysis of the 1:1 VPDMS:DVB
copolymer coated on KBr powder. The DRIFTS cell was heated at a rate of 1 °C/min from a
temperature of 100 °C to 700 °C (the maximum operating temperature of our DRIFTS cell is 800
°C) since the melting temperature of KBr is ~734°C. The IR absorbance spectra were recorded at
20 °C temperature intervals, or approximately every 20 min. After every temperature rise interval
of 100 °C, the increase in temperature was stopped and the samples were allowed to “soak” at that
temperature for 1 hr. During the “soak” pyrolysis period under isothermal conditions, the IR
absorbance spectra were again recorded every 10 min. Comparing the spectra for the starting (not
46

pyrolyzed) polymer coated onto the KBr powder in Figure 3-4 to the corresponding FTIR spectra
of the same polymer coated on a Si wafer in Figure 3-3, it can be concluded that the major bands
in both spectra are the same. They include the Si-C stretching band at ~800 cm
-1
(band i in Figure
3-3), the phenyl stretching band at 1100 cm
-1
(band iii), the methyl symmetric stretching band at
1260 cm
-1
(band iv), the Si-H peak at approximately 2100 cm
-1
(band vi), and the backbone C-H
stretching bands between 2920 and 3000 cm
-1
(band vii). The various benzene ring C-H and C=C
stretching bands of pDVB are located between 1400 and 1600 cm
-1
(band v).
From the spectra in Figure 3-4, it can be concluded that there are no significant changes
that take place in the structure of the polymer until a pyrolysis temperature of ~300 °C is reached.
Between a temperature of 300 °C and 400 °C, the Si-H stretching band (band vi, ~2100 cm
-1
)
decreases significantly. In addition, the methyl stretching band (band iv) and the Si-C6H5 stretching
bands (band iii) both decrease. The various stretching bands associated with pDVB (band v)
decrease as well, likely indicating that some of the rings have started to cleave or decompose.
Straus et al.
114
who previously studied the pyrolysis of pDVB also found that at temperatures
between 300 °C and 400 °C, the carbon bonds in the pDVB ethylene chain break to form
monomers, dimers, and trimers.
By a pyrolysis temperature of 500 °C, the stretching bands for most of the organic groups
have substantially decreased, and there is also a significant decrease in the ethylene band at 2920
cm
-1
(band vii). The disappearance of these bands represents a similar behavior to that we have
previously seen during the pyrolysis of organosilicon polymers, where the majority of the
functional groups cleave at pyrolysis temperatures greater than 400 °C. Schiavon et. al.
115
, during
pyrolysis experiments with cyclical polysiloxane composites using DRIFTS, also found that the
majority of the C-H bands disappeared in this temperature range.

47

Figure 3-4. DRIFTS data during pyrolysis for the copolymer coated on KBr. The dashed lines
represent: (i) Si-C stretching, (iiD) Si-(CH2)n-Si stretching, (iii) Si-C5H6 stretching, (iv)
symmetric Si-CH3 stretching, (v) pDVB C-H and C=C stretching bands, (vi) Si-H stretching, and
(vii) C-H hydrocarbon bonds from polyethylene bonds and aromatic rings.

By a pyrolysis temperature of ~600 °C, most of the phenyl (band iii) and methyl stretching
vibration (band iv) bands have completely disappeared, and the ethylene stretching band has
largely vanished as well. Through this pyrolysis process, the Si-C stretching band around 800 cm
-
1
(band i) has remained prominent, likely, indicating that the silicon-carbon bond has stayed intact
in the structure. The final structure that is observed after a pyrolysis temperature of 700 °C consists
48

of a Si-C network with small amounts of Si-(CH2)n-Si (band iiD) being present as well. There are,
however, residual organic functional groups remaining in the spectra, as indicated by the presence
of methyl and polyethylene stretching bands.
We have also studied, via DRIFTS, the pyrolysis of copolymer films deposited on BaF2
powder. Using BaF2 instead of KBr is advantageous because its melting temperature is 1368
o
C,
and, thus, it can withstand higher pyrolysis temperatures. Figures 3-5, 3-6, 3-7, and 3-8 show the
results from the pyrolysis of BaF2 powders coated with pVPDMS, 1:1 VPDMS:DVB, 2:1
VPDMS:DVB, and 1:2 VPDMS:DVB polymers. The range of pyrolysis temperatures in these
Figures is broader than those in Figure 3-4, ranging up to 800 °C, the limit of the DRIFTS cell.
Similarly with the spectra of the polymer-coated KBr powder, the DRIFTS spectra of the
polymer-coated BaF2 powders show the same peaks as those in the FTIR spectra of the same
polymers coated on Si wafers, shown in Figure 3-3. For all four polymers, there were no major
changes in the spectra up to a pyrolysis temperature of ~300 °C. In the spectra taken at a
temperature of 400 °C, we observe a reduction in the size of several bands, including the Si-H
stretching band (band vi) at ~2100 cm
-1
and the phenyl Si-C6H5 stretching band (band iii) at ~1100
cm
-1
. In addition, there is a slight reduction in the methyl stretching band at 1260 cm
-1
(band iv)
and of the backbone C-H bands around 2980 cm
-1
(band vi). Moreover, for the 1:1 and 1:2
VPDMS:DVB copolymers, the pDVB-related stretching band ~1350 cm
-1
(band v) is significantly
reduced, and there is also a slight increase in the band around 1000 cm
-1
(band iiD), assigned to
the Si-(CH2)n-Si stretching vibration. An increase of the band assigned to Si-(CH2)n-Si between
400 °C and 600 °C was also observed by Breuning
112
during the heat treatment of a polysilazane
polymer, and it can be attributed to the rearrangement of bonds to form carbon bridges between
silicon atoms. The higher incorporation of DVB monomer in the 1:1 and 1:2 VPDMS:DVB
49

copolymers seems to have led to a larger increase in the formation of hydrocarbon bridges, while
no such bridges are observed forming for the 2:1 VPDMS:DVB copolymer which has less DVB
incorporated in its structure. We did not observe the formation of such bridges for the pVPDMS
polymer, so it can be concluded that the presence of DVB has an important role in the formation
of these hydrocarbon bridges.
In the spectra taken at a pyrolysis temperature of 600 °C, the backbone and the methyl
stretching bands (bands vii and iv, respectively) have significantly decreased in size. The size of
the aromatic stretching bands (band v) has decreased as well. For the 1:1 and 1:2 VPDMS:DVB
copolymers, these bands are much more prominent than their counterparts in the pVPDMS and
2:1 VPDMS:DVB copolymer spectra. However, by a pyrolysis temperature of 700 °C, these bands
have largely disappeared from the spectra of all polymers studied. In the spectra taken at the
temperature of 800 °C, the bands for Si-(CH2)n-Si (band iiD) substantially decrease for the 1:1 and
2:1 VPDMS:DVB copolymers, indicating the conversion of these groups into a Si-C network.
However, for the 1:2 copolymer, the Si-(CH2)n-Si band is still quite prominent, again indicating
the impact the increased presence of DVB in the copolymer precursor has on the composition of
the resulting ceramic. The film from pyrolysis at 800 °C consists, largely, of a Si-C network
structure with a few functional groups along with residual Si-(CH2)n-Si segments remaining in the
network. These results agree with the KBr DRIFTS results, and they confirm that a Si-CX-type of
material is produced at elevated pyrolysis temperatures.

50

Figure 3-5. DRIFTS data for the pVPDMS-coated BaF2. The dashed lines represent: (i) Si-C
stretching, (iii) Si-C5H6 stretching, (iv) symmetric Si-CH3 stretching, (v) pDVB C-H and C=C
stretching bands, (vi) Si-H stretching, and (vii) C-H hydrocarbon bonds from polyethylene bond
and aromatic rings.

51

Figure 3-6. DRIFTS data for the 1:1 p(VPDMS-coDVB)-coated BaF2. The dashed lines
represent: (i) Si-C stretching, (iiD) Si-(CH2)n-Si stretching, (iii) Si-C5H6 stretching, (iv)
symmetric Si-CH3 stretching, (v) pDVB C-H and C=C stretching bands, (vi) Si-H stretching, and
(vii) C-H hydrocarbon bonds from polyethylene bond and aromatic rings.
52

Figure 3-7. DRIFTS data for the 2:1 p(VPDMS-coDVB)-coated BaF2. The dashed lines
represent: (i) Si-C stretching, (iiD) Si-(CH2)n-Si stretching, (iii) Si-C5H6 stretching, (iv)
symmetric Si-CH3 stretching, (v) pDVB C-H and C=C stretching bands, (vi) Si-H stretching, and
(vii) C-H hydrocarbon bonds from polyethylene bond and aromatic rings.
53

Figure 3-8. DRIFTS data for the 1:2 p(VPDMS-coDVB)-coated BaF2. The dashed lines
represent: (i) Si-C stretching, (iiD) Si-(CH2)n-Si stretching, (iii) Si-C5H6 stretching, (iv)
symmetric Si-CH3 stretching, (v) pDVB C-H and C=C stretching bands, (vi) Si-H stretching, and
(vii) C-H hydrocarbon bonds from polyethylene bond and aromatic rings.

To better visualize the changes that the polymer goes through as it is being pyrolyzed, we
baselined the IR spectra and performed an area-under-the-curve analysis to study the rates of
disappearance of various functional groups. This analysis assumes the applicability of Lambert’s
Law
116
, meaning that the concentration of the functional groups in the material is proportional to
the absorbance of their corresponding individual peaks. This law has been shown previously
117,118

to be valid for a variety of organosilicon polymer materials, and we have no reason to believe that
54

it does not apply to the organosilicon-type of materials studied here as well. The spectra area peaks
were analyzed using Spectragryph and Origin Peak Fitting. The area between 2800 and 3100 cm
-
1
required peak fitting, as the peaks between 2800 and 3000 cm
-1
are indicative of the C-H
stretching for the sp
3
carbons in the polymer backbone, while the peaks between 3000-3100 cm
-1
are indicative of the sp
2
carbon C-H stretching of the aromatic rings from the DVB and Si-C6H5.
The points collected during the “soak” periods, each spaced 100 °C apart, were averaged out to a
single point.
Figure 3-9 shows the peak area for each functional group derived from the DRIFTS spectra.
In the spectra, significant decreases in the areas of the functional groups did not occur until a
pyrolysis temperature of approximately 200 °C. At ~280 °C (for the pVPDMS) and ~320 °C (for
the 2:1 VPDMS:DVB copolymer), we begin to see a large decrease in the area of band v (aromatic
C-H and C=C stretching bands), signifying, likely, the decomposition of the benzene rings. For
the 1:1 and 1:2 VPDMS:DVB copolymers, the decrease of the band occurs at a higher a pyrolysis
temperature, ~400 °C. We attribute this to the cross-linking ability of the DVB that helps keep the
polymer intact at higher pyrolysis temperatures by slowing down the decomposition of the
aromatic rings.
In the pVPDMS, 2:1 VPDMS:DVB, and 1:2 VPDMS:DVB copolymer spectra, the
reduction of the aromatic rings area is subsequently followed by an increase in the area of the
methyl (band iv) and polyethylene chain peaks (band iiD) in all of the copolymer spectra, which
is indicative of possibly rearrangement to form hydrocarbon bridges between the Si atoms as well
as loose hydrocarbon functional groups. The 1:1 VPDMS:DVB copolymer, in comparison, had a
small amount of phenyl loss before 400 °C, but by a pyrolysis temperature of 420 °C, the peak
areas for the phenyl, methyl, and ethylene groups had all started diminishing. By a pyrolysis
55

temperature of 500 °C, both the methyl and ethylene areas in all of the polymers had started
decreasing, with the 1:2 VPDMS:DVB copolymer having the highest pyrolysis temperature before
these groups started decreasing. After this pyrolysis temperature, there is a sharp decrease in both
the ethylene and methyl groups for all of the polymer films. Previously, we have seen that this is
the temperature at which organic functional groups had started to cleave from organosilicon
polymers. It is important to note that the decrease of the methyl and ethylene peak areas is delayed
in the 1:2 VPDMS:DVB polymer, likely due to the high level of cross-linking compared to the
other polymer films.
Past a temperature of 500
o
C, the peak areas of all functional groups continue to decrease,
but the Si-C stretching bands stays relatively high. This is indicative of the polymer changing from
a largely organosilicon structure into an Si-C type of network structure. At the pyrolysis
temperature of 800 °C, only about 20% or less of the phenyl groups (band v) have remained in the
structure. The ethylene band remains very prominent in the spectra, of all polymers ranging
between 40 to 60% of the original area. Among all polymers, the 1:2 VPDMS:DVB copolymer
retained the most functionality, with approximately 60% of the methyl and ethylene groups
remaining. This can be attributed to the larger amount of Si-(CH)n-Si remaining in the structure,
as indicated by the DRIFTS results.

56

57

58

Figure 3-9. Functional group peak area change as a function of pyrolysis temperature for different
polymer films: a) pVPDMS b) 1:1 p(VPDMS-co-DVB) c) 2:1 p(VPDMS-co-DVB) and d) 1:2
p(VPDMS-co-DVB).

Finally, to confirm the eventual transition of the polymers into a ceramic material, the
polymer-coated powders were pyrolyzed ex-situ at a temperature of 1000 °C for 2 hr. in a graphite
furnace. For these experiments, after placing the polymer samples inside, the furnace was first
evacuated using a mechanical vacuum pump to remove any potential contaminants, and then
flowing Ar was injected into the furnace to ensure that pyrolysis takes place under inert conditions.
Post-pyrolysis, the powders were placed again in the DRIFTS chamber to carry out the analysis,
with uncoated BaF2 used as the background. The FTIR results of the polymer-coated powders
pyrolyzed at the higher temperature obtained in the DRIFTS cell are shown in Figure 3-10. The
spectra for the pVPDMS and the 1:1, 2:1, VPDMS:DVB copolymers only show a strong
59

absorption band between 700 and 900 cm
-1
, which is the primary anti-symmetric Si-C absorption
band, as also reported by other groups
119
studying SiC with the DRIFTS technique. Specifically,
the spectra contain no bands around 3000 cm
-1
, which indicates that the long-chain carbon and
aromatic bonds have been removed by this temperature. The 1:2 spectra also showed primarily a
rich SiC environment, but also contained a small absorption band around 1100 cm
-1
, which is
indicative of the presence of a network of Si-(CH2)n-Si bridges
102
.

Figure 3-10. DRIFTS IR spectra of polymer-coated BaF2 powders pyrolyzed at 1000 °C.

Table 3-3. Atomic Composition of the Pyrolyzed Polymer Films from EDS Analysis
Sample
Atomic Composition (%)
C/Si Ratio Si/O Ratio
Carbon Silicon Oxygen Other
pVPDMS 72.7 19.2 8.1 0 3.8 2.4
1:1 VPDMS:DVB 72.7 19.3 8.0 0 3.8 2.4
2:1 VPDMS:DVB 70.6 20.6 8.8 0 3.4 2.3
1:2 VPDMS:DVB 75.5 18.5 6.0 0 4.1 3.1

60

In order to determine the change in atomic composition of the polymer films, the polymer-
coated wafers were placed in a tube furnace and pyrolyzed under flowing Ar at 1200 °C. Table 3-
3 shows the atomic composition of the pyrolyzed films derived from the EDS analysis. As
expected, there was a significant decrease in the carbon content of these films when compared to
that (see Table 3-2) of the starting polymer films. The differences become more obvious when one
compares the (C/Si) atomic ratio of the original films to the corresponding ratio of the pyrolyzed
ones. The significant decrease in the (C/Si) ratio agrees with the DRIFTS results, which suggest a
large portion of the methyl and ethylene functional groups are evolved from the film as the
pyrolysis temperature is increased. The (Si/O) atomic ratio differences between the original
polymer films and their pyrolyzed counterparts are, on the other hand, significantly smaller as can
be seen from comparing the two Tables, with the ratio remaining almost invariant for the 1:1
p(VPDMS-co-DVB) and 2:1 p(VPDMS-co-DVB copolymers. The EDS results suggest that the
final ceramic is a silicon oxycarbide (SiOxCy) material, though it is likely that a good fraction of
its carbon content is free-standing carbon.

Thermogravimetric Analysis of Polymer Pyrolysis
Thermogravimetric Analysis (TGA) measurements can be used to determine the mass loss
of materials and their stability during heating as a function of temperatures, and it is useful in
determining the kinetics of polymer decomposition. For the TGA experiments in this study, the
polymers were coated on BaF2 powder and then pyrolyzed under flowing Ar. The heating program
consisted of the following steps: First heating at a rapid rate of 5 °C/min to a temperature of 100
°C and “soaking” there for 2 hr, followed then by rapid heating to reach 300 °C, with a “soak”
61

period there of 1 hr; this was then followed by several steps each consisting of a slow 100 °C
temperature rise followed by an 1 hr “soak” period until a final pyrolysis temperature of 900 °C
was reached, after which the sample was cooled down to room temperature.

Figure 3-11. TGA graph showing the fraction of decomposition of the polymer films deposited
on BaF2: a) pVPDMS b) 1:1 VPDMS:DVB p(VPDMS-co-DVB) c) 2:1 VPDMS:DVB
p(VPDMS-co-DVB) and d) 1:2 VPDMS:DVB p(VPDMS-co-DVB).

From the mass loss of the polymer during the TGA experiments, one can calculate the
polymer decomposition fraction, α, at each time and corresponding temperature from the equation
below, where mi, mt, and mf are the initial mass, the mass at time t, and the final mass of the
sample, respectively.
(1)

𝛼 =
𝑚 𝑖 − 𝑚 𝑡 𝑚 𝑖 − 𝑚 𝑓
62

The results of the experiments are shown in Figure 3-11, which shows the decomposition
fraction as a function of time. There was minimal mass loss before a pyrolysis temperature of 300
°C, with the 2:1 VPDMS:DVB copolymer experiencing the highest degree of decomposition at
~10%. This is in agreement with the DRIFTS experiments, where we observed minor losses in Si-
H and phenyl groups, and the small change in mass can, potentially, be attributed to those losses.
There were no major changes during the “soaking” period at 300 °C, but the loss in mass
accelerated quickly after the temperature was raised beyond 300 °C. This is consistent with the
scission of the aromatic rings from both the DVB and VPDMS components of the polymers, as
observed from the DRIFTS results. The biggest change in α occurred between 400 °C and 600 °C.
This large mass loss correlates well with the degradation of the aromatic groups, and the cleavage
of the polyethylene backbone and the methyl groups, as seen in the DRIFTS results. At a pyrolysis
temperature of 500 °C, all four polymers had an α of ~0.5.
In the temperature range between 500 °C and 700 °C, the pVPDMS polymer had the
highest increase in α, while the 1:2 VPDMS:DVB polymer had the slowest. At these pyrolysis
temperatures, the DRIFTS results showed a significant decrease in the aromatic, methyl, and
ethylene groups in the pVPDMS polymer. In comparison, the 1:2 VPDMS:DVB polymer initially
had a significant loss of aromatic groups, followed by a low loss period between 500 °C and 600
°C, and finally an accelerated aromatic group loss after 600 °C. This indicates that the aromatic
groups stayed intact in the 1:2 VPDMS:DVB polymer at higher pyrolysis temperatures than for
all the other polymers, likely due to a higher degree of cross-linking. By a pyrolysis temperature
of 900 °C, α for all films is above 0.9 and it also increases slowly during the “soaking” period.
This observation is again consistent with the DRIFTS data that indicate that changes continue to
happen in this temperature range albeit at a slower pace. These results show that the polymer
63

pyrolysis process occurs in three distinct stages: Small changes for pyrolysis temperatures below
~350 °C, the main scission reactions taking place between 350 and 600 °C, with the final reactions
occurring after 600 °C.
To more clearly visualize the changes that take place during pyrolysis, we plot in Figure
3-12 the derivative of weight, mt, change with respect to time as a function of time. For all four
polymers, there are two primary peaks observed in the derivative of weight change graph: A peak
taking place at ~18,000 sec, just after the beginning of the “soak” period at 400
o
C, and another at
22,500 sec after the beginning of the “soak” period at 500 °C. Based on the DRIFTS data, the first
peak corresponds to the scission of the aromatic rings from the polymer network structure, while
the second peak corresponds to the loss of the methyl and ethylene groups after the aforementioned
Kumada rearrangement of the silicon-carbon bridges. So, the derivative of weight change allows
us to confirm that the primary scission reactions for all of the polymers occur between 400 °C and
600 °C. There are smaller size peaks at later times during the pyrolysis process, which correspond
to more subtle changes in the material’s structure as it continues to evolve into a Si-C type of
material.

64

Figure 3-12. Weight change loss during polymer pyrolysis under TGA experiments for polymer
films a) pVPDMS b) 1:1 VPDMS:DVB p(VPDMS-co-DVB) c) 2:1 VPDMS:DVB p(VPDMS-
co-DVB) and d) 1:2 VPDMS:DVB p(VPDMS-co-DVB).

Ceramic Yield
To determine the ceramic yield of the samples, the polymers were deposited on porous SiC
tablets and subsequently pyrolyzed. The fabrication technique for these tablets can be found
elsewhere.
64
The mass of these tablets was measured before polymer deposition, after deposition,
and after pyrolysis. The equation for calculating the ceramic yield is shown below, and the
measured ceramic yields for the four polymers are shown in Table 3-4.

𝐶𝑒𝑟𝑎𝑚𝑖𝑐 𝑦𝑖𝑒𝑙𝑑 =
𝑊𝑒𝑖𝑔 ℎ𝑡 𝑜𝑓 𝐶𝑒𝑟𝑎𝑚𝑖𝑐 𝐺𝑒𝑛𝑒𝑟𝑎𝑡𝑒𝑑 𝑃𝑜𝑙𝑦𝑚𝑒𝑟 𝐷𝑒𝑝𝑜𝑠𝑖𝑡𝑒 𝑑 =
𝑃𝑜𝑠𝑡 −𝑃𝑦𝑟𝑜𝑙𝑦𝑠𝑖𝑠 𝑊𝑒𝑖𝑔 ℎ𝑡 −𝑃𝑟𝑒 −𝐷𝑒𝑝𝑜𝑠𝑖𝑡𝑖𝑜𝑛 𝑊𝑒𝑖𝑔 ℎ𝑡 𝑃𝑜𝑠𝑡 −𝐷𝑒𝑝𝑜𝑠𝑖𝑡𝑖𝑜𝑛 𝑊𝑒𝑖𝑔 ℎ𝑡 −𝑃𝑟𝑒 −𝐷𝑒𝑝𝑜𝑠𝑖𝑡𝑖𝑜𝑛 𝑊𝑒𝑖𝑔 ℎ𝑡 (2)

65

It can be seen from Table 3-4, that the ceramic yield for pVPDMS was the lowest among
the four polymers studied, less than 10%. Likewise, it can be seen that the 1:1 and 1:2
VPDMS:DVB copolymers had relatively low yield as well, with yields of approximately 13%.
The polymer with the highest ceramic yield post-pyrolysis was the 2:1 VPDMS:DVB copolymer,
with a yield of above 21%. This then gives credence to the hypothesis that cross-linking the
polymer prior to its pyrolysis does, indeed, help with retention of the polymer mass during
pyrolysis and the generation of a greater quantity of the inorganic film (compare the pVPDMS
ceramic yield with that of 2:1 VPDMS:DVB copolymer). There is, however, an optimal DVB
content in the original polymer, with too high of a DVB to VPDMS ratio resulting in a lower
ceramic yield. This can be attributed to the pDVB polymer chains disintegrating during the earlier
stages of pyrolysis.

Table 3-4. Ceramic yield of various polymer films post-pyrolysis
Sample Ceramic Yield (%)
pVPDMS 9.09
2:1 VPDMS:DVB 21.74
1:1 VPDMS:DVB 12.50
1:2 VPDMS:DVB 13.04

Film Thickness Change
Profilometry was used to analyze the changes in the thicknesses of the thin films as a
function of pyrolysis temperature. Copolymer films with thicknesses of 10 µm, as measured by
the interferometer, were deposited on Si wafers with electric tape placed on one side of the wafer
in order to create an edge for the profilometry measurements. Before pyrolysis, the polymer film
66

thicknesses were measured by a stylus profilometer using the edge of the polymer film. The films
were then heated in a tube furnace under flowing Ar to set temperatures at a heating rate of 2
°C/min, and they had a “soak” period of 2 hr at the desired set temperature. After the “soaking”
period, the wafers were cooled to room temperature at a cooling rate of 3 °C/min before they were
removed from the tube furnace. The film thickness of the pyrolyzed samples was then measured
using profilometry. The equation for calculating the thickness change is shown below, and the
thickness change for each polymer film as a function of temperature is shown in Figure 3-13.

𝑇 ℎ𝑖𝑐𝑘𝑛𝑒𝑠𝑠 𝑐 ℎ𝑎𝑛𝑔𝑒 =
𝑇 ℎ𝑖𝑐𝑘𝑛𝑒𝑠𝑠 𝑜𝑓 𝑝𝑜𝑙𝑦𝑚𝑒𝑟 𝑝𝑜𝑠𝑡 −𝑑𝑒𝑝𝑜𝑠𝑖𝑡𝑖𝑜𝑛 𝑇 ℎ𝑖𝑐𝑘𝑛𝑒𝑠𝑠 𝑜𝑓 𝑝𝑜𝑙𝑦𝑚𝑒𝑟 𝑝𝑟𝑒 −𝑑𝑒𝑝𝑜𝑠𝑖𝑡𝑖𝑜𝑛 𝑥 100 (3)

There were no significant changes in the polymer film thickness when heated from room
temperature to 200 °C. However, by a pyrolysis temperature of 300 °C, the thicknesses of the
pVPDMS and 2:1 VPDMS:DVB copolymer films had decreased significantly, with thicknesses
being equal to ~71% and ~83% of the original bulk polymer thickness, respectively. In
comparison, the 1:1 and 1:2 VPDMS:DVB copolymer films showed significantly less change in
thicknesses, with both retaining over 90% of their original thickness. This can be attributed to the
fact that higher incorporation of DVB overall leads to higher cross-linking, which helps with
polymer mass retention at lower pyrolysis temperatures until the scissioning of the DVB moieties
from the polymer commences. The most significant change in the thickness of all pyrolyzed films
occurred between pyrolysis temperatures of 300 °C and 400 °C. All film thicknesses dropped to
less than 40% of their initial bulk polymer film thicknesses, with the largest being the thickness of
the 1:2 VPDMS:DVB copolymer at ~35% of its original value and the smallest being the one of
the 2:1 VPDMS:DVB copolymer film at ~21%. After a pyrolysis temperature of 400 °C, all of the
67

films showed continued loss in thickness until a pyrolysis temperature of 900 °C, with the
pVPDMS film showing the highest thickness retention between 500 °C and 900 °C, and the 2:1
VPDMS:DVB copolymer showing the least thickness retention in the same range. By a pyrolysis
temperature of 900 °C, the thickness of all the polymer films had reduced to between 8 and 10 %
of the original bulk polymer thickness, with the pVPDMS showing the highest retention at 11.75%
and the 1:1 and 2:1 VPDMS:DVB copolymers showing the least retention at ~9% of their original
bulk film thickness.

68

Figure 3-13: Thickness change of the various polymer films as a function of temperature for a)
the pVPDMS and the 2:1 VPDMS:DVB copolymer; b) the 1:1 VPDMS:DVB and the 1:2
VPDMS:DVB copolymers.

69

3.5 Conclusion

In this project, we studied the deposition of pre-ceramic polymers on inorganic powder
substrates and their subsequent pyrolysis to form ceramic films. Specifically, we systematically
studied the deposition of the copolymer (p(VPDMS-co-DVB)) via PECVD at various reactor
monomer feed compositions on two different powder substrates (KBr and BaF2) and their
subsequent pyrolysis for temperatures up to 1000
o
C.
We have selected in our studies a monomer (VPDMS) and a crosslinking agent (DVB) for
preparing the pre-ceramic polymer materials with no oxygen in their structure, which makes them
appropriate to use to fabricate Si-C type films upon pyrolysis. However, some minor quantity of
oxygen gets incorporated into the structure of the polymer film either during the preparation stage
or, most likely, during the sample transfer from the PECVD chamber into the pyrolysis furnace
(current efforts in the group focus on the design and construction of a PECVD system that
encompasses in situ pyrolysis, thus avoiding sample exposure to atmospheric conditions).
Analysis of the structural changes of the polymer as it undergoes pyrolysis indicates that at least
some of that initial oxygen is retained in the structure to eventually form a SiOxCy ceramic.
The DRIFTS technique was used to study in situ the mechanism, to track the fate
of the various polymer functional groups, and to identify the various structural rearrangements that
take place during the pyrolysis process. The TGA method was used to study the mass loss of the
film during pyrolysis. Combining the TGA and DRIFTS analysis results helped to shed additional
light into the mechanism by which the original polymer precursor converts into the final ceramic
via pyrolysis. We also performed ex-situ profilometry and ceramic yield measurements of the films
in order to determine the change in film properties post-pyrolysis. We found that the film
70

thicknesses decreased by up to 90% at 900 °C, confirming the results of the DRIFTS and TGA
experiments.

3.6 Acknowledgements

The authors acknowledge the support of the National Science Foundation (Award# CMMI-
2012196). EDS data were acquired at the Core Center of Excellence in Nano Imaging at the
University of Southern California.

71

4 Chapter Four: Time-Resolved Operando Analysis of the Pyrolysis of
a PECVD-Deposited Siloxane Polymer Using a Combined DRIFTS-
MS System

4.1 Abstract

Silicon-type thin films, made of silica, silicon carbide (SiC), or oxycarbide, find use as
membranes and electronic sensors, and in semiconductor and solar energy applications.
Previously, we studied
100
the preparation of nanoporous silica membranes via deposition of
poly(1,3,5,7-tetravinyl-1,3,5,7-tetramethylcyclotetrasiloxane) (pV4D4) films onto SiC
macroporous substrates via initiated chemical vapor deposition (iCVD) and their subsequent
controlled-atmosphere pyrolysis. Here, we utilize a different method, plasma-enhanced chemical
vapor deposition (PECVD), to deposit thin pV4D4 films onto a variety of substrates at significantly
higher deposition rates than iCVD and employ a number of experimental techniques to
comprehensively investigate the mechanism of conversion of these films into silica ceramics via
controlled-atmosphere pyrolysis. The aim of these studies is to better understand the impact of
preparation conditions on the structure and properties of the resulting ceramic films. The
experiments are coupled with complementary molecular simulations of the pyrolysis process that
employ a reactive force field (ReaxFF). This has allowed better understanding at the molecular
level of the processes that take place during the conversion, via pyrolysis, of the pV4D4 polymer
into a silica ceramic.

72

4.2 Introduction

Silicon-based thin films, such as silica and silicon carbide (SiC), and oxycarbide, are
useful in a variety of applications due to their versatile physicochemical properties, chemical and
thermal stability, and structural stability.
7,48,120
These organosilicon thin films have been used as
di-electric materials
121,122,123
, anti-reflective coatings
124,125
, anti-fouling layers
126,127
, and in
molecular separations
128
, amongst other applications. There are several different techniques for
depositing such films on substrates, including spin-coating
129
, dip-coating
17,18
, and chemical
vapor deposition (CVD)
21,22
. CVD, in particular, allows the application of these films with
conformal coverage on complex substrate surfaces
130
and without the use of toxic
solvents
131,132,133
.
Among the different CVD methods, plasma-enhanced chemical vapor deposition
(PECVD) is an appealing choice for a variety of applications. In PECVD, a power source (AC or
DC) is used to generate free radicals in a plasma atmosphere, which then react with precursor gas
molecules to grow a film on a substrate surface.
134
Compared to conventional CVD, PECVD
utilizes lower deposition temperatures and offers relatively high deposition rates.
135
Previously,
we deposited via initiated chemical vapor deposition (iCVD) a siloxane-type polymer, poly(tetra-
vinyl-tetra-methyl cyclotetrasiloxane) (pV4D4), onto a SiC porous support and pyrolyzed it in an
inert atmosphere to prepare a silica thin membrane film.
100
pV4D4 is a versatile polymer that
finds various applications, including in the biomedical device
136
and dielectric materials
137
fields.
In this paper, we use, instead, low-energy PECVD to deposit the pV4D4 polymer onto a variety
of substrates, including silicon (Si) wafers and barium fluoride (BaF2) powder, and study its
pyrolysis to produce silica films.
73

We analyze the changes in chemical structure of the pV4D4 film during pyrolysis
employing a time-resolved, operando combined Diffuse Reflectance Infrared Fourier Transform
Spectroscopy (DRIFTS) and Mass Spectrometry (MS) system. We combine the DRIFTS-MS
investigations with parallel thermogravimetric analysis (TGA) studies of the pyrolysis process,
together with Energy-dispersive X-Ray Spectroscopy (EDS) analysis of the resulting pyrolyzed
materials. Furthermore, we complement the experimental studies with molecular simulations of
the polymer pyrolysis process to help gain better fundamental insight into the phenomena that
take place during the conversion of the polymer into a ceramic.
DRIFTS is a powerful technique that allows one to study in situ changes in the bonding
environment of the polymer structure as it is being pyrolyzed.
138
Combining the DRIFTS system
with a MS analyzer (also known as residual gas analyzer (RGA)) allows one to analyze the various
stable gas species emitted from the polymer as it is being pyrolyzed, while simultaneously
monitoring changes in its solid-state chemical composition and structure.
139
Operando DRIFTS-
MS systems have been used to study catalytic reactions. Wang et al., for example, studied the
kinetics of the CO2 methanation reaction over Ru/Al2O3 catalysts using such a system.
140
Ochoa
et al. used the technique to study the oxidation of ethanol over mixed oxide catalysts to help
elucidate the different reaction pathways.
141
In the polymer pyrolysis technical area, White
142

used a combined DRIFTS-MS system to study the thermal decomposition of polystyrene, finding
that it decomposed shortly after 500 °C, but we know of no prior study of the pyrolysis of
siloxane-type polymers to form inorganic materials. pV4D4 is an ideal model system for such a
study, as it has a highly crosslinked siloxane structure and studying its pyrolysis could yield
interesting insight into siloxane chemistry. Increasing the cross-linking in the precursor polymer
has been shown to increase the ceramic yield during film pyrolysis.
69

74

In this study, we combine the experimental studies of pV4D4 pyrolysis with a
complementary computational analysis with the aim to gain information, at the atomistic level,
that is difficult to acquire experimentally under real-world conditions. In our efforts, we utilize a
reactive molecular dynamics (RMD) approach that allows large-scale simulations of chemical
events, like polymer pyrolysis, at a fraction of the computational cost of QM methods.
143
We
employ ReaxFF, a popular reactive forcefield that is compatible with several different software
packages, including LAAMPS. In ReaxFF, interatomic distance is used to describe bond order
formalism, which can then be used to predict the breaking of bonds that are present and the
formation of new ones
37
, thus helping bridge the gap between QC methods and empirical force-
fields by simplifying the former. Our group has previously used ReaxFF to study the pyrolysis of
allyl-hydridopolycarbosilane (AHPCS) into SiC.
33
Chenoweth et al.
38
modeled the pyrolysis of
polydimethylsiloxane (PDMS) and found that at lower pyrolysis temperatures (up to 900 K), it
depolymerized and formed cyclosiloxane oligomers. Similarly, Chen et al.
144
studied the
decomposition of hexamethyldisiloxane into smaller chemical moieties, finding that the scission
of Si-C bonds preceded the defragmentation of C-H and Si-O bonds, and that the main products
of this pyrolysis were methane (CH4) and linear siloxanes.
In summary, this paper introduces a methodology that can elucidate the reaction
mechanism of polymer pyrolysis by combining observations of the solid-state bond changes in the
polymer, evolved gas phase species analysis, and RMD molecular simulations. Having better
fundamental understanding of the molecular events that take place during pre-ceramic polymer
pyrolysis can help reveal how parameters such as monomer selection and the experimental
conditions, including the heating rate and temperature of pyrolysis, affect the final ceramic
structure and properties. This, in turn, can help experimentalists optimize monomer selection and
75

experimental pyrolysis protocols programs and, in the long-term, reduce energy costs during scale-
up operations for inorganic materials preparations.

4.3 Experimental

Deposition Method
The PECVD deposition was performed inside a custom-made, pancake-shaped vacuum
reactor (48 mm in height, 250 mm in diameter, GVD Corporation) with a thermally-cooled stage.
The 1,3,5,7-tetravinyl-1,3,5,7-tetramethylcyclotetrasiloxane (V4D4) (Gelest, Inc., used as
received without further purification) monomer was loaded into a stainless-steel jar that was held
under vacuum and heated to 50 °C. The line connecting the monomer jar to the reactor was heated
to 65 °C in order to prevent any adsorption or condensation of the monomer in the line. The
substrates were placed on the stage of the reactor and kept at a temperature of 50 °C via a
recirculating chiller (Thermo Scientific Haake A25). The V4D4 flow rate was kept constant at 1.2
sccm using a needle valve, and the nitrogen (N2) gas flow rate into the reactor was kept constant
at 40 sccm via a mass flow controller (MFC, MKS Type 1479A). A rotary vane vacuum pump
(Edwards E2M40) was employed to keep the reactor under vacuum condition, at 185 mTorr, via a
throttle-valve (MKS 153D) by active feedback control from a capacitance manometer (MKS
622C01TDE Baratron). An external radio-frequency (RF) plasma generator (Diener, 13.56 MHz,
100 W) was connected to a nichrome filament array (Omega Engineering, 80%/20% Ni/Cr) to
ignite the plasma in order to deposit the polymer film. The power supplied via the plasma generator
was kept constant at 25 W in order to induce a deposition rate of ~100 nm/min. Si wafers (Wafer
World, 100 mm) were thoroughly cleaned with solvents and dried in air before being placed onto
76

the reactor stage. In order to monitor the thickness of the pV4D4 film in-situ, a helium-neon laser
interferometer (Industrial Fiber Optics) was used on a Si wafer. For the DRIFTS and TGA
experiments, PECVD was used to deposit 10 𝜇 m of pV4D4 onto BaF2 (Alfa Aesar, 40 mesh)
powder. Prior to the deposition, the BaF2 powder was dried in an oven at 100 °C for 1 h in order
to remove any remaining moisture, and then it was evenly distributed on a Si wafer using a clean
razor blade.

Ex-situ FTIR Analysis
pV4D4-coated Si wafers (see Sect. 2.1) were placed in the center of a tube-furnace
(Lindberg/Blue, Model STF55433C). The furnace was then purged with ultra-high purity (UHP)
argon (Ar) at 300 sccm for approximately 1 h, and the samples were then heated, at a heating rate
of 3 °C/min, to the desired pyrolysis temperature where they were held for a 2 h hour “soak” period
before being cooled down to ambient room temperature at a rate of 3 °C/min. Transmission IR
spectra of both the as deposited as well as the pyrolyzed films on the wafers were collected using
Fourier transform infrared (FTIR) spectroscopy (Nicolet iS10, Thermo Scientific), with a clean
bare Si wafer used as the background. To collect the monomer spectra, a drop of V4D4 was placed
in between two Si wafers and the spectra were collected, with two Si wafers being used as the
background. The SpectraGryph-1.2 software was utilized to analyze the FTIR (and also the
DRIFTS, see Sect. 2.3) spectra.

DRIFTS-RGA
The changes in the bonding environment of the polymer film were studied in-situ using
DRIFTS (COLLECTOR II, Thermo Scientific). The purge line for the DRIFTS cell was connected
77

to a mass analyzer (RGA200, Stanford Research Systems) in order to analyze the composition of
gas components inside the DRIFTS chamber. The pressure of the vent line and the amount of gas
entering the RGA were controlled using a series of needle valves. The apparatus is shown in Figure
4-1. The polymer-coated powder was placed inside the DRIFTS sample cup and was leveled using
a razor blade, as recommended by the instrument manual. Ar gas with a constant flow rate of 10
sccm was then introduced into the cell to purge the chamber for approximately 1 h to remove any
residual air remaining in the system. Prior to taking measurements, the RGA was allowed to reach
a stable baseline. The sample in the DRIFTS cell was heated at a rate of 20 °C/min to 900 °C, and
the IR absorbance spectrum was recorded at every 100 °C interval. During the experiment, the
pressure of the purge line from the DRIFTS cell was kept at a minimum of 0.1 psig in order to
induce a positive pressure and to ensure that ambient air did not enter through the vent lines.

Figure 4-1. Schematic of the combined DRIFTS-RGA apparatus

Thermogravimetric Analysis (TGA)
Thermogravimetric analysis (Cahn Instruments, 12130-01) was used to measure in-situ the
mass loss of the pV4D4 polymer during pyrolysis. For that, approximately 100 mg of the pV4D4-
78

coated BaF2 powder (prepared as described above in Section 2.1) was placed inside the TGA
sample cup, and high-purity Ar gas was then allowed to purge the system at a flow rate of 30 sccm
for 20 min. For the TGA test, the chamber was heated to 100 °C at a heating rate of 3 °C/min and
held at that temperature for 2 h in order to move any residual moisture. The sample was then heated
to 900 °C at a heating rate of 3 °C/min before being cooled down to room temperature at a rate 5
°C/min.

X-ray Energy Dispersive Spectroscopy (EDS)
A plasma focused ion beam (PFIB)/scanning electron microscopy (SEM) system (Thermo
Scientific, Helios G4 PFIB UXe DualBeam FIB/SEM) with an attached x-ray energy dispersive
spectroscopy (EDS) accessory was used to investigate the atomic composition of the polymer
films. Data were collected with a voltage of 5 kV and a 1.6 nA beam current. The working distance
of the EDS detector was set to 5.5 mm and 1,000,000 counts were collected for each sample.

ReaxFF Algorithm/Work-Flow
The PV4D4 system was created through open source IQMol Software and relaxed using
the Universal Forcefield (UFF) available within the IQMol package.
145
We randomly reacted the
vinyl bonds in a head-to-tail polymerization so that approximately 52% of the vinyl bonds were
reacted for the system. The cross-linked structure thus created was repeated periodically to yield a
system with approximately 2000 atoms. The polymer system was first initiated with 3 pV4D4
complexes of 15 monomers each, which each had the requisite vinyl reactivity. The system was
then compressed using NVT-MD to a density of 1.22 gm/cc. The compressed system thus obtained
was pyrolyzed using NVT-MD at a timestep of 0.1 fs per step at various temperatures from 300K
79

to 2500K. The heating rates for the pyrolysis process were 1 K/ps, 2 K/ps and 3 K/ps. We dumped
the files every 1 ps and observed the fragments formed at the given step using depth first search
analysis on the molecular graph formed using the nearest bonded neighbor cutoff criteria for
various atoms.

4.4 Results and Discussion

pV4D4 Deposition via PECVD and Ex-situ Pyrolysis as Studied by Transmission FTIR
We have found that PECVD deposits pV4D4 films at a much higher deposition rate than
what was previously attained with iCVD (~100 nm/min vs. ~10 nm/min) under comparable reactor
conditions. This has benefits in terms of shortening the deposition time, and thus the costs, for the
preparation of these films for potential commercial uses, but also makes it more convenient to
grow thicker polymeric films on substrates to test and experiment with techniques such as
DRIFTS, TGA, and RGA, thus improving analytical accuracy.
Table 4-1. IR bands shown in Figures 4.2-4.4
Regions IR Bands Corresponding to Bonds
i C-H stretching from polyethylene bonds, asymmetric CH 3 stretching
ii Modes related to vinyl bonds
iii Si-CH3 bending
iv Si-O-Si bonding environments
v Si-C stretching, rocking in Si-CH 3
vi Si-O-Si out of plane deformations

80

Figure 4-2 shows the transmission FTIR spectrum of the pV4D4 polymer deposited via
PECVD on a Si wafer. For comparison in the same Figure, we also show the FTIR spectra of the
V4D4 monomer. For easy reference, Table 4-1 shows the various bands and their assignments to
functional groups observed. There are several prominent bands which are present both in the
V4D4 monomer and the pV4D4 polymer. These include bands in the region between 1000 and
1100 cm
-1
, where one finds a broad band corresponding to the Si-O-Si bending (~1060 cm
-1
) and
the Si-O ring stretching (~1065 cm
-1
) bands (region iv). In addition, there are several bands
corresponding to the Si-C bond in both the V4D4 and pV4D4: Si-CH3 bending (~1250-1270 cm
-
1
) (region iii), Si-C asymmetric rocking (~790 cm
-1
) (region v), and Si-C rocking (~750 cm
-1
)
(region v).
146
There is also a sharp peak at 2960 cm
-1
which corresponds to asymmetric CH3
stretching in Si-CH3. There are several bands, however, in the monomer spectrum corresponding
to the vinyl bonds (peaks ii) that are not present in the polymer spectrum. They include the bands
at ~960 and ~1400 cm
-1
corresponding to the wagging and bending modes of CH2, and the band
at ~1600 cm
-1
corresponding to the C=C stretching in vinyl bonds.
67
In addition, there is a small
band at ~3060 cm-1 associated with C-H stretching in CH=C in vinyl groups.
147
These bands
corresponding to the vinyl bonds located at 960, 1400, 1600, and 3060 cm
-1
(peaks ii) are not
present in the polymer spectrum, and furthermore, there is a rise in the bands at 2870 and 2920
cm
-1
corresponding to the C-H stretching in aliphatic polyethylene backbone structures (region i)
formed during the polymerization of V4D4.
67
It is important to note that PECVD typically grows
polymer films by random radicalization of the precursor monomer, leading to random
recombination reactions and cross-linking between cleaved moieties.
148
This contrasts with
targeted deposition methods such as iCVD, which promotes film growth by radicalizing more
81

labile bonds such as vinyl groups. Nevertheless, it can be seen from the IR spectra that the use of
a lower plasma power promotes the chain growth through the vinyl bonds.

Figure 4-2. Spectra of the V4D4 monomer and the pV4D4 polymer deposited via PECVD on a
Si wafer
We have used FTIR, ex-situ, to characterize the structure of polymer films pyrolyzed at
different temperatures. As noted in Section 2.2, Si wafers, chosen due to their high IR
transmittance, were used as the substrates to deposit dense, 3 µm thick pV4D4 polymer films using
PECVD. The polymer-coated Si wafers were then pyrolyzed as described in Section 2.2. The IR
spectra of the pyrolyzed films were immediately collected using transmission FTIR. Figure 4-3
shows the FTIR spectra of the as deposited pV4D4 film as well as of the polymer films pyrolyzed
at different temperatures. Figure 4-3a shows the FTIR spectra extending from 400 to 3100 cm
-1
,
and Figure 4-3b and Figure 4-3c show the expanded views of the regions of the spectra
corresponding to the polyethylene and methyl bands, respectively.
82

FTIR analysis of the pV4D4 sample pyrolyzed at 300 °C shows no significant difference
in the peak intensity from the as-deposited pV4D4 sample, thus indicating very little change in its
chemical structure. By a pyrolysis temperature of 400 °C, the peaks at 2860 cm
−1
and 2920 cm
−1
,
corresponding to the C-H in -CH2-CH2- backbone bonds (region i), are slightly reduced, due to
cleavage of some of the polyethylene bonds. The methyl stretching bands (region iii), on the other
hand, remain intact at 400 °C with none of the related peaks in Figures 4-3b and 4-3c being
reduced. The band at 1065 cm
-1
(SiO2R2) still predominates the structure, but bands at 1028 cm
-1
(Si-O-Si) and 1120 cm
-1
(SiO3R, silsesquioxane cage structure) are also present.
100

Figure 4-3. a) Ex-situ IR absorbance spectra showing the transformation of the PECVD-deposited
pV4D4 polymer during pyrolysis; b) expanded view of the IR region from 2750 to 3100 cm
-1

showing the polyethylene stretching band; c) expanded view of the IR region from 1200 to 1300
cm
-1
showing the methyl stretching band.

83

This is probably due to the fact that the pyrolysis of pV4D4 was carried out in an inert
environment, with fewer oxygen atoms being available to enable the network rings to form
silsesquioxane cages. At a pyrolysis temperature of 500 °C, the ethylene peaks at 2860 and 2920
cm
-1
(region i) are significantly reduced, but the neighbouring methyl peak is only slightly reduced.
Furthermore, the Si-CH3 stretching band transitions from a single peak at 1260 cm
-1
into a peak
doublet at 1262 cm
-1
and 1275 cm
-1
. We have previously shown that the splitting of this peak
relates to the formation of some silsesquioxane (SiO3R) groups.
100
The spectra of the pV4D4
sample pyrolyzed at 600 °C shows a broad band between 950 cm
-1
and 1250 cm
-1
(region iv),
indicating the presence of a mixture of Si-O-Si, silsesquioxane, and Si-O-C bonds.
149
There is also
a peak between 400 cm
-1
and 500 cm
-1
(region v)

that represents the Si-O-Si out-of-plane
deformation typically found in silica structures.
150
By this temperature, the polyethylene bands at
2860 and 2920 cm
-1
(region i) have disappeared; the methyl band (region iii)

has also been
significantly reduced, and has also shifted to a wavelength of 1275 cm
-1
, indicating that the
pyrolyzed polymer has more of a silsesquioxane rather than a network-type structure. At a
pyrolysis temperature of 700 °C, the methyl peak has completely disappeared, but the broad band
at 1150 cm
-1
(region iv), likely indicating the presence of Si-O-C in the structure, still remains. At
800 °C, the Si-O-Si band (region iv)

continued to change into a silica structure, as indicated by a
peak shift to 1065 cm
-1
, whereas the silsesquioxane and Si-O-C shoulders were significantly
reduced. There is no significant change happening in the structure when going from 800 °C to 900
°C, indicating that the polymer has largely finished pyrolyzing by 800 °C.

84

pV4D4 Pyrolysis Studied with the Integrated DRIFTS-RGA system
For these experiments, pV4D4 films with a thickness of 10 µm were deposited on BaF2
powder via PECVD and the resulting polymer-coated powders were then placed inside the cell
compartment of the DRIFTS instrument which was connected with a MS analysis system (RGA).
The pV4D4 polymer films were then pyrolyzed in situ by raising the sample temperature (up to
900
o
C) in a linear fashion (10 °C/min) with the DRIFTS instrument used to monitor the IR spectra
of the films during pyrolysis and the RGA employed for monitoring various gaseous species
evolved.
85

Figure 4-4. DRIFTS data showing the IR absorbance spectra for pV4D4 films at different
pyrolysis temperatures
86

Figure 4-5. Functional group peak area change as a function of pyrolysis temperature

Figure 4-4 shows the DRIFTS-IR spectra for the pV4D4 polymer as it undergoes pyrolysis
at various temperatures (see Table 4-1 for the functional groups corresponding to the various major
IR bands) and Figure 4-5 shows the change in the DRIFTS spectra peak area for the polyethylene
and methyl bonds. The DRIFTS spectrum of the as-deposited polymer has, as expected, bands that
are similar with those found in the FTIR spectrum of the pV4D4 polymer grown on the Si wafer,
see Figure 4-3: A broad band between 950 cm
-1
and 1100 cm
-1
that corresponds to the asymmetric
Si-O-Si stretching (region iv),
151,152,153
, a sharp, prominent band at approximately 1250 cm
-1
to
1270 cm
-1
that corresponds to the Si-CH3 band (region iii), C-H stretching bands between 2800
and 3000 cm
-1
corresponding to the polyethylene backbone formed during the polymerization and
the CH3 group (region i), and two small bands in the region between 700 cm
-1
and 800 cm
-1

corresponding to the Si-C bending and rocking modes (region v).
87

As the pyrolysis temperature increased, we found that there were only minimal changes in
the IR spectra until a temperature of 400 °C was reached, which is similar to the behavior of the
pV4D4 on Si wafers samples pyrolyzed in a tube-furnace (Figure 4-3). Most notably, the areas the
polyethylene backbone (region i) and methyl group (region iii) bands did not noticeably decrease,
see Figure 4-5, which indicates that the pV4D4 chemical structure stayed largely intact up to that
pyrolysis temperature; this observation is in line with the discoveries of other researchers
investigating polysiloxane polymer pyrolysis using similar techniques.
59
By 500 °C, there was a
small reduction in the polyethylene stretching (region i) and the methyl (region iii) peaks. As the
pyrolysis temperature increased past 500 °C, as Figure 4-5 shows, there was a significant drop in
the polyethylene peak area, but this was not the case with the methyl peak. There was, however, a
decrease in the methyl band area beginning around 600 °C (Figure 4-5), which agrees with the
observation by Narisawa
70
, who found that polysiloxanes had their methyl groups removed at ~600
°C. Both the methyl and the polyethylene peak areas have decreased appreciably by a temperature
of 700 °C, and the area of the Si-C bending and rocking peaks (region v) has been largely reduced
by this temperature as well. By a pyrolysis temperature of 800 °C, most of the peaks corresponding
to the hydrocarbon bonds (regions i, iii, and v) have largely disappeared from the spectra, with the
peaks corresponding to Si-O-Si, at around 1050 cm
-1
and 600 cm
-1
being the only peak remaining.
These are the characteristic IR peaks for amorphous silica.
150
Increasing the pyrolysis temperature
to 900 °C did not significantly change the DRIFTS spectra, as the majority of the functional groups
have already been cleaved by then. The DRIFTS results with the pV4D4-coated BaF2 powders are
very similar to the observations made with the polymer samples coated on Si wafers and pyrolyzed
in a tube-furnace. This shows that the substrate on which these polymer films are deposited has
88

little impact on their structure, and also that the pyrolysis conditions in the DRIFTS chamber
resemble those in a conventional tube-furnace.

Figure 4-6. Partial pressures of gaseous species as detected by RGA

89

During pV4D4 pyrolysis, the RGA instrument attached to the DRIFTS chamber detected
several gas phase species, including hydrogen (H2), methane (CH4), ethylene (C2H4), ethane
(C2H6), propene (C3H6), and propane (C3H8), which were emitted from the polymer being
pyrolyzed. The increases in the gas phase species concentration detected by the RGA are shown
in Figure 4-6 in terms of the ratio of the partial pressure of the species P measured by the MS
detector divided by the partial pressure P0 of the same species detected prior to the start of the
pyrolysis. It can be seen from the results presented in Figure 4-6 that gas species evolution
primarily took place in two temperature ranges, namely between 400 °C and 500 °C and in between
600 °C and 700 °C. In the first stage of gas species evolution, we observed a large increase in the
partial pressure for C2H4, C2H6, and C3H8. The partial pressures for C2H6 and C3H8 increase
sharply, peaking at ~500 °C, and they quickly diminish by the time the pyrolysis temperature has
reached 600 °C. Plawsky et. al
154
reported that, during the pyrolysis of an organo-silicate polymer,
C2H6 is thermodynamically favored to form at temperatures lower than 600 °C, and it is likely
generated from free CH3 radicals recombining. This first stage of gas phase species evolution is
associated with the scissioning and loss of the polyethylene polymer backbone due to pyrolysis,
as can also be seen in the DRIFTS and ex-situ FTIR spectra. In previous models of polysiloxane
polymer pyrolysis
154
, it has been seen that the scissioning of -CH2-CH2- in the polymer backbone
leads to the formation of C2H4. Wilson et. al
59,155
also found via MS analysis that during the
pyrolysis of polysiloxane polymers, C2H4 species were detected at temperatures between 400 °C
and 650 °C. In our experiments (Figure 4-6) in comparison to C2H6 and C3H8, the increase in C2H4
gas phase concentration is more gradual with C2H4 being emitted in the gas phase starting with a
temperature of 200 °C and its concentration remaining relatively constant throughout the whole
pyrolysis experiment. This is indicative that ethylene is continually being formed not only due to
90

reformation of the -CH2-CH2- groups but also by the removal and recombination of methyl groups
followed by atomic hydrogen abstraction.
The second stage involves the evolution of H2 and CH4 (in addition to the constant
evolution of C2H4 that continues from the first stage) The increase in partial pressure for CH4
commenced at a temperature of ~425 °C and peaks at a time when the temperature is around 650-
700 °C. Such CH4 evolution behavior has been reported in other MS studies of polysiloxane
polymer pyrolysis
154,156
and has been attributed to the homolytic cleavage of Si-CH3 bonds at
elevated temperatures. This is also consistent with the DRIFTS data presented in Figures 4-4 and
4-5 that indicate the loss of methyl groups in the same range of temperatures.
The increase in H2 partial pressure commences at a temperature of ~ 500
o
C and peaks at a
temperature of ~750
o
C. During their study of the pyrolysis of polysiloxane gels, Bahloul-Hourlier
et. al
156
found that increase in the rate of evolution of CH4 and H2 species at these temperatures
was attributed to the formation Si-CH2-Si and Si-CH2-CH2-Si hydrocarbon bridges and their
subsequent decomposition. The delayed H2 evolution is in contrast with other lower temperature
H2 evolution observed with other Si-type materials
156
, and can be attributed to the lack of Si-H
bonds in the pV4D4 polymer. The higher temperature H2 evolution, which peaks around 750 °C
in Figure 4-6, was previously reported
59,154,155
and was attributed to the scission of C-H bonds
which are, typically, the second strongest bonds in siloxane-type polymers after the Si-O bonds,
and are known to be cleaved only at temperatures above 600 °C.
156
The cleavage and
recombination of methyl radicals, followed by atomic hydrogen abstraction, is likely responsible
for the secondary peak observed during ethylene evolution, see Figure 4-6.
91

While the formation of larger molecules such as C4H10 and C6H6 are thermodynamically
favored during pyrolysis, they were not detected by the RGA in our study. We hypothesize this to
be due to the slow diffusion of -CH2-CH2- groups in the solid state that kinetically limits the ability
of the cleaved polyethylene fragments to recombine into larger MW alkane or cyclic molecules.

TGA Study of pV4D4 Pyrolysis
TGA was also used to measure the mass loss of the pV4D4 polymer during pyrolysis as a
function of time (and corresponding temperature). The mass loss data are complementary to the
FTIR and RGA results in the effort to better delineate the mechanism of pV4D4 pyrolysis. For the
TGA experiments, as with the DRIFTS/RGA experiments, we utilized BaF2 powders coated with
5 µm of pV4D4. Prior to using the polymer-coated powder for the TGA experiments, it was dried
at 100 °C in a tube-furnace for 1 h under flowing Ar to remove any residual moisture. The
experimental heating protocol for the TGA experiments consisted of raising the sample
temperature in a linear fashion (3 °C/min) to 100 °C and letting the sample stand (soak) at this
temperature for 1 h to remove potential volatile contaminants, including moisture. Subsequently,
the sample temperature was raised linearly to 900 °C. We carried out three TGA runs at different
heating rates of 5, 10, and 15 K/min in order to determine the effect of heating rate, if any, on the
observed behavior.
The mass loss behavior of the pV4D4 polymer can be quantified via the fraction of
decomposition (α) defined by Equation 4 below, where mi, mt, and mf are the initial mass, mass at
time t, and final mass of the sample, respectively.

92

(4)

Figure 4-7. Fraction of decomposition versus temperature for three different heating rates

Figure 4-7 shows the change in α as a function of temperature for the three different heating
rates. Before a temperature of 400 °C, there was a gradual increase of α (with no difference
observed between the three heating rates) up to a value of 0.15. This is, likely, due to evolution of
small amounts of Si-H, and small amounts of ethylene groups being cleaved off, as also indicated
by the RGA results. For temperatures between 400 °C and 700 °C, α increased from 0.15 to 0.9,
indicating that this is range where most of the polymer decomposition occurred. This temperature
range also corresponds to the temperature range for which the DRIFTS and RGA results indicated
that the majority of the polyethylene and methyl groups were evolved from the polymer structure.
In the temperature range between ~385 °C and ~600 °C, α increased in approximately linear
𝛼 =
𝑚 𝑖 − 𝑚 𝑡 𝑚 𝑖 − 𝑚 𝑓
93

fashion for all 3 heating rates, from ~0.15 to ~0.7. This is the temperature range in which the
DRIFTS results indicated that most of the decomposition of the polyethylene backbone took place.
Past 600 °C and up to a temperature of 765 °C, α continued to increase linearly albeit at a slower
rate from ~0.7 to ~0.97, the increase corresponding to the loss of methyl groups as determined by
the DRIFTS/RGA results. Past a temperature of 800 °C, α continues to increase but very slightly,
indicating that by this temperature, the polymer had largely finished pyrolyzing.

Figure 4-8. Derivative of weight change during pV4D4 film pyrolysis
To better understand the changes that are occurring in the polymer structure, we calculated
the derivative of weight change versus temperature for the 3 K/min heating rate TGA test. The
results in Figure 4-8 indicate the presence of three primary peaks. The first peak, centered around
200 °C, corresponds to very minor scissoning of the polyethylene backbone and methyl groups, as
seen in Figure 4-5. The largest peak, centered around 500 °C, corresponds to the scission of the -
94

CH2-CH2- bands from the polyethylene backbone, as also seen in the DRIFTS and ex-situ FTIR
results. These reactions contributed to a large mass loss due to the evolution of C2/C3 species such
as C2H4, C2H6, and C3H8, as shown by the RGA data. The second peak, centered around 700 °C,
corresponds to the loss of methyl functional groups from the polymer as well as hydrogen
abstraction from labile C-H bonds. The mass loss corresponding to this peak was smaller in
comparison to the larger peak due to the majority of the carbon functional groups from the
polyethylene having already been scissioned off the polymer. This small peak correlates to the
formation of CH4 and H2 species, as also seen by the RGA.

Atomic Composition Analysis of Materials from pV4D4 Pyrolysis
EDS was used to analyze the atomic composition of the pyrolyzed films. The films were
prepared as previously described in Section 4.1. We estimated the electron beam penetration depth
to be ~1 μm, so to ensure that there was minimal effect on the EDS results from the Si wafer
substrate, the thicknesses of all of the polymer films were measured using a stylus profilometer
(AMBiOS XP-2) to ensure they were at least 1.5 μm thick. Prior to the EDS tests, the as prepared
films were sputtered with a palladium/platinum coating to increase their conductivity. Table 4-2
shows the atomic % elemental composition (At%) of the pyrolyzed films from the EDS analysis.
Table 4-2. At% of the pV4D4 films pyrolyzed at various temperatures
Composition Deposited 300 °C 400 °C 500 °C 600 °C 700 °C 800 °C 900 °C
C 61.1 57.1 56.1 45.4 29 29.1 3 2.4
O 23.5 23.4 23.5 31.9 45.4 47.7 64.2 67.5
Si 15.4 19.6 20.3 22.7 25.6 23.2 32.8 30.1
95

For each film, we analyzed three different spots, each at a distance of at least 1 mm apart
from the others aiming to determine film uniformity. For each spot, three different measurements
were made, and their average values were then compared with those of the other two spots. Good
film uniformity was noted. The At% values reported in Table 4-2 represent the average among the
three spots. The results from the EDS experiments are consistent with the observations from the
DRIFTS-RGA part of the study (Section 4.2). There is little change in the pyrolyzed film’s atomic
composition until a major transformation begins to happen between 400 °C and 500 °C. At this
temperature, there is a ~10 % reduction in the carbon content of the pyrolyzed pV4D4 film.
Between a temperature of 500 and 600 °C, there is an additional decrease of ~16 %. These two
reductions in the carbon content of the film are, likely, due to the polyethylene -CH2-CH2-
functional groups being removed from the structure, as indicated also by the DRIFTS studies
(Figures 4-4 and 4-5). Between 600 and 800 °C, there is a substantial decrease in the carbon content
down to 3%. This is consistent with the DRIFTS/RGA results that indicate large shifts in the
chemical structure of the film involving the scission of methyl groups from the polymer, the rapid
evolution of H2 and CH4, and the transition of the polymer into a silica film. Further pyrolysis at a
temperature to 900 °C showed negligible changes in the atomic composition of the material,
signifying that the transition of the polymer into a silica ceramic was complete by a temperature
of 800 °C. Such a conclusion is also supported by the FTIR and DRIFTS observations.

4.6 Computational Studies
The ReaxFF reactive forcefield for pV4D4 was developed by further retuning the
parameters
38
from a previous parameterized forcefield for a PDMS polymer system. The ground-
truth data-set for this training effort contained the equation of states for bond and angle scans for
96

various configurations, as shown in Figures 4-13 and 4-14 in the Supplementary Materials
Section. Comparison of ReaxFF and DFT obtained bond and angle scans indicate that the ReaxFF
values are in good agreement with the calculated DFT values. DFT calculations were performed
using QCHEM code with B3LYP functional with 6-31G** basis set. We trained a total of 49
parameters for this data-set which comprised of two-body, off-diagonal and three-body terms. The
two-body C-C, C-Si, O-Si bond terms were reparametrized from the earlier parameterization along
with the Si-C and Si-O off-diagonal terms to obtain a good representation of bond-scans with
respect to the DFT obtained bond-scan. Further, retuning of the C-Si-C, C-Si-O and O-Si-O three-
body angle terms was performed to have an exact representation of the closed ring structure of
pV4D4 polymer.
Figure 4-9 shows the simulation results of the carbon atomic content in the pV4D4 polymer
clusters as a function of temperature, expressed as the % fraction of the original number of carbon
atoms that still remain in the polymer structure, ATC%, as it is being pyrolyzed. We carried three
different pyrolysis simulations that varied from each in the rate (1 K/ps, 2 K/ps, and 3 K/ps) by
which we heated up the polymer structure. The differences in behavior among the three simulations
is qualitatively similar to the experimental behavior of mass loss (see Figure 4-7) observed when
heating the polymer sample with different heating rates. For all of three simulations, the three
polymer chain structures stayed intact, and the decrease in ATC% was due to the loss of carbon
functional groups.
The experimental results discussed above show that as the pyrolysis temperature increased,
the carbon At% of the polymer significantly decreased, and the polymer transitioned into a Si-O-
C structure. The simulation also showed that as the polymer was heated, starting from a
temperature of 300 K the ATC% also decreased significantly. Interestingly, the number of silicon
97

and oxygen atoms in the polymer structure remained constant, indicating that neither element was
removed from the polymer structure during pyrolysis. This is consistent with our RGA studies, in
which we did not observe any Si- or O-containing gas phase species. It is, likely, due to the fact
that the Si-O bond dissociation energy is known to be significantly higher than the bond
dissociation energies for the other bonds found in this siloxane-type polymer.
157,158
Figure 4-10
shows the mass of the pyrolyzed polymer, expressed as 100X(m/m0), where m is the mass of the
polymer at any given time (and corresponding temperature of pyrolysis) and m0 its initial mass
prior to the start of pyrolysis. As expected, the behavior is similar to the one for ATC% in Figure
4-9.

Figure 4-9. Carbon content of the polymer strands as a function of pyrolysis temperature

98

Figure 4-10. Mass loss of the polymer strands as a function of pyrolysis temperature

In order to better understand the dynamic changes that take place during pyrolysis, in the
simulations we “tracked down” the number of Si-C bonds that remain in the simulation box. In
Figure 4-11, we plot the fraction of the original bonds that remain in the polymer structure as it is
being pyrolyzed at various temperatures. As the temperature increases past 700 K, there is a steady
decrease of the Si-C bonds in the system, which are known to be more labile than the tight cyclic
Si-O bonds.
35
Such a decrease is attributed to the disintegration of the polyethylene bands, and
also to scissioning of the methyl groups from the polymer.
In the simulations we also tracked down the number of stable gas phase species in the
simulation box, which are shown in Figure 4-12. The simulations indicate the formation of H2,
CH4, and C2, C3 and C4 hydrocarbon species, all of which (other than the C4 species) were also
observed experimentally, as noted previously. Further, the simulations do not detect the presence
of C5 or C6 hydrocarbon species, in line with the experimental observations. We attempt no
99

quantitative comparisons here between the simulations and the RGA compositional data, as the
experimental set-up is a flow system and the simulations employ a close system. Given such
differences, it is quite remarkable that even a good qualitative agreement is attained.

Figure 4-11. Temperature evolution of the Si-C bonds during the pyrolysis of pV4D4
100

101

Figure 4-12. Evolution of the compounds observed during the pyrolysis process for the a) 1
K/ps, b) 2 K/ps, and c) 3 K/ps heating rate simulations.

4.5 Conclusion

In this study, we developed a methodology to investigate the pyrolysis of a PECVD-
deposited pV4D4 film. In situ DRIFTS was used to analyze the changes in the solid-state bonding
environment, and time-resolved MS was employed to monitor the concentration of gas phase
species evolved as the pV4D4 polymer was being pyrolyzed. Combining these two in situ
techniques allowed us to establish that pV4D4 pyrolysis takes place in two distinct stages. The
first stage is associated with the cleavage of the more labile polyethylene backbone formed during
the polymerization and leads to the release of C2 and C3 species such as ethylene and propane. The
second stage is associated with the cleavage of the methyl groups, and with hydrogen abstraction
from C-H bonds, which then leads to the formation of CH4 and H2.
102

The DRIFTS and time-resolved MS techniques were also complemented with TGA
measurements that further helped to confirm the two-stage mechanism of pV4D4 pyrolysis. EDS
was used to provide an elemental analysis of the pyrolyzed polymer samples. The EDS results are
in line with the FTIR/DRIFTS measurements and confirm that polymer pyrolysis is complete by
a temperature of ~800 °C to form a silica material with a very minor amount of carbon remaining
in its structure. A computational model using ReaxFF was developed and utilized to provide
further fundamental insight into the molecular processes that lead to the formation of inorganic
silica films from siloxane-type, pre-ceramic polymer precursors. Good qualitative agreement
between simulations and experiments was found.

4.6 Acknowledgements

The authors acknowledge the support of the National Science Foundation (Award# CMMI-
2012196). The EDS data were acquired at the Core Center of Excellence in Nano-Imaging at the
University of Southern California.

103

4.7 Supplementary Materials Section

Figure 4-13. Comparison of the ReaxFF and DFT bond dissociation energies for the following
bonds: a) Si-O, b) Si-CH3, c) C-CH3, and d) Si-Vinyl.
104

Figure 4-14. Comparison of the ReaxFF and DFT bond angle distortion energies for the
following bonds: a) Si-O-Si, b) O-Si-O, c) C-Si-C, and d) O-Si-C.

105

5 Chapter Five: Conclusion

5.1 Closing Remarks

In this Thesis, we presented novel fabrication methods for producing ceramic films. In
Chapter 2, we showed that we were able to fabricate microporous silica films on macroporous
ceramic substrates for gas separation applications through a sequential process of iCVD and
subsequent pyrolysis.
100
Because the iCVD process uses a low energy input and mild reactor
conditions, the functionality of the monomer is retained in the resulting polymer and etching of
the film is avoided. Additionally, the iCVD process has the capability to be scaled-up in a
manufacturing process using the “roll-to-roll” process.
58

However, the iCVD method, due to its relatively low deposition rates observed and
incorporation of peroxide initiators, may not be appropriate for the preparation of pre-ceramic
polymers leading upon pyrolysis to SiC. We addressed this concern in Chapter 3, where we used
PECVD to deposit an organosilicon copolymer, and further studied its pyrolysis to form a ceramic.
Our PECVD technique uses a low-energy plasma power to ionize an inert gas and initiate a
polymerization reaction. The low energy employed for this PECVD deposition largely targeted the
vinyl bonds in the monomer to initiate the polymerization reaction but, nevertheless, resulted in a
high deposition rate.
In Chapter 4, we discussed the fabrication of silica and silicon oxycarbide films using
PECVD deposition of a siloxane polymer and subsequent pyrolysis. We used a combined
DRIFTS-MS system in order to analyze the solid-state bonding environment and evolved gas
species in-situ, allowing us to elucidate the mechanisms of the pyrolysis.

106

5.2 Future Research Directions

5.2.1 Preparation of SiC Membranes via In-situ PECVD Deposition and Pyrolysis

In Chapter 3, we demonstrated the ability to deposit an organosilicon polymer via PECVD
and subsequently pyrolyze it to form a SiC-type film. However, the polymer film still contained a
certain fraction of oxygen in its structure, and the resulting ceramic ended up being an oxycarbide
rather than a pure SiC film. It is, therefore, important to explore ways in the future to reduce oxygen
contamination both during the film deposition and the pyrolysis steps. Currently, the process is
sequential, calling for the transfer of the polymer-coated substrate from the iCVD and PECVD
reactor chambers to the tube furnace where the pyrolysis takes place. This transfer step exposes
the substrate to the ambient atmosphere, potentially resulting in the contamination of the sample.
Specifically, oxygen may adsorb on the polymer film surface and react with Si moieties during
pyrolysis, thus forming Si-O bonds. In order to conduct the pyrolysis step in situ, and thus avoid
exposing the polymer films to room air, we have built a system (as shown in Figure 5-1) that allows
for heating inside the reactor to temperatures up to 1000
o
C. In the future, the pyrolysis step will
be conducted in situ inside this iCVD/PECVD reactor, thus avoiding the transfer step and potential
film contamination. Carrying out the film deposition and pyrolysis steps in the same chamber will
allow us to directly correlate the properties of the final ceramic with the characteristics and
structure of the deposited precursor polymer film without undue concerns about potential film
modification.
The reactor already has a vacuum function, that allows for easy purging with inert gas and
subsequently pumping down to low pressures for conducting the pyrolysis step under vacuum
conditions. However, the current pumping system employs a mechanical pump attaining a modest
vacuum in the ~0.1 Torr range. To avoid potential oxygen contamination during the deposition
107

step, we plan in the future the use of a higher vacuum environment by employing a turbomolecular
pump in place of the current mechanical pump system.

Figure 5-1. Schematic of the modified iCVD reactor with in-situ heating apparatus

5.2.2 Testing of the Membranes in a Membrane Steam Reforming Reactor

An important application for the SiC films prepared in this research is to use them as
membranes in gas phase reactive separation applications.
3,4
In Chapter 2, we presented some of
our preliminary efforts on preparing Si-type membranes by depositing a dense polymer film on
the surface of a macroporous substrate via iCVD and subsequent pyrolysis. A future aim of this
project will be to deposit and pyrolyze these polymer films on porous cylindrical SiC ceramic
substrates using the in-situ PECVD technique following by subsequent pyrolysis. The separation
characteristics of the resulting membranes will then be tested in a SMR-MR. We have recently
constructed a lab-scale membrane reactor that can employ these cylindrical membranes for use in
SMR tests. This reactor can also function as a packed bed reactor (PBR), and we have performed
initial PBR experiments in order to determine the kinetic parameters of the commercial nickel-
based catalyst (Reformax 300 LTD, Clariant) we will be using in our future MR experiments.
108

In the future, we plan to test the membranes’ ability to reliably separate H2 from reformate
mixtures (that also contain other species like CO, CO2, CH4 and H2O) under temperature and
pressure conditions relevant to the proposed application. SMR reactors, typically, employ extreme
temperatures and pressures and, therefore, it is important for us to ensure that the membranes can
withstand these harsh process conditions. For that purpose, the membranes will be rigorously
tested over an extended period of time to ensure that they do not degrade. We believe that because
vapor-phase techniques, like iCVD and PECVD, conformally coat substrates, there will be less
defects and pinholes in the resulting membrane films. Therefore, we expect that the separation
factor and permeance of these membranes will be higher than those of membranes prepared using
solution-phase techniques, and that they will also exhibit better mechanical and chemical stability.

109

6 References

1
Gallucci, F., Fernandez, E., Corengia, P., and van Sint Annaland, M., Recent Advances on Membranes
and Membrane Reactors for Hydrogen Production, Chem. Eng. Sci., 2013, 92, 40–66.
2
Yun, S., and Oyama, S.T., Correlations in Palladium Membranes for Hydrogen Separation: A Review, J.
Membr. Sci., 2011, 375, 1-2, 28-45.
3
Elyassi, B., Sahimi, M., and Tsotsis, T.T., Silicon Carbide Membranes for Gas Separation Applications,
J. Membr. Sci., 2007, 288, 290-297.
4
Elyassi, B., Sahimi, M., and Tsotsis, T.T., A Novel Sacrificial Interlayer-based Method for the
Preparation of Silicon Carbide Membranes, J. Membrane Sci., 2008, 316, 73-79.
5
Elyassi, B., Deng, W., Sahimi, M., and Tsotsis, T.T., Effect of Porous and Non-porous Fillers on the
Performance of Silicon Carbide Membranes, Ind. Eng. Chem. Res., 2013, 52, 10269-10275.
6
Iulianelli, A., Liguori, S., Wilcox, J., and Basile, A., Advances on Methane Steam Reforming to Produce
Hydrogen Through Membrane Reactors Technology: A Review, Catal. Rev. Sci. Eng., 2016, 58, 1-35.
7
Elyassi, B., Inorganic Membranes, Encyclopedia of Chemical Processing, 2009, 1–16.
8
Uemiya, S., Brief Review of Steam Reforming Using a Metal Membrane Reactor, Top. Catal., 2004, 29,
1-2, 79–84.
9
Fernandes, F. A. N. and Soares, A. B., Modeling of Methane Steam Reforming in a Palladium
Membrane Reactor, Lat. Am. Appl. Res., 2006, 36, 3, 155–161.
10
Xu, X., Yang, W., Song, C., Liu, J., and Lin, L., Hydrogen Separation by Zeolite Membranes, ACS
Div. Fuel Chem. Prepr., 2003, 48, 1, 284–285.
11
Parsley, D., Ciora, R. J., Flowers, D. L., Laukaitaus, J., Chen, A., Liu, P. K. T., Yu, J., Sahimi, M.,
Bonsu, A., and Tsotsis, T. T., Field Evaluation of Carbon Molecular Sieve Membranes for the Separation
and Purification of Hydrogen from Coal- and Biomass-Derived Syngas., J. Membr. Sci., 2014, 450, 81–
92.
12
Lin, Y. S., Kumakiri, I., Nair, B. N., and Alsyouri, H., Microporous Inorganic Membranes., Sep. Purif.
Methods, 2002, 31, 2, 229–379.
13
Shelekhin, A. B., Dixon, A. G., and Ma, Y. H., Theory of Gas Diffusion and Permeation in Inorganic
Molecular Sieve Membranes, AIChE J., 1995, 41, 1, 58–67.
14
Jimenez, G., Dillon, E., Dahlmeyer, J., Garrison, T., Garrison, T., Darkey, S., Wald, K., Kubik, J.,
Paciulli, D., Talukder, M., et al. A Comparative Assessment of Hydrogen Embrittlement : Palladium and
Palladium-Silver Subjected to Hydrogen Absorption / Desorption Cycling, Adv. Chem. Eng. Sci., 2016,
6, 246–261.
110

15
Khatib, S. J., Oyama, S. T., de Souza, K. R., and Noronha, F. B., Review of Silica Membranes for
Hydrogen Separation Prepared by Chemical Vapor Deposition, In Membrane Science and Technology,
2011, 25–60.
16
Satvekar, R., Phadatare, M. R., Tiwale, B. M., and Pawar, S. H., Influence of Silane Content on the
Optical Properties of Sol Gel Derived Spin Coated Silica Thin Films, Int. J. Basic Appl. Sci., 2012, 4,
468–476.
17
Saputra, R. E., Astuti, Y., and Darmawan, A., Hydrophobicity of Silica Thin Films : The Deconvolution
and Interpretation by Fourier-Transform Infrared Spectroscopy, Spectrochim. Acta Part A Mol. Biomol.
Spectrosc., 2018, 199, 12–20.
18
Khalib, N. A. Z. Fabrication of Silica Ceramic Membrane via Sol-Gel Dip-Coating Method at Different
Nitric Acid Amount., IOP Conf. Ser. Mater. Sci. Eng., 2018, 290, 1-6.
19
Boffa, V., Blank, D. H. A., and Elshof, J. E., Hydrothermal Stability of Microporous Silica and Niobia
– Silica Membranes, J. Memb. Sci., 2008, 319, 1-2, 256–263.
20
Seidel, S., Riche, C., Gupta, M., Chemical Vapor Deposition of Polymer Films. Encycl. Polym. Sci.
Technol., 2011.
21
Iarikov, D. D., Hacarlioglu, P., and Oyama, S. T., Amorphous Silica Membranes for H2 Separation
Prepared by Chemical Vapor Deposition on Hollow Fiber Supports, 1st ed., Elsevier B.V., 2011, Vol. 14.
22
Khatib, S., and Oyama, S. T., Silica Membranes for Hydrogen Separation Prepared by Chemical Vapor
Deposition (CVD), Sep. Purif. Technol., 2013, 111, 20–42.
23
Tenhaeff, W., and Gleason, K., Initiated Chemical Vapor Deposition of Polymeric Thin Films: iCVD
and oCVD, Adv. Funct. Mater., 2008, 18, 7, 979-992.
24
Seidel, S., Kwong, P., and Gupta, M., Simultaneous Polymerization and Solid Monomer Deposition for
the Fabrication of Polymer Membranes with Dual-Scale Porosity, Macromolecules, 2013, 46, 2976-2983.
25
Kwong, P., Seidel, S., and Gupta, M., Solventless Fabrication of Porous-on-Porous Materials, ACS
Appl. Mater. Interfaces, 2013, 5, 9714-9718.
26
Seidel, S., and Gupta, M., Systematic Study of the Growth and Morphology of Vapor Deposited Porous
Polymer Membranes, J. Vac. Sci. Technol. A, 2014, 32, 041514.
27
Seidel, S., Dianat, G., and Gupta, M., Formation of Porous Polymer Coatings on Complex Substrates
Using Vapor Phase Precursors, Macromol. Mol. Eng., 2016, 301, 371-376.
28
Dianat, G., Seidel, S., De Luna, M., and Gupta, M., Vapor Phase Fabrication of Hydrophilic and
Hydrophobic Asymmetric Polymer Membranes, Macromol. Mol. Eng., 2016, 301, 1037-1043.
29
Dianat, G., and Gupta, M., Sequential Deposition of Patterned Porous Polymers Using
Poly(dimethylsiloxane) Masks, Polym., 2017, 126, 463-469.
111

30
Dianat, G., Movsesian, N., and Gupta, M., Process-Structure-Property Relationships for Porous
Membranes Formed by Polymerization of Solid Monomer by a Vapor-Phase Initiator, Macromol.,
2018, 51, 10297-10303.
31
Chan, K., and Gleason, K., A Mechanistic Study of Initiated Chemical Vapor Deposition of Polymers:
Analyses of Deposition Rate and Molecular Weight, Macromol., 2006, 39, 11, 3890-3894.
32
Coclite, A., Gleason, K., Initiated PECVD of Organosilicon Coatings: A New Strategy to Enhance
Monomer Structure Retention, Plasma Process Polym., 2012, 9, 425-434
33
Naserifar, S., L. Liu, L., Goddard III, W.A., Tsotsis, T.T, and Sahimi M., Towards a Process-Based
Molecular Model of SiC Membranes. II. Reactive Dynamics Simulation of the Pyrolysis of a Polymer
Precursor to Form Amorphous SiC, J. Phys. Chem. C, 2013, 117, 3320-3329.
34
Naserifar, S., Tsotsis, T.T., Goddard III, W.A., and Sahimi M., Towards a Process-Based Molecular
Model of SiC Membranes. III. Prediction of Transport and Separation of Binary Gaseous Mixtures Based
on the Atomistic Reactive Force Field, J. Membrane Sci., 2015, 473, 85-93.
35
Naserifar, S., Goddard III, W.A., Tsotsis, T.T., and Sahimi M., First Principle-Based Multiparadigm,
Multiscale Strategy for Simulating Complex Materials Processes with Applications to Amorphous SiC
Films, J. Chem. Phys., 2015, 142, 174703.
36
Naserifar, S., L. Liu, L., Goddard III, W.A., Tsotsis, T.T, and Sahimi M., Towards a Process-Based
Molecular Model of SiC Membranes. I. Development of a Reactive Force Field, J. Phys. Chem. C, 2013,
117, 3308-3319.
37
van Duin, A.C.T, Dasgupta, S., Lorant, F., and Goddard, W. A., ReaxFF: a Reactive Force Field for
Hydrocarbons, J. Phys. Chem. A, 2001, 105, 9396-9409.
38
Chenoweth, K., Cheung, S., van Duin, A.C.T., Goddard, W.A. and Kober, W., Simulations on the
Thermal Decomposition of a Poly-(dimethylsiloxane) Polymer Using the ReaxFF Reactive Force Field,
JACS, 2005, 127, 7192-7202.
39
Chenoweth, K., van Duin, A. C. T., and Goddard, W. A., ReaxFF Reactive Force Field for Molecular
Dynamics Simulations of Hydrocarbon Oxidation, J. Phys. Chem. A, 2008, 112, 1040–1053.
40
Chenoweth, K., van Duin, A. C. T., Persson, P., Cheng, M., Oxgaard, J., and Goddard, W. A.,
Development and Application of a ReaxFF Reactive Force Field for Oxidative Dehydrogenation on
Vanadium Oxide Catalysts, J. Phys. Chem. C, 2008, 112, 14645–14654.
41
Nielson, K. D., van Duin, A. C. T. Van, Oxgaard, J., Deng, W., and Goddard, W. A., Development of
the ReaxFF Reactive Force Field for Describing Transition Metal Catalyzed Reactions, with Application
to the Initial Stages of the Catalytic Formation of Carbon Nanotubes, J. Phys. Chem. A, 2005, 109, 493–
499.
42
Ostadhossein, A., Kim, S., Cubuk, E. D., Qi, Y., and van Duin, A. C. T., Atomic Insight into the
Lithium Storage and Diffusion Mechanism of SiO2/Al2O3 Electrodes of Lithium Ion Batteries: ReaxFF
Reactive Force Field Modeling, J. Phys. Chem. A, 2016, 120, 2114–2127.
112

43
Valentini, P., Schwartzentruber, T., and Cozmuta, I., Molecular Dynamics Simulation of Sticking on Pt
(111) Using the Ab Initio Based ReaxFF Reactive Force Field, J. Chem. Phys., 2010, 133, 84703.
44
Bedrov, D., Smith, G. D., and van Duin, A. C. T., Reactions of Singly-Reduced Ethylene Carbonate in
Lithium Battery Electrolytes: A Molecular Dynamics Simulation Study Using the ReaxFF, J. Phys.
Chem. A, 2012, 116, 2978–2985.
45
Song, C., Liu, Q., Ji, N., Kansha, Y., and Tsutsumi, A., Optimization of Steam Methane Reforming
Coupled with Pressure Swing Adsorption Hydrogen Production Process by Heat Integration, Appl.
Energy 2015, 154, 392–401.
46
da Silva Veras, T., Mozer, T. S., da Costa Rubim Messeder dos Santos, D., da Silva César, A.
Hydrogen: Trends, Production and Characterization of the Main Process Worldwide. Int., J. Hydrogen
Energy, 2017, 42, 4, 2018–2033.
47
Tsuru, T., Morita, T., Shintani, H., Yoshioka, T., and Asaeda, M., Membrane Reactor Performance of
Steam Reforming of Methane Using Hydrogen-Permselective Catalytic SiO2 Membranes., J. Membr.
Sci., 2008, 316, 1-2, 53–62.
48
Ockwig, N. W., and Nenoff, T. M., Membranes for Hydrogen Separation, Chem. Rev., 2007, 4078–
4110.
49
Beurden, P., On The Catalytic Aspects Of Steam Reforming Methane - A Literature Survey, ECN
2004, 1–27.
50
Gallucci, F., Comite, A., Capannelli, G., and Basile, A., Steam Reforming of Methane in a Membrane
Reactor: An Industrial Case Study, Ind. Eng. Chem. Res., 2006, 45, 9, 2994–3000.
51
Baker, R. W., Membrane Technology and Applications, Third John Wiley & Sons: West Sussex, 2012.
52
Nagasawa, H., Minamizawa, T., Kanezashi, M., Yoshioka, T., and Tsuru, T., High-Temperature
Stability of PECVD-Derived Organosilica Membranes Deposited on TiO2, Sep. Pur. Technol., 2014, 121,
13-19.
53
Nagasawa, H., Minamizawa, T., Kanezashi, M., Yoshioka, T., and Tsuru, T., Microporous Organosilica
Membranes for Gas Separation Prepared via PECVD Using Different O / Si Ratio Precursors, J. Membr.
Sci., 2015, 489, 11–19.
54
Baxamusa, S. H., Im, S. G., and Gleason, K. K., Initiated and Oxidative Chemical Vapor Deposition: A
Scalable Method for Conformal and Functional Polymer Films on Real Substrates, Phys. Chem. Chem.
Phys., 2009, 11, 26 , 5227–5240.
55
Gupta, M., and Gleason, K. K. Initiated Chemical Vapor Deposition of Poly(1H,1H,2H,2H-
Perfluorodecyl Acrylate) Thin Films, Langmuir, 2006, 22, 24, 10047–10052.
56
Cheng, C., and Gupta, M., Surface Functionalization of 3D-Printed Plastics via Initiated Chemical
Vapor Deposition, Beilstein J. Nanotechnol., 2017, 8, 1, 1629–1636.
113

57
Haller, P. D., Flowers, C. A., and Gupta, M., Three-Dimensional Patterning of Porous Materials Using
Vapor Phase Polymerization, Soft Matter, 2011, 7, 2428–2432.
58
Cheng, C., and Gupta, M., Roll-to-Roll Surface Modification of Cellulose Paper via Initiated Chemical
Vapor Deposition, Ind. Eng. Chem. Res., 2018, 57, 34, 11675–11680.
59
Wilson, A. M., Zank, G., Eguchi, K., Xing, W., Yates, B., and Dahn, J. R., Polysiloxane Pyrolysis.
Chem. Mater., 1997, 9, 7, 1601–1606.
60
Burns, G. T., Taylor, R. B., Xu, Y., Zangvil, A., and Zank, G. A., High-Temperature Chemistry of the
Conversion of Siloxanes to Silicon Carbide, Chem. Mater., 1992, 4, 6, 1313–1323.
61
Gao, H., Wang, H., Zhao, Z., Niu, M., Su, L., and Wei, Y., Reactive Dynamics Simulation Study on the
Pyrolysis of Polymer Precursors to Generate Amorphous Silicon Oxycarbide Structures., J. Phys. Chem.
C, 2018, 122, 10, 5767–5773.
62
Sarkar, S., Gan, Z., An, L., and Zhai, L., Structural Evolution of Polymer-Derived Amorphous SiBCN
Ceramics at High Temperature, J. Phys. Chem. C, 2011, 115, 50, 24993–25000.
63
Deng, W., Yu, X., Sahimi, M., and Tsotsis, T. T., Highly Permeable Porous Silicon Carbide Support
Tubes for the Preparation of Nanoporous Inorganic Membranes, J. Membr. Sci., 2014, 451, 192–204.
64
Dabir, S., Deng, W., Sahimi, M., and Tsotsis, T.T., Fabrication of Silicon Carbide Membranes on
Highly Permeable Supports, J. Membr. Sci., 2017, 537, 239–247.
65
ThermoSpectraTech. COLLECTOR
TM
II User’s Manual.
66
Raj, R., Pederiva, L., Narisawal, M., and Soraru, G. D., On the Onset of Fracture as a Silicon-Based
Polymer Converts into the Ceramic Phase, J. Am. Ceram. Soc., 2018, 102, 3, 924–929.
67
Trujillo, N. J., Wu, Q., and Gleason, K. K. Ultralow Dielectric Constant
Tetravinyltetramethylcyclotetrasiloxane Films Deposited by Initiated Chemical Vapor Deposition
(iCVD), Adv. Funct. Mater., 2010, 20, 4, 607–616.
68
Lubguban, J. , Rajagopalan, T., Mehta, N., Lahlouh, B., Simon, S. L., and Gangopadhyay, S., Low-k
Organosilicate Films Prepared by Tetravinyltetramethylcyclotetrasiloxane, Appl. Phys., 2002, 92, 2,
1033–1038.
69
Carlsson, D. J., Cooney, J. D., Gauthier, S., and Worsfold, D. J., Pyrolysis of Silicon-Backbone
Polymers to Silicon Carbide, J. Am. Ceram. Soc., 1990, 73, 2, 237–241.
70
Narisawa, M., Silicone Resin Applications for Ceramic Precursors and Composites, Mater., 2010, 3, 6,
3518–3536.
71
Ryan, E. T., Gates, S. M., Cohen, S. A., Ostrovski, Y., Adams, E., Virwani, K., and Grill, A., Effect of
Low-Frequency Radio Frequency on Plasma-Enhanced Chemical Vapor Deposited Ultra Low-κ
Dielectric Films for Very Large-Scale Integrated Interconnects, Appl. Phys., 2014, 115, 144107.
114

72
Mantz, R. A., Jones, P. F., Chaffee, K. P., Lichtenhan, J. D., Gilman, J. W., Ismail, I. M. K., and
Burmeister, M. J., Thermolysis of Polyhedral Oligomeric Silsesquioxane ( POSS ) Macromers and POSS
- Siloxane Copolymers, Chem. Mater., 1996, 8, 6, 1250–1259.
73
Burkey, D. D., and Gleason, K. K., Temperature-Resolved Fourier Transform Infrared Study of
Condensation Reactions and Porogen Decomposition in Hybrid Organosilicon-Porogen Films, J. Vac.
Sci. Technol., 2004, 22, 1, 61–70.
74
Burkey, D. D., and Gleason, K. K., Structure and Mechanical Properties of Thin Films Deposited from
1,3,5-Trimethyl-1,3,5- Trivinylcyclotrisiloxane and Water, J. Appl. Phys., 2003, 93, 9, 5143–5150.
75
Samitier, J., Marco, S., Morante, J. R., Estreve-Tinto, J., and Bausells, J., Analysis by FT-IR
Spectroscopy of SiO2-Polycrystalline in Micromechanics : Stress Measurements, Sensors and Actuators,
1992, 32, 347–353.
76
Sanaeishoar, H., Sabbaghan, M., and Mohave, F., Synthesis and Characterization of Micro-Mesoporous
MCM-41 Using Various Ionic Liquids as Co-Templates, Microporous Mesoporous Mater. 2015, 217,
219–224.
77
Shokri, B., Firouzjah, M. A., and Hosseini, S. I., FTIR Analysis of Silicon Dioxide Thin Film
Deposited by Metal Organic-Based PECVD, In 19th international symposium on plasma chemistry
society, Bochum, Germany, 2009, pp 1–4.
78
Babonneau, F., Thorne, K., and Mackenzie, J. D., Dimethyldiethoxysilane/Tetraethoxysilane
Copolymers: Precursors for the Silicon-Carbon-Oxygen System, Chem. Mater., 1989, 1, 5, 554–558.
79
Mabboux, P., and Gleason, K. K., Chemical Bonding Structure of Low Dielectric Constant Si : O : C :
H Films Characterized by Solid-State NMR., J. Electrochem. Soc., 2005, 152, 1, F7–F13.
80
Alexander, M. R., Short, R. D., and Jones, F. R., A Study of HMDSO/O2 Plasma Deposits Using a
High-Sensitivity and -Energy Resolution XPS Instrument : Curve Fitting of the Si 2p Core Level, Appl.
Surf. Sci., 1999, 137, 1-4, 179–183.
81
O'Hare, L., Parbhoo, B., and Leadley, S. R., Development of a Methodology for XPS Curve-Fitting of
the Si 2p Core Level of Siloxane Materials, Surf. Interface Anal., 2004, 36, 10, 1427–1434.
82
Belot, V., Corriu, R., Leclercq, D., Mutin, P. H., and Vioux, A., Thermal Reactivity of
Hydrogenosilsesquioxane Gels, Chem. Mater. 1991, 3, 1, 127–131.
83
Ivashchenko, L., Vasin, A., Ivashchenko, V., Ushakov, M., and Rusavsky, A. Blue Light Emission
from PECVD Deposited Nanostructured SiC, Mater. Res. Soc. Symp. Proc., 2006, 910.
84
Yakimova, R., Petoral, R., Yazdi, G., Vahlberg, C., Spetz, A., and Uvdal, K., Surface Functionalization
and Biomedical Applications Based on SiC, J. Phys. D: Appl. Phy., 2007, 40, 6435.
85
Bourenane, K., Keffous, A., Nezzal, G., Bourenane, A., Boukennous, Y., and Boukezzata, A., Influence
of Thickness and Porous Structure of SiC Layers on the Electrical Properties of Pt/SiC-pSi and Pd/SiC-
pSi Schottky Diodes for Gas Sensing Purposes, Sens. Actuators B: Chem., 2008, 129, 2, 612-620.
115

86
Kolar, F., Machovic, V., Svitilova, J., and Borecka, L., Structural Characterization and Thermal
Oxidation Resistance of Silicon Oxycarbides Produced by Polysiloxane Pyrolysis, Mater. Chem. Phys.,
2004, 86, 1, 88-98.
87
Eray, E., Boffa, V., Jorgensen, M., Magnacca, G., and Candelario, V., Enhanced Fabrication of Silicon
Carbide Membranes for Wastewater Treatment: From Laboratory to Industrial Scale, J. Membr. Sci.,
2020, 606, 118080.
88
Mourhatch, R., Tsotsis, T., and Sahimi, M., Network Model for the Evolution of the Pore Structure of
Silicon-Carbide Membranes During Their Fabrication, J. Membr. Sci., 2010, 356, 1, 138-146.
89
Liu, D., Hu, P., Chen, G., and Han, W., Fabrication of Silicon Carbide Coating on Mullite Fiber-
reinforced Mullite Matrix Composite via Dip-Coating, Ceram. Int., 2018, 44, 2, 2584-2586.
90
Nechanicky, M., Chew, K., Sellinger, A., and Laine, R., α-Silicon Carbide/β-Silicon Carbide
Particulate Composites via Polymer Infiltration and Pyrolysis (PIP) Processing using Polymethylsilane,
J. Eur. Ceram. Soc., 2000, 20, 4, 441-451.
91
Henry, A., Hassan, J., Bergman, J., Hallin, C., and Janzen, E., Thick Silicon Carbide Homoepitaxial
Layers Grown by CVD techniques., Chem. Vap. Deposition, 2006, 12, 475-482.
92
Pedersen, H., Leone, S., Kordina, O., Henry, A., Nishizawa, S., Koshka, Y., and Janzen, E. Chloride-
Based CVD Growth of Silicon Carbide for Electronics Applications, Chem. Rev., 2012, 112, 4, 2434-
2453.
93
Chen, E., Du, G., Zhang, Y., Qin, X., Lai, H., and Shi, W., RF-PECVD Deposition and Optical
Properties of Hydrogenated Amorphous Silicon Carbide Thin Films, Ceram. Int., 2014, 40, 9791-9797.
94
Flannery, A., Mourlas, N., Storment, C., Tsai, S., Tan, S., Heck, J., Monk, D., Kim, T., Gogoi, B., and
Kovacs, G., PECVD Silicon Carbide as a Chemically Resistant Material for Micromachined Transducers,
Sens. Actuator A Phys., 1998, 70, 48-55.
95
Cogan, S., Edell, D., Guzelian, A., Liu, Y., and Edell, R., Plasma-enhanced Chemical Vapor Deposited
Silicon Carbide as an Implantable Dielectric Coating, J. Biomed. Mater. Res., 2003, 67A, 3, 856-867.
96
Pages, X., Rouessac, V., Cot, D., Nabias, G., and Durand, J., Gas Permeation of PECVD Membranes
Inside Alumina Substrate Tubes, Sep. Purif. Technol., 2001, 25, 399-406.
97
Huran, J. Valovic, A., Kobzec, A.P., Balalykin, N.I., Kucera, M., Hascik, S., Malinovsky, L., and
Kovacova, E., Structural and Physical Characteristics of PECVD Nanocrystalline Silicon Carbide Thin
Films, Physics Procedia, 2012, 32, 303-307.
98
Wei, J., Chen, B., Poenar, D., Lee, Y., and Iliescu, C., Low-stress PECVD Amorphous Silicon Carbide
Layers for Biomedical Application, Proc. SPIE, Micro- and Nanotechnology: Materials, Processes,
Packaging, and Systems IV, 2009, 72690C.
99
Eom, J., Kim, Y., and Raju, S., Processing and Properties of Macroporous Silicon Carbide Ceramics: A
Review, J. Asian Ceram. Soc., 2013, 1, 220-242.
116

100
Nguyen, B., Dabir, S., Tsotsis, T., and Gupta, M., Fabrication of Hydrogen-Selective Silica
Membranes via Pyrolysis of Vapor Deposited Polymer Films, Ind. Eng. Chem. Res., 2019, 58, 15190-
15198.
101
Martinu, L., Zabeida, O., and Klemberg-Sapieha, J.E., Plasma-enhanced Chemical Vapor Deposition
of Functional Coatings, in: Handb. Depos. Technol. Film. Coatings, Elsevier, 2010, pp. 392–465.
102
Freitas, A., Maciel, C., Rodrigues, J., Ribeiro, R., Delgado-Silva, A., and Rangel, E., Organosilicon
Films Deposited in Low-Pressure Plasma From Hexamethyldisiloxane – A Review, Vacuum, 2021, 194,
110556.
103
Daniel, A., Pen, C., Archambeau, C., and Reniers, F., Use of a PECVD-PVD Process for the
Deposition of Copper Containing Organosilicon Thin Films on Steel, Appl. Surf. Sci., 2009, S82-S85.
104
Bae. I, Cho, S., Lee, S., Kim, Y., and Boo, J., Growth of Plasma-Polymerized Thin Films by PECVD
Method and Study on Their Surface and Optical Characteristics, Surf. Coat. Technol., 2005, 193, 142-
146.
105
Cheng, C., and Gupta, M., Solvent-Free Synthesis of Selectively Wetting Multilayer and Janus
Membranes, Adv. Mater. Interfaces, 2020, 7, 20, 2001103.
106
Zhao, J., Wang, M., and Gleason, K.K., Stabilizing the Wettability of Initiated Chemical Vapor
Deposited (iCVD) Polydivinylbenzene Thin Films by Thermal Annealing, Adv. Mater. Interfaces, 2017,
4, 1700270.
107
Moni, P., Mohr, A., and Gleason, K., Growth Rate and Cross-Linking Kinetics of Poly(divinyl
benzene) Thin Films Formed via Initiated Chemical Vapor Deposition, Langmuir, 2018, 34, 6687-6696.
108
Ermakova, E., Sysoev, S., Nikulina, L., Tsyrendorzhieva, I., Rakhlin, V., and Kosinova, M.,
Trimethyl(phenyl)silane – a Precursor for Gas Processes of SiCx:H Film Deposition: Synthesis and
Characterization, Mod. Electron. Mater., 2015, 1, 114-119.
109
Kobayashi, S., IR Spectroscopic Study of Silicon Nitride Films Grown at a Low Substrate
Temperature Using Very High Frequency Plasma-Enhanced Chemical Vapor Deposition, World
J. Condens. Matter Phys., 2016, 6, 287-293.
110
Dos Santos, D., Vechiato-Filho, A., Pesqueria, A., Guiotti, A., Rangel, E., da Cruz, N., and Goiato, M.,
Effect of Nonthermal Plasma Treatment on the Surface of Dental Resins Immersed in Artificial Saliva, J.
Polym. Eng., 2016, 36, 8, 785-793.
111
Rangel, R., Cruz, N., and Rangel, E., Role of the Plasma Activation Degree on Densification of
Organosilicon Films, Mater., 2020, 13, 1, 25.
112
Breuning, T., Study of Pyrolysis of Polysilazane Precursor in Si-C-(O,N) System, J. Anal. Appl.
Pyrolysis., 1999, 49, 1-2, 43-51.
113
Moene, R., Makkee, M., and Mouljin, J.A., High Surface Area Silicon Carbide as Catalyst Support
Characterization and Stability, Applied Catalysis A: General, 1998, 167, 321-330.
117

114
Straus, S., and Madorsky, S., Thermal Stability of Polydivinylbenzene and of Copolymers of Styrene
With Divinylbenzene and With Trivinylbenzene, J. Res. Natl. Bur. Stand A Phys. Chem., 1961, 243-248.
115
Schiavon, M.A., Radovanovic, E., and Yoshida, I., Microstructural Characterization of Monolithic
Ceramic Matrix Composites from Polysiloxane and SiC Powder, Powder Technol., 2001, 232-241.
116
Shi, Z., Sanguinito, S., Goodman, A., Jessen, K., and Tsotsis, T.T., Investigation of Mass Transfer and
Sorption in CO2/Brine/Rock Systems via In Situ FT-IR, Ind. Eng. Chem. Res., 2020, 59, 20181-20189.
117
Rich, S., Mille, V., Vivien, C., Godey, S., and Supiot, P., Kinetics of RPECVD Organosilicon Polymer
Post-treatment in a N 2/O 2 Microwave Plasma Remote Afterglow, Plasma Process Polym, 2010, 7, 9-10,
775-784.
118
Zhang, L., Ljazouli, R., Lefaucheux, P., Tillocher, T., Dussart, R., Mankelevich, Y., Marneffe, J.,
Gendt, S., and Baklanov, M., Damage Free Cryogenic Etching of a Porous Organosilica Ultralow-k Film,
Electrochem. Solid-State Lett., 2013, 2, N5.
119
Tsuge, A., Uwamino, Y., and Ishizuka, T., Determination of Silicon Dioxide in Silicon Carbide by
Diffuse Reflectance Infrared Fourier Transform Spectrometry, Appl. Spectrosc., 1986, 40, 3, 310-313.
120
Yi, B., Wang, S., Hou, C., Huang, X., Cui, J., and Yao, X., Dynamic Siloxane Materials: From
Molecular Engineering to Emerging Applications, Chem. Eng. J., 2021, 405, 127023.
121
Fujii, T., Yokoi, T., Hiramatsu, M., and Nawata, M., Low Di-electric Constant Film Formation by
Oxygen-radical Polymerization of Laser-evaporated Siloxane, J. Vac. Sci. Technol. B, 1997, 15, 746.
122
Qi, H., Wang, X., Zhu, T., Li, J., Xiong, L., and Liu, F., Low Dielectric Poly(imide siloxane) Films
Enabled by a Well-Defined Disiloxane-Linked Alkyl Diamine, ACS Omega, 2019, 4, 26, 22143-22151.
123
Wang, Z., Wang, H., Mitra, A., Huang, L., and Yan, Y., Pure-Silica Zeolite Low-K Dielectric Thin
Films, Adv. Mater., 2001, 13, 10, 746-749.
124
Katagiri, K., Yamazaki, S., Inumaru, K., and Koumoto, K., Anti-reflective Coatings Prepared via
Layer-by-layer Assembly of Mesoporous Silica Nanoparticles and Polyelectrolytes, Polym. J., 2015, 47,
190-194.
125
Biswas, P. K., Devi, P., Chakraborty, P., Chatterjee, A., Ganguli, D., Kamath, M., and Joshi, A.,
Porous Anti-reflective Silica Coatings with a Spectral Coverage by Sol-gel Spin Coating Technique, J.
Mater. Sci. Lett., 2003, 22, 181-183.
126
Dowling, D., Nwankire, C., Riihimaki, M., Keiski, R., and Nylen, U., Evaluation of the Anti-fouling
Properties of nm Thick atmospheric Plasma Deposited Coatings, Surf. Coat. Technol., 2010, 205, 5,
1544-1551.
127
Zhang, H., and Chiao, M., Anti-fouling Coatings of Poly(dimethylsiloxane) Devices for Biological and
Biomedical Applications, J. Med. Biol. Eng., 2015, 35, 143-155.
128
Roualdes, S., Sanchez, J., and Durand, J., Gas Diffusion and Sorption Properties of Polysiloxane
Membranes Prepared by PECVD, J. Membr. Sci., 2002, 198, 299-310.
118

129
Hall, D.B., Underhill, P., and Torkelson, J.M., Spin Coating of Thin and Ultrathin Polymer Films,
Polym. Eng. Sci., 1998, 38, 12, 2039-2045.
130
Gleason, K.K., CVD Polymers: Fabrication of Organic Surfaces and Devices, 2015, John Wiley &
Sons.
131
Ozaydin-Ince, G., Coclite, A.M., and Gleason, K.K., CVD of Polymeric Thin Films: Applications in
Sensors, Biotechnology, Microelectronics/organic Electronics, Microfluidics, MEMS, Composites and
Membranes, Reports on Progress in Physics, 2011, 75, 1, 016501.
132
Coclite, A.M., et al., 25th Anniversary Article: CVD Polymers: A New Paradigm for Surface
Modification and Device Fabrication, Adv. Mater., 2013, 25, 38, 5392-5423.
133
Gleason, K.K., Overview of Chemically Vapor Deposited (CVD) Polymers, CVD Polymers:
Fabrication of Organic Surfaces and Devices, 2015, 1-11.
134
de Freitas, A.S., Maciel, C.C., Rodrigues, J.S., Ribeiro, R.P., Delgado-Silva, A.O. and Rangel, E.C.,
Organosilicon Films Deposited in Low-pressure Plasma from Hexamethyldisiloxane—A review,
Vacuum, 2021, 194, 110556.
135
Sreenivasan, R. and Gleason, K. K., Overview of Strategies for the CVD of Organic Films and
Functional Polymer Layers, Chem. Vap. Deposition, 2009, 15, 77-90.
136
Chen, N., Kim, D. H., Kovacik, P., and Sojoudi, H., Annu. Rev. Chem. Biomol. Eng. 2016, 7, 373.
137
Shen, B., Wang, S., and Tenhaeff, W., Ultrathin Conformal Polycyclosiloxane Films to Improve
Silicon Cycling Stability, Sci. Adv., 2019, 5, 7.
138
Larkin, P., Chapter 3 - Instrumentation and Sampling Methods, In Infrared and Raman Spectroscopy,
2011, 27-54.
139
Thompson, J. M., Mass spectrometry, 2018, Pan Stanford.
140
Wang, X., Hong, Y., Shi, H., and Szanyi, J., Kinetic Modeling and Transient DRIFTS-MS Studies of
CO2 Methanation over Ru/Al2O3 Catalysts, J. Catal., 2016, 343, 185-195.
141
Ochoa, J., Trevisanut, C., Millet, J., Busca, G., and Cavani, F., In Situ DRIFTS-MS Study of the
Anaerobic Oxidation of Ethanol over Spinel Mixed Oxides, J. Phys. Chem. C., 2013, 117, 45, 23908-
23918.
142
White, R., Thermal Analysis by Diffuse Reflectance Fourier Transform Infrared Spectroscopy-Mass
Spectrometry, J. Anal. Appl. Pyrolysis, 1991, 18, 325-339.
143
Nomura, K., Kalai, R., Nakano, A., Rajak, P., and Vashista, P., RXMD: A Scalable Reactive
Molecular Dynamics Simulator for Optimized Time-to-solution., SoftwareX, 2020, 11, 100389.
144
Chen, S., Liu, C., Li, Q., Liu, Y., Xin, L., and Yu, W., A ReaxFF-based Molecular Dynamics Study of
the Pyrolysis Mechanism of Hexamethyldisiloxane, J. Mol. Liq., 2002, 356, 119026.
119

145
Shao, Y., Gan, Z., Epifanovsky, E., Gilbert, A. T. B., Wormit, M., Kussmann, J., Lange, A. W., Behn,
A., Deng, J., Feng, X… and Head-Gordon, M. Advances in molecular quantum chemistry contained in
the Q-Chem 4 program package, Mol. Phys., 2015, 113, 184-215.
146
Gates, S. M., Neumayer, D. A., Sherwood, M. H., Grill, A., Wang, X., and Sankarapandian, M.,
Preparation and Structure of Porous Dielectrics by Plasma Enhanced Chemical Vapor Deposition, J.
Appl. Phys., 2007 101, 094103.
147
Kherroub, D., Belbachir, M., and Lamouri, S., Activated bentonite (Maghnite-H+) as Green Catalyst
for Ring-opening Polymerization of 1,3,5,7-tetravinyltetramethylcyclotetrasiloxane, Res. Chem.
Intermed., 2017, 43,10.
148
Hamedani, Y., Macha, P., Bunning, T. J., Naik, R. R., & Vasudev, M. C., Plasma-Enhanced Chemical
Vapor Deposition: Where we are and the Outlook for the Future, Chemical Vapor Deposition - Recent
Advances and Applications in Optical, Solar Cells and Solid State Devices, edited by Sudheer Neralla,
IntechOpen, 2016.
149
Oh. T., and Choi, C., Comparison between SiOC Thin Films Fabricated by Using Plasma Enhanced
Vapor Deposition and SiO2 Thin Films by Using Fourier Transform Infrared Spectroscopy, J. Korean
Phys. Soc., 2010, 56, 4, 1150-1155.
150
Ellerbrock, R., Stein, M., and Schaller, J., Comparing Amorphous Silica, Short-range-ordered Silicates
and Silicic Acid Species by FTIR, Sci Rep., 2022, 12, 11708.
151
Sassin, M., Long, J., Wallace, J., and Rolison, D., Routes to 3D Conformal Solid-state Dielectric
Polymers: Electrodeposition Versus Initiated Chemical Vapor Deposition, Mater. Horiz., 2015, 2, 502.
152
Choi, J., Seong, H., Pak, K., and Im., S., Vapor-phase Deposition of the Fluorinated Copolymer Gate
Insulator for the P-type Organic Thin-film Transistor, J. Inf. Disp., 2016, 17, 2, 43-49.
153
Ondo, D., Loyer, F., Werner, F., Leturcq, R., Dale, P., and Boscher, N., Atmospheric-Pressure
Synthesis of Atomically Smooth, Conformal, and Ultrathin Low-k Polymer Insulating Layers by Plasma-
Initiated Chemical Vapor Deposition, ACS Appl. Polym. Mater., 2019, 1, 3304-3312.
154
Plawsky, J., Wang, F., and Gill, W., Kinetic Model for the Pyrolysis of Polysiloxane Polymer to
Ceramic Composites, AIChe J., 2002, 481, 10, 2315-2323.
155
Wilson, A.M., Zank, G., Eguchi, K., Xing, W., Yates, B., and Dahn, J.R., Pore Creation in Silicon
Oxycarbides by Rinsing in Dilute Hydrofluoric Acid, Chem. Mater., 1997, 9, 2139-2144.
156
Bahloul-Hourlier, D., Latournerie, J., and Dempsey, P., Reaction Pathways During the Thermal
Conversion of Polysiloxane Precursors into Oxycarbide Ceramics, J. Eur. Ceram. Soc., 2005, 25, 979-
985.
157
Walsh, R., Bond Dissociation Energy Values in Silicon-Containing Compounds and Some of Their
Implications, Acc. Chem. Res., 1981, 14, 246-252.
158
Dean, J. (Eds.), Lange's Handbook of Chemistry, McGraw-Hill, New York, 14th edition, (1992).

Abstract (if available)

Abstract The efficient separation of H2 under steam reforming conditions is important for the development of the clean energy industry and has helped drive inorganic membrane research for several decades. Silicon-based ceramic materials have shown great potential as nanoporous membranes for gas separations due to their high temperature resistance and excellent chemical and mechanical stability. In our work, we demonstrate the ability to fabricate highly permeable silicon-based membranes using chemical vapor deposition (CVD) and subsequent pyrolysis. Chapter 1 of this Thesis introduces the initiated chemical vapor deposition (iCVD) and plasma-enhanced chemical vapor deposition (PECVD) techniques, and the motivation for using polymer films as precursors for the preparation of ceramic materials. Chapter 2 demonstrates the fabrication of a silica membrane through the iCVD deposition of a polymer onto a ceramic substrate and its subsequent pyrolysis. We demonstrate the ability to conformally deposit on a porous ceramic substrate a cyclic siloxane polymer called pV4D4, and we pyrolyze the film in order to form a thin silica membrane. Chapter 3 demonstrates the fabrication of a SiC-type film through the deposition of an organosilicon polymer using PECVD and subsequent pyrolysis. We evaluate how changing the monomer ratio in during the preparation of a precursor copolymer film affects the structure of the resulting pyrolyzed ceramic. Chapter 4 discusses the deposition of pV4D4 via PECVD and its subsequent pyrolysis using a unique DRIFTS-MS experimental system that allows for the in-situ characterization and analysis of the resulting materials. We also perform complementary computational simulations modeling the pyrolysis of the pV4D4 in order to gain further insight into the reaction mechanism. Chapter 5 provides closing remarks on the pyrolysis studies completed to date and discusses future research directions.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Controlling polymer film patterning, morphologies, and chemistry using vapor deposition

PDF

Vapor phase deposition of dense and porous polymer coatings and membranes for increased sustainability and practical applications

PDF

Vapor phase deposition of polymers in the presence of low vapor pressure liquids

PDF

Understanding structural changes of polymer films during vapor phase deposition

PDF

Scale-up of vapor-phase deposition of polymers: towards large-scale processing

PDF

Inorganic/organic hybrid materials and grafted coatings via vapor phase deposition processes

PDF

Fabrication of polymer films on liquid substrates via initiated chemical vapor deposition: controlling morphology and composition

PDF

Novel processing of liquid substrates via initiated chemical vapor depostion

PDF

Simultaneous monomer deposition and polymerization at low substrate temperatures for the formation of porous polymer membranes

PDF

Fabrication of functional porous membranes via polymerization of solid monomer by a vapor-phase initiator

PDF

Formation of polymer gels, films, and particles via initiated chemical vapor deposition onto liquid substrates

PDF

Continuum and pore netwok modeling of preparation of silicon-carbide membranes by chemical-vapor deposition and chemical-vapor infiltration

PDF

Selective deposition of polymer coatings onto structured surfaces

PDF

The patterning of polymer thin films on porous substrates via initiated chemical vapor deposition

PDF

A process-based molecular model of nano-porous silicon carbide membranes

PDF

Fabrication of nanoporous silicon oxycarbide materials via a sacrificial template technique

PDF

Experimental studies and computer simulation of the preparation of nanoporous silicon-carbide membranes by chemical-vapor infiltration/chemical-vapor deposition techniques

PDF

Chemical vapor deposition of graphene: synthesis, characterization, and applications

PDF

Chemical vapor deposition of graphene and two-dimensional materials: synthesis, characterization, and applications

PDF

Fabrication of silicon carbide sintered supports and silicon carbide membranes

Asset Metadata

Creator Nguyen, Bryan Hoang (author)

Core Title Fabrication of silicon-based membranes via vapor-phase deposition and pyrolysis of organosilicon polymers

School Viterbi School of Engineering

Degree Doctor of Philosophy

Degree Program Chemical Engineering

Degree Conferral Date 2023-05

Publication Date 02/01/2023

Defense Date 01/11/2023

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag chemical vapor deposition,DRIFTS,iCVD,OAI-PMH Harvest,PECVD,pyrolysis

Format theses (aat)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Gupta, Malancha (committee chair), Nakano, Aiichiro (committee chair), Tsotsis, Theodore (committee chair)

Creator Email bnguyen828@gmail.com,bryanngu@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-oUC112724692

Unique identifier UC112724692

Identifier etd-NguyenBrya-11460.pdf (filename)

Legacy Identifier etd-NguyenBrya-11460

Document Type Dissertation

Format theses (aat)

Rights Nguyen, Bryan Hoang

Internet Media Type application/pdf

Type texts

Source 20230206-usctheses-batch-1006 (batch), University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the author, as the original true and official version of the work, but does not grant the reader permission to use the work if the desired use is covered by copyright. It is the author, as rights holder, who must provide use permission if such use is covered by copyright. The original signature page accompanying the original submission of the work to the USC Libraries is retained by the USC Libraries and a copy of it may be obtained by authorized requesters contacting the repository e-mail address given.

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA

Repository Email cisadmin@lib.usc.edu