University of Southern California Dissertations and Theses

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

(USC Thesis Other)

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

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

Simultaneous Monomer Deposition and Polymerization at Low Substrate
Temperatures for the Formation of Porous Polymer Membranes

Doctoral Dissertation
Scott Seidel
August 2015

University of Southern California
Mork Family Department of Chemical Engineering and Materials Science
Los Angeles, CA, USA

ii

Committee Members
Dr. Malancha Gupta
Dr. Noah Malmstadt
Dr. Aiichiro Nakano
iii

Executive Summary
In this work we present a novel method to produce porous polymer coatings and
membranes using a vapor phase polymerization technique called initiated chemical vapor
deposition (iCVD). The premise of this technique is the simultaneous deposition and
polymerization of a monomer species, which is achieved by increasing the monomer partial
pressure above its saturation pressure when the substrate temperature is below the freezing point
of the monomer. We explore the mechanism of this process, detail the ability to control the
morphology and chemistry, and extend this technique towards complex substrates found in many
applications.
This document is broken into eight parts. Chapter 1 introduces the reader to different
forms of vapor phase polymerization and specifically the details of iCVD. The advantages of
these vapor phase techniques are directly compared to the limitations of solution phase
techniques traditionally used to fabricate porous polymers. Chapter 2 details the experimental
procedures and characterization techniques used in these studies. Chapter 3 explores the
requirements needed to form porous polymer with additional information into both the monomer
deposition and monomer polymerization steps. Chapter 4 and 5 examine the degree to which the
morphology and chemistry, respectively, can be controlled and tailored towards specific
applications. Chapter 6 extends the technique to porous and complex substrates, which
traditionally represent the most difficult substrates to coat with porous polymer. Chapter 7
introduces future and ongoing work on exploring the full range of capabilities of the system.
Chapter 8 concludes the document by discussing the impact of the work on the vapor phase
polymerization community and on several specific applications.
iv

General Acknowledgements
I would like to thank everyone who has helped me throughout my research career. I had
my first exposure to polymer research as a freshman in the Department of Macromolecular
Science and Engineering at Case Western Reserve University, particularly with Dr. Christoph
Weder and Dr. David Schiraldi. I would like to thank Dr. Malancha Gupta, my committee
members Dr. Noah Malmstadt and Dr. Aiichiro Nakano, and my colleagues for their input and
suggestions. I would also like to thank the NSF for supporting me through the NSF GRFP and
USC for supporting me through the Viterbi Dean’s Fellowship.

v

Table of Contents
Executive Summary ....................................................................................................................... iii
General Acknowledgements .......................................................................................................... iv
Table of Contents ............................................................................................................................ v
List of Figures ............................................................................................................................... vii
List of Tables ............................................................................................................................... xiii
Chapter 1: Introduction ................................................................................................................... 1
1.1 Overview ............................................................................................................................... 1
1.2 Vapor Phase Polymerization Techniques ............................................................................. 3
1.3 Initiated Chemical Vapor Deposition ................................................................................... 6
1.4 Advantages of Vapor Phase Polymerization ....................................................................... 10
1.5 Porous Polymers ................................................................................................................. 14
Chapter 2: Experimental ............................................................................................................... 19
2.1 Materials and Supplies ........................................................................................................ 19
2.2 Procedures ........................................................................................................................... 20
2.3 Characterization .................................................................................................................. 25
Chapter 3: Fabrication of Porous Polymers .................................................................................. 27
3.1 Requirements ...................................................................................................................... 27
3.2 Mechanism .......................................................................................................................... 33
Chapter 4: Morphological Control ................................................................................................ 42
4.1 Deposition Time.................................................................................................................. 42
4.2 Substrate Temperature ........................................................................................................ 46
4.3 Partial Pressure.................................................................................................................... 49
4.4 Surface Energy .................................................................................................................... 51

vi

Chapter 5: Chemical Control ........................................................................................................ 53
5.1 Generality ............................................................................................................................ 53
5.2 Chemical Modification ....................................................................................................... 56
Chapter 6: Applications ................................................................................................................ 64
6.1 Complex Substrates ............................................................................................................ 64
6.2 Porous Substrates ................................................................................................................ 67
Chapter 7: Future Work ................................................................................................................ 75
Chapter 8: Conclusions ................................................................................................................. 80
References ..................................................................................................................................... 81

vii

List of Figures
Figure 1. Number of citations per given year related to the chemical vapor deposition of
polymers. Publication search performed through Web of Knowledge. ........................... 1
Figure 2. Schematic of a typical iCVD reactor. Polymerization occurs through the
adsorbed monomer (M) on the substrate due to the cleaving of the initiator (I)
into free radicals (•). ......................................................................................................... 7
Figure 3. Photographs of Kimwipes® before (left) and after coating with PPFDA (right). ........ 12
Figure 4. Flow chart showing the different structures possible (SEM images) in vapor
phase polymerization. Monomer partial pressures above the monomer saturation
pressure and substrate temperatures below the freezing point of the monomer are
required for porous film formation. ............................................................................... 28
Figure 5. FTIR spectra of the MAA monomer, the sample before and after sublimation,
and the PMAA thin film reference. The strong peak attributed to C=C stretching
of the monomer at 1636 cm
-1
(dashed line) is no longer present after sublimation,
at which point the porous film is free of MAA monomer and is chemically
identical to the thin PMAA reference film. ................................................................... 30
Figure 6. Schematic of porous film fabrication showing the growth on a molecular and
macroscopic scale. Cross-sectional SEM images showing the large-scale pores
between (left) and small-scale pores within (right) the microstructures. ....................... 31
Figure 7. Mercury porosimetry data showing the pore size distribution. ..................................... 32
Figure 8. a) Mass and b) thickness data for MAA deposition at a range of substrate
temperatures and MAA flow rates. ................................................................................ 35
viii

Figure 9. Optical cross-sectional images show that the substrate temperature has a large
impact of the morphology on the depositing MAA. ...................................................... 36
Figure 10. a) Mass and b) thickness of the composite samples at a range of substrate
temperatures and MAA flow rates. ................................................................................ 38
Figure 11. a) Thickness of the final porous PMAA membrane samples at a range of
substrate temperatures and MAA flow rates. b) Thickness comparison of the
monomer, composite, and polymer samples at a 120 mg/min MAA flow rate
showing the thickness loss of the samples caused by the heated filament and
removal of the MAA template. ...................................................................................... 39
Figure 12. SEM images showing the effect of adding a post-deposition polymerization step
to the structure and cohesiveness of the porous coatings at short deposition times. ..... 41
Figure 13. a) Top-down SEM images with inset cross-sectional SEM images showing the
microstructure and growth of PMAA porous films as a function of deposition
time. Graphs showing b) thickness growth rate and c) large-scale pore size. ............... 44
Figure 14. Top-down (a-e) and cross-section (f-j) images SEM images of PMAA samples
grown for 20 minutes at a,f) 7.1, b,g) 9.5, c,h) 15.3, d,i) 19.5, and e,j) 40.5
µm/min, resulting in the following respective thicknesses: 142, 190, 306, 390,
and 810 µm. Unlabelled scale bars represent 200 µm. .................................................. 45
Figure 15. Top-down SEM images comparing a) a 920 µm sample grown for 60 minutes
(15.3 µm/min) to b) an 810 µm sample grown for 20 minutes (40.5 µm/min). ............ 45
Figure 16. Porous film thickness growth rate increases with decreasing substrate
temperature. ................................................................................................................... 47
ix

Figure 17. Top-down optical images showing the morphology change caused by varying
the substrate temperature at a constant MAA flow rate of 80 mg/min. ......................... 47
Figure 18. Cross-sectional SEM images showing the morphology of the porous polymer
membranes at varying MAA flow rates and substrate temperatures. ............................ 48
Figure 19. Higher magnification SEM images show the presence of small-scale pores for
the high MAA flow rate samples at a) -20 °C and b) 0 °C. ........................................... 48
Figure 20. Porous film thickness growth rate increases with increasing monomer partial
pressure. ......................................................................................................................... 50
Figure 21. Cross-sectional SEM images of porous films resulting from pump down at a) 80
mTorr and b) 315 mTorr (with inert nitrogen flow), respectively. ................................ 50
Figure 22. a) Optical top down image showing the (2) widespread porous PMAA on the
bare silicon substrate compared to (1) isolated growth on the fluorinated
substrate. b)Top-down and c) cross-sectional SEM images of the sample from the
outlined region. .............................................................................................................. 52
Figure 23. a) FTIR spectra of the NIPAAm monomer, the sample before and after
sublimation, and the PNIPAAm thin film reference. The strong peak attributed to
C=C stretching of the monomer at 1622 cm
-1
(dashed line) is no longer present
after sublimation, at which point the porous film is free of NIPAAm and is
chemically identical to the thin PNIPAAm films. (b) SEM images showing top-
down and inset cross-sectional images of PNIPAAm porous films. ............................. 54
Figure 24. (a) Photograph and (b) SEM images of a free-standing P(MAA-co-EGDA)
porous film. .................................................................................................................... 55
x

Figure 25. (a) FTIR spectra of a PMAA sample and a P(MAA-co-MAN) sample at 47 wt%
MAN. (b) Zoomed-in area showing peak fitting analysis of the P(MAA-co-
MAN) sample................................................................................................................. 58
Figure 26. Plot of the MAN content as a function of annealing time. .......................................... 59
Figure 27. (a) Water contact angles on the membranes as a function of the MAN content.
(b) SEM images showing the loss of structure that occurs at conversions above 60
wt% MAN. ..................................................................................................................... 60
Figure 28. Dissolution time as a function of the MAN content and pH. ...................................... 61
Figure 29. (a) Reaction scheme of DAP with P(MAA-co-MAN) and (b) the FTIR spectra
confirming crosslinking. The vertical dotted lines represent the N-H bending of
the amine (1545 cm
-1
) and the C=O stretching of the amide (1636 cm
-1
) groups of
DAP................................................................................................................................ 63
Figure 30. Vertically-oriented Stainless steel tubing (a) before and (b) after porous coating
as seen inside the reactor from above. (c) Conformal coating on horizontally-
oriented tube. (d) Porous PMAA coverage on internal surfaces of tube radii of
0.88 mm, 3.26 mm, 5.64 mm, 8.81 mm. (e) SEM micrograph showing porous
PMAA on the top external surface of the substrate with a high magnification
inset. ............................................................................................................................... 65
Figure 31. Schematic representation and optical images of the fabrication of patterned
porous-on-porous materials. .......................................................................................... 68
Figure 32. Schematic of membrane structure with scanning electron micrographs showing
the a) top-down membrane morphology with an inset at higher magnification, b)
tapered edge, c) exterior cross-section, and d) interior cross-section. ........................... 70
xi

Figure 33. The distribution of the diameters of the large-scale pores for membranes
fabricated at deposition times of 10, 20, and 30 minutes. .............................................. 71
Figure 34. a) Top-down scanning electron micrograph of porous P(MAA-co-EGDA)
membrane. b) Intensity plot of dyed P(MAA-co-EGDA) membranes. Mixture of
crystal violet and ponceau S flowing through a paper-based microfluidic device
in the c) absence and d) presence of a porous P(MAA-co-EGDA) membrane. ............ 72
Figure 35. (a) Top-down and (b) cross-sectional SEM images of gauze and (c) top-down
and (d) cross-sectional SEM images of P(MAA-co-MAN) membrane with 38
wt% MAN on gauze....................................................................................................... 74
Figure 36. Photograph of low partial pressures (above the saturation pressure) attempting
to achieve stable monomer deposition. The arrows show (from left to right):
widespread porous polymer, nonporous thin polymer film, porous polymer fractal
patterns, and bare silicon wafer which was masked with polyimide tape. .................... 76
Figure 37. Preliminary results on producing a capping layer without relying on monomer
liquid condensation. Top-down SEM images at a) low and b) high magnification
show the partial covering of a thin nonporous layer. c) The cross-sectional image
shows the nonporous film folded over (arrow) on top of a thick porous layer. ............. 77
Figure 38. A capped membrane produced by inducing liquid monomer condensation
during the growth process. Top-down SEM images showing a) low magnification
(with cracks) and b) high resolution views of the capping layer and a c) cross-
sectional image showing the porous structure capped with a nonporous layer
(arrow)............................................................................................................................ 78
xii

Figure 39. Top-down SEM image of part of the P(MAA-co-EGDA) sample which had no
condensing crosslinker. .................................................................................................. 79

xiii

List of Tables
Table 1. Mercury intrusion porosimetry data. .............................................................................. 31
Table 2. Time Study and Resultant Growth Rates of PMAA Porous films ................................. 43
Table 3. Reactor conditions and resultant growth rates varying substrate temperatures. ............. 47
Table 4. Reactor conditions and resultant growth rates varying MAA partial pressures. ............ 49

1

Chapter 1: Introduction
1.1 Overview
Chemical vapor deposition is a workhorse technology in the coating industry and there is
a growing interest in depositing polymers by these means. Over the past twenty years, citations
relating to chemical vapor deposition of polymers have increased more than one hundred-fold
1

(Figure 1) leading to many significant advances in the field. One of these advancements was the
development of the iCVD technology which improved upon the existing techniques by providing
faster deposition rates while maintaining full functional retention of the precursors. The iCVD
technique has been used to a wide-range of applications including sensors,
2
drug delivery,
3
and
superhydrophobic surfaces.
4

Figure 1. Number of citations per given year related to the chemical vapor deposition of polymers. Publication
search performed through Web of Knowledge.

2

The chemical vapor deposition of polymer films eliminates the need for organic solvents
and thereby offers a safer and cleaner alternative to liquid-phase processing. Developments in
these techniques allow vapor phase polymerization onto temperature-sensitive substrates and
three-dimensional surfaces. These deposition techniques can be used to modify the surfaces of
curved, pillared, and porous materials with hydrophilic, responsive, and conductive films. The
polymer chains can either be physisorbed or chemisorbed onto the surface and crosslinking is
possible by copolymerization. Current and future work involves using these techniques for the
modification of microfluidic devices, implants, paper, and for applications in diagnostics,
filtration, and solar cells.
The ability to extend these vapor phase technologies from dense to porous coatings
would provide many advantages over traditional solution phase methods. These solution-phase
methods typically cannot be used to uniformly coat substrates with complex three-dimensional
surfaces such as textiles and membranes due to surface tension effects. In addition, the use of
organic solvents leads to difficulties with disposal and increased costs. The goal of this work is
to develop a technique which avoids these limitations.
3

1.2 Vapor Phase Polymerization Techniques
Due to the wide range of vapor phase polymerization techniques that are used both in
academia and industry, only a handful will be discussed below. These techniques have varying
amounts of overlap with the iCVD technology, which will be discussed in detail in the following
section. The purpose of the detailing these techniques is to understand the implications of our
research on the broader chemical vapor deposition community.
Parylenes, or poly(p-xylylenes), were the earliest polymers to be successfully deposited
using vapor phase polymerization techniques. Parylene has been commercially synthesized via
the Gorham process since 1966.
5
The polymerization occurs in a segmented reactor that contains
three sequential zones for sublimation, pyrolysis, and deposition. First, [2.2]paracyclophane
precursors are vaporized at 70 – 90 °C in the sublimation zone. The precursors then pass into a
pyrolysis zone kept at 600 – 800 °C where they are cleaved into p-quinodimethanes. The
temperature in the pyrolysis zone must be carefully controlled to retain the functionality of the
monomer.
6
The p-quinodimethane molecules are sent into a deposition chamber where they
adsorb and polymerize onto a substrate that is kept at room temperature.
7,8,9
Common deposition
rates range between 30 – 100 nm/min,
10,11
and the molecular weights of the polymers are
typically above 100,000 g/mol.
12,13
The moderate substrate temperatures allow temperature-
sensitive materials to be coated.
14
An inert carrier gas is often used to help transport the
molecules through the reactor. A wide variety of parylene-based polymers can be produced for
biological applications,
10,11
dielectrics,
15
and microfluidics
16
by synthesizing novel precursors
with pendant functionalities.
4

A chemical vapor deposition system known as gas phase deposition polymerization
(GDP),
17,18
gas phase assisted surface polymerization (GASP),
19
or vapor phase assisted surface
polymerization (VASP)
20
utilize an H-shaped glass tube reactor near the saturation pressure of
the monomer. Liquid phase monomer is added to the other leg of the vessel opposite to the
substrate.
19
A solution of initiator is dried onto the surface, after which the tube is sealed under
vacuum and placed in a 40 – 60 °C oven.
19,20
The process follows typical free radical
polymerization. The resulting polymer films are composed of high molecular weight chains
ranging from 185,000 to 741,000 g/mol for poly(methyl methacrylate) (PMMA)
17
and 35,400 to
114,000 g/mol for poly(2,2,3,3,3-pentafluoropropyl methacrylate) (PPFMA).
18
This method is
especially suited for academic use without the need for expensive equipment.
Plasma enhanced chemical vapor deposition (PECVD) is commonly used to make
electrical components on the laboratory and industrial scale. Two PECVD reactor designs are the
most widespread. The first utilizes a bell-jar shape with internal parallel-plate electrodes while
the second is tubular shaped with external coils or ring electrodes.
21
The monomer molecules are
broken into free radicals using a high energy plasma source. However, the high energy supplied
by the plasma can destroy the chemical functionality of the monomers and leads to etching. The
PECVD technology is governed by a complex free radical mechanism that is much more
complicated than initiator-based free radical polymerization and causes damage to occur to
functional groups during deposition.
22
Common issues in PECVD processing include dangling
bonds, double bonds, side reactions, loss in functionality, and uncontrolled crosslinking.
46
There
are several steps that can be taken to prevent this damage. Decreasing the W/FM parameter
(where W is the power of the plasma, F is the flow rate, and M is the precursor molecular
weight) can lead to improved films. Pulsing of the plasma or positioning the substrate
5

downstream of the plasma region can also be employed to limit the destruction of the functional
groups. For example, pulsating the plasma in PECVD can be used to produce films with
increased functionality ( ∼80%), however the growth rates are slow (<15 nm/min).
23,46

6

1.3 Initiated Chemical Vapor Deposition
Initiated chemical vapor deposition is a vapor phase free radical polymerization
technique that produces functional, conformal films using mild reactor conditions. Thin films can
be produced on any flat, curved, or porous substrate in a rapid, one-step reaction. The iCVD
technique requires very low energy, allowing the functional moieties of the monomers to remain
fully intact during the deposition process. The iCVD process is beneficial for a breadth of
applications and the process has also been scaled up in the form of a roll-to-roll reactor, which is
an important step toward commercial viability.
24

The iCVD process is typically performed in a pancake-shaped vacuum chamber as shown
in Figure 2. Monomer, initiator, and crosslinker molecules are introduced into the reactor
through heated lines that are connected to heated jars. Commercially available monomers are
generally used which eliminates the need to synthesize precursors. Flow rates are controlled
using mass flow controllers or by modulating the temperatures of the jars. The exhaust line to the
pump is located on the opposite side of the reactor to create a uniform flow over the substrate.
The top of the reactor is typically composed of quartz or glass which allows for visual inspection
during depositions and real time thickness measurements via interferometry. A resistively-heated
filament array is positioned above the substrate. Filaments composed of tantalum, tungsten,
stainless steel, and nichrome have been used.
25
The temperature of the stage is typically
maintained between 10 °C – 40 °C using a recirculated water chiller.
26
The pressure in iCVD
systems is kept constant throughout the deposition, generally between 100 –500 mTorr.
7

Figure 2. Schematic of a typical iCVD reactor. Polymerization occurs through the adsorbed monomer (M) on the
substrate due to the cleaving of the initiator (I) into free radicals (•).

The iCVD process proceeds via a free radical polymerization mechanism. The initiator
molecules decompose into primary free radicals when contacting the heated filament array.
Monomer molecules adsorb to the cooled substrate and chain propagation occurs on the surface.
No significant polymerization occurs in the gas phase.
27
A major difference between
polymerization in the iCVD process and traditional solution phase free radical polymerization is
that the temperature at which the initiator decomposes is independent of the polymerization
temperature. Thermally-decomposable initiators can be used while simultaneously coating
temperature-sensitive substrates such as fabrics or paper. A variety of initiators have been used
in the iCVD process including triethylamine (TEA),
28
perfluorooctane sulfonyl fluoride
(PFOS),
29
di-tert-amyl peroxide (TAPO),
30
and di-tert-butyl peroxide (TBPO).
31
The filament
temperatures required for pyrolysis are approximately 550 °C for TEA, 435° – 550 °C for PFOS,
285° – 310 °C for TAPO, and 180° – 250 °C for TBPO. The most common initiator used is
TBPO due to its low decomposition temperature since excessively high filament temperatures
can lead to partial decomposition of the monomer.
28
The filament temperature has a moderate
8

effect on the deposition rate with higher filament temperatures leading to increased deposition
rates.
30

The most important parameter in iCVD is the concentration of the monomer on the
surface of the substrate. The monomer concentration at the surface of the substrate is monotonic
with the value of P
m
/P
sat
where P
m
is the partial pressure of the monomer and P
sat
is the saturation
pressure. P
sat
is dependent on the temperature of the substrate and can be estimated using the
Clausius-Clapeyron equation. The concentration of the monomer at the surface can be increased
by either increasing the flow rate of the monomer or by decreasing the temperature of the
substrate. A P
m
/P
sat
value greater than one will lead to condensation on the substrate while P
m
/P
sat

values ranging between 0.4 – 0.7 are ideal.
30
This range allows a few monolayers to be present at
the surface for excellent deposition rates. The major type of termination reaction also depends on
the value of P
m
/P
sat
. At low values of P
m
/P
sat
, excess initiator will lead to primary radical
termination as the principle form of chain termination. However, at high values of P
m
/P
sat
, the
growing chains terminate via coupling or disproportionation reactions.
32

Lau and Gleason conducted a systematic study using various alkyl acrylates and found
the deposition rates and molecular weights of the polymers were highly dependent on the
concentration of the monomer at the surface of the substrate.
30
Lowering the stage temperature
increases the concentration of monomer at the surface of the substrate and thereby increases the
molecular weight and the deposition rate. Molecular weights are usually measured using gel
permeation chromatography (GPC).
27,33
For example, Lau et al. polymerized low molecular
weight poly(N-butyl acrylate) ranging from 2,000 to 15,000 g/mol.
32
Gupta et al. reported
molecular weights of 42,800 to 177,300 g/mol for poly(1 H, 1 H, 2 H, 2 H-perfluorodecyl
acrylate) (PPFDA)
34
and Bose et al. produced poly(2-hydroxyethyl methacrylate) (PHEMA) with
9

extremely high molecular weights of approximately 820,000 g/mol.
35
In all cases, the molecular
weight and deposition rate increased as the concentration of the monomer at the surface was
increased. Although deposition rates from 20 – 100 nm/min are most common,
25
rates as high as
375 nm/min and 1.5 µm/min have been observed for the polymerization of PFDA
34
and
HEMA,
35
respectively. Other factors such as rate coefficients of initiation, propagation, and
termination have a less significant effect on the deposition rate and molecular weight than the
monomer concentration at the surface.
30
Although no clear trend emerges for the effects of
monomer concentration on the polydispersity index, the values for the acrylates studied by Lau
et al. were between 1.3 and 2.2, which agrees with solution phase free radical polymerization
results.
30

10

1.4 Advantages of Vapor Phase Polymerization
Prior to vapor phase polymerization technologies, nonporous polymer films were created
by solution phase processes such as spin coating, solution casting, and dip coating. Surface
tension and gravitational effects limit these methods from conformally coatings substrates with
complex three-dimensional surfaces such as textiles and membranes. In addition, the use of
organic solvents leads to environmental, economic, and safety concerns. Solution phase
techniques rely on the miscibility of monomer, polymer, and other precursors in solvents and
therefore the amount of crosslinker and generality for a range of chemical moieties are limited.
Converting to a vapor phase polymerization system eliminates the need for solvents and
therefore makes the polymerization process safer and cleaner.
Vapor phase polymerization techniques have attracted considerable interest because the
lack of solvents in these techniques allows for the production of conformal nonporous films on a
wide variety of substrates. Typical vapor phase polymerization methods do not rely on solvents
because the precursors are kept below their respective saturation pressures in order to achieve
surface adsorption of the precursors to produce conformal polymer films.
36,37
An advantage of
using a solventless technique for surface modification is that typically insoluble polymers, such
as poly(1H, 1H, 2H, 2H-perfluorodecyl acrylate) (PPFDA)
34
and poly(furfuryl methacrylate),
38

can be deposited and incompatible monomers can be copolymerized.
39
The lack of solvents in
conjunction with the mild stage temperatures allows successful deposition of functional coatings
onto sensitive substrates for practical applications in textiles and medicine.
40,41
The mild stage
temperature allows sensitive substrates to be coated that would otherwise char or damage in
processes that require high temperatures.
11

In addition to sensitive substrates, many complex three-dimensional substrates such as
microstructures can be successfully coated. The surface modification of high aspect ratio
microstructures is important for the fabrication of microelectronic components,
42
membranes,
43,44

and biosensors.
45
It is difficult to uniformly coat complex three-dimensional surfaces using
solution phase polymerization due to inherent surface tension effects. In contrast, substrates with
complex topographies can be coated using CVD techniques. For example, the iCVD process has
been used to uniformly coat features such as trenches,
46,47
fibrous paper,
48
membranes,
49
multi-
wall carbon nanotubes,
50
and particles
51
since there are no clogging or dewetting issues in vapor
phase polymerization. For example, a trench coated using CVD exhibits a uniform layer of
polymer along the top, bottom, and sidewalls, whereas the trench coated via spin coating only
had polymer at the bottom of the trench due to surface tension effects.
46
The ability to
successfully deposit functional coatings onto temperature-sensitive substrates has many practical
applications in textiles and medicine.
40,41,52
In the iCVD process, the substrate is kept around
room temperature using a recirculating chiller. The mild stage temperature in the iCVD process
allows temperature-sensitive substrates to be coated that would otherwise char or damage in
other VDP or solution-phase techniques. For example, Kimwipes® were successfully coated
with poly(1H,1H,2H,2H-perfluorodecyl acrylate) (PPFDA) using iCVD without damaging the
underlying substrate (Figure 3). There was no damage during the coating process and the
hydrophobic coating repelled the water drops. The solventless nature of CVD processes is a
critical advantage over solution-phase polymerization.

12

Figure 3. Photographs of Kimwipes® before (left) and after coating with PPFDA (right).

Full retention of functional moieties is one of the most significant advantages of iCVD
over other CVD processes such as PECVD. In PECVD, the growth rate can be increased by
increasing the energy supplied to the system; however, the higher energy has a greater chance of
damaging the functional moieties on the polymer film. Improvements to the PECVD process
have been made by reducing the amount of power inputted to the system by pulsing the plasma
on and off, called pulsed PECVD or PPCVD. Films with increased functionality ( ∼80%) have
been produced by sacrificing the growth rate.
23
In contrast, films with 100% functionality can be
produced without compromise using iCVD. The iCVD process requires less power (65 mW/cm
2
)
than the PPCVD process (127 mW/cm
2
) and the deposition rates are much higher for iCVD
(>100 nm/min) than PPCVD (<15 nm/min).
46,53
The amount of control possible in CVD processes such as iCVD allow easy modification
to fit the requirements of the desired application, such as the need for a specific functionality or
crosslinking. The functionality of the film can be tuned by simply changing the precursor. New
films are constantly being deposited as novel precursors become commercially available. Films
with controllable crosslinking can be produced by flowing an additional precursor during the
deposition. If the precursor is a divinyl monomer, the crosslinking will occur during
13

polymerization. If a precursor containing reactive functional groups (such as an epoxide ring) is
introduced, crosslinking can be induced using heat or radiation after deposition has occurred.
54,55

The degree of crosslinking can be tailored by controlling the flow rate of the precursor. Several
notable crosslinkers used in the iCVD process are glycidyl methacrylate (GMA), ethylene glycol
diacrylate (EGDA), and ethylene glycol dimethacrylate (EGDMA). Mao et al. deposited
copolymers of GMA with 2,2,3,3,4,4,5,5,6,6,7,7-dodecafluoroheptyl acrylate (DFHA) and
PFEMA.
54
Once the deposition was complete, the film was placed in a 200°C vacuum oven to
break the epoxide rings and crosslink the film. The surface energies of the copolymer films
remained similar to the fluorinated homopolymer films despite being composed of 30-80%
GMA. Surface energy and XPS studies suggested that the fluorinated moieties existed in higher
concentrations at the surface of the film. The addition of GMA crosslinker improved the
modulus of the film by more than 7-fold despite a negligible effect on the surface energy. Two
examples of a divinyl crosslinker used in gas-phase polymerizations are EGDA and
EGDMA.
35,56
EGDA is commonly used for crosslinking PHEMA to make hydrogels both by
PECVD
57
and by the iCVD process.
58
The addition of crosslinker decreases the solubility and
increases the mechanical robustness of the PHEMA films.

14

1.5 Porous Polymers
Polymer films have led to significant advances in a wide range of applications such as
sensors,2 dielectric materials,
59
and water-repellent surfaces.
34
Porosity can be introduced into
these films in order to significantly increase their surface area and improve their usefulness;
60,61

for example, porous coatings greatly improve resolution in chemical and biological
sensors.
62,63,64
Adding porosity also helps to extend polymers towards applications such as
catalytic supports
65
and microreactors
66
, tissue scaffolding,
67,68,69
drug delivery,
70,71,72
and
separations.
73,74
Another consideration for a variety of applications is the need to use chemically
or physically sensitive substrates such as in microelectronics

or complex, nonplanar substrates
such as in tissue engineering.2
,75
The combination of flexibility, functionality, biocompatibility,
and simplicity in processing makes polymers a versatile option for many applications where
inorganic materials are insufficient.
76

The common methods for producing porous polymers require the use of solvents, which
both expose the substrates to potentially damaging conditions and make nonplanar coatings
difficult. For example, one of the most widely cited techniques in recent years is the breath figure
technique, which relies on solvent evaporation and water vapor condensations to produced
hexagonally-packed pores.
77,78,79
Other techniques for producing porous polymer materials are
solution casting and particulate leaching (SCPL),
80
cryopolymerization,
81,82,83
thermally induced
phase separation,
84,85,86
and copolymer self-assembly.
87,88,89
Only moderate success has been
achieved to use these techniques to conformally coat nonplanar substrates.
90,91
For example,
porous polymer has been formed on top of a sacrificial liquid layer, which is used to smooth out
a rough surface before being removed. However, this is only successful with certain star
15

polymers with a T
g
below 48 °C.
92
In addition, for some applications, such as dielectric
materials, it would be beneficial to fabricate the porous coatings in situ to form hierarchical
structures.
93,94

SCPL utilizes solid porogens to produce the desired porosity. In SCPL, insoluble
porogens, typically salts or sugars, are added to a polymer solution and then cast and allowed to
dry. Then the porogen is dissolved away, using a solvent (typically water) that will not affect the
polymer, creating regular porosity. For example Mikos et al
80
dissolved poly(L-lactic acid) in
chloroform and used sodium chloride, sodium tartrate, or sodium citrate salts to act as the
porogen. The polymer-salt composite was then immersed in water to dissolve and remove the
salt. Median pore diameters ranged from 29 to 116 μm depending on the size of the salt used.
This method provides relatively uniform closed pores in a simple solution phase technique.
One of the most diverse ways to produce porosity is create a stable polymer solution
before inducing phase separation (IPS) by relying on evaporation (EIPS), nonsolvents (NIPS),
vapors (VIPS), or thermal changes (TIPS).
84,85,86
For example in the TIPS method, polymer is
dissolved in an organic solvent at an elevated temperature, which increases the quality of the
solvent and makes it compatible with a wide range of polymers. The solution is then cast on a
cold substrate, which reduces the solubility of the system and causes phase separation.
Evaporation of the solvent leads to solidification of the porous structure. Lo et al
85
dissolved
poly(L-lactic acid) in 55 °C phenol or 85 °C naphthalene and cast these solutions onto a metal
plate. They achieved pore sizes between 20 and 500 μm by modifying the concentration and
molecular weight of their polymer precursor. In a slightly different method, Caneba et al
86
made
porous poly(methyl methacrylate) by dissolving the polymer in sulfolane before cooling the
16

solution to below 27 °C, the freezing point of the solvent, and then allowing the frozen solvent to
sublimate completely.
Another method to achieve porous structures is to use copolymer self-assembly. The
block copolymer solution is cast onto a surface where one of two techniques is used to produce
pores. In one case, the block copolymer film is allowed to form before being exposed to an
etching step which removes one of the blocks. This results in very regular nanometer-sized pores
that make them a very active area of research.
95,96,97,98
In the second case, placing the solution
casted film in humid conditions during the solvent evaporation leads to a secondary assembly
step.
89
Micrometer-sized pores are formed by the condensation of the water vapor onto the
surface of the evaporating polymer solution. The pores are forms when the water droplets, which
are enveloped by the polymer solution during the process, are evaporated. The end result is very
regular, hexagonally packed honeycomb pores. For example, Widawski et al.
87
observed that
poly(styrene-b-paraphenylene) produced 0.2 – 10 μm pores by solution casting the copolymer in
carbon disulphide under humid conditions and Yabu and Shimomura
89
prepared a fluorinated
copolymer with pore sizes between 0.5 – 50 μm using the same method.
Cryopolymerization is the process of polymerizing monomer solutions at low
temperatures and utilizes frozen solvent as the porogen. Cryopolymerization differs from the
previous techniques because the precursor is monomer instead of the polymer. Typically the
monomers are commercially available monomers and are mixed with crosslinkers and either
initiators or catalysts prior to freezing. Depending on the system, water and ethylene glycol are
also sometimes added. Polymerization is initiated before the temperature is reduced to freeze the
solvent to allow the solvent crystals to act as a porogen. However, this process usually takes
several hours to a day for significant conversions. For example, Savina et al
99
produced 5 – 100
17

μm pores in poly(2-hydroxyethyl methacrylate) by reacting 2-hydroxyethyl methacrylate with N,
N’-methylenebis(acrylamide), N, N, N’, N’-tetramethylethylenediamine, and ammonium
persulfate at -12 °C for 20 hours. Methacrylic acid and acrylic acid monomers can be
polymerized into porous structures by a similar method by using the monomer and UV radiation
as shown my Bamford et al.
100,101
However, this method is specific to these monomers,
maximum monomer conversion is approximately 30% and takes several days to complete, and
morphologies across the sample vary widely.
These techniques offer a wide range of porosity, pore size, and chemical functionality;
however, the need to meet solubility requirements restricts the generality and the degree of
control afforded by these methods. Increasingly, fabrication processes are moving towards
solventless methods in order to lessen the environmental and economic impact associated with
the use of organic solvents.
102,103,104
For example, controllable parylene nanostructures have been
created by flowing the precursors at oblique angles to the substrate.
105,106
High local partial
pressures lead to shadowing effects and the creation of angled pillars that can easily be made into
more complicated shapes, such as helices, by rotating the substrate.
107

Recently, two methods have been developed to produce porous materials via iCVD. Both
of these methods rely on increasing precursors past their respective saturation points, whereas
traditional thin film formation relies on maintaining the monomer partial pressure below the
saturation pressure. The first method is the basis of this dissertation and will be discussed at
length in the following chapters. For the second method, Anthamatten and coworkers used
simultaneous liquid monomer condensation and polymerization in conjunction with a porogen
species to produce their porous films.
108,109
In this technique, initiator, monomer, crosslinker, and
18

porogen are all introduced in the vapor phase and the condensation of the monomer and porogen
leads to the formation of porous polymer on the substrate by a phase separation mechanism.

19

Chapter 2: Experimental
2.1 Materials and Supplies
Methacrylic acid (MAA) (Aldrich, 99 %), N-isopropylacrylamide (NIPAAm) (Aldrich, 97
%), ethylene glycol diacrylate (EGDA) (Aldrich, 90 %), t-butyl peroxide (TBPO) (Aldrich, 98
%), 1,3-diaminopropane (Aldrich 99%), toluidine blue O (Aldrich, 80%), crystal violet (Aldrich,
90%), ponceau S (Aldrich, 75%), pH 4, pH 6, pH 7, and pH 8 buffer (BDH, ACS grade) were
used as received without further purification. Silicon wafers (Wafer World), chromatography
paper (Whatman, No. 1), medical gauze (Rite Aid), and stainless steel tubes (McMaster-Carr)
were utilized as substrates.
The iCVD system vacuum chamber (GVD corporation) is a custom-designed pancake-
shaped reactor (250 mm diameter, 48 mm height) evacuated by a rotary vane pump (Edwards
E2M40). The pressure is kept constant using a throttle valve (MKS 153D) and measured using a
capacitance manometer (MKS Baratron 622A01TDE). Precursors flows are controlled using
needle-valves or mass flow controllers (MKS 1479A). A nichrome filament array (Omega
Engineering, 80 %/20 % Ni/Cr) is used to cleave the initiator into free radicals. Substrates are
placed on top of a thermoelectric cooler (TE Technology). The thermoelectric cooler is placed on
the chamber stage which temperature is maintained by a water-based recirculating chiller. The
thermoelectric cooler was then equilibrated to the desired temperature by controlling the input
current via a DC power supply (Volteq HY3010D).

20

2.2 Procedures
The general procedure for fabrication porous PMAA membranes was the following.
TBPO was maintained at 23 °C and fed into the chamber at a flow rate of 7.8 mg/min using a
mass flow controller. This flow was used to maintain the desired reactor pressure. A nichrome
filament array was then resistively heated to 250 °C, which thermally cleaved the initiator into
free radicals. The thermoelectric cooler was then equilibrated to the desired temperature by
controlling the input current of the power supply. The system was allowed to equilibriate for at
least one minute. Finally, the MAA was introduced into this system. The MAA source jar was
maintained at 23 °C and the flow rate was controlled by using a needle valve to meter the flow.
In order to measure accurate flow rates for the TBPO and MAA species, we weighed the
precursor jars before and after each deposition. After the reaction proceeded, the precursor flow
was stopped, the filament array was turned off, and the sample was allowed to pump down at 0 –
5 °C until all the excess monomer sublimated, which was determined to be when the system
returned to base pressure.
Overall, TEC temperatures ranged between -25 °C and 10 °C, reactor pressures ranged
between 150 mTorr and 450 mTorr, MAA flow rates varied between 45 mg/min and 120
mg/min, and the deposition time varied from 6 seconds to 90 minutes. To study the structural
properties of the porous films as a function of the deposition time, the substrate temperature was
maintained at -10 °C, the reactor pressure was maintained at 315 mTorr, and the deposition times
ranged from 2 minutes to 90 minutes. To study the structural properties of the porous films as a
function of monomer partial pressure, the substrate temperature was maintained at -10 °C, the
deposition time was 20 minutes, and the reactor pressure ranged from 220 mTorr to 410 mTorr.
To study the structural properties of the porous films as a function of the substrate temperature,
21

the reactor pressure was maintained at 315 mTorr, the deposition time was 20 minutes, and the
substrate temperature ranged from 0 °C to -15 °C.
To study MAA deposition in the absence of polymerization, MAA and TBPO precursors
were introduced at a pressure of 300 mTorr and substrate temperatures between -25 °C and 10
°C with the filament unheated. The MAA flow rate was 45 mg/min, 80 mg/min, or 120 mg/min
while the TBPO flow rate was 7.8 mg/min. In this case, only the physical deposition of MAA
occurred as no free radicals were present to initiate polymerization. The TBPO flow was
included to maintain the same conditions that we used to fabricate porous PMAA membranes.
After the deposition, the sample was quickly imaged using a camera to measure the thickness. In
order to account for any small variations in wafer size, mass deposition was reported as a mass
per unit area. The monomer deposition on the silicon wafer was weighed using a balance and the
area of the silicon wafer was measured with digital calipers. For mass per unit area and thickness
measurements, each data point represents one membrane.
The pressure dependence of the freezing point of MAA was determined experimentally.
A thermocouple immersed in 5 mL of liquid MAA in a glass Petri dish was used to measure the
temperature at which the phase transition was visually observed at 150 mTorr, 600 mTorr, and
760 Torr, with the pressure maintained by inert nitrogen. The MAA liquid was cooled down
from 23 °C using the thermoelectric cooler until the formation of solid crystals appeared. The
freezing point of MAA appeared to be 16 ± 1 °C and was independent of pressure over the range
studied. The temperature was cycled between 14 °C and 18 °C several times to test
reproducibility. For the case of NIPAAm, the literature value of the freezing point (63 ± 1 °C)
was used and was assumed to be independent of pressure over the range studied.
22

For the fabrication of poly(N-isopropylacrylamide) (PNIPAAm) porous films, the
monomer was maintained at a flow rate of 0.5 sccm, the substrate temperature was maintained at
30 °C, the reactor pressure was maintained at 315 mTorr, and the deposition time was 20
minutes. For the fabrication of poly(methacrylic acid-co-ethylene glycol diacrylate) (P(MAA-co-
EGDA)) porous films, the MAA and EGDA monomer flow rates were maintained at 4.0 sccm
and 0.1 sccm, respectively, the substrate temperature was maintained at -10 °C, the reactor
pressure was maintained at 315 mTorr, and the deposition time was 20 minutes.
Thin nonporous PMAA and PNIPAAm reference films were fabricated for controls. The
reference polymer films were deposited onto silicon wafers under conditions that prevented solid
monomer deposition such that thin (~ 400 nm), smooth films were formed instead of porous
films. The PMAA reference was deposited at a substrate temperature of 20 °C whereas the
PNIPAAm reference was deposited at a substrate temperature of 30 °C.
For the fabrication of P(MAA-co-MAN) membranes, samples were placed in a
conventional oven (VWR) at 185 ± 2 °C for 5 – 5000 minutes. A nitrogen-purged vacuum oven
was used for the annealing of the membranes on medical gauze in order to avoid oxidation of the
gauze. For the fabrication of crosslinked samples, the P(MAA-co-MAN) membranes on silicon
wafers were placed in a closed container with 1 mL of 1,3-diaminopropane (DAP) at room
temperature for 3 – 72 hours. Samples were placed under vacuum in a desiccator overnight to
remove excess DAP. Control silicon wafers were included to ensure that any DAP condensate
was removed prior to analysis. Control PMAA membranes were also included to confirm that the
DAP primarily reacted through the MAN moieties.
23

To produce substrates with patterned surface energy, silicon wafers were masked with
polyimide tape and then coated with 30 nm of dense poly(1H,1H,2H,2H-perfluorodecyl acrylate)
films using the iCVD technique. The coating was formed at 30 mTorr reactor pressure and a 30
°C stage temperature with 1H,1H,2H,2H-perfluorodecyl acrylate and TBPO flow rates of 11.7
mg/min and 13.9 mg/min, respectively. The tape was then removed and porous poly(methacrylic
acid) was deposited onto the patterned substrate using the above flow rates and substrate
temperatures.
Paper-based microfluidic devices were fabricated by printing wax (Xerox Phaser 8560N)
onto chromatography paper and subsequently heating the paper to 180 °C for 3 minutes to melt
the wax through the depth of the paper.
110,111
The ability of porous P(MAA-co-EGDA)
membranes to selectively separate cationic analytes was analyzed by depositing porous P(MAA-
co-EGDA) membranes for 10 minutes at the inlet of paper-based microfluidic channels using a
stainless steel mask, as described above. Three µL of a buffered pH 6 solution containing 2
mg/mL crystal violet and 0.25 mg/mL ponceau S was then applied to the inlet of the channel and
allowed to flow through the device in ambient conditions. Separation of the dyes was performed
using three separate channels to confirm the result.
Dyeing of the porous P(MAA-co-EGDA) membranes was performed by immersing the
samples in a 0.001 wt% toluidine blue O in buffered pH 8 solution for 12 hours. The membranes
were then washed three times by immersion in a buffered pH 8 solution for a total of 36 hours.
The membranes were then allowed to dry in ambient conditions and scanned using a color
desktop printer (HP Deskjet F4480). The images were converted to gray scale and line intensities
were gathered using ImageJ. The intensities were normalized against unmodified
chromatography paper treated with toluidine blue O in the same manner as described above.
24

Intensities were averaged over 2 samples with 2 scans per sample taken perpendicular to each
other.

25

2.3 Characterization
The chemical composition of the porous films was confirmed using FTIR (Thermo
Nicolet iS10). The samples were removed from the reactor and analyzed immediately. The
spectra were baseline corrected and normalized based on the carbonyl peak. For peak
deconvolution, the spectra were imported into CasaXPS software (version 2.3.15) for peak fitting
analysis. Weight percentages were determined from the relative areas of the MAA and MAN
carbonyl peaks and their respective molecular weights of 86.06 g/mol and 154.10 g/mol.
Dynamic light scattering (Wyatt DynaPro Titan) was used to estimate the molecular weight of a
PMAA membrane formed at a -20 °C substrate temperature.
The structural properties of the porous films were studied using a scanning electron
microscope (SEM) (JEOL-7001, JEOL-6610, Topcon Aquila) at a 10 kV accelerating voltage.
Gold or platinum was sputtered onto the samples for 30 seconds prior to imaging. SEM images
were used to measure the thickness of the samples and to estimate the pore size. For the pore size
analysis, the images were processed using ImageJ (version 1.44p) as recommended by NIST,
112

and pore sizes represent pore diameters under the assumption of circular pores. The estimates
were made from an area of approximately 1 mm
2

from top-down scanning electron micrographs
of the interior of the membranes
Mercury intrusion porosimetry was performed using an AutoPore IV 9500
(Micromeritics) for more accurate measurements the pore size distribution, porosity, mean pore
diameter, total intrusion volume, and total pore area of the membranes. A penetrometer volume
of 5.2 mL was used and the pressure was increased from a filling pressure of 0.99 psia up to
50000 psia.
26

Water contact angle measurements were performed using a contact angle goniometer
(Ramé-hart 290-F1) with drop sizes of 10 μL, where error bars represent the standard deviation
of 3 measurements on a given sample. The dissolution time of the membranes was studied using
an overhead webcam (IPEVO Ziggi HD) and ImageJ software (version 1.44p). While soaking in
either pH 4 or pH 7 buffer, images of the membranes were taken every five minutes and
analyzed. The dissolution time was determined as the point at which there was a 50% reduction
in the brightness value in the ImageJ analysis corresponding to the loss of the majority of the
membrane mass.
For the monomer deposition study, the sample was quickly imaged using a camera to
measure the thickness. The monomer deposition on the silicon wafer was weighed using a
balance and the area of the silicon wafer was measured with digital calipers. In order to account
for any small variations in wafer size, mass deposition was reported as a mass per unit area.

27

Chapter 3: Fabrication of Porous Polymers
3.1 Requirements
In traditional thin film processing, the monomer partial pressure is usually kept between
40 % and 70 % of its saturation pressure to achieve high growth rates while maintaining film
uniformity. Exceeding the saturation pressure causes the monomer vapor to condense as a liquid
or deposit as a solid depending on the substrate temperature. While simultaneous condensation of
liquid monomer and polymerization results in nonporous films,
108
here we show that
simultaneous deposition of solid monomer and polymerization can be used to fabricate porous
polymer films.
113
Thus for the production of porous films, two requirements must be met: the
monomer partial pressure must be above its saturation pressure at the given substrate temperature
and this substrate temperature must be below the freezing point of the monomer, as shown in
Figure 4. In order to achieve a larger range and better control of the substrate temperature, the
iCVD system was modified with a thermoelectric cooler (TEC). Substrate temperatures as low as
-20 °C were reached using the TEC, which allowed for the monomer partial pressure and
substrate temperature requirements could be achieved for a variety of monomers.
28

Figure 4. Flow chart showing the different structures possible (SEM images) in vapor phase polymerization.
Monomer partial pressures above the monomer saturation pressure and substrate temperatures below the freezing
point of the monomer are required for porous film formation.

Poly(methacrylic acid) (PMAA) was used as a model system for much of this work due
to the relatively high freezing point of methacrylic acid (MAA) and the pH-responsive nature of
PMAA,
114,115,116
which make the polymer both easy to process and functional. The freezing point
of MAA was experimentally determined to be insignificantly affected by pressure under all
reaction conditions used in this study, and therefore was considered to be approximately 16 °C ±
1 °C. The substrate temperatures were varied between 0 °C and -20 °C and the saturation
29

pressures of MAA at these substrate temperatures were calculated using the Clausius–Clapeyron
relation.
34
Note that the substrate temperatures are below the freezing point of MAA and the
monomer partial pressures are above the monomer saturation pressures under all experimental
conditions, satisfying the requirements for successful porous film fabrication.
The general procedure for the fabrication of porous polymers using the iCVD technique
is as follows. First the TBPO is introduced into the system and used to equilibrate the reactor at
the desired pressure. Then the filament is heated to 250 °C and the TEC is set to the desired
substrate temperature. The MAA is introduced which leads both to solid monomer deposition on
the substrate and the partial polymerization of the MAA over the duration of the given deposition
time. At this point the sample consists of both MAA and PMAA. Then all the flows are stopped
and the reactor is allowed to pump down to base pressure with the substrate between 0 – 5 °C in
order to sublimate the remaining MAA. The removal of the unreacted monomer was detected by
significant outgassing during the pump down period, leading to a 20 – 50 mTorr increase above
the base pressure of the reactor. The outgassing persisted for several minutes, after which the
porous film was assumed to be free of monomer. To further verify that the outgassing was
caused by the sublimation of solid monomer, Fourier transform infrared spectroscopy (FTIR)
was used to study the chemical composition of the porous films before and after pump down.
Figure 5 shows that the spectrum of the sample prior to pump down is similar to the spectrum of
pure MAA monomer, indicating that a large amount of unreacted monomer is present in the
sample. After pump down, the peak attributed to the C=C stretching at 1636 cm
-1
is no longer
present, which confirms the removal of unreacted monomer. The spectrum of the final PMAA
porous film matches reference PMAA thin films, confirming polymerization. The expected
absorbance peaks were observed for PMAA:
117
broad O-H stretching between 2500-3300 cm
−1
,
30

asymmetric −CH
3
stretching at 2993 cm
−1
, asymmetric −CH
2
− stretching at 2944 cm
−1
,
carboxylic acid C=O stretching at 1699 cm
−1
, −CH
3
and −CH
2
− deformation at 1455 cm
−1
, and
−CH
3
deformation at 1389 cm
−1
.

Figure 5. FTIR spectra of the MAA monomer, the sample before and after sublimation, and the PMAA thin film
reference. The strong peak attributed to C=C stretching of the monomer at 1636 cm
-1
(dashed line) is no longer
present after sublimation, at which point the porous film is free of MAA monomer and is chemically identical to the
thin PMAA reference film.

The simultaneous solid monomer deposition and partial polymerization leads to
membranes with dual-scale porosity, where large-scale pores were formed during the deposition
process and small-scale pores were formed when the remaining monomer was sublimated. We
demonstrate that the solid monomer serves as a template for polymerization and therefore the
final structure of the membranes can be tuned by controlling the physical deposition of the
monomer. Although the sublimation step caused a reduction in the thickness of the sample due to
the loss of the structural support, the overall morphology is derived from by the monomer
template.
31

Figure 6. Schematic of porous film fabrication showing the growth on a molecular and macroscopic scale. Cross-
sectional SEM images showing the large-scale pores between (left) and small-scale pores within (right) the
microstructures.

Mercury porosimetry was used to characterize the pore size and porosity of the membranes both
before and after annealing. Due to the dual-scale porosity of the membranes, the pore size
distribution (Figure 7) is very broad with large-scale pores roughly from 5 microns to greater
than 200 microns in size and small-scale pores roughly 100 nm to 5 microns in size, which is
consistent with our SEM images, such as those in Figure 6. The membranes were found to be
highly porous before and after annealing, with porosity values above 90 %, total intrusion
volumes above 8.8 mL/g, and total pore areas above 71.9 m
2
/g (Table 1).
Table 1. Mercury intrusion porosimetry data.

Porosity
(%)
Mean Pore
Diameter (µm)
Total Intrusion
Volume (mL/g)
Total Pore
Area (m
2
/g)
PMAA 90.8 17.8 12.2 71.9

32

Figure 7. Mercury porosimetry data showing the pore size distribution.

33

3.2 Mechanism
Porous polymer formation was caused by two processes that are occurring
simultaneously: solid monomer deposition and polymerization. In the absence of polymerization,
the MAA deposits as solid interconnected pillar-like microstructures that become larger over the
course of the deposition. These microstructures result from unstable growth caused by deposition
conditions far from the monomer saturation point.
118
These microstructures act as a template
upon which the polymerization can occur and can be visually seen as a structured opaque film
during the deposition process. They have void spaces between them on the order of tens to
hundreds of microns in diameter. This film sublimated completely during the pump down period,
leaving bare wafer and indicating that the deposited film was composed of solid monomer. Due
to the smooth nature of the walls of the microstructures, we assume that significant vapor phase
nucleation during the majority of the porous film growth is not occurring.
119

The solid monomer serves as a template for polymerization and therefore understanding
the behavior of the physical deposition of the monomer is critical for understanding the
membrane growth process.
120
The iCVD process is typically used to make thin dense films by
maintaining the monomer partial pressure (P
m
) below the saturation pressure (P
sat
). Gleason and
coworkers found that the P
m
/P
sat
ratio is the most important parameter governing the deposition
rate for traditional iCVD processes.
30,121,34
At values less than one, the concentration of monomer
adsorbed onto the surface of the substrate monotonically increases with the P
m
/P
sat
ratio
according to the Brunauer-Emmett-Teller (BET) isotherm and leads to an increase in the growth
rate of the dense polymer films. In contrast, our technique requires physical monomer deposition
and therefore we are using P
m
/P
sat
ratios above one. At these unconventional conditions, our
initial experiments indicated that relying on the P
m
/P
sat
ratio alone does not adequately explain
34

our results. We chose to deconvolute the P
m
/P
sat
ratio by independently studying the effects of
the monomer flow rate and substrate temperature.
We first explored how the mass and thickness of the deposited monomer are affected by
the substrate temperature and monomer flow rate in the absence of polymerization (i.e. the
initiator was flowing, but the filament was not turned on). We found that the mass of monomer
deposition is highly dependent on the flow rate of MAA in the system as seen by an approximate
four-fold increase in the mass per unit area when the flow rate is increased from 45 mg/min to
120 mg/min (Figure 8a). It is important to note that the small variation of the monomer partial
pressure caused by increasing the flow rate does not significantly impact the mass of monomer
deposition. In contrast to the flow rate study, the amount of monomer deposition only shows a
slight decrease when increasing the substrate temperature from -25 °C to 10 °C. The large
dependence on the monomer flow rate compared to the substrate temperature indicates that the
deposition process under this range of conditions is limited by the amount of monomer
introduced into the reactor. We see a larger reduction in the mass of monomer deposition at
substrate temperatures above 5 °C due to incomplete coverage on the substrate. In the case of the
45 mg/min and 10 °C sample, no solid MAA deposits because the MAA partial pressure is
approximately equal to its saturation pressure.
35

Figure 8. a) Mass and b) thickness data for MAA deposition at a range of substrate temperatures and MAA flow
rates.

We observed a different trend when studying the effects of the substrate temperature and
monomer flow rate on the sample thickness. Although we continued to see a high dependence on
the flow rate of MAA, we now observed a strong decrease in the thickness when increasing the
temperature of the substrate, especially at high monomer flow rates (Figure 8b). For example, at
a MAA flow rate of 120 mg/min, there is a large thickness reduction of approximately 70 %
when increasing the substrate temperature from -20 °C to 0 °C despite the deposited mass
36

remaining nearly constant. By analyzing the images of the MAA deposition (Figure 9), we found
that these large thickness differences are due to significant variations in the morphologies. The -
20 °C substrates had widespread nucleation and consisted of tall, slender MAA microstructures
that grow vertically on the substrate. The lower substrate temperature reduces the surface
diffusion of MAA and increases the shadowing effects that cause microstructure formation. As
the substrate temperature is increased, nucleation is reduced, individual microstructures become
less-defined, and monomer deposition begins to resemble two-dimensional growth. Above 5 °C,
the combination of a mass reduction as discussed previously with lower nucleation results in
web-like monomer growth that does not completely cover the substrate.

Figure 9. Optical cross-sectional images show that the substrate temperature has a large impact of the morphology
on the depositing MAA.

In order to determine how the monomer deposition is affected during simultaneous
polymerization, we conducted the same experiments as above except we heated the filament to
250 °C to break the TBPO into free radicals to initiate polymerization. We will refer to these
37

samples as the composite samples due to the presence of both the MAA monomer and the
PMAA polymer. The composite results show the same trends as the monomer deposition study
(Figure 10). The mass of monomer deposition has a large dependence on the MAA flow rate and
a slight dependence on the substrate temperature, while the thickness is significantly dependent
on both parameters. However when the values of the two studies are compared, there is a slight
decrease in the mass and a significant decrease in the thickness in the composite samples. These
reductions are consistent with additional heat from the filament array warming the top surface of
the sample. We tested for effects from polymerization by replacing the initiator with an inert
nitrogen gas such that we could heat the filament array without initiating polymerization. We
found that this system behaved similarly to when polymerization is occurring, indicating that the
heat from the filament array has a more significant impact on the monomer deposition than the
polymerization process.
38

Figure 10. a) Mass and b) thickness of the composite samples at a range of substrate temperatures and MAA flow
rates.

The final step in the formation of the membranes is the sublimation of the unpolymerized
monomer remaining in the system. Figure 11a shows the porous polymer membrane thickness as
a function of substrate temperature at various MAA flow rates. As expected, higher MAA flow
rates and lower substrate temperatures result in significantly thicker samples. The monomer,
composite, and polymer samples at a 120 mg/min MAA flow rate are compared in Figure 11b.
Compared to the monomer system, there is a thickness reduction due to the heat of the filament
39

array in the composite samples as discussed above. There is also a large thickness loss when
comparing the composite samples to the porous PMAA membranes. This is likely caused by the
compression of the membrane once the structural support of the unpolymerized MAA is
removed by sublimation. Despite these thickness losses, it is clear that the trends in the porous
polymer membrane thickness are determined by the deposition of the MAA template.

Figure 11. a) Thickness of the final porous PMAA membrane samples at a range of substrate temperatures and
MAA flow rates. b) Thickness comparison of the monomer, composite, and polymer samples at a 120 mg/min MAA
flow rate showing the thickness loss of the samples caused by the heated filament and removal of the MAA
template.
40

As stated earlier, this technique is able to produce hundreds of microns of porous PMAA
over the course of twenty minutes due to the solid MAA deposition creating high porosities.
113

Although this allows for relatively quick and easy fabrication of thick free-standing membranes,
the high deposition rate makes it difficult to create thinner porous coatings. We found that we
needed approximately 10 minutes of simultaneous MAA deposition and partial polymerization in
order to form enough polymer to achieve a cohesive material, which limited thicknesses to above
several hundreds of microns. In order to overcome this issue for the creation of thinner porous
coatings, the amount of polymerization needed to be significantly increased. Due to the nature of
vapor phase processing, we were able to introduce or remove precursors at will during the
deposition. We therefore decided to deconvolute the time allowed for the physical deposition of
MAA and the time allowed for polymerization. This allowed us to study the ability to increase
the polymerization after the deposition process is complete by continuing to flow free radicals
into the system. We found that if we continued to introduce free radicals to the system while the
MAA was sublimating, we could get much more cohesive coatings. For example in Figure 12,
coatings formed after 6 seconds, 30 seconds, and 2 minutes of deposition with and without an
additional 10 minutes of polymerization are compared. The sample with the additional
polymerization exhibited a better cohesiveness therefore resulted in a better coating. These
results will be discussed further in Chapter 7.
41

Figure 12. SEM images showing the effect of adding a post-deposition polymerization step to the structure and
cohesiveness of the porous coatings at short deposition times.

42

Chapter 4: Morphological Control
4.1 Deposition Time
Morphological control is essential in order to understand the range of structures that can
be produced. One way to vary the morphology is to vary the deposition time. Top-down and
cross-sectional images of the porous films were taken to analyze the thickness growth rate and
the diameter of the large-scale pores formed between the microstructures as deposition time
increased from 2 – 90 minutes (Table 2), as shown in Figure 13. During the growth process, the
monomer deposits into pillar-like microstructures due to the monomer being well above the
saturation pressure. Variations in deposition cause roughness in the deposited surface and
shadowing effects cause those variations to grow into microstructures.
122
The sample deposited
for 2 minutes is a relatively even porous film whereas well-defined microstructures are grown by
20 minutes. We would like to study two processes related to this microstructure growth. First, we
want to explore the closing of the void space between the microstructures at high deposition
times. Second, we look to investigate whether we can prevent the formation of microstructures
throughout the entire deposition to produce a smoother porous film.
The thickness growth rate increased to 15.3 ± 1.2 µm/min within 20 minutes and
remained constant thereafter. The large-scale pore size increased from 8 ± 5 µm for porous films
formed after 2 minutes to 61 ± 7 µm for porous films formed after 20 minutes due to the merger
of smaller microstructures into fewer, thicker microstructures. After 20 minutes, the pore size
decreased as densification of the porous film began to dominate. The densification process
occurs by filling the void space between the microstructures. Densification of vapor deposited
porous structures has been widely studied for the process of frost formation,
123,124
where a
reduction in growth rate induces densification as additional vapor deposits within the porous
43

structure rather than contributing toward an increase in thickness. This effect is caused by a
thermal gradient between the cold substrate surface and the warmer top surface of the depositing
species. Initially, our system did not appear to undergo the same type of densification based on
the growth rate staying constant during the process. However, from our previous studies we
know that the growth rate of the final membrane may be convoluted by either the monomer
deposition or monomer polymerization, which are occurring simultaneously. We know that at
short deposition times, there is not enough polymerization to maintain a full height of the
structure. This will lead to vastly underestimated growth rates at short time and likely have
hidden the evidence that the thickness growth rate is slowing with time and will eventually stop.
At this point in our system, we hypothesize that the temperature of the top surface of the MAA
will be warm enough that its saturation pressure will equal the partial pressure in the reactor. In
this case, lowering the substrate temperature or increasing the partial pressure should allow for a
thicker membrane without densification.

Table 2. Time Study and Resultant Growth Rates of PMAA Porous films
Sample
Deposition
Time
(min)
Substrate
Temperature
(°C)
Reactor
Pressure
(mTorr)
Monomer
Partial
Pressure
(mTorr)
Thickness
Growth
Rate
(μm/min)
T1 2 -10 315 268 10.2 ± 4.2
T2 10 -10 315 268 10.8 ± 3.5
T3 20 -10 315 268 15.3 ± 1.2
T4 30 -10 315 268 15.4 ± 2.3
T5 60 -10 315 268 15.4 ± 4.4
T6 90 -10 315 268 15.3 ± 1.7

44

Figure 13. a) Top-down SEM images with inset cross-sectional SEM images showing the microstructure and growth
of PMAA porous films as a function of deposition time. Graphs showing b) thickness growth rate and c) large-scale
pore size.

In order to test this theory, we compared the morphologies of the previous time series to a
new series with varying growth rates. In the latter series, the growth rate was varied by varying
the reactor pressure in order to match the height of the former series while maintaining a
deposition time of 20 minutes. The morphology of the former study showed that the
densification process started to occur at around 20 minutes, which correlated to approximately
300 µm. In contrast, analysis of the latter study shows no significant densification even at
thicknesses as high as 810 µm (Figure 14). SEM images of the top and cross-section show that
the microstructures and void space are larger with increasing monomer partial pressures.
45

Specifically, comparing the high growth rate 810 μm sample to a similar thickness porous film
produced at a lower deposition rate (920 μm, 60 minutes) shows signification morphology
differences (Figure 15), which supports that the morphology is significantly impacted when the
saturation pressure of the growing MAA surface approached the partial pressure in the reactor.

Figure 14. Top-down (a-e) and cross-section (f-j) images SEM images of PMAA samples grown for 20 minutes at
a,f) 7.1, b,g) 9.5, c,h) 15.3, d,i) 19.5, and e,j) 40.5 µm/min, resulting in the following respective thicknesses: 142,
190, 306, 390, and 810 µm. Unlabelled scale bars represent 200 µm.

Figure 15. Top-down SEM images comparing a) a 920 µm sample grown for 60 minutes (15.3 µm/min) to b) an 810
µm sample grown for 20 minutes (40.5 µm/min).
46

4.2 Substrate Temperature
The growth of the porous films can be tuned by varying the parameters which most
significantly affect solid monomer deposition: monomer partial pressure and substrate
temperature.
125
In this study we determined the effects of the substrate temperature (Table 3) on
the growth (Figure 16) and morphology of the porous PMAA samples. An optical microscope
was used to image top-down views (Figure 17) and SEM was used to image cross-sections
(Figure 18) of the membranes. We found that the low temperature samples are composed of
well-defined, high aspect ratio pillar-like microstructures similar to those seen optically in the
monomer deposition study. Increasing the substrate temperature leads to less defined
microstructures and two-dimensional growth. Once the substrate temperature is above 5 °C, the
membranes do not completely cover the entire substrate due to the limited nucleation and web-
like growth of the MAA template. Under all conditions, the membranes exhibit two sets of pores
that occur by different mechanisms. A larger set of pores forms from the void space between the
growing microstructures of the monomer template, which are on the order of tens to hundreds of
microns. The smaller pores form when the unreacted MAA is sublimated after the deposition is
complete and are on the scale of hundreds of nanometers to several microns (Figure 19).
Dynamic light scattering was used to determine the molecular weight of a sample formed at a
substrate temperature of -20 °C to confirm the presence of high molecular weight polymer at this
low substrate temperature. The molecular weight was found to be 148 ± 23 kDa, which is
comparable to the molecular weight of polymer formed during iCVD depositions near room
temperature.
51

47

Table 3. Reactor conditions and resultant growth rates varying substrate temperatures.
Sample
Deposition
Time
(min)
Substrate
Temperature
(°C)
Reactor
Pressure
(mTorr)
Monomer
Partial
Pressure
(mTorr)
Thickness
Growth
Rate
(μm/min)
S1 20 0 315 268 12.2 ± 1.5
S2 20 -5 315 268 12.6 ± 1.2
S3 20 -10 315 268 15.3 ± 1.2
S4 20 -15 315 268 23.4 ± 0.8

Figure 16. Porous film thickness growth rate increases with decreasing substrate temperature.

Figure 17. Top-down optical images showing the morphology change caused by varying the substrate temperature at
a constant MAA flow rate of 80 mg/min.
48

Figure 18. Cross-sectional SEM images showing the morphology of the porous polymer membranes at varying
MAA flow rates and substrate temperatures.

Figure 19. Higher magnification SEM images show the presence of small-scale pores for the high MAA flow rate
samples at a) -20 °C and b) 0 °C.

49

4.3 Partial Pressure
Increasing the monomer partial pressure (Table 4) results in a larger driving force of
monomer deposition and hence an increase in thickness growth rate. For example, Figure 20a
shows a monotonic increase in the thickness growth rate from 7.1 ± 1.3 µm/min to as high as
40.5 ± 1.9 µm/min as the monomer partial pressure was increased from 187 mTorr to 349 mTorr,
respectively, while keeping all other variables constant. Decreasing the substrate temperature
also increases the driving force of monomer deposition. Figure 20b shows that the thickness
growth rate monotonically increased from 12.2 ± 1.5 µm/min to 23.4 ± 0.8 µm/min as the
substrate temperature was decreased from 0 °C to -15 °C, respectively, while keeping all other
variables constant. Our results demonstrate that the thickness growth rate can be tuned even over
modest pressure and temperature ranges.

Table 4. Reactor conditions and resultant growth rates varying MAA partial pressures.
Sample
Deposition
Time
(min)
Substrate
Temperature
(°C)
Reactor
Pressure
(mTorr)
Monomer
Partial
Pressure
(mTorr)
Thickness
Growth
Rate
(μm/min)
P1 20 -10 220 187 7.1 ± 1.3
P2 20 -10 265 226 9.5 ± 0.4
P3 20 -10 315 268 15.3 ± 1.2
P4 20 -10 360 306 19.5 ± 2.4
P5 20 -10 410 349 40.5 ± 1.9

50

Figure 20. Porous film thickness growth rate increases with increasing monomer partial pressure.

The reactor pressure under which sublimation occurs was varied in order to study its
effect on the final porous film structure. These experiments were performed under the same
conditions with the pump down pressure either at 80 mTorr (fully open to vacuum; no nitrogen
flow) or maintained at the deposition pressure of 315 mTorr (with nitrogen flow). In both cases,
the monomer was fully sublimated within 30 minutes. We observed no structural differences in
either the void spaces found between the microstructures (large-scale pores) or the small-scale
pores found within the microstructures (Figure 21). Therefore the porous films formed for our
remaining experiments were pumped down with no nitrogen flow for simplicity.

Figure 21. Cross-sectional SEM images of porous films resulting from pump down at a) 80 mTorr and b) 315 mTorr
(with inert nitrogen flow), respectively.
51

4.4 Surface Energy
Our studies demonstrate that PMAA membrane growth and structure is controlled by the
deposition of the MAA monomer template. We wanted to use this relationship to pattern the
membrane growth by selectively depositing monomer in certain locations on the substrate. One
way to control nucleation is to vary the surface energy of the substrate. For example, Vo et al.
observed morphological differences when growing aluminum-doped zinc oxide crystals from the
vapor phase on substrates with varying surface energies.
126
In addition, Tang et al. found that
fluorinated surfaces improved the anti-frosting performance of surfaces.
127
We explored how the
surface energy of our substrates impacted our final polymer membranes by patterning a low
surface energy fluorinated polymer, poly(1H,1H,2H,2H-perfluorodecyl acrylate) (PPFDA), on
the surface of silicon wafers. The surface energy of the PPFDA coating is 13.6 mN/m
128
whereas
the surface energy of the silicon wafer is approximately 100 mN/m.
129
The fabrication of PMAA
membranes on these substrates results in a significant morphology change at the interface of the
fluorinated and non-fluorinated boundary. On the bare silicon surface, the membrane formation
occurs in the same manner as described above, leading to complete coverage of porous PMAA
on the silicon surface. In contrast, the fluorinated substrate experiences minimal MAA
nucleation, which leads to a dense as opposed to a porous PMAA coating over a majority of the
area. These two morphologies are evident in the patterned sample shown in Figure 22. The
success of this patterning is dependent on the relative area of PPFDA present, as well as the flow
rate of the MAA. When there is a minimal amount of bare silicon or the MAA flow rate is
sufficiently high, widespread MAA nucleation occurs and leads to the production of porous
PMAA membranes on the PPFDA surface.
52

Figure 22. a) Optical top down image showing the (2) widespread porous PMAA on the bare silicon substrate
compared to (1) isolated growth on the fluorinated substrate. b)Top-down and c) cross-sectional SEM images of the
sample from the outlined region.

53

Chapter 5: Chemical Control
5.1 Generality
Porous films composed of poly(N-isopropylacrylamide) (PNIPAAm) were fabricated to
show the generality of our technique. were fabricated. PNIPAAm has many potential uses due to
its temperature-responsive hydrophilicity.
130,131
The N-isopropylacrylamide monomer has a
freezing point of 63 ± 1 °C,
132
which allows for easy processing even at modest substrate
temperatures. FTIR analysis (Figure 23a) confirmed that porous film formation occurred by the
same mechanism as the PMAA porous films. The spectrum of the sample prior to pump down is
similar to the spectrum of pure NIPAAm, indicating that a large amount of unreacted monomer
is present in the sample. After pump down, the peak attributed to the C=C stretching at 1622 cm
-
1
is no longer present which confirms the removal of the unreacted monomer. The spectrum of
the final PNIPAAm porous film matches reference PNIPAAm thin films, confirming
polymerization. The expected peaks were observed for PNIPAAm:
117,133
N-H stretching at 3292
cm
−1
, N-H overtone at 3078 cm
−1
, asymmetric −CH
3
stretching at 2972 cm
−1
, asymmetric −CH
2
−
stretching at 2932 cm
−1
, symmetric −CH
3
stretching at 2875 cm
−1
, secondary amide C=O
stretching at 1649 cm
−1
, N-H bending and C-N stretching at 1539 cm
−1
, −CH
3
and −CH
2
−
deformation at 1457 cm
−1
, and −CH
3
deformation at 1386 and 1367 cm
−1
. Top-down and cross-
sectional SEM images of the PNIPAAm porous films (Figure 23b) exhibited a similar porous
structure and the same dual-scale porosity as the PMAA porous films. The PNIPAAm porous
films exhibited thinner microstructures which are potentially due to differences in monomer
crystallization and are the subject of further investigation.
54

Figure 23. a) FTIR spectra of the NIPAAm monomer, the sample before and after sublimation, and the PNIPAAm
thin film reference. The strong peak attributed to C=C stretching of the monomer at 1622 cm
-1
(dashed line) is no
longer present after sublimation, at which point the porous film is free of NIPAAm and is chemically identical to the
thin PNIPAAm films. (b) SEM images showing top-down and inset cross-sectional images of PNIPAAm porous
films.

Our method allows for the fabrication of porous copolymer films which represents a
useful method for modifying porous film properties, including for increased mechanical strength
and chemical resistance.
53
The only constraint for producing copolymer porous films is that at
least one of the monomers must satisfy the two requirements described previously for porous
film formation. For example, poly(methacrylic acid-co-ethylene glycol diacrylate) (P(MAA-co-
EGDA)) porous films were formed by simultaneously flowing MAA and the cross-linker
ethylene glycol diacrylate (EGDA) under conditions in which only MAA deposited as solid
55

monomer by keeping the EGDA partial pressure below its saturation pressure. The lack of solid
EGDA monomer ensured that the structure of the porous films was controlled by the solid MAA
deposition, leading to the formation of porous films that exhibited a similar structure to the
PMAA porous films (Figure 24). The EGDA precursor was incorporated into the final porous
film via polymerization, which improved structural integrity and made the porous film insoluble
in water. The long range order and cohesiveness of the porous film is also demonstrated.

Figure 24. (a) Photograph and (b) SEM images of a free-standing P(MAA-co-EGDA) porous film.

56

5.2 Chemical Modification
PMAA is one of the most versatile polymers for aqueous-based applications due to its
hydrophilicity and pH-responsive acidic pendant groups that have a modest pKa value (5.7) and
can be deprotonated in order to change the properties of the polymer.
134
However, in aqueous
media PMAA generally needs to be chemically modified or grafted to the substrate in order to
prevent its dissolution.
135
One method of modifying PMAA is to utilize an intramolecular
condensation reaction that converts the water soluble methacrylic acid (MAA) moieties into less
soluble methacrylic anhydride (MAN) moieties upon annealing at temperatures approaching 200
°C.
136,137
This leads to the formation of poly(methacrylic acid-co-methacrylic anhydride)
(P(MAA-co-MAN)), in which the MAN content, and thereby the solubility of the polymer, can
be controlled. This intramolecular condensation reaction has previously been studied for
nonporous PMAA thin films, and extending this process to porous membranes would allow for
their use in aqueous media while maintaining the pH-responsive nature imparted by the MAA
moieties.
PMAA membranes were synthesized by simultaneously depositing and partially
polymerizing MAA monomer. The subsequent sublimation of unreacted MAA led to membranes
with dual-scale porosity. The large-scale pores form during the deposition process by the void
space present between the solid MAA monomer structures. The small-scale pores form by the
removal of unreacted monomer during the sublimation step. The average thickness of the
resulting PMAA membranes was analyzed by scanning electron microscopy (SEM) to be 320 ±
30 μm. The membranes were converted to P(MAA-co-MAN) by a post-polymerization
annealing process at 185 ± 2 °C for 5 – 5000 minutes. We utilized Fourier transform infrared
57

(FTIR) spectroscopy to quantitatively determine the amount of MAN in the membranes as a
function of the annealing time by comparing the relative areas of the carbonyl peaks of the MAA
and MAN moieties (Figure 25).
33
Peak fitting was necessary in order to effectively differentiate
the four desired peaks. The carbonyl peaks for PMAA appear at 1700 cm
-1
and 1735 cm
-1
and
represent hydrogen bonded and non-hydrogen bonded environments, respectively.
136
For
P(MAA-co-MAN), two additional carbonyl peaks located at 1756 cm
-1
and 1802 cm
-1
are
present. The location and relative intensity of these peaks are consistent with 6-membered MAN
rings formed by the reaction between adjacent intramolecular MAA moieties in a head-to-tail
configuration.
136
We do not observe any significant interchain anhydride linkages in our samples
which is consistent with literature sources.
138
These interchain linkages have a unique infrared
absorbance peak due to carbonyl stretching centered at 1743 cm
-1
that is not present in any of our
samples, irrespective of the annealing time.
136
We found that the MAN content increased linearly
with the log of the annealing time until reaching a plateau of 67 ± 3 wt% at extended annealing
times (Figure 26).
58

Figure 25. (a) FTIR spectra of a PMAA sample and a P(MAA-co-MAN) sample at 47 wt% MAN. (b) Zoomed-in
area showing peak fitting analysis of the P(MAA-co-MAN) sample.

59

Figure 26. Plot of the MAN content as a function of annealing time.

We studied the wettability of the membranes and found that they transitioned from
hydrophilic to hydrophobic as the MAN content was increased as expected. The contact angle of
water on the membranes varied from a fully wetted state at less than 18 wt% MAN to a
maximum contact angle of approximately 140 ° at 60 wt% MAN (Figure 27a). The high contact
angles were due to a combination of the high content of hydrophobic MAN groups and the rough
surface of the membranes. The contact angles dropped to approximately 75 ° at a MAN content
above 60 wt%. SEM analysis of these membranes showed that a significant loss of structure
occurred at this point (Figure 27b). It is likely that this loss of structure was due to the glass
transition temperature (T
g
) of the copolymer dropping below the annealing temperature of 185 ±
2 °C as more MAN was incorporated. The T
g
of the copolymer at 60 wt% MAN was estimated
to be 184 °C using the Fox equation,
139
where the T
g
of PMAA and PMAN are 228 °C and 159
°C, respectively.
140
Although we found that longer annealing times at temperatures as low as 150
°C could be used to avoid the T
g
from dropping below the annealing temperature, the total
annealing time was significantly increased, which is consistent with prior literature.
138
60

Figure 27. (a) Water contact angles on the membranes as a function of the MAN content. (b) SEM images showing
the loss of structure that occurs at conversions above 60 wt% MAN.

We studied the dissolution behavior of the membranes as a function of the MAN content
and pH (Figure 28) and found that below 18 wt% MAN, the copolymer membranes dissolved
readily in both pH 4 and pH 7 buffers. Between 18 and 50 wt%, the membranes exhibited pH-
dependent dissolution times with membranes dissolving slower in pH 4 buffer than pH 7 buffer,
as expected since PMAA is less soluble in acidic media. For example, P(MAA-co-MAN)
membranes with 32 wt% MAN dissolved in pH 4 buffer in approximately 3 hours compared to 5
minutes in pH 7 buffer. Above 50 wt%, we found that the membranes were insoluble in both pH
buffers for over 4 months. At these high conversions, the pH 4 buffer had no observable effect
on the membranes whereas the pH 7 buffer caused the membranes to swell into a hydrogel.
61

Figure 28. Dissolution time as a function of the MAN content and pH.

The presence of MAN moieties in the P(MAA-co-MAN) membranes allows for further
chemical modification due to the susceptibility of the bonds to nucleophilic attack.
[137]
We were
interested in using a bifunctional nucleophile in order to preserve the wetting properties of the
membranes while being able to render them insoluble in aqueous media. This post-
polymerization modification is a valuable alternative to in situ crosslinking
113
because the latter
is difficult to control during membrane formation due to the condensation of crosslinking
monomers at the low substrate temperatures used. We were able to chemically modify P(MAA-
co-MAN) membranes by exposing them to strongly nucleophilic 1,3-diaminopropane (DAP)
vapors, which have two primary amines available for reaction (Figure 29a). The use of DAP
vapors ensures that the entire fabrication process remains all-dry, which is important for the
preservation of the porous morphology of the membranes since solvents may dissolve, swell, or
otherwise alter the original membrane morphology. We found that successful crosslinking of the
P(MAA-co-MAN) membrane with 32 wt% MAN occurred when exposed to DAP vapors at
room temperature. An FTIR spectrum of a membrane after 72 hours of exposure to DAP
62

confirmed its incorporation by the addition of the N-H bend of the primary amine peak at 1545
cm
-1
and the C=O stretch of the amide peak 1636 cm
-1
in conjunction with the loss of area of the
MAN peaks at 1756 cm
-1
and 1802 cm
-1
(Figure 29b). We found that exposing the membranes to
DAP vapors for as little as 3 hours produced membranes which were insoluble in both pH 4 and
pH 7 buffers. Upon soaking membranes in pH 4 buffer, the crosslinked membranes remained
unswollen and insoluble for over 4 months while the uncrosslinked membranes dissolved in
approximately 3 hours. In comparison, upon soaking in pH 7 buffer, the crosslinked membranes
swelled into a hydrogel while the uncrosslinked P(MAA-co-MAN) membranes dissolved
completely within 5 minutes. Although we chose to use the crosslinking DAP molecule for the
post-polymerization modification, this solvent-free modification process can be extended to
other functionalities in order to expand the applications of the membranes.
63

Figure 29. (a) Reaction scheme of DAP with P(MAA-co-MAN) and (b) the FTIR spectra confirming crosslinking.
The vertical dotted lines represent the N-H bending of the amine (1545 cm
-1
) and the C=O stretching of the amide
(1636 cm
-1
) groups of DAP.

64

Chapter 6: Applications
6.1 Complex Substrates
In order to understand the advantages and limitations of applying these coatings to 3D
substrates, we used stainless steel tubing as a model substrate. The tubing provided both curved
and angled features and internal and external surfaces, while the use of stainless steel allowed us
to minimize effects of thermal gradients. One of the major advantages of using a vapor phase
technique is the ability to coat complex substrates and avoid the surface tension, gravitational,
and solubility effects which limit solution phase processing. However, our technique to produce
porous polymer coatings is complicated by the fact that it relies on a combination of both
chemical vapor deposition (partial polymerization to PMAA) and physical vapor deposition
(solid MAA deposition). While chemical vapor deposition techniques (such as iCVD) exhibit
extreme conformality, there are limitations that arise with utilizing a physical vapor deposition
mechanism. Typically, physical vapor deposition techniques such as evaporation-deposition,
sputtering, and electron beam lithography are line-of-sight techniques which are governed by a
beam of evaporated source material that flows towards the substrate.
141,142
The deposition rate
and coverage of the coating is dependent on the geometry of the reactor and location of the
substrate, including the angle of incidence and the distance of the substrate from the source.
However since our technique uses high vapor pressure monomer reactants, we are able to
maintain a significant partial pressure of the monomer and only deposit on the areas that we
selectively cool down. We hypothesize that this allows our technique to not be line-of-sight with
respect to the reactant inlets and achieve more conformal coverage than traditional physical
vapor deposition processes.
65

We tested our hypothesis by coating our 3D substrates and examining the surfaces
through optical and SEM techniques. We found that the external surfaces coated extremely well
and conformally, irrespective of whether there were curved or angled features and the orientation
of the substrate (Figure 30). SEM analysis shows that the coating consisted of microstructures
with dual-scale porosity which is consistent with the morphology on flat surfaces. However, the
internal surface was not completely covered with porous polymer even though the entire surface
is fully accessible to precursor vapors. We hypothesize that although solid MAA did not deposit
on the surface to act as a template for the porous coating, there was enough MAA present to
polymerize as a dense PMAA film.

Figure 30. Vertically-oriented Stainless steel tubing (a) before and (b) after porous coating as seen inside the reactor
from above. (c) Conformal coating on horizontally-oriented tube. (d) Porous PMAA coverage on internal surfaces of
tube radii of 0.88 mm, 3.26 mm, 5.64 mm, 8.81 mm. (e) SEM micrograph showing porous PMAA on the top
external surface of the substrate with a high magnification inset.
66

The radius of the substrate was varied in order to understand the impact of substrate
dimensions on the coverage of the porous PMAA coatings. Tube radii from 0.88 mm to 8.81 mm
and 12 mm long were positioned vertically and the coatings were analyzed by optical and SEM
techniques. In all cases, the external surfaces were conformally coated including both curved and
angled features. The length of conformality on the internal surface varied and the results are
shown in (Figure 30d). Unsurprisingly, the smaller the radius of the tube, the less area of the
internal surface was coated with porous polymer. This indicates that the degree of conformality
is based on the local MAA partial pressure when monomer depleted due to deposition is taken
into account.

67

6.2 Porous Substrates
In order to fabricate the patterned polymer membranes, a thermoelectric cooler was
placed inside of an iCVD vacuum chamber to achieve cold substrate temperatures (-10 °C) in
order to satisfy the two requirements previously discussed: the partial pressure of the monomer
must be greater than its saturation pressure and the temperature of the substrate must be less than
the freezing point of the monomer. Chromatography paper was used as a porous substrate.
143
The
paper was placed on top of the thermoelectric cooler and covered with a stainless steel mask in
order to pattern the deposition of the polymer, as shown in Figure 31. During the deposition,
polymer visibly grew on both the mask and the exposed paper. Following the deposition, the
unreacted frozen monomer was sublimed until the system returned to base pressure leaving
behind porous polymer membranes patterned directly onto the chromatography paper. While
circular membranes were fabricated for demonstration, this technique can be extended to other
shapes by simply varying the mask.
68

Figure 31. Schematic representation and optical images of the fabrication of patterned porous-on-porous materials.

We examined the structure of the polymer membranes at deposition times of 10, 20, and
30 minutes using scanning electron microscopy (SEM). Lower deposition times (< 2 minutes)
did not result in a cohesive polymer layer. For deposition times in the range of 10 to 30 minutes,
the membrane thickness had a linear growth rate of 32.2 ± 2.8 µm/min. The radii of the base
(polymer-paper interface) of the membranes (2.8 ± 0.1 mm) were in relatively good agreement
with the radius of the mask (2.7 mm) for all deposition times, confirming successful patterning
of the membranes. The radii of the top (polymer-air interface) of the membranes (2.3 ± 0.1 mm)
were slightly smaller than the radius of the mask. The tapered structure of the membranes is
attributed to shadowing effects during deposition caused by the mask and the growing polymer,
which has been similarly observed in other vapor phase deposition systems including the oblique
69

angle deposition of parylene
107
and the sputtering of metals,
144
leading to non-uniform deposition
onto the substrate.
Morphologically, the interior of the membranes were composed of randomly aligned
pillar-like microstructures and displayed similar dual-scale porosity as the membranes deposited
onto planar substrates with large-scale pores on the order of tens of microns in diameter and
small-scale pores on the order of hundreds of nanometers in diameter (Figure 32). The tapered
edges of the membranes displayed a dense polymer layer that partially bridged individual
microstructures. The interior cross-sectional image showed that this dense layer was restricted to
the edge of the membranes while the interior remained highly porous. This dense layer likely
forms as a result of the shadowing effects at the interface between the mask and the growing
polymer, which prevents the deposition of additional solid monomer and the formation of
microstructures but still allows adsorbed monomer vapor to polymerize across the surface.
Relative to the deposition of membranes onto flat surfaces, the deposition onto porous substrates
was more disordered leading to a broad size distribution of the large-scale pores (Figure 33). The
distribution of the pore diameters was independent of deposition time and membrane fabrication
was reproducible with consistent morphology and an average pore diameter of approximately 30
± 4 µm.
70

Figure 32. Schematic of membrane structure with scanning electron micrographs showing the a) top-down
membrane morphology with an inset at higher magnification, b) tapered edge, c) exterior cross-section, and d)
interior cross-section.
71

Figure 33. The distribution of the diameters of the large-scale pores for membranes fabricated at deposition times of
10, 20, and 30 minutes.

While the SEM images showed no polymer deposition outside of the patterned area, we
dyed crosslinked PMAA membranes with toluidine blue O
145
to further study whether the
regions of chromatography paper covered by the mask were coated with polymer. The polymer
was crosslinked to prevent dissolution during dyeing. In order to fabricate the crosslinked
poly(methacrylic acid-co-ethylene glycol diacrylate) (P(MAA-co-EGDA)) membranes, EGDA
was added into the reactor at a partial pressure below its saturation pressure to ensure that it did
not affect the morphology of the porous membranes (Figure 34). The P(MAA-co-EGDA)
membranes were fabricated at deposition times of 10, 20, and 30 minutes and analyzed to
determine the locations of polymer deposition. The intensity plots of the polymer membranes
show a radius and tapered edge which is consistent with the structure of the membranes as
observed by SEM. Outside of the area of the polymer membrane, the intensity quickly returns to
72

the baseline for all deposition times, indicating that there is negligible deposition of polymer in
the areas covered by the mask. The absence of polymer under the mask is attributed to the
depletion of monomer from the vapor phase as a result of its deposition, preventing the diffusion
of monomer vapor throughout the paper substrate. The absence of polymer allows our fabrication
technique to be used in applications where patterning of polymer deposition is essential, such as
in sensing and lab-on-a-chip applications.

Figure 34. a) Top-down scanning electron micrograph of porous P(MAA-co-EGDA) membrane. b) Intensity plot of
dyed P(MAA-co-EGDA) membranes. Mixture of crystal violet and ponceau S flowing through a paper-based
microfluidic device in the c) absence and d) presence of a porous P(MAA-co-EGDA) membrane.
73

As an example of the utility of these patterned porous-on-porous materials, we
demonstrated their use as filters for paper-based microfluidic devices. Paper-based microfluidic
devices are a new type of point-of-care device,
146,147
but they generally lack the functionality
available to pressure-driven microfluidic devices.
148
For instance, the ability of unmodified
chromatography paper to separate analytes is limited, as shown by its inability to separate a
mixture of crystal violet and ponceau S (Figure 34c). However, the incorporation of a P(MAA-
co-EGDA) porous polymer membrane patterned onto the inlet of a paper-based microfluidic
device allows for the selective separation of crystal violet as a model cationic analyte from a
mixture containing the anionic dye ponceau S (Figure 34d). This separation is due to the
increased electrostatic attraction between the electronegative P(MAA-co-EGDA) and the
electropositive crystal violet, which is further enhanced by the high surface area provided by the
porous polymer membrane.
Our fabrication technique allows for the deposition of functional membranes onto
complex substrates for use in applications such as wound dressings. For example in Figure 35,
we demonstrate the successful fabrication of P(MAA-co-MAN) membranes with 38 wt% MAN
on gauze. The use of a gauze substrate compared to silicon wafer leads to no significant
differences in the membrane growth rate with thicknesses around 300 µm. In addition, the MAN
content after annealing was consistent with values found previously. Gauze is a common and
inexpensive wound dressing, but it does not actively help to quicken recovery and is a poor
absorbent.
149
The addition of a membrane allows for the ability to deliver drugs, increase exudate
absorbance, and improve the overall rate of healing.
150,151
74

Figure 35. (a) Top-down and (b) cross-sectional SEM images of gauze and (c) top-down and (d) cross-sectional
SEM images of P(MAA-co-MAN) membrane with 38 wt% MAN on gauze.

75

Chapter 7: Future Work
We want to continue to develop this technology to determine its full capabilities. By
understanding the morphological differences created with variations in the deposition driving
force, we can tailor the coatings to best match the application in mind. One major extension we
plan to study in our system is the creation of nonporous capping layers during porous film
production. Membranes can be broken into two major types: those that utilize porous structures
and those which use nonporous films for their application. The latter type of membrane is termed
asymmetric because a thin nonporous polymer layer is situated on top of a porous layer, which is
used for structural support. The membrane separates chemical species based on the permeation
of that species through the nonporous polymer layer. In order to keep a high throughput, the
thickness of the nonporous polymer film needs to be thin and well controlled. There are
significant mechanical and compatibility advantages in forming the two layers out of the same
material and for having a single cohesive structure.
One potential necessity for the production of asymmetric membranes is the ability to
achieve stable deposition that lacks defined microstructures. Research into the transition from
unstable to stable deposition has been studied in various inorganic deposition systems.
152,153
In
those systems, their high substrate temperatures make stable deposition much more likely due to
increased surface diffusion. In the case of stable deposition, the growing composite of monomer
and polymer will appear to be adding in complete layers, with no void space left within its
volume. On the other hand, unstable deposition will see the growth of microstructures and will
contain void space even before the monomer is sublimated away. Increasing the instability of the
deposition will cause the microstructures become thinner and more numerous.
76

We first tested a low partial pressure sample (150 mTorr, where saturation pressure is
133 mTorr) and found that these low partial pressures resulted in fractal patterns such as seen in
Figure 36. These patterns were common in partial pressure in close proximity to the saturation
pressure and when the monomer was introduced into the reactor chamber slowly. These patterns
are likely caused by monomer adding to a small number of nucleation points instead of further
nucleation occurring. Therefore, it might be necessary to explore conditions to achieve fast and
widespread nucleation, but to deposit for short times. This could allow for a smooth and cohesive
solid monomer template upon which to polymerize.

Figure 36. Photograph of low partial pressures (above the saturation pressure) attempting to achieve stable monomer
deposition. The arrows show (from left to right): widespread porous polymer, nonporous thin polymer film, porous
polymer fractal patterns, and bare silicon wafer which was masked with polyimide tape.

The hypothesis is that the polymerization of and on top of the top surface of a relatively
smooth deposition should lead to a nonporous capping layer. This can be done by flowing
initiator after stopping the deposition of the monomer, which we term as post-deposition
polymerization. For a preliminary study, we deposited MAA for only 1 minute before stopping
the monomer flow and allowing polymerization to occur for 25 minutes (free radicals are present
77

for the entire time). We used a 1 minute deposition time to attempt to have stable monomer
deposition, before any microstructures form. The SEM images show that a partial nonporous
capping layer was formed over much of the sample (Figure 37). However, it appears the surface
of the growing monomer template still had significant roughness, leading to holes in the capping
layer. We can explore variations to the monomer deposition and the post-deposition
polymerization time to see if we can achieve a full capping layer under these conditions.

Figure 37. Preliminary results on producing a capping layer without relying on monomer liquid condensation. Top-
down SEM images at a) low and b) high magnification show the partial covering of a thin nonporous layer. c) The
cross-sectional image shows the nonporous film folded over (arrow) on top of a thick porous layer.

An important advantage of fabricating capped porous films by this technique is that the
skin and porous layers are one cohesive structure. Because the monomer partial pressure can be
modified in real time during the fabrication process, the polymer chains continue to grow and
there is no interface on the molecular scale. Vapor phase polymerization allows for real time
control of the monomer partial pressure that solution phase techniques cannot match. This allows
polymer skins to no longer be restricted to the top of the porous layer. The skins could be put
78

anywhere in the middle of the porous film by growing an additional porous layer on top of the
skin. Multiple skins could also be fabricated by switching between the regimes several times.
A second method to create a camping later is to induce monomer condensation in
addition to deposition in order to produce a smooth surface upon which to polymerize. We were
able to produce a P(MAA-co-EGDA) capped membrane by flowing in EGDA above its
saturation pressure. Since the substrate temperature was above its freezing point, the EGDA
condense on the substrate and allowed for a smooth top layer (Figure 38). SEM imaging of the
membrane clearly shows the presence of a nonporous capped layer on top of a thick porous
layer. We believe the cracking is due to the polymerization occurring while the sample is swelled
with liquid EGDA. When the monomers are sublimated or evaporated, the deswelling leads to
cracks in the final film.

Figure 38. A capped membrane produced by inducing liquid monomer condensation during the growth process.
Top-down SEM images showing a) low magnification (with cracks) and b) high resolution views of the capping
layer and a c) cross-sectional image showing the porous structure capped with a nonporous layer (arrow).

79

In this sample, another interesting note is that the thick porous layer does not appear to
have any microstructures, even though similar conditions without EGDA produce
microstructures. We directly compared an area of high EGDA condensation to an area with little
to no EGDA condensation (Figure 39) and found that the microstructure formation is prevented
under the presence of a significant amount of liquid. The sample without significant EGDA
condensation clearly contains microstructures. The high gradient of EGDA condensation across
the sample is due to the small flow (< 0.4 sccm) of EGDA that is introduced into the system.
This large gradient across the sample is not apparent with MAA because it is introduced at a high
flow rate. The necessity to condense monomer to form the capping layer is useful for some
situations, but the inclusion of monomer condensation removes some of the advantages of
utilizing a solventless technique.

Figure 39. Top-down SEM image of part of the P(MAA-co-EGDA) sample which had no condensing crosslinker.

80

Chapter 8: Conclusions
In conclusion, we demonstrate a novel technique for producing porous polymer films
using vapor phase polymerization. By increasing the monomer partial pressure above its
saturation pressure and reducing the substrate temperature below the freezing point of the
monomer, simultaneous solid monomer deposition and polymerization occur leading to the
formation of microstructures. The void spaces between these microstructures create large-scale
pores within the porous film. When the solid monomer is sublimated, small-scale pores form
within the microstructures leading to porous polymer films with dual-scale porosity. Control
over the thickness growth rate and pore size is achieved by varying the process parameters and
the generality of this fabrication process is shown by using two stimuli-responsive polymers as
model systems to generate both homopolymer and copolymer porous films. By using a vapor
phase technique to produce porous polymer films, solvent compatibility and surface tension
effects are eliminated, allowing for the fabrication of novel tailored porous polymers.
These processing conditions represent an untapped area of study and have the potential to
create next-generation highly tailored porous polymer films.
154
The lack of solvents eliminates
solubility requirements, increasing the control over the chemical composition and allowing for
the production of stimuli-responsive and copolymer porous films. The simplicity of using only
monomer precursors as a porogen allows this technique to be compatible with roll-to-roll
processing and a variety of vapor phase technologies previously discussed. In addition, the
fundamentals of this technique are not exclusive to the iCVD technology. Therefore it may be
possible to similarly modify PECVD, parylene deposition, and other chemical vapor deposition
techniques to produce similar materials. This would allow for the further extension of chemical
vapor deposition techniques to a wide range of new and exciting applications.
81

References

1) Web of Knowledge Database Search. Accessed 14 April 2015.
2) Tenhaeff, W. E.; McIntosh, L. D.; Gleason, K. K.; Adv. Funct. Mater. 2010, 20, 1144.
3) McInnes, S. J. P.; Szili, E. J.; Al-Bataineh, S. A.; Xu, J.; Alf, M. E.; Gleason, K. K.; Short, R.
D.; Voelcker, N. H.; ACS Appl. Mater. Interfaces 2012, 4, 3566.
4) Ma, M.; Gupta, M.; Li, Z.; Zhai, L.; Gleason, K. K.; Cohen, R. E.; Rubner, M. F.; Rutledge,
G. C.; Adv. Mater. 2007, 19, 255.
5) Gorham, W. F.; J. Polym. Sci., Part A: Polym. Chem. 1966, 4, 3027.
6) Lahann, J.; Langer, R.; Macromolecules 2002, 35, 4380.
7) Schmidt, C.; Stumpflen, V.; Wendorff, J. H.; Hasenhindl, A.; Gronski, W.; Ishaque, M.;
Greiner, A.; Acta Polym. 1998, 49, 232.
8) Vaeth, K. M.; Jensen, K. F.; Chem. Mater. 2000, 12, 1305.
9) Vaeth, K. M.; Jackman, R. J.; Black, A. J.; Whitesides, G. M.; Jensen, K. F.; Langmuir 2000,
16, 8495.
10) Lahann, J.; Klee, D.; Hoecker, H.; Macromol. Rapid Commun. 1998, 19, 441.
11) Lahann, J.; Klee, D.; Pluester, W.; Hoecker, H.; Biomaterials 2001, 22, 817.
12) Fortin, J. B.; Lu, T.-M.; Chem. Mater. 2002, 14, 1945.
13) Beach, W. F.; Macromolecules 1978, 11, 72.
14) Simon, P.; Mang, S.; Hasenhindl, A.; Gronski, W.; Greiner, A.; Macromolecules 1998, 31,
8775.
82

15) Dabral, S.; Zhang, X.; Wu, X. M.; Yang, G.-R.; You, L.; Lang, C. I.; Hwang, K.; Cuan, G.;
Chiang, C.; Bakhru, H.; Olson, R.; Moore, J. A.; Lu, T.-M.; McDonald, J. F.; J. Vac. Sci.
Technol., B 1993, 11, 1825.
16) Lahann, J.; Balcells, M.; Lu, H.; Rodon, T.; Jensen, K. F.; Langer, R.; Anal. Chem. 2003, 75,
2117.
17) Yasutake, M.; Hiki, S.; Andou, Y.; Nishida, H.; Endo, T.; Macromolecules 2003, 36, 5974.
18) Yasutake, M.; Andou, Y.; Hiki, S.; Nishida, H.; Endo, T.; J. Polym. Sci., Part A: Polym.
Chem. 2004, 42, 2621.
19) Andou, Y.; Yasutake, M.; Jeong, J.-M.; Nishida, H.; Endo, T.; Macromol. Chem. Phy. 2005,
206, 1778.
20) Andou, Y.; Jeong, J.-M.; Hiki, S.; Nishida, H.; Endo, T.; Macromolecules 2009, 42, 768.
21) Biederman, H.; Slavinska, D.; Surf. Coat. Technol. 2000, 125, 371.
22) Yasuda, H.; Luminous Chemical Vapor Deposition and Interface Engineering, Marcel
Dekker, New York, 2005, p. 64.
23) Tarducci, C.; Schofield, W. C. E.; Badyal, J. P. S.; Brewer, S. A.; Willis, C.; Chem. Mater.
2002, 14, 2541.
24) Gupta, M.; Gleason, K. K.; Thin Solid Films 2006, 515, 1579.
25) Martin, T. P.; Lau, K. K. S.; Chan, K.; Mao, Y.; Gupta, M.; O’Shaughnessy, W. S.; Gleason,
K. K.; Surf. Coat. Technol. 2007, 201, 9400.
26) Seidel, S.; Riche, C.; Gupta, M.; Chemical Vapor Deposition of Polymer Films. Encycl.
Polym. Sci. and Technol. 2011.
27) Chan, K.; Gleason, K. K.; Macromolecules 2006, 39, 3890.
28) Chan, K.; Gleason, K. K.; Chem. Vap. Deposition 2005, 11, 437.
83

29) Lewis, H. G. P.; Caulfield, J. A.; Gleason, K. K.; Langmuir 2001, 17, 7652.
30) Lau, K. K. S.; Gleason, K. K.; Macromolecules 2006, 39, 3688.
31) Mao, Y.; Gleason, K. K.; Langmuir 2004, 20, 2484.
32) Lau, K. K. S.; Gleason, K. K.; Macromolecules 2006, 39, 3695.
33) Mao, Y.; Gleason, K. K.; Langmuir 2006, 22, 1795.
34) Gupta, M.; Gleason, K. K.; Langmuir 2006, 22, 10047.
35) Bose, R. K.; Lau, K. K. S.; Biomacromolecules 2010, 11, 2116.
36) Limb, S. J.; Gleason, K. K.; Edell, D. J.; Gleason, E. F.; J. Vac. Sci. Technol. A 1997, 15,
1814.
37) Liger, M.; Rodger, D. C.; Tai, Y. C.; In Micro Electro Mechanical Systems, Proceedings of
the 16th IEEE International Conference, Kyoto, Japan, January 2003, 602.
38) Chen, G.; Gupta, M.; Chan, K.; Gleason, K. K.; Macromol. Rapid Commun. 2007, 28, 2205.
39) O’Shaughnessy, W. S.; Mari-Buye, N.; Borros, S.; Gleason, K. K.; Macromol. Rapid
Commun. 2007, 28, 1877.
40) Quinn, C. P.; Pathak, C. P.; Heller, A.; Hubbell, J. A.; Biomaterials 1995, 16, 389.
41) Belkas, J. S.; Munro, C. A.; Shoichet, M. S.; Johnston, M.; Midha, R.; Biomaterials 2005,
26, 1741.
42) Mayer, T. M.; de Boer, M. P.; Shinn, N. D.; Clews, P. J.; Michalske, T. A.; J. Vac. Sci.
Technol., B 2000,18, 2433.
43) Wavhal, D.S.; Fisher, E. R.; J. Membr. Sci. 2002, 209, 255.
44) Yu, H.-Y.; Hu, M.-X.; Xu, Z.-K.; Plasma Processes Polym. 2005, 2, 627.
45) Chan-Park, M. B.; Yan, Y.; Neo, W. K.; Zhou, W.; Zhang, J.; Yue, C. Y.; Langmuir 2003,
19, 4371.
84

46) Tenhaeff, W. E.; Gleason, K. K.; Adv. Funct. Mater. 2008, 18, 979.
47) Baxamusa, S. H.; Gleason, K. K.; Thin Solid Films 2009, 517, 3536.
48) Kwong, P.; Flowers, C. A.; Gupta, M.; Langmuir 2011, 27, 10634.
49) Gupta, M.; Kapur, V.; Pinkerton, N. M.; Gleason, K. K.; Chem. Mater. 2008, 20, 1646.
50) Lau, K. K. S.; Bico, J.; Teo, K. B. K.; Chhowalla, M.; Amaratunga G. A. J.; Milne, W. I.;
McKinley, G. H.; Gleason K. K.; Nano Letters 2012, 3, 1701.
51) Lau, K. K. S.; Gleason, K. K.; Adv. Mater. 2006, 18, 1972.
52) Martin, T. P.; Sedransk, K. L.; Chan, K.; Baxamusa S. H.; Gleason, K. K.; Macromolecules
2007, 40, 4586.
53) Chan, K.; Gleason, K. K.; Langmuir 2005, 21, 8930.
54) Mao, Y.; Gleason, K. K.; Macromolecules 2006, 39, 3895.
55) Mao, Y.; Felix, N. M.; Nguyen, P. T.; Ober, C. K.; Gleason, K. K.; J. Vac. Sci. Technol. B
2004, 22, 2473.
56) Prasitsilp, M.; Siriwittayakorn, T.; Molloy, R.; Suebsanit, N.; Siriwittayakorn, P.;
Veeranondha, S.; J. Mater. Sci.: Mater. Med. 2003, 14, 595.
57) Pfluger, C. A.; Carrier, R. L.; Sun, B.; Ziemer, K. S.; Burkey, D. D.; Macromol. Rapid
Commun. 2009, 2, 126.
58) Marí-Buyé, N.; O’Shaughnessy, S.; Colominas, C.; Semino, C. E.; Gleason, K. K.; Borrós,
S.; Adv. Funct. Mater. 2009, 9, 1276.
59) Trujillo, N. J.; Wu, Q.; Gleason, K. K.; Adv. Func. Mater. 2010, 20, 607.
60) Wu, D.; Xu, D.; Sun, B.; Fu, R.; He, H.; Matyjaszewski, K.; Chem. Rev. 2012, 112, 3959.
61) Ma, H.; Hao, J.; Chem. Soc. Rev. 2011, 40, 5457.
62) Yang, J.-S.; Swager, T. M.; J. Am. Chem. Soc. 1998, 120, 11864.
85

63) Shen, Y.; Liu, Y.; Zhu, G.; Fang, H.; Huang, Y.; Jiang, X.; Wang, Z. L.; Nanoscale 2013, 5,
527.
64) Audouin, F.; Fox, M.; Larragy, R.; Clarke, P.; Huang, J.; O’Connor, B.; Heise, A.;
Macromolecules 2012, 45, 6127.
65) Mitchell, D. T.; Lee, S. B.; Trofin, L.; Li, N.; Nevanen, T. K.; Soederlund, H.; Martin, C. R.;
J. Am. Chem. Soc. 2002, 124, 11864.
66) Haswell, S. J.; Skelton, V.; Trends Anal. Chem. 2000, 19, 389.
67) Cameron, N. R.; Polymer 2005, 46, 1439.
68) Zoppi, R. A.; Contant, S.; Duek, E. A. R.; Marques, F. R.; Wada, M. L. F.; Nunes, S. P.;
Polymer 1999, 40, 3275.
69) Beattie, D.; Wong, K. H.; Williams, C.; Poole-Warren, L. A.; Davis, T. P.; Barner-Kowollik,
C.; Stenzel, M. H.; Biomacromolecules 2006, 7, 1072.
70) Shaikh, R. P.; Pillay, V; Choonara, Y. E.; du Toit, L. C.; Ndesendo, V. M. K.; Bawa, P.;
Cooppan, S.; AAPS PharmSciTech 2010, 11, 441.
71) Wang, S.; Microporous Mesoporous Mater. 2009, 117, 1.
72) Hsu, Y.-Y.; Gresser, J. D.; Trantolo, D. J.; Lyons, C. M.; Gangadharam, P. R. J.; Wise, D.
L.; J. Biomed. Mater. Res. 1997, 35, 107.
73) Nevanen, T. K.; Soederlund, H.; Martin, C. R.; Science 2002, 296, 2198.
74) Pinnau, I.; Koros, W. J.; J. Appl. Polym. Sci. 1991, 43, 1491.
75) Yang, R.; Gleason, K. K.; Langmuir, 2012, 28, 12266.
76) Thomas, A.; Kuhn, P.; Weber, J.; Titirici, M.-M.; Antonietti M.; Macromol. Rapid Commun.
2009, 30, 221.
77) Escalé, P.; Rubatat, L.; Billon, L.; Save, M.; European Polymer Journal 2012, 48, 1001.
86

78) Cai, Y.; Newby, B.-M. Z.; Langmuir 2009, 25, 7628.
79) Park, J. S.; Lee, S. H.; Han, T. H.; Kim, S. O.; Adv. Func. Mater. 2007, 17, 2315.
80) Mikos, A. G.; Thorsen, A. J.; Czerwonka, L. A.; Bao, Y.; Langer, R.; Winslow, D. N.;
Vacanti, J. P.; Polymer 1994, 35, 1068.
81) Oxley, H. R.; Corkhill, P. H.; Fitton, J. H.; Tighe, B. J.; Biomaterials 1993, 14, 1064.
82) Perez, P.; Plieva, F.; Gallardo, A.; Roman, J. S.; Aguilar, M. R.; Morfin, I.; Ehrburger-Dolle,
F.; Bley, F.; Mikhalovsky, S.; Galaev, I. Y.; Mattiason, B.; Biomacromolecules 2008, 9,
66.
83) Xue, W.; Hamley, I. W.; Huglin, M. B.; Polymer 2002, 43, 5181.
84) Matsuyama, H.; Yuasa, M.; Kitamura, Y.; Teramoto, M.; Lloyd, D. R.; J. of Membr. Sci.
2000, 179, 91.
85) Lo, H.; Ponticiello, M. S.; Leong, K. W.; Tissue Eng. 1995, 1, 15.
86) Caneba, G. T.; Soong, D. S.; Macromolecules 1985, 18, 2538.
87) Widawski, G.; Rawiso, M.; Francois, B.; Nature 1994, 369, 387.
88) Jenekhe, S. A.; Chen, X. L.; Science 1999, 283, 372.
89) Yabu, H.; Shimomura, M.; Chem. Mater. 2005, 17, 5231.
90) Connal, L. A.; Qiao, G. G.; Soft Matter, 2007, 3, 837.
91) Jiang, X.; Zhang, T.; Xu, L.; Wang, C.; Zhou, X.; Gu, N.; Langmuir 2011, 27, 5410.
92) Connal, L. A.; Vestberg, R.; Gurr, P. A.; Hawker, C. J.; Qiao, G. G.; Langmuir 2008, 24,
556.
93) Ding, S.; Wang, P.; Wan, X.; Zhang, D. W.; Wang, J.; Lee, W. W.; Mater. Sci. Eng. B 1999,
83, 130.
94) Schwodiauer, R.; Bauer, S.; Rev. Sci. Instrum. 2002, 73, 1845.
87

95) Olayo-Valles, R.; Lund, M. S.; Leighton, C.; Hillmyer, M. A.; J. Mater. Chem. 2004, 14,
2729.
96) Thurn-Albrecht, T.; Steiner, R.; DeRouchey, J.; Stafford, C. M.; Huang, E.; Bal, M.;
Tuominen, M.; Hawker, C. J.; Russell, T. P.; Adv. Mater. 2000, 12, 787.
97) Urbas, A. M.; Maldovan, M.; DeRege, P.; Thomas, E. L.; Adv. Mater. 2002, 14, 1850.
98) Okumura, A.; Nishikawa, Y.; Hashimoto, T.; Polymer 2006, 47, 7805.
99) Savina, I. N.; Cnudde, V.; D’Hollander, S.; Van Hoorebeke, L.; Mattiasson, B.; Galaev, I.
Y.; Prez, F. D.; Soft Matter 2007, 3, 1176.
100) Bamford, C. H.; Jenkins, A. D.; Ward, J. C.; J. Polymer Sci. 1960, 48, 37.
101) Bamford, C. H.; Eastmond, G. C.; Ward, J. C.; Proc. R. Soc. A 1963, 271, 357.
102) Lee, E. S.-Q.; Rangaiah, G. P.; Ind. Eng. Chem. Res. 2009, 48, 7662.
103) Sharma, S.; Chua, Y. C.; Rangaiah, G. P.; Mater. Manuf. Process. 2011, 26, 431.
104) Figoli, A.; Marino, T.; Simone, S.; Di Nicolo, E.; Li, X.-M.; He, T.; Tornaghi, S.; Drioli, E.;
Green Chem. 2014, 16, 4034.
105) Demirel, M. C.; Colloids Surf. A 2008, 321, 121.
106) Demirel, G.; Malvadkar, N.; Demirel, M. C.; Thin Solid Films 2010, 518, 4252.
107) He, M.; Wang, P.-I; Lu, T.-M.; Langmuir 2011, 27, 5107.
108) Tao, R.; Anthamatten, M.; Langmuir 2012, 28, 16580.
109) Tao, R.; Anthamatten, M.; Macromol. Rapid Commun. 2013, 34, 1755.
110) Lu, Y.; Shi, W.; Jiang, L.; Qin, J.; Lin, B.; Electrophoresis 2009, 30, 1497.
111) Carrilho, E.; Martinez, A. W.; Whitesides, G. M.; Anal. Chem. 2009, 81, 7091.
112) NIST Index of Imaging Processing Labs. http://www.nist.gov/lispix/imlab/ (accessed Sept.
15, 2012).
88

113) Seidel, S.; Kwong, P.; Gupta, M.; Macromolecules 2013, 46, 2976.
114) Jeong, B.; Gutowska, A.; Trends in Biotechnol. 2002, 20, 305.
115) Gustafson, R. L; Lirio, J. A.; J. Phys. Chem. 1965, 69, 2849.
116) Ito, Y.; Inaba, M.; Chung, D.-J.; Imanishi, Y.; Macromolecules 1992, 25, 7313.
117) Stuart, B. H.; Infrared Spectroscopy: Fundamentals and Applications, 1st ed.; John Wiley &
Sons, Ltd.: West Sussex, England, 2004, pp. 45-93.
118) Demirel, M. C.; So, E.; Ritty, T. M.; Naidu, S. H.; Lakhtakia, A.; J. Biomed. Mater. Res. B
2007, 81B, 219.
119) Swihart, M. T.; Curr. Opin. Colloid Interface Sci. 2003, 8, 127.
120) Seidel, S.; Gupta, M.; J. Vac. Sci. Technol. A 2014, 32, 4, 041514.
121) Ozaydin-Ince, G.; Gleason, K. K.; J. Vac. Sci. Technol. A 2009, 27, 1135.
122) Robbie, K.; Brett, M. J.; J. Vac. Sci. Technol. A. 1997, 15, 1460.
123) Na, B.; Webb, R. L.; Int. J. Heat Mass Transfer 2004, 47, 925.
124) Kandula, M.; Int. J. Heat Mass Transfer 2011, 54, 3719.
125) Van Leeuwen, C.; J. Crystal Growth 1979, 46, 91.
126) Vo, T. T. T.; Ho, Y.-H.; Lin, P.-H.; Tai, Y. CrystEngComm 2013, 15, 6695.
127) Tang, L.; Li, Y.; Wu, X.; Shan, X.; Wang, W.; Polym. Adv. Technol. 2004, 15, 39.
128) Haller, P. D.; Bradley, L. C.; Gupta, M.; Langmuir 2013, 29, 11640.
129) Tang, Z.; Shi, T.; Liao, G.; Liu, S.; Microelectron. Eng. 2008, 85, 1754.
130) Dong, L.; Agarwal, A. K.; Beebe, D. J.; Jiang, H.; Nature 2006, 442, 551.
131) Alf, M. E.; Godfrin, P. D.; Hatton, T. A.; Gleason, K. K.; Macromol. Rapid Commun.
2010, 31, 2166.
132) Nomura, R.; Nakano, T.; Yamada, Y.; Matsuda, H.; J. Org. Chem. 1991, 56, 4076.
89

133) Pan, Y. V.; Wesley, R. A.; Luginbuhl, R.; Denton, D. D.; Ratner, B. D.;
Biomacromolecules 2001, 2, 32.
134) Porasso, R. D.; Benegas, J. C.; van den Hoop, M. A. G. T.; J. Phys. Chem. B 1999, 103,
2361.
135) Zhou, S.; Chu, B.; J. Phys. Chem. B 1998, 102, 1364.
136) Grant, D. H.; Grassie, N.; Polymer 1960, 1, 125.
137) Uhrich, K. E.; Cannizzaro, S. M.; Langer, R. S.; Shakesheff, K. M.; Chem. Rev. 1999, 99,
3181.
138) Ho, B.-C.; Lee, Y.-D.; Chin, W.-K.; J. Polym. Sci. Pol. Chem. 1992, 30, 2389.
139) Fox, T. G.; Flory, P. J.; J. Appl. Phys. 1950, 21, 581.
140) Sigma-Aldrich. Thermal Transitions of Homopolymers: Glass Transition & Melting Point.
https://www.sigmaaldrich.com/content/dam/sigma-aldrich/docs/Aldrich/General
_Information/thermal_transitions_of_homopolymers.pdf, accessed: November, 2014.
141) Cale, T. S.; Raupp, G. B.; J. Vac. Sci. Technol. B, 1990, 8, 1242
142) Kelkar, S. S.; Wolden, C. A.; J. Vac. Sci. Technol. B, 2014, 32, 052001.
143) Kwong, P.; Seidel, S.; Gupta, M. ACS Appl. Mater. Interfaces 2013, 5, 9714.
144) Rossnagel, S. M.; Nichols, C.; Hamaguchi, S.; Ruzic, D.; Turkot, R.; J. Vac. Sci. Technol.
B 1996, 14, 1819.
145) Ma, Z.; Kotaki, M.; Ramakrishna, S.; J. Membr Sci. 2006, 272, 179.
146) Martinez, A. W.; Phillips, S. T.; Butte, M. J.; Whitesides, G. M.; Angew. Chem. Int. Ed.
2007, 46, 1318.
147) Yetisen, A. K.; Akram, M. S.; Lowe, C. R.; Lab Chip, 2013, 13, 2210.
90

148) Focke, M.; Kosse, D.; Müller, C.; Reinecke, H.; Zengerle, R.; von Stetten, F.; Lab Chip
2010, 10, 1365.
149) C. T. Hess, Wound Care, 5th ed, Lippincott Williams & Wilkins, Philadelphia, PA, USA
2005.
150) Boateng, J. S.; Matthews, K. H.; Stevens, H. N. E.; Eccleston, G. M.; J. Pharm. Sci. 2008,
97, 2892.
151) Corkhill, P. H.; Hamilton, C. J.; Tighe, B. J.; Biomaterials 1989, 10, 3.
152) Iwamoto, A.; Yoshinobu, T.; Iwasaki, H.; Phys. Rev. Lett. 1994, 72, 4025.
153) Ko, D. Y. K.; Seno, F.; Phys. Rev. B 1994, 50, 17583.
154) Nie, Z.; Kumacheva, E.; Nat. Mater. 2008, 7, 277.

Abstract (if available)

Abstract In this work we present a novel method to produce porous polymer coatings and membranes using a vapor phase polymerization technique called initiated chemical vapor deposition (iCVD). The premise of this technique is the simultaneous deposition and polymerization of a monomer species, which is achieved by increasing the monomer partial pressure above its saturation pressure when the substrate temperature is below the freezing point of the monomer. We explore the mechanism of this process, detail the ability to control the morphology and chemistry, and extend this technique towards complex substrates found in many applications.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

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

PDF

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

PDF

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

PDF

Understanding structural changes of polymer films during vapor phase deposition

PDF

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

PDF

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

PDF

Novel processing of liquid substrates via initiated chemical vapor depostion

PDF

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

PDF

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

PDF

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

PDF

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

PDF

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

PDF

Selective deposition of polymer coatings onto structured surfaces

PDF

Biogas reforming: conventional and reactive separation processes and the preparation and characterization of related materials

PDF

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

PDF

Controlled synthesis of large grain single crystal graphene by chemical vapor deposition

PDF

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

PDF

Silicon carbide ceramic membranes

PDF

Gallium arsenide based semiconductor laser design and growth by metal-organic chemical vapor deposition

PDF

Synthesis of high-quality nanoparticles using microfluidic platforms

Asset Metadata

Creator Seidel, Scott (author)

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

School Viterbi School of Engineering

Degree Doctor of Philosophy

Degree Program Chemical Engineering

Publication Date 06/30/2015

Defense Date 05/22/2015

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

Tag chemical vapor deposition,coatings,membranes,OAI-PMH Harvest,polymers

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Gupta, Malancha (committee chair), Malmstadt, Noah (committee member), Nakano, Aiichiro (committee member)

Creator Email sseidel@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c3-583630

Unique identifier UC11300595

Identifier etd-SeidelScot-3525.pdf (filename),usctheses-c3-583630 (legacy record id)

Legacy Identifier etd-SeidelScot-3525.pdf

Dmrecord 583630

Document Type Dissertation

Format application/pdf (imt)

Rights Seidel, Scott

Type texts

Source 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 a...

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