b» D! l
a .41. n . c u U I263.S~'.§flm

:ipiw.

 

th» in. v. .1 ¢,Q\‘....P£C

11

1.

 

 

 

 

 

a. .munriu 9: Ar
. .. .
u»...<.w$w.

.3.

F.

x... n ‘
It. 1: 2:! E
13:31:. : 55.2.
v. n: 3.1%. I

L 8

O.

4-00.31”! I
7” o. I! .r

 

 

 

 

 

 

   

4. a .\
Jad‘vduuh
It" fits
a." a.
t « q- (E .
.f.»L.fi.m
:xr: on“. .

.3»? 1ftn¢fi§§
. if e: .21... r. y
.514»: .Ic 1:.
. .~ 9‘ i

11...}: .1
If!!! 2,. .

 

2 .
317....
x 9.. v.
«flarinqifieuvlat
1. x» . ‘ u
‘14 .unfilIKhtml; .3
.QI p.11 .3

300}

This is to certify that the
dissertation entitled

Poly(ethylene oxide) based polymer brushes:
Synthesis, properties and applications

presented by

Ying Zheng

has been accepted towards fulfillment
of the requirements for the

PhD. degree in Chemistry

 

 

flgé’ééa

l/Major Professor’s Signature

ZS 4q9¢,5'/‘ 2.006

 

Date

MSU is an Affirmative Action/Equal Opportunity Institution

'—_———— -——- 4%,, —— 7

 

LIBRARY
Michigan State
University

 

 

 

 

PLACE IN RETURN Box to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE

DATE DUE

DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2/05 p2/CIRC/Date0ue.indd-p.1

 

 

POLY(ETHYLENE OXIDE) BASED POLYMER BRUSHES:
SYNTHESIS, PROPERTIES AND APPLICATIONS

By

Ying Zheng

A Dissertation
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
DOCTOR OF PHILOSOPHY
Department of Chemistry

2006

ABSTRACT
POLY(ETHYLENE OXIDE) BASED POLYMER BRUSHES:
SYNTHESIS, PROPERTIES AND APPLICATION
By

Ying Zheng

Surface-initiated atom transfer radical polymerization (ATRP) of poly (ethylene
glycol) methyl ether methacrylate (PEGMA) macromonomers from gold substrates leads
to thin films of comb-like polymer brushes. Polymerization of these monomers in water
resulted in essentially constant polymerization rates and yielded coatings with thicknesses
as high as 400 nm. Polymers with side chains consisting of 22-23 ethylene oxide
repeating units aligned to form crystalline polymers analogous to their non-brush analogs.
However, polarized optical microscopy confirmed their crystallization as two-
dimensional spherhulites, and the film morphologies, crystallization rates and lamellae
orientations all depended on the film thickness. The crystallization rate decreased with
the film thickness, as reflected by the film morphology changes. An Avrami analysis of
growth rates indicated heterogeneous nucleation and one-dimensional growth. Both
AFM images and reflectance FT-IR spectra of the films were consistent with crystalline
lamellae preferentially oriented normal to the surface for brushes with thicknesses above
100 nm. As the film thickness decreased, the favored orientation shifted to crystalline

lamellae parallel to the surface, with the PEO side-chains oriented perpendicular to the

 

 

 

surface. The lamellae orientations are related to polymer-surface interactions as well as to

the “comb-like” polymeric structure.

Polymerizations were optimized to grow comb-like polymer brushes on flat
surfaces and on silica nanoparticles. Kinetic studies of the growth of polymethacrylates
with (PEG) polyethylene glycol segments of various lengths identified optimum
polymerization condition for the growth of thick poly(PEGMA) films on planar
substrates. These materials have potential utility as membranes, electrolytes for lithium
ion batteries, and other applications.

Crosslinked brushes were grown from gold substrates by copolymerization of a
macromonomer, poly(ethylene glycol) methyl ether methacrylate (PEGMA), and a
crosslinker, bisphenol A ethoxylate dimethacrylate (BisA-EDMA) using surface-initiated
ATRP. The crosslink density of the film was estimated from FT-IR measurements and
from the aqueous swelling of the films, as measured by in-situ ellipsometry. These films
show increased dimensional stability with cross-link while keeping high swelling in
water. These crosslinked films are attractive for their mechanical stability and their

ability to resist protein adsorption in biological applications.

Comb-like and crosslinked polymer brushes containing PEO segments were
grafted on the surface of porous alumina membrane supports and the perrneabilities of
C02 and H2 through the membranes were studied. These films are highly permeable to
CO; and selectively permeate CO; with a C02 permeation coefficient as high as 300
Barrer and a COZ/Hz selectivity of 14 respectively for poly(PEGMA) film. The Cog/gas

permeability decreased for shorter PEG segments and for highly crosslinked films.

To My Famny

-iv-

a. ‘m’g

 

 

1.15 ..

ACKNOWLEDGMENTS

First, I would like to express my deep appreciation to my advisor, Professor
Gregory Baker. Thank you for your guidance and constant encouragement, which built
my confidence and enabled me to achieve my Ph.D. at Michigan State University. The
polymer science knowledge, strict scientific attitude and initiative in research that I have
learned from you will be invaluable in my future career. Professor Babak Borhan not
only taught me good spectroscopic techniques and organic chemistry, but he also helped
me in the first step of my Ph.D. studies in Baker’s group. I greatly appreciate him for his
trust and help. I learned from Professor Merlin Bruening a sincere approach to science
and critical insight. I also like to thank Professor Peter Wagner for being my committee
member.

I would like to give my thanks to all Baker group members: JB, Bao, Ping, Qin,
Leslie, F eng, quei, Jon, DJ, Erin and Sampa for useful discussions and for help in my
research and my life. I also learned good surface science from Dr. Wenxi, and I enjoyed
working with Lei, Tiffany, Matt and Anagi. I would like to thank all my friends for
making my life in MSU enjoyable and memorable. Special thanks to Lisa Dillingham for
her help during my stay in chemistry department, which made my life easier.

Finally, I want to thank my family. They have always been so supportive in every
stage of my life. My life would not have been the same without their love and support.

My dearest parents and my brother deserve my biggest thanks and love.

TABLE OF CONTENTS

Page
List of Tables ....................................................................................... ix
List of Figures ....................................................................................... x
List of Schemes .................................................................................... xvii
List of Abbreviations ............................................................................. xviii
Chapter I. Introduction to polymer brushes ................................................. 1
Synthesis of polymer brushes by surface-initiated polymerization ........................ 4
Surface-initiated living anionic and living cationic polymerizations ............... 5
Surface-initiated ring opening metathesis polymerization (ROMP) ................ 9
Surface-initiated ring opening polymerization (ROP) .............................. 12
Surface-initiated controlled radical polymerization (CRP) ......................... l4
Atom Transfer Radical Polymerization (ATRP) ........................................... 17
ATRP kinetics ............................................................................ 19
ATRP components ....................................................................... 20
ATRP of PEGMA ........................................................................ 3O
Polymer brushes formed by surface-initiated ATRP ....................................... 32
Methods for the controlled growth of polymers from surfaces ..................... 32
Controlled ATRP from surfaces of particles .......................................... 36
Physicochemical properties of polymer brushes ........................................... 38
Physical states of polymer brushes .................................................... 38
Stimuli-responsive polymer brushes ................................................... 40
Applications of polymer brushes ............................................................ 47
Nanotechnology—Patteming of polymer brushes ................................... 47
Biological application in protein resistance and cell adhesion ..................... 49
Polymer brushes grafied membranes ................................................... 51
References ....................................................................................... 53
Chapter II. Crystallization of Polymer Brushes ............................................ 61
Introduction ...................................................................................... 61
Experimental Section ........................................................................... 66
Materials .................................................................................. 66
Synthesis of initiator [Br-C(CH3)2-COO(CH2)118);] ............................... 66
Preparation of immobilized initiators on Au surfaces .............................. 67
Surface-initiated polymerization of PEGMA ........................................ 68

-vi-

Solution ATRP of PEGMA ............................................................ 68

Characterization methods ............................................................... 69
Crystallization studies ................................................................... 69
Results and discussion ......................................................................... 70
Preparation of polymer brushes ........................................................ 70
Morphology of crystalline polymer brushes .......................................... 75
Spectroscopic characterization ......................................................... 82
Crystallization kinetics of polymer brushes .......................................... 90
Conclusions .................................................................................... 100
References ..................................................................................... 101

Chapter 111. Synthesis of Comb-like Polymer Brushes on Planar and
Nanoparticulate Surfaces104

Introduction .................................................................................. 104
Experimental ................................................................................. 106
Results and Discussion ..................................................................... 112
Surface-initiated ATRP of PEGMA on Au ......................................... 112
Rapid grth of comb-like polymer brushes on the surface ..................... 119
Synthesis of PEG-functionalized nanoparticles .................................... 124
Growth of polymer brushes on silica wafers ....................................... 124
Synthesis of comb-like polymers on the surface of fumed silica ................ 127
Summary and Outlook ..................................................................... 132
References .................................................................................... l 33
Chapter IV. Crosslinked Polymer Brushes Containing PEO Segments .............. 135
Introduction .................................................................................. 135
Experimental ................................................................................. 136
Materials ................................................................................. l 36
Characterization methods .............................................................. 136
Synthesis of crosslinked polymer brushes .......................................... 137
Determination of cross-link density of polymer brushes .......................... 138
Results and Discussion ..................................................................... 140
Synthesis of crosslinked polymer brushes .......................................... 140
Characterization of crosslinked polymer fihns ..................................... 143
Swelling behavior of crosslinked copolymer films ................................ 148
Conclusion and Outlook ................................................................... 150
References ................................................................................... 1 5 1

-vii-

Chapter V. Polymer Membranes for Gas Separation .................................... 152

Background .................................................................................. 1 52
Experimental ................................................................................. 158
Materials ................................................................................. 1 5 8
Characterization methods .............................................................. 1 5 8

Anchoring initiators on the surface of porous alumina membrane support. . 1 59
Surface-initiated ATRP of . poly(PEGMA) films using CuBr/CuBrszpy

catalyst ................................................................................... 1 59
Polymerization of PEGMA using CuBr/Me4Cyclam/CuBr2(anbpy)2
catalyst complexes ..................................................................... 160
Growth of crosslinked copolymer film .............................................. 160
Gas permeation measurements ....................................................... 161
Results and Discussion ..................................................................... 163
Synthesis of and characterization of polymer fihns ................................ 163
Pure gas permeation of poly(PEGMA) brushes .................................... 165

Pure gas permeation through crosslinked polymer films [poly (PEGMA-co-
Summary and Outlook ..................................................................... 177
References .................................................................................... 179

- viii -

LIST OF TABLES

Table Page
Table 2.1. Crystalline PEO absorption band assignment ...................................... 84

Table 2.2. Avrami exponent n determined from 2-D spherulite growth monitored by in
situ polarized optical microscopy ................................................................. 94

Table 3.1. Reaction conditions for polymerization of PEGMA(22-23) and PEGMA(4-5)

from the surface of fumed silica particles ...................................................... 129
Table 3.2. Results of TGA control experiments ............................................... 130
Table 3.3. DSC results of poly(PEGMA) silica particle composites ........................ 131
Table 5.1 Physical properties of H2, C02, and N2 gases ..................................... 154
Table 5.2. Representative examples of pure gas C02 permeability and COz/Hz selectivity
in PEG-containing polymer ..................................................................... 157
Table 5.3. Summary of gases permeation through poly(PEGMA) films .................. 171

Table 5.4. Gas permeability coefficients and average selectivities for crosslinked
poly(PEGMA-co-BisA-EDMA) membrane ................................................... 176

-ix-

LIST OF FIGURES

Figure Page

Figure 1.1. Polymer brushes on flat and curved surfaces ....................................... 1

Figure 1.2. “Grafting to” (a) and “grafting from” (b) methods for the generation of
polymer brushes. Adapted with permission from J. Am. Chem. Soc. 1999, 121, 1016-
1022. Copyright ©1999 American Chemical Society ........................................... 3

Figure 1.3. Examples of polymer brushes grown by living anionic polymerization. . . . . ....7

Figure 1.4. Scheme showing the surface-initiated polymerization of 2—oxazolines for the
preparation of the amphiphilic nanocomposites, with 2-ethyl—2-oxazoline and N,N-di-n-
octadecylamine used as monomer and terminating agent, respectively. Adapted with
permission from Macromolecules, 2001, 34, 1606-1611. Copyright © 2001 American
Chemical Society ..................................................................................... 8

Figure 1.5. Polymer brushes grown by living cationic polymerization ........................ 9

Figure 1.6. Surface-initiated ROMP applied to the synthesis of norbornene-based
polymer brushes ..................................................................................... 10

Figure 1.7. Graphic representation of ROMP initiated by DPN and a monomer coated
AFM tip. Reprinted with permission from Angew. Chem, Int. Ed. 2003, 42, 4785-4789.

Copyright © 2003 Wiley Intersciences ........................................................... 11
Figure 1.8. PCL and PLA brushes formed by surface-initiated ROP ........................ 12
Figure 1.9. Synthesis of poly(HEMA)—g-(polylactide) ......................................... 13
Figure 1.10. Synthetic strategy for the anionic ring-opening polymerization of ethylene
oxide initiated from a silica surface modified by Al(OiPr)3/GPS ............................ 14
Figure 1.11. Synthesis of OEG containing polymer brushes on the silica surface. . . . . 17
Figure 1.12. Structure of the [CuI (bpy)2]+ cation in [CuI (bpy)2]+[ClO4]’ .................. 26
Figure 1.13. Structure ofthe [Cu’ (bpy) x2] (X=Br, c1) dimer .............................. 26
Figure 1.14. Molecular structure of [CuI(HI\/ITETA)]+[Cu‘C12]_ ............................. 27

Figure 1.15. Structure of the [Cu"(bpy)2Br]+ cation in [Cu"(bpy)2Br]+[Br]- ............... 27
Figure 1.16. Molecular structure of [Cu"(HMTETA)Br]+[Br]’ .............................. 28
Figure 1.17. General synthetic route to polymer brushes on gold and silica substrates... 33

Figure 1.18. Schematic illustration of linear, comb-like and crosslinked polymer
brushes ................................................................................................ 35

Figure 1.19. Synthesis of mixed PMMA/PS brushes from an asymmetric di-functional
initiator-terminated SAM (Y—SAM) by combining ATRP and NMRP techniques.
Reprinted with permission from J. Am. Chem. Soc. 2004, 126, 6124-6134. Copyright ©
2004 American Chemical Society ................................................................ 36

Figure 1.20. Poly(PEGMA) and PMEMA brushes grown from silica nanoparticles ...... 37

Figure1.21. Preparation of comb-coil polymer brushes on the surface of silica
nanoparticles ......................................................................................... 38

Figure 1.22. Synthetic scheme of side chain LCP brushes by surface-initiated ATRP. . ..39
Figure 1.23. Speculative model for nanopattem formation in di-block copolymers ....... 41
Figure 1.24. Reversible response of triblock copolymer brushes to different solvents. . ..41

Figure 1.25. Tapping mode AFM height images of PDMS-b-PS-b-PDMSA brushes after
the following treatments: (left) after immersion in toluene and drying with nitrogen,
(right) after immersion in toluene, gradual addition of hexane and drying with nitrogen.
Reprinted with permission from Macromol. Chem. Phys. 2004, 205, 411. Copyright ©
2004 Wiley Intersciences .......................................................................... 43

Figure 1.26. 1H NMR spectra of PAA/PS particles dispersed in (a) CDCl3, (b) DMF-d7,
and (c) CD3OD. A drop of DMF-d7 was added into the pa1ticles prior to CD0; and
CD3OD to increase the concentration of the dispersed nanoparticles. Reprinted with
permission from J. Am. Chem. Soc. 2005, 127, 6248-6256. Copyright © 2005 American
Chemical Society ................................................................................... 44

Figure 1.27. Diagram showing reversible formation of intermolecular hydrogen bonding
between PNIPAAM chains and water molecules (left) and intrarnolecular hydrogen
bonding between the C=O and N-H groups in PNIPAAM chains (right) below and above
the LCST. This mechanism is proposed to explain the thermally responsive wettability of
a PNIPAAM thin film .............................................................................. 45

-xi-

Figure 1.28. Surface-roughness-enhanced wettability of a PNIPAAM-modified surface. a)
The relationships between groove spacing (D) of rough surfaces and the water contact
angle at 25 ° C (A), and at 40 C (I). b) Water drop profile for thermally responsive
switching between super-hydrophilicity and super-hydrophobicity of a PNIPAAM-
modified rough surface with a groove spacing of about 6 pm, at 25 °C and 40 °C. The
water contact angles are ~O ° and 149.3 °, respectively. 0) The temperature (T)
dependence of water contact angles for PNIPAAM thin films on a rough substrate with
groove spacing of about 6 pm (A) and on flat substrate (I). (1) Water contact angles at
two different temperatures for a PNIPAAM-modified rough substrate with a groove
spacing of 6 pm. Half cycles: 20 °C; and integral cycles: 50 °C. Reprinted with
permission from Angew. Chem, Int. Ed. 2004, 43, 357-360. Copyright © 2004 Wiley
Intersciences ........................................................................................ 46

Figure 1.29. (top) Preparation of surface-confined PNIPAAM polymer brush
nanopattems by combining “nanoshaving” and surface-initiated ATRP using a surface-
tethered thiol initiator. (Bottom) Contact mode AF M height images (20 um X 20 um) and
corresponding typical height profiles of a PNIPAAM brush line nanopattem imaged at
room temperature in (a) air, (b) MQ-grade water, and (c) a mixture of MeOH/water (1:1,
v:v). Reprinted with permission from Nano Lett. 2004, 4, 373-376. Copyright © 2004
American Chemical Society ....................................................................... 48

Figure 1.30. A) a PEGMA polymerized by ATRP from a micropattemed gold surface. (B)
NIH 3T3 fibroblast cells cultured on patterns of adsorbed fibronectin (20 um circles and
40 um stripes). Reprinted with permission from Biomacromolecules, 2005, 6, 2427-2448.
Copyright © 2005 American Chemical Society ......................................... 50

Figure 1.31. Preparation of polymer brushes grafted to porous membranes and an
illustration of the mechanism for regulating water and solute permeation through a
polymer brush grafted membrane and a hydrogel membrane ................................. 51

Figure 2.1. Architecture of comb-like poly(PEGMA) brushes grown from a surface. . 63

Figure 2.2. Reflectance FTIR spectra of (a) an initiator monolayer, and (b) an amorphous
114 nm thick poly(PEGMA) film. Spectrum b was taken immediately after its synthesis.
Trace (c) shows a poly(PEGMA) film of the same fihn thickness after crystallization at
room temperature .................................................................................... 73

Figure 2.3. Evolution of the ellipsometric film thickness with time for the ATRP of

PEGMA at room temperature from initiators anchored to gold surfaces. Conditions:
[CuBr] = 1 mM, [CuBrz] = 0.3 mM, [bpy] = 3 mM, and [M]:[H20] = 2:1 (v/v) ........... 74

-xii-

Figure 2.4. Optical micrographs of poly(PEGMA) brushes grown from gold surfaces and
viewed through crossed polarizers: (a) and (b) show the crystallization of a 240 nm at 10
°C and at 5 °C, respectively; (c) and (d) show the crystallization of 148 nm and 76 nm,
films, respectively at 5 °C. Images in this dissertation are presented in color .............. 76

Figure 2.5. AFM images of a 108 nm crystalline polymer brush film. At left are the
height images and at right are the corresponding 3-D views. In panel a, the scan distance
is 4 x 4 pm, with a 2 distance of 25.44 nm. In panel b, the scan distance is 5 X 5 pm, and
the 2 distance is 30.42 nm. Images in this dissertation are presented in color .............. 77

Figure 2.6. AFM images of crystalline polymer brushes. (3) shows data for a 310 nm
thick film and (b) shows data for a the 122 nm film. Both image are 5 x 5 pm, and the
heights of the 310 and 122 nm films are ~20 and 41 nm, respectively. Images in this
dissertation are presented in color ................................................................ 79

Figure 2.7. (Left) AFM height images (5 x 5 pm) of crystalline polymer brushes with
thickness of: (a) 92nm, 2 distance is 24.49nm; (b) 70nm, 2 distance is 31.46nm; (c) 35nm,
2 distance is 10.97nm. (Right) Phase images of crystalline polymer brushes with
thickness of: (a) 92nm, 2 distance is 475mv; (b) 70nm, 2 distance is 919.13mV; (c) 35nm,
2 distance is 561mv. Images in this dissertation are presented in color ..................... 80

Figure 2.8. AFM height images of a crystalline film with thickness of 55 nm. Scan
distance of the left image is 5 x 5 pm with a 2 distance of 15 nm. The scan distance of the
right image is 1 x 1 pm with a z distance of 15 nm. Images in this dissertation are
presented in color ................................................................................... 81

Figure 2.9. Crystalline structure of PEO with top view and side view ...................... 83

Figure 2.10. Reflectance FTIR spectra of poly(PEGMA) brushes of various thicknesses; a)
13 nm;b) 35 nm; e) 70m; (1) 92 nm; e) 177 nm;f)240nm .............................. 86

Figure 2.11. Lamellae orientations in the films of poly-PEGMA: (a) Chain axis of PEO
side chains is oriented perpendicular to the substrate, lamellae parallel to the surface; (b)
Chain axis of PEO side chains is parallel to the substrate, edge-on lamellae ............... 89

Figure 2.12. Optical micrographs showing the growth of spherulites in a 240 nm thick
film at 23 i 2 °C. The samples are viewed using crossed polarizers. The scale bar
represents 100 um. Images in this dissertation are presented in color ....................... 92

Figure 2.13 a) The evolution of spherulite radii with time for the crystallization of
poly(PEGMA) brush films of various thickness at room temperature. b) Spherulite
grth rates measured as a function of the poly(PEGMA) film thickness .................. 93

- xiii -

Figure 2.14. Linear fit of the Avrami equation to the experimental data for crystalline
polymer brushes with different film thickness .................................................. 95

Figure 2.15. ‘H-NMR spectrum of poly(PEGMA) ............................................. 97

Figure 2.16. Reflectance FT-IR spectra of 56 and 290 nm thick poly(PEGMA) films that
were spin-coated on the gold surfaces ........................................................... 98

Figure 2.17.Crystal growth rates measured for spin-coated poly(PEGMA) and
poly(PEGMA) brushes of comparable film thickness (~300 nm) ............................ 99

Figure 3.1 Evolution of the ellipsometric film thickness with time for the grth of
poly(PEGMA)(22-23) from Au. Conditions: [M]:[H20] = 3:2 (v/v). (a) [CuBr] =10 mM,
[CuBrz] = 5 mM, [bpy]= 30 mM. (b) [CuCl] =10 mM, [CuBrz] = 5 mM, [bpy]= 30 mM.
The curves are a guide to the eye ................................................................ 113

Figure 3.2. Evolution of the ellipsometric film thickness with time for the growth of
poly(PEGMA)(8-9) from Au. Conditions: (a) [CuBr] = 28.5 mM, [CuBrz] = 9.6 mM,
[bpy] = 85.5mM; (b) [CuCl] = 28.5 mM, [CuBrz] = 9.6 mM, [bpy] = 85.5 mM. The

curves are a guide to the eye ..................................................................... 114

Figure 3.3. Evolution of the ellipsometric film thickness with time for the grth of
poly(PEGMA)(22-23) films from Au. [M]:[H20] = 3:2 (v/v). (a) [CuBr] = 0.7 mM,
[CuBrz] = 0.21 mM, [bpy] = 2.1 mM; (b) [CuBr] = 2 mM, [CuBrz] = 0.6 mM, [bpy] = 6
mM; (c) [CuBr] = 5 mM, [CuBrz] = 1.5 mM, [bpy] = 15 mM; (d) [CuBr] = 10 mM,
[CuBrz] = 3 mM, [bpy] = 30 mM ............................................................... 116

Figure 3.4. Evolution of the ellipsometric film thickness as a function of time for the
polymerization of poly(PEGMA)(8-9). [M]:[[HZO] = 1:1.(a) [CuBr] = 28.5 mM, [CuBrz]
= 9.6 mM, [pr] = 85.5 mM. (b) [CuBr] = 3 mM, [CuBrz] = 1 mM, [pr] = 9 mM. The
curves are a guide to the eye ..................................................................... 117

Figure 3.5. Evolution of the ellipsometric film thickness with time for the polymerization
of PEGMA(8-9) at different monomer concentrations. Conditions: [CuBr] = 28.5 mM,
[CuBrz] = 9.6 mM, [bpy] = 85.5 mM. (a) [M]:[HzO]=2:1 and (b) [M]:[[HZO]=I:1. The
curves are a guide to the eye ..................................................................... 118

Figure 3.6. Evolution of the ellipsometric fihn thickness with time for the polymerization
of PEGMA(4-5) using Me4Cyclam as the ligand. Conditions: [M]:[HzO+DMF] = 1:1;
[CuBr] = 2 mM, [CuBrz/(anbpy)2] = 0.6 mM, [Me4Cyclam] = 6 mM ................... 120

Figure 3.7. Evolution of the ellipsometric film thickness with time for the polymerization
of PEGMA(4-5) using HMTETA as the ligand. Conditions: [M]/[HzO+DMF] = 1:1,

-xiv-

[CuBr] = 2 mM, [CuBrz] = 0.6 mM, [HMTETA] = 6 mM. The data points are the average
of two independent polymerizations ............................................................ 121

Figure 3.8. Evolution of the ellipsometric film thickness with time for the polymerization
of PEGMA(4-5) using HMTETA“) and Me4Cyclam (I) as ligands. The data are re—
plotted from Figures 3.6 and 3.7 ................................................................. 123

Figure 3.9. Evolution of the ellipsometric film thickness with time for the polymerization
of PEGMA(4-5). Conditions: [M]:[H20] = 2:1, (a) [CuBr] = 2 mM, [CuBrz] = 0.6 mM,

[bpy] = 6 mM; (b) [CuBr] = 1 mM, [CuBrz] = 0.3 mM, [bpy] =3mM ..................... 126
Figure 3.10. FT-[R spectra of (a) initiator anchored fumed silica particles; (b)
poly(PEGMA)(22-23) grown from fumed silica particles .................................... 128
Figure 3.11. TGA analysis of: (a) bare fumed silica; (b) initiator-anchored fumed silica;
(c) poly-PEGMA (22-23) attached fumed silica .............................................. 130
Figure 3.12. TGA analysis of (a) bare fumed silica; (b) initiator-anchored fumed silica; c)
poly(PEGMA)(4-5) attached fumed silica ...................................................... 132
Figure 4.1 Crosslinking and branching in films containing BisA-EDMA ................. 139

Figure 4.2 Reflectance FT-IR spectra of polymer films grown from an initiator anchored
planar Au surface:(a) a poly(PEGMA) film and (b) a crosslinked film with 30 mol%
crosslinker in the feed ............................................................................. 142

Figure 4.3. Evolution of film thickness with time for the copolymerization of PEGMA
and BisA-EDMA. The overall concentration of monomer and crosslinker for all runs is
0.4 M. Catalyst: [CuBr] = 1mM; [CuBrz] = 0.3mM; [bpy] = 3mM. The inset shows the
mole % composition of BisA-EDMA in the monomer feed solution ....................... 144

Figure 4.4. The relationship between the feed composition and the measured film. . 145

Figure 4.5. The fraction of crosslinkers in which both two double bonds polymerized.
Each point is the average of at least three independent samples ............................ 146

Figure 4.6. Crosslink density vs. the mole percentage of crosslinker in feed ............ 147

Figure 4.7. a) (Top) thickness of dry and water-swollen films as a firnction of the mole
percentages of crosslinker in the monomers feed solution; b) (bottom) the relationship
between the normalized swelling of films and the fihn crosslink density ................. 149

Figure 5.1. Apparatus for gas permeation measurements .................................... 162

-xv-

Figure 5.2. Synthesis of composite membranes with polymer skins ....................... 164

Figure 5.3. Transmittance FT-IR spectrum of poly(PEGMA)(22-23) and poly(PEGMA)
(4-5) grafted to porous alumina membrane supports .......................................... 165

Figure 5.4. Single-gas fluxes of CO2 and H2 through fihns prepared by polymerization of
PEGMA from porous alumina as a function of trans-membrane pressure drop. The outlet
pressure was 1 atm and the measurements were performed at room temperature. . . . . ....166

Figure 5.5. FESEM image of the filtrate side of a porous alumina after grth of poly-
PEGMA film from the surface .................................................................. 167

Figure 5.6. Single-gas fluxes of CO2 and H2 through poly-PEGMA grafted membrane as
a function of trans-membrane pressure drop. The initially crystalline film was soaked in
water, and then measurements were taken after the film was dried ........................ 168

Figure 5.7. Single-gas fluxes of CO2 and H2 through films prepared by polymerization
of PEGMA with 8-9 ethylene oxide units from porous alumina as a function of trans-
membrane pressure drop .......................................................................... 169

Figure 5.8. F ESEM image of the filtrate side of porous alumina after growth of (a) poly-
PEGMA(8-9) fihn and (b) poly-PEGMA(4-5) film .......................................... 170

Figure 5.9. Single-gas fluxes of CO2 (I), H2 (9) and N2 (A) through cross-linked co-
polymer films prepared by polymerization of PEGMA with a cross-linker (BisA-EDMA)
from porous alumina as a function of trans-membrane pressure drop. For the film from
50mol% crosslinker, the flow rate of N2 is too slow to be measured ....................... 173

Figure 5.10. FESEM images (cross-sectional) of porous alumina (0.02 pm surface pore
diameter) after grth of crosslinked films, which were polymerized from the monomer
solution containing (a) 30mol% crosslinker; (b) 50 mol% crosslinker; (c) 80 mol%
crosslinker; and (d) 100 mol% crosslinker (a poly(BisA-EDMA) film) ................... 174

-xvi-

LIST OF SCHEMES

Scheme Page

Scheme 1.1. The general mechanism of controlled radical polymerization and some

typical examples .................................................................................... 15
Scheme 1.2. The mechanism of ATRP of poly(ethylene glycol) methyl ether

methacrylate (PEGMA) ............................................................................ 18
Scheme 1.3. Representative monomers used for ATRP ....................................... 21
Scheme 1.4. Representative examples of initiators used for ATRP .......................... 23
Scheme 1.5. Examples of ligands often used for copper-mediated ATRP .................. 25

Scheme 1.6. Synthetic approaches used to incorporate PEO segments in polymers 31

Scheme 2.1. Synthetic procedure for preparation of initiator on Au substrate ............. 68
Scheme 2.2. Growth of poly(PEGMA) films from gold ....................................... 72
Scheme 3.1. Surface-initiated ATRP of PEGMA from gold surfaces ...................... 105
Scheme 3.2. Synthetic route to (11-(2-bromo-2-methyl)propionyloxy) undecyldimethyl-

chlorosilane ......................................................................................... 109
Scheme 3.3. Growth of comb-like PEO polymer brushes fi'om SiO2 surfaces ........... 125

Scheme 3.4. Modification of silica particles by surface-initiated ATRP of PEG-
methacrylates ...................................................................................... 127

Scheme 4.1. Synthetic route to crosslinked copolymer films grown from gold
substrates ........................................................................................... 141

Scheme 5.1. Structures of representative PEG (meth)acrylates and PEG di(meth)acrylates
used in this study and reported in the literature for membrane separations ................ 156

- xvii -

AFM

AGET
ATRP
BisA-EDMA

bpy

uCP

CRP
DMAEMA
DMF

DP

dNbpy
anbpy
DSC

DPE

DPN
EDTA
EGDMA
FESEM
FT -IR
GMA

GPC

GPS

HEA
HEMA
HMTETA
kp
kt
LCP

LCST
LDPE
Me4Cyclam
MegTREN
MA

LIST OF ABBREVIATIONS

atomic force microscopy

activator generated by energy transfer

atom transfer radical polymerization
bisphenol A ethoxylate dimethacrylate (Mn =1 ,700g/mol)
bipyridine (or 2,2’-dipyridyl)

micro—contact printing

controlled radical polymerization

dimethyl amino ethyl methacrylate

N, N-Dimethylformamide

degree of polymerization
4,4’-di(5-nonyl)-2,2’-bipyridine
4,4’-di(n-nonyl)-2,2’-bipyridine

differential scanning calorimetry
diphenylethylene

dip pen nanolithography
ethylenediaminetetraacetic acid

ethylene glycol dimethacrylate

field emission scanning electron microscopy
Fourier transform infrared spectroscopy
glycidyl methacrylate

gel permeation chromatography
3-glycidyloxypropy1 trimethoxysilane
2-hydroxy ethyl acrylate

2-hydroxy ethyl methacrylate

1 , 1 ,4,7,1 0, 10-hexamethyltriethylenetetramine

propagation rate

termination rate

liquid crystalline polymer

lower critical solution temperature

low density polyethylene

1 ,4,8,1 1-tetramethyl-1 ,4, 8,1 l-tetraazacyclotetradecane
tris[2-(dimethylamino)ethyl]amine
methyl acrylate

methyl methacrylate

number average molecular weight

weight average molecular weight
nitroxide mediated radical polymerization
nuclear magnetic resonance

N-propyl (2-pyridyl) methanimine
oligo(ethylene glycol)

- xviii -

 

PAA
PCL
PDI
PDMS
PDMSA
PEO
PEG
PEGMA

PEGDA
PEGDMA
PGMA

PI

PLA
PMEMA
PMDETA
PMMA
PNIPAAM
PS

PtBA

Rp

Rs

RAFT
ROP
ROMP
SAM
TEMPO

TGA
THF

poly(acrylic acid)

poly(e-caprolactone)

polydispersity index calculated as MW/Mn
poly(dimethoxysiloxane)
Poly(3-(dimethoxymethylsilyl)propyl acrylate)
poly(ethylene oxide)

poly(ethylene glycol)

poly(ethylene glycol) methyl ether methacrylate,
Mn= 300, 475, and 1100 g/mol
poly(ethylene glycol) diacrylate
poly(ethylene glycol) dimethacrylate
poly(glycidyl methacrylate)

polyisoprene

poly(lactic acid) or poly(lactide)
poly((2-N-morpholino)ethyl methacrylate)
N, N, N’, N”, N”-pentamethyl diethylene triamine
poly(methyl methacrylate)
poly(N-isopropylacrylamide)

polystyrene

poly(tert-butyl acrylate)

rate of polymerization

radius of gyration

reverse addition-fragmentation transfer
ring opening polymerization

ring opening metathesis polymerization
self-assembled monolayer

2, 2, 6, 6-tetramethyl-1-piperidyloxy

glass transition temperature

thermal gravimetric analysis
tetrahydrofuran

melting point

-xix-

Chapter I

Introduction to Polymer Brushes

Thin polymer films are useful for modifying surface properties. Common
methods for the deposition of polymers on surfaces include Spin and Spray coating,
however, physically adsorbed films are easily displaced by good solvents. In contrast, the
chemical stability and mechanical robustness of polymer brushes have led to their
widespread use for surface modification. The term “polymer brush” generally refers to an
ensemble of polymer chains with each chain bound by one end to a surface or interface.
The high tethering density forces the chains to stretch away from the surface or interface,

resulting in a brush height greater than the polymer radius of gyration (Rgfl’2

221122216

Figure 1.1. Polymer brushes on flat and curved surfaces.

In general, polymer brushes are covalently bound to surfaces by reaction of
functional groups on polymer chains with surface reactive sites, or by selective
adsorption of block copolymer segments that strongly interact with the surface. Non-
covalently attached polymer brushes also suffer from poor stability in good solvents. This
discussion will emphasize covalently bound polymer brushes.

-1-

Strategies for anchoring polymer brushes onto surfaces can generally be
categorized as “grafting to” and “graftingfrom” methods (Figure 1.2).3 The “grafting to”
approach refers to tethering preformed polymer chains onto surfaces, which facilitates
formation of polymer brushes with well-defined architectures. However, this method is
limited in that it usually results in low polymer grafting densities and low film thickness
due to the steric hindrance of previously attached polymer chains. In contrast, the
“grafting fiom” method involves attaching initiators on the surface, and initiating
polymerization of monomers from the surface. Since the only limit to the polymer chain
propagation is the diffusion of monomers to the surface reactive sites, this method usually
generates thicker polymer films with higher grafting densities than the “grafting to”
approach.4

The “grafting fiom” method is also called surface-initiated polymerization. In
order to achieve a high degree of control over the polymer brush architecture, the use of
controlled polymerizationss’ 6 to provide well-defined polymers with low polydispersity
(PDI) is highly desirable. Controlled surface-initiated polymerizations offer several
advantages when forming well-defined dense and thick films including: 1) good control
over the film thickness by controlling the polymerization time, 2) simplified separation
and purification issues since the active radicals are confined on the surface and no
polymers are formed in solution, eliminating the need for exhaustive extraction,7 3)
tunable grafting densities by controlling the surface coverage of the initiator,8 4) the
ability to simultaneously grow polymer chains from multifunctional cores or surfaces,
and 5) the ability to control brush architectures, ranging from linear tethered polymers,

graft (comb-like) copolymers, hyperbranched (star-like or dendritic) polymers, to cross-

a) “Grafting to” b) “Grafting from”

M M
M
M

R. in i
............. l Propagating
chain ends

L
Layer

thickness I I H 2,

, ”Ill _. SAM

average distance between
grafting points

 

Figure 1.2. “Grafting to” (a) and “grafting from” (b) methods for the generation of
polymer brushes. Adapted with permission from J. Am. Chem. Soc. 1999, 121, 1016-
1022. Copyright ©1999 American Chemical Society.
linked polymer networks. Surface-initiated polymerization also provides good control
over the “vertical” compositions of tethered linear polymers (normal to the surface)
through the growth of block copolymers. Polymer brushes with varying compositions and
dimensions have been prepared by their controlled growth from flat surfaces (gold, silica
wafers, and glass slides), non—planar surfaces (gold, silica, latex particles, magnetic
beads, polystyrene particles, carbon nanotubes), porous supports, chromatography media,
and other materials."‘10
The physical properties of polymer brushes largely depend on the grafting
densities, polymer chain lengths and flexibility, as well as the brush architecture. Polymer
brushes have potential applications in many areas and they can greatly affect surface
properties such as wettability and adhesion. Polymer brushes function as “smart” surfaces

that are responsive to the environmental stimuli.”l3 barriers for chemical etching,”"6

layers resistant to adsorption of cells and biomolecules,l7 dielectric layers in electronic
device,18 and other applications.

Atom transfer radical polymerization (ATRP) is one of the most popular
controlled polymerization methods due to its simplicity and functional group tolerance.”
2' ATRP also is a versatile technique for preparing well-defined (co)polymer films. In
addition, ATRP of various monomers from preformed firnctional surfaces or colloidal
initiators has provided polymer brushes with precise dimensions and functionalities. The
nanoscale features or nano-objects in these films are typically characterized by various

techniques, such as ellipsometry, contact angle measurements, and atomic force

microscopy (AF M).

Synthesis of polymer brushes by surface-initiated polymerization

Surface-initiated polymerization generally includes anchoring initiator on the
surface followed by the polymerization of monomer from the initiator. Initiators can be
anchored to surfaces using well-known chemistry including the formation of self-
assembled monolayers (SAM), such as thiols self-assembled on gold surfaces, silylation
of silicon or glass surfaces, or by plasma treatment of substrates to directly form initiating
species on polymer surfaces.

Different types of surface-initiated polymerizations have been reported, including

22-27 28-30 2

anionic, cationic, conventional free radical?" 3 ring opening metathesis

”'36 and ring-opening polymerization (ROP).”39 Of these,

polymerization (ROMP),
controlled radical polymerization (CRP) is prominent due to its broad functional group

tolerance compared to anionic and cationic polymerizations, and its ability to form

polymers with high molecular weight and a PDI usually less than 1.5. Conventional

-4-

anionic polymerizations are the prototype living polymerization, but anionic
polymerizations from surfaces have slow growth rates (usually days to generate a 100 nm
film), and the polymerization system is sensitive to oxygen, moisture and impurities.
Living cationic polymerization also can generate complex polymer architectures and
polymers with low PDIs. However, the film growth rates for surface initiated cationic
polymerizations are even slower than their anionic analogs, and also quite sensitive to
moisture and other impurities. Much faster polymerization rates and films of predictable
thicknesses are provided by traditional free radical polymerizations from surfaces,
however free radical polymerization is incompatible with formation of certain
architectures such as block copolymers. By maintaining a low concentration of active
radicals in the system to suppress termination reactions, CRPs provide the advantages
characteristic of living systems such as anionic and cationic polymerization. ATRP is
tolerant of many functional groups and impurities, and because it is relatively easy to
implement, it is the most popular CRP methodology for the controlled synthesis of
polymer brushes. ROP has been used for preparing many commercially important
polymers such as poly(s-caprolactone) (PCL) and polylactide (PLA). Strained cyclic
monomers such as norbomenes are polymerized by ROMP. Some examples of surface-
initiated polymerizations will be provided before the discussion of surface-initiated

ATRP.

Surface-initiated living anionic and living cationic polymerizations

Due to its lack of chain transfer and termination reactions, anionic
polymerizations provide the best possible control over polymer architectures. Surface-

initiated anionic polymerization has been used to grow polymer brushes from various

-5-

small particles such as silica gel, carbon black, and silicon wafers. However, the
challenge of using anionic polymerization is its stringent polymerization conditions, slow
film growth rates, and the limited range of monomers that are compatible with anionic
polymerizations. Jordan et al.24 used lithiated biphenyl SAMS and initiated the
polymerization of styrene from gold substrates, obtaining uniform 18 nm thick films after
a three-day reaction. Later, Advincula et al.22 activated diphenylethylene (DPE) SAMS
with n-butyllithium to initiate anionic polymerization of styrene. A 5-day reaction gave a
polystyrene (PS) film only 23 nm thick. However, the formation of polystyrene-block-
polyisoprene (PS-b-PI) copolymer brushes demonstrated the living nature of the
polymerization. Polymer brushes have also been grown from DPE derivatives anchored
through quaternary ammonium tethers to clay surfaces. Zheng et a1.25 grafted
polybutadiene brushes to the surface of PS nanoparticles, and reported that the grafting
density and degree of polymerization greatly affect the dispersion of nanoparticles in
matrices and the mechanical properties of nanocomposites. F oster26 used DPE-containing
SAMS to generate polyisoprene-block-poly(ethylene oxide) (PI—b-PEO) block copolymer
brushes from silicon surfaces, creating a hydrophilic surface by introducing the PEO

block. Again, only thin films (24 nm) were obtained. Recently, Advincula et al.27

grew
14 nm thick PEO homopolymers and polyisoprene-block—poly(methyl methacrylate) (PI-
b-PMMA) copolymer brushes on gold surfaces. From these examples, we see that the
traditional monomers used for anionic polymerizations can be applied to surface-initiated

polymerization, however, obtaining thin films after long reaction times restricts broader

application of anionic polymerization to the synthesis of polymer brushes.

Figure 1.3. Examples of polymer brushes grown by living anionic
polymerization.

H. .i- . .. .E?
””“l ”5 “ES; Si“ 7L

WCHz-CH+
Adapted from Jordan et al. J Am. Chem. Soc. 1999 121 1016 1022

b)

Adapted from Foster et al. Macromolecules 2002 35 9964 9974

c)

AU S—CHZ n-BuLi

 

 

S‘CHZ

 

O

 

phosphonium base

Adapted from Advincula et al. J. Polym. Sci. A .' Polym. Chem 2006 44 769

Little work has been reported on the growth of polymer brushes using cationic
polymerization. Ulman et al. reported surface-initiated cationic polymerization of 2-
oxazolines from gold coated wafers28 and gold nanoparticle surfaces.29 Figure 1.4 shows
the synthetic pathway used to obtain amphiphilic polymer brushes using this chemistry.

i

orf ‘

O
O
Et—(Oj
N
6802CF3 Et Et 7?
0(soz-cr=3)2
S

H

HOH

O

.40 >—~

n-1

—b —>

 
  

    

  
 
 

 

—-b
Au
nanoparticle ©
N(CH2)1—7CH3
< (CHfi-CHa Hydrophobic
17 alkyl shell
H_N(CH2)1—70H3 E>_N
Terra—cm o ".1 E Hydrophilic
17 polymer shell Metal
3-D core
SAM

Figure 1.4. Scheme showing the surface-initiated polymerization of 2-oxazolines for the
preparation of the amphiphilic nanocomposites, with 2-ethyl-2-oxazoline and N,N-di-n-
octadecylamine used as monomer and terminating agent, respectively. Adapted with
permission from Macromolecules, 2001, 34, 1606-1611. Copyright © 2001 American
Chemical Society.

Zhao et al.40 synthesized 30 nm thick PS brushes in one hour from cumyl methyl
ether groups attached to SAMS on silicates. The reaction was carried out at -78 °C to

suppress the chain transfer reactions characteristic of cationic polymerizations.

0\ W b 9H3
O’Si_©'C-CD3 = rMMQ—c CH —c1-r
/ . 1m ,+ 2 I 2;,

 

CH3
Di-tert—butylpyridine

 

Figure 1.5. Polymer brushes grown by living cationic polymerization.

Surface-initiated ring opening metathesis polymerization (ROMP)

Surface-initiated ROMP is well suited for the polymerization of functionalized
norbomenes, and as shown in Figure 1.6, has been demonstrated on Au, SiO2 and Si
surfaces using Grubbs-type catalysts (Cat-1 and Cat-2) linked to surfaces by various
molecules (SAM-1, 2, 3). Whitesides et al.33 used a surface-immobilized Grubbs-type
ROMP catalyst to grow functionalized polynorbomenes. They reported the rapid,
controlled growth of 90 nm thick brushes polymer brushes in 30 min, and also formed
diblock copolymer brushes by exposing the brushes to a second monomer solution. The
same strategy was utilized by Moon et al.34 to form brushes from polymers with poly(p-
phenylene ethynylene) Side chains for chemical sensing applications. Although very thin
films (10 nm) were obtained, the brush fluorescence was brighter than comparable spin-

cast films, which was attributed to reduced aggregation of polymer chains.

Figure 1.6. Surface-initiated ROMP applied to the synthesis of norbomene-based
polymer brushes

 

 

 

 

 

 

 

 

 

 

/—_\
O Q“ “
Cln. R'u— . Tm / / Men
PCY3
Cat-1 Cat-2 SAM-1 SAM-2 SAM-3
3)
F1. _ / _ /
SAM-1 r—Q-ISi Cat-1! . T0331
SI ”OH Si O -——> SI 0 /
\ \ R
F0H
— R R - H 8' CB
/ - 9 I( )3
O
bO‘filMi' 0M6 003H17
______, Si O / n r _ .
\ [R] —
— C3H170

Adapted from Whitesides et al. Macromolecules 2000, 33, 2793-2795 and Moon
et al. Macromolecules 2002, 35, 6086-6089.

b)

F

1. PCI5

M Cat-1 NRU Lb
Si > St —* Si T 3‘
2. XMQCH2CH=CH2 RU

__ L.

 

 

 

 

 

 

 

 

 

 

—

Adapted from Grubbs, R. H. et al. Langmuir 2001, 17, 1321-1323.

-10-

Grubbs et al.35 directly attached initiators through a Si-C bond to a Si (111)
surface, providing a link between the polymer and a Si wafer without the insulating SiO2
layer. By varying the monomer concentrations, they obtained brush thickness ranging
from 0.9 nm to 5 pm.

ROMP has been coupled with other techniques to form nanopattemed
functionalized surfaces. Mirkin et al.36 combined dip-pen nanolithography (DPN) and

ROMP to form polymer brush arrays on the nanometer length scale. The novelty of this

 

approach is that combinatorial polymer brush arrays can be formed by controlling the tip

and substrate reaction time and different monomers can be used for polymerization. r

2:23.21 41 _. edited

Ru
l
v /?\\ /f\. Si &

 

 

 

sro2
Constant tip-substrate Different monomer
contact time . .
Different tlp—
Ru Ru substrate
contact time Ru Ru
R R
R R1
Ph n Ph D
I I Ph D Ph n
l I
/Sli\ /Sli\ - '
/&\ /&\

 

 

 

 

Figure 1.7. Graphic representation of ROMP initiated by DPN and a monomer coated
AFM tip. Adapted with permission from Angew. Chem, Int. Ed. 2003, 42, 4785-4789.
Copyright © 2003 Wiley Intersciences.

-11-

Surface-initiated ring opening polymerization (ROP)

Surface-initiated ROP has been applied to the synthesis of poly(N-
propionylethyleneimine), PCL, PLA, and polyglutamate brushes.3 Figure 1.8 shows
examples of the polymerization of e-caprolactone and L-lactide initiated from SAMS

terminated with ethylene glycol.

o
O
O
a OH I 0
) PhCHZOH 0% t...

EtaAl
”IOTO
o o O : H
b) XH ’ 0 '
Sn(Oct)2 Xi/‘H/ rm;

X=O,NH

 

I

 

Figure 1.8. PCL and PLA brushes formed by surface-initiated ROP.

Husseman et a1.37 grew PCL from gold using aluminum alkoxide-catalyzed ROP.
Brushes up to 70 nm thick formed in a few hours using a diethylaluminum alkoxide
based catalyst; the concurrent formation of polymers in solution required exhaustive
washing of the substrates to remove adsorbed polymer. SAMS terminated with
oligo(ethylene glycol) (OEG) was also used to initiate polymerization of lactide,38
however, the growth rate was slow, with a three day polymerization at 40 °C yielding a

12 nm thick film. The use of amine-terminated SAMS on SiO2 were also investigated, and

-12-

70 nm thick PLA brushes were grown in three days at 80 °C. Kim et al. reported the
polymerization of lactide from poly(2-hydoxyethy1 methacrylate) (poly(HEMA)) brushes
at 95 °C using Sn(2-ethylhexanoate)2 as the catalyst.41 In this example, the terminal OH
groups on the poly(HEMA) side chains initiated the ROP of lactide. Despite the use of a
gold substrate, the film was stable at the polymerization temperature due to partial

crosslinking of poly(HEMA) as a result of transesterification.

 

 

Au R CHZ— 0 AU R CHz—C Br
CO2CH2CH2CH n <f=0 n
PHEMA Sn(0ct)2. H2
Toluene, 95°C $.42 Po H O H
e_l&_¢_o—“_¢—c H
l CH3 CH1
PHEMA-g—PLA

Figure 1.9. Synthesis of p0ly(HEMA)-g—(polylactide).

Joubert et (11.39 recently reported the first example of grafting PEO to silica
nanoparticles using surface-initiated ROP of oxirane monomers. The Silica particles were
first treated with pre-hydrolyzed 3-glycidyloxypropyl trimethoxysilane (GPS) to form
particles with grafted alcohols, which acted as a co-initiator/chain-transfer reagent in the

polymerization.

-13-

 
 

(€014: OH

RD
/S'\ AI'OR /
= ': ‘

W Grafted PEO chain

urn” hydrolysable PEO chain

Free PEO chains

Figure 1.10. Synthetic strategy for the anionic ring-opening polymerization of ethylene
oxide initiated from a silica surface modified by Al(OiPr)3/GPS.

Surface-initiated controlled radical polymerization (CRP)

Over the past decade, a variety of CRP methods have been developed and
blossomed. A major difference between traditional free radical polymerization and CRP
is the lifetime of active radicals. In traditional radical polymerizations, active radicals are
formed, add to monomers (propagate), and terminate. In CRP, the low concentration of
active radicals greatly extends the lifetime of a growing polymer chain due to a dynamic
equilibrium between the dormant species and active radicals. Although the details of the
various types of controlled radical polymerizations are different, the general principle is
still the same.19 As shown in Scheme 1.1, the covalent bond of the dormant species is
homolytically and reversibly cleaved by physical (heat or light) or chemical (metal
catalysts) methods to generate a reactive radical, which adds to monomers to form

polymer chains. The low concentration of growing radicals is maintained by fast

-14-

reversion to the dormant species, which greatly reduces the chances of bimolecular

radical coupling and disproportionation reactions. This equilibrium exists in every

propagation step resulting in uniform polymer chains.

(1)

(2)

(3)

Stimuli

(physical, chemical)

 

 

 

Dormant Active
Polymer
CH3 M" Xn Lm CH3
.... CHz—c—CI -_ MCH2_('; . cw":1 xn I.m
c=o c::=o
OCHs OCH3
A o
MCHz-CH-O'N ..——:_ MCHz-CH ' O"N

McHz—CH—s—p-Pn .. ~R -___—* MCHréH R—s—g—Ph
s S

Scheme 1.1. The general mechanism of controlled radical polymerization and

some typical examples.

Metal-catalyzed living radical polymerization, (ATRP)20 (1), nitroxide mediated

polymerization (NMP)5 (2), and reverse addition fragmentation transfer (RAFT)42 (3) are

the most extensively used controlled radical polymerizations since they provide good

control over the polymer architecture, are compatible with a broad range of monomers

and polymerization conditions, and the chemicals used in the polymerization are readily

-15-

accessible. Each method has its own advantages as well as limitations.43 NMP has been
successfully applied only to styrene and styrene derivatives, and usually provides poor
control in the polymerization of acrylates and methacrylates. High temperatures are
usually required when the 2,2,6,6-tetramethyl-1-piperinyloxy (TEMPO) free radical is
used for NMP, because of the strength of the C-O bond between TEMPO and the
growing chain end. In addition, alkoxy amines are usually difficult to remove from the
chain end, which limits its conversion to other functional groups. The advantage of NMP
is that it does not use metal catalysts, which often cannot be completely removed from
polymerizations. In theory, RAFT can be applied to all radical polymerizations. However,
high temperatures (>140 °C) are required for RAFT of vinyl esters, and the residual
thioester end groups can cause odor problems in low molecular weight species.
Development of improved transferable groups for RAFT is desirable. ATRP has been
applied to largest number of monomers and over a wide temperature range (-20 to 130
°C). Since the polymer end group is a halogen, it can be easily converted to other
functional groups. Some tolerance to oxygen and compatibility with water makes ATRP
probably the most popular CRP method. However, the biggest challenge for ATRP is to
solve the challenge of removing/recycling the metal catalysts, which impedes its wide
application in biotechnology and for industrial syntheses of polymers. A detailed
discussion of ATRP appears in the following section.

Andruzzi et al.44 reported the use of surface-initiated NMP to grow polymer
brushes with pendant OEG segments from a silica surface. The monomers were styrene
derivatives with pendent OEG units, and were polymerized at 125 °C for 48 h. The as-

forrned polymer brushes were used to study protein adsorption and cell adhesion.

~16-

602 %(OCH20H2),,OCH3

 

3-a n= 3
34) n: 7
3 1
la) 2, styrene —
ib) 2, 3-a, b yo—b "
. I (OCH20HzanCHa
8102 -?i
3
(OCH2CHzinOCHa n = 3, 7
n = 3, 7

Figure 1.11. Synthesis of tethered OEG containing polymer brushes on silica surfaces.

Atom Transfer Radical Polymerization (ATRP)

The general mechanism of ATRP is Shown in Scheme 1.2. The terminal halogen
atom of the initiator (R-X) is extracted by a metal catalyst, such as a CuI complex, during
initiation to form a reactive radical. This process occurs with a rate constant of km and
the reverse reaction occurs with a rate constant kdmt. The reactive radical will add to
monomers to form a polymer just as in traditional free radical polymerization, with the
propagation rate kp. The advantage of ATRP is the low concentration of reactive radicals
in the polymerization system, which is maintained by the equilibrium between the
reactive radical and dormant species (K,q = km /kdcact). Since kdcact > km, the equilibrium

favors the dormant species. Termination (kt) also occurs in ATRP, mainly by bimolecular

-17-

coupling or disproportionation. However, if the radical concentration is sufficiently low,
termination can be ignored. Therefore, a successful ATRP will yield polymers with high

molecular weight and low PDI.

k
R-X + Cu(l)X/bpy .:i—-‘E‘__ R' + Cu(l|)lebpy
kdeact \‘x‘
“\‘k‘

kp monomer “ termination

 

 

R-CHyC—X am CH3
I —’
{3:0 i C“ )x’bpy kdeacr R-CH2-C::° + CU(||)X2/bpy
(OCH2CH27n30CH3 9:0
(OCHZCHZ‘F‘OCHg

kp ll monomer

Polymer

Scheme 1.2. The mechanism of ATRP of poly(ethylene glycol) methyl ether
methacrylate (PEGMA).

-13-

A T RP kinetics

For copper-mediated ATRP, the polymerization rate (RP) can be described as45

K..iMiIIi. i951— Eq. 1

R = k
P [XCuII]

P

where ch = k ac. / kdcact. The PDI can be expressed as

, k I
M” =1+ °[ 1" u (3—1) Eq.2
M kdcact[xcu ] P

n

 

 

In Eq. 1, RP is first order with respect to the solution concentrations of monomer [M] and
the Cu1 catalyst. The initial build up of the Cu” species eventually leads to control over
the polymerization as the equilibrium between Cu1 and Cu" is established. Kcq depends on
the monomer structure and can be adjusted by the choice of the metal and ligands in the
catalysts. A high value for ch gives a faster rate of polymerization, but if the value of Keq
is too high, ATRP simply becomes a conventional redox-initiated radical polymerization
(de30. ~ 0) with substantial termination. If ch is too small, the polymerization is too slow
to be useful. Eq. 2 indicates that the PDI (MW/Mn) decreases with conversion, p. For a
given monomer, a decrease in the kp/kdmt ratio through fast deactivation of the catalyst
will result in lower PDI. In addition, a higher concentration of deactivator (Cuu) will help
control the polymerization (narrow molecular weight distribution), with the decrease of
polymerization rate. Therefore, in some examples of Cu-catalyzed ATRP, small amounts
of Cu" halides are added to the system in the initial stage of polymerization to control the

polymerization.

-19-

A T RP components

Monomers

ATRP has successfully been employed on styrenes, acrylates, methacrylates,
acrylonitriles, (meth)acrylamide, and other monomers. Scheme 1.3 lists representative
monomers successfully polymerized using ATRP. Monomer structure is one of the
determining factors of kp, since it defines the stability of the radicals formed after the
addition of growing radicals. Each monomer also has its own equilibrium constant (Keq =
kact/kdew), which defines its polymerization rate in ATRP. Therefore, each monomer has
unique optimum polymerization conditions. The monomers used in this study are
PEGMA macromonomers, methacrylates which contain pendent ethylene oxide
segments. The novelty of these macromonomers is partly due to the incorporation of the
biologically relevant PEO segments, and that they also directly afford comb-shaped
polymers with regular and densely attached branches. Furthermore, crosslinked polymers
can be prepared by homopolymerization of divinyl monomers or copolymerization of

mono and divinyl groups.” 47

-20-

Scheme 1.3. Representative monomers used for ATRP

a) Styrenes

 

R= CH3, tBu x=Br, c1, F

b) Acrylates

O
\

10 33:0

OH

c) Methacrylates

rk
f)
erofiofk

MMA HEMA PEGMA

__ o ==)=o ==>=o

O O thi
k— /> I n 2n

CmF2m+1
BA

0 02 02
%7 NMe2 0
0:13==
GMA DMAEMA EGDMA

-21-

Initiators

The diversity of ATRP initiators is larger than for other controlled
polymerizations. Scheme 1.4. lists representative examples of initiators. A good ATRP
initiator should be sufficiently reactive to ensure that all chain ends start to grow
polymers at the same time, and the equilibrium in the initiation step should favor the
dormant species in order to reduce the chance of Side reactions. Substituent groups on
initiators stabilize the radical in the order of CN > C(=O)R > C(=O)OR > Ph > Cl > Me,
with or-haloesters the most common initiators used in surface-initiated ATRP. The
general order in bond strength is R-Cl > R-Br > R-1, and while alkyl iodides are efficient
initiatiors, bromo and chloro or-haloesters are the most often used Since alkyl iodides

require special precautions to avoid side reactions in ATRP.

Catalysts and ligands
(a) Cu based ATRP

A variety of transition metal complexes have been applied in ATRP, such as
Mo,48 Re,49 Ru,50’ 5‘ Fe,52'54 Rh,55 Ni,56 57 Pd,58 and Cu 21’ 59 metal complexes. The
following discussion will emphasize copper-based ATRP since Cu mediated ATRP was
used in this study. Cu complexes are versatile ATRP catalysts, with styrenes,

(meth)acrylates, amides and acrylonitriles all successfully polymerized by ATRP.

-22-

Scheme l.4. Representative examples of initiators used for ATRP

a) Halogenated alkanes and benzylic halides

R X CHC'z O O
RCCI3 x c1 or
R = H. Cl. Br. R = H, CH3, X = Br, CI
Ph, CF3, 002CH3, X = Br. 0'
CnHZn+1
b) 01- Haloesters and their derivatives
0 O O o O 0
Br '2 Br

0 O O

o
HO 8 B \ B
\/\o ' Vokr ' V\o/“\r r wok“

The choice of ligands used in ATRP catalysts can be used to adjust the
equilibrium between the dormant and reactive radicals, and to solubilize and stabilize the
metal catalyst in the reaction media at different temperatures. For copper-based ATRP,
nitrogen ligands work best, with sulfur, oxygen or phosphorous ligands less effective,
possibly due to the inappropriate electronic effects or unfavorable binding constants. The
electronic and steric effects of ligands greatly affect the catalyst reactivity. Significant

steric hindrance at the metal center or the use of ligands with strong electron-withdrawing

-23-

groups decreases catalyst reactivity. Scheme l.5 shows examples of ligands used for
copper-based ATRP. Bipyridine (bpy) has been the ligand of choice for copper-mediated
ATRP of styrene and (meth)acrylic monomers. Cu/bpy complexes perform well in
aqueous media, but their limited solubility in organic solvents reduces their control over
the polymerization. To improve the solubility of the catalysts in organic solvents and
maintain better control over the polymerization, bpy is often replaced by an alkyl-
substituted analog 4,4’-di(n-nonyl)-2,2’-bipyridine (anbpy). The activity of nitrogen
ligands usually increases with the number of coordinated sites (N4 > N3 > N2 >> N1), and
decreases with the number C atoms that link nitrogen (C2 > C3 > C4). Multidentate
ligands, such as tris[2-(dimethylamino)ethyl]amine (MegTREN) and 1,1,4,7,10,10-
hexamethyltriethylenetetramine (HMTETA) have been successfully used for the ATRP

of styrene and methacrylates.

(b) Structures of CuI complexes

Cationic CuI complexes prefer a tetrahedral or square planar configuration with a
tetradentate ligand or with two bidentate ligands. Cul complexes with bidentate ligands
such as bpy are generally represented as [CuI (bpy)2]+[Y]'. For Cul Br and CuI C1 salts,
the counter ion [Y]' can also be the [Cu1 Br2 ]' and [CuI Cl2]' anions, respectively. In the
solid state, the typical geometry of the [Cu1 (bpy)2]+ cation is a distorted tetrahedron, as
shown in Figure 1.12.60 The average Cu-N bond length ranges from 1.985 to 2.057 A and
is not greatly affected by the counter ion or substituents on the bpy ligand. The dihedral

angle is in the range of 80-83° due to the restricted geometry of the bpy ligands.

-24-

Scheme 1.5. Examples of ligands often used for copper-mediated ATRP.

a) Bidentate ligands (N2)

CW -N \N/ -N \N/ -~ N-Pr

bpy dNbpy anbpy

b) Tridentate ligands (N3)

R

' \
R, (\N/N ,R —N/——\N— i \ N
R’N N‘R QT) ’ N

z_
\ /

R=H, DETA
R=CH3, PMDETA

c) Tetradentate Ligands (N4)

R. /_\ R
R m R N N’ F183
C: :3 f 3 ,_ N
R’U R R7“ “TR ' R‘R
R=H, Cyclam [2:1-1' TETA R=H, TREN

R=CH3, M84 Cyclam R=CH3, HMTETA

-25-

NPI'PMI
'i
i \ N | \
/N N /
R=H, BPMA

R=(CH2)7CH3 BPMOA

/\ N \
’N ri/
/N
\l
TPMA

R=CH3,M86TREN

  

“v“

Figure 1.12. Structure ofthe [Cu‘ (bpy)2]+ cation in [Cu' (bpy)2]+[ClO4]'.

Bridging is observed in Cu'X (X = halide) (Figure 1.13). In [Cul (bpy) Br2], each
CuI center has a distorted tetrahedral geometry and is coordinated to the two nitrogen

atoms of a bpy ligand and two bridging bromides.

\ /\/x\/N\ /

Cu

/ \ / W \N/ \

Figure 1.13. Structure of the [Cu1 (bpy) X2] (X=Br, Cl) dimer.

The structures of Cu1 /bpy complexes in solution also have been studied. In non-
polar media, CuIX/bpy complexes were proposed to exist predominantly as the bridged

complexes shown in Figure 1.13. However, in polar media, the predominant complexes

are [Cu' (bpyirriYI.

-26-

Cu/HMTETA complexes are very reactive catalysts and have been used in my
research. The X-ray structure of [Cu'(HMTETA)2]+[Cu'Cl2]' complex (Figure 1.14)

indicates Cu1 in a distorted tetrahedral geometry.

9 9

Figure 1.14. Molecular structure of [Cu‘(HMTETA)]+[Cu'C12]‘.

(c) Structures of Cun complexes

It is generally accepted that the most common Cu"X2/bpy complexes are
[Cu"(bpy)2X]+[X]-. As shown in Figure 1.15, [Cu"(bpy)2X]+ adopts a trigonal

bipyramidal geometry.

 

Figure 1.15. Structure of the [Cu"(bpy)2Br]+ cation in [Cu”(bpy)2Br]+[Br]_.

-27-

Cu(II)X2 complexes with tetradentate ligands such as HMTETA
([Cu"(HMTETA)X]+[X]-) involve penta-coordination. As shown in Figure 1.16, the Cu
center coordinates with four nitrogen atoms of the ligand and the donor halogen atom,

forming a square pyramidal geometry.

 

Figure 1.16. Molecular structure of [Cu"(HMTETA)Br]+[Br]‘.

Water as a polymerization solvent

A variety of solvents have been used for ATRP, including toluene, anisole,
diphenyl ether, ethyl acetate, acetone, DMF, alcohols, and water.20 Generally, the use of
polar solvents improves the solubility of Cu" complexes and favors homogeneous, well-
controlled ATRP. Water has been increasingly employed for ATRP of polar, water
soluble monomers having hydroxyl, oxyethylene, amino, ammonium, and carboxylate
functional groups. Water also can greatly accelerate the polymerization rate. The first
aqueous copper-mediated ATRP was reported by Matyjaszewski et al.,61 the

polymerization of 2-hydroxyethyl acrylate (HEA) in aqueous media (50:50 v:v) using a
-23-

CuBr/bpy catalyst at 90 °C. A 12 hour polymerization resulted in 87% conversion of
monomer to polymer. Later, Armes et al.62’ 63 reported the remarkably fast ATRP of
PEGMAS in water, 90% conversion in 30 min at 20 °C, while retaining control over the
polymerization (PDI < 1.3). The fast polymerization rate was attributed to the formation
of mononuclear CuI species in water. Huang et al. described the rapid growth of

poly(HEMA) brushes in water using a catalyst prepared from 30 mol% CuCl, CuBr2, and

bpy-64

Water used as a co-solvent accelerates polymerizations, but there are no reports of
the catalyst structure in water. However, it has been proposed that [CuI (bpy)2]+[X]' is the
predominant complex in polar media, and is believed to be the reactive catalyst. There is
also the possibility of solvent and monomer coordination to the catalyst. Haddleton65
studied the effect of water on the polymerization of PEGMA(8-9) using a CuBr/N-(n-
propyl)-2-pyridylmethanimine (NPrPMl) complex. They proposed the competitive
coordination of water and ligand to the copper center, which results in an increase in
polymerization rate. Later, Jonsson66 studied the ATRP of PEGMA(7-8) in aqueous
media, and calculated the propagating radical concentration from the redox properties of
the alkyl halide and Cu complex. The calculated radical concentration in water is 3 orders
of magnitude higher than in acetonitrile, which contributes to the fast ATRP kinetics in

water.

-29-

A TRP ofPEGMA

There are two general strategies for incorporating PEO segments in polymer
structures via ATRP. In the macroinitiator approach”72 shown in Scheme 1.6., an a)-
hydroxy group of PEO is converted into an ATRP initiator by coupling with an or-
bromoacyl bromide or or-bromocarboxylic acid. This scheme provides block copolymers.
The macromonomer approach is the direct polymerization of a PEG-containing monomer
such as a PEG methacrylate. This scheme provides densely grafted comb polymers with
PEG teeth. The macromonomer approach is a convenient way to form macromolecular
structures with PEO segments. Haddleton et al.73 investigated ATRP of PEGMA (Mn =
480 g/mol) in toluene mediated by Cu'Br/NPrPMI. Using an on-line lH-NMR analysis,
they observed a fast polymerization rate, 90% conversion in 1 hour at 90 °C. These
reaction conditions are harsh and incompatible with polymerizations from gold since the
Au-S bond is unstable above 60 °C. Armes et al. 62’ 63 studied ATRP of PEGMAs with ~8
and 45 repeating units of ethylene oxide in water, and also found that polymerization rate
was remarkably fast (90% conversion in 20 min at 20 °C), but still controlled (PDI < 1.3).
Apparently, water greatly accelerated the polymerization rate, as discussed above. The
structure of PEGMA plays an important role in accelerating polymerization since the
sterically congested PEGMA inhibits radical-radical coupling, which reduces the
termination rate (kt). Decreases in k with increased monomer Size also has been observed
for alkyl-substituted methacrylate esters. For example, kt for undecenyl methacrylate is

0.6 x 1045 L/(mol ~ 5), 1/40 of the rate for methyl methacrylate (25.5 x 1045 L/mol - s).

-30-

Scheme l.6. Synthetic approaches used to incorporate PEO segments in polymers.

(a) Macroinitiator approach

0 macroinitiator polymer

.01-019

B
r Br

O
‘ ATRP
/Of/\O)r: ' /O~(/\O)):‘\(Br

(b) Macromonomer approach

 

ll

 

O macromonomer

O
0 V0)?" R’OMO
R’ Br > o O
O CuX, Ligands P )-
O

m

 

Ito et al.74 reported aqueous ATRP of two styrene-functionalized PEG
macromonomers (Mn = 2300 g/mol). They reported that kp /k1 was ~0.25 in water, and
proposed inter-micellar polymerization to explain the high molecular weight they
obtained. However, no evidence for micellar polymerization was found in Arme’s
system. Later, Jonsson et al.66 estimated kp to be 3.6 x 10 3 M‘l s'1 for PEGMA (Mn = 480
g/mol) in water by correlating the apparent polymerization rate with the redox potential
(El/2) of copper complex. The high overall polymerization rate was ascribed to
complexation of ethylene glycol groups to copper, displacing some of the diamine

ligands to form a very reactive metal catalyst.

-31-

The above examples of PEGMA polymerization provided low molecular weights,
with degree of polymerizations (DPS) ~10-33. A recent paper by Matyjaszewski et al.75
described aqueous AGET (activator generated by electron transfer) ATRP of PEGMA
homopolymers and random copolymers with DPS >240 and PDIS <1.3.

Given the results Of PEGMA polymerization in solution, we expect surface-
initiated ATRP of these monomers to provide thick polymer films on surfaces. The

surface-initiated ATRP of PEGMA is elaborated in Chapters 11 and III.

Polymer brushes formed by surface-initiated ATRP

An advantage of the ATRP system is the facile process by which substrates are
converted into initiators. Functionalization of the target substrate uses commercially
available or-haloesters or benzyl halides and their derivatives, rather than the multi-step
syntheses required for alkoxyarnines and xanthates used in other controlled
polymerizations. Not surprisingly, ATRP is the most commonly used controlled
polymerization technique, and has been applied to the growth of well-defined polymer
brushes on gold, Silicon wafers, inorganic particles/colloids, organic latexes, and a variety

of other substrates.

Methods for the controlled growth of polymers from surfaces.

As described earlier, the preparation of polymer brushes requires attachment of an
initiator on the surface followed by ATRP of vinyl monomers from the substrate. Figure

1.19 depicts the general scheme for polymer brush formation on flat surfaces.

-32-

0
OH R‘
1) Hs/‘H 2) BIJKF , (R'=H,CH3)
m C 3 -
Br
S/HOW‘JQ
m 0

Or I
O

Svl’toJKK ATRP
m Br

Polymer brushes
OH R R1 m

OH R*= benzyl halides, sulfonyl
OH halides, ct-haloesters, etc.

R= c1. OCH3, R,= c1. OCH3, CH,

 

 

Figure 1.17. General synthetic route to polymer brushes on gold and silica substrates.

The main challenge in surface-initiated ATRP from flat surfaces is the inherent
low concentration of initiators in the system, and therefore, a low concentration of the
Cull radical deactivator. At low CuII concentrations, ATRP may resemble a redox
process, an uncontrolled polymerization dominated by termination. The addition of
deactivator in the form of Cu11 can compensate for the low CuII concentration, or
sacrificial initiator can be added. In the case of added sacrificial initiator, termination of
radicals in solution will generate a sufficient concentration of deactivator to afford
control over the polymerization. The addition of “free” initiator to the system is not only
beneficial for controlled synthesis of polymer brushes, but also provides a way to
characterize polymer brushes. Analysis of polymers formed in solution can be used to
estimate the molecular weight of tethered polymer chains, since several experiments have
shown that the polymer chains cleaved from surfaces have molar masses and PDIS

-33-

Similar to the polymers formed in solution. The DP of tethered polymer chains can be
estimated from the concentrations of consumed monomers and sacrificial initiator,
assuming all initiators were used.

DP = A [M]/[initiator]

However, a significant drawback of adding sacrificial initiator is the formation of
polymer in solution, which can be very difficult to remove from surfaces, especially
during the synthesis of crosslinked polymer brushes. The alternative, adding a Cu"
deactivator, is particularly useful in controlling polymer brush growth. Using 30 mol%
CuBr2 as a deactivator in HEMA polymerizations, Huang et al. obtained 400 nm thick
films compared to 200 nm films without CuBr2.64 The results are sensitive to the CuBr2
concentration; the limiting film thickness using 5% CuBr2 (relative to the Cul
concentration) was slightly >200 nm. A related study on the polymerization kinetics of
methyl acrylate (MA) defined 0.1 x 10'3 M as the CuI concentration that yields the
thickest films when the Cu11 concentration was fixed at 30 mol%.76 The Cu1 concentration
was varied from 40 mM to 2.5 x 10‘4 M in this study. Diluting the catalyst concentration

also helps control the growth of polymer brushes.

Brush architectures generated by surface-initiated A T RP.

The versatility of surface-initiated polymerizations makes it possible to form a

variety of brush architectures on flat surfaces, including block copolymer, binary polymer

77, 78 79, 80

brushes, comb-like polymer brushes, hyperbranched polymers, and crosslinked
polymer brushes,81 as shown in Figure 1.18. Huang et al. described the facile synthesis of
crosslinked polymer brushes on gold by surface-initiated ATRP.81 Chapter III will

-34-

describe studies of comb-like polymer brushes with PEO segments grown on different

substrates, and crosslinked polymer brushes will be described in Chapter IV.

Homo or block Comb-like Crosslinked

Figure 1.18. Schematic illustration of linear, comb-like and crosslinked polymer

brushes.

Polymers grown by ATRP retain a halogen at the terminus of the growing
polymer chain. As described earlier, these sites can be reactivated to initiate the
formation of block copolymers. As shown by Zhao et al.82 these Sites also can be
permanently deactivated. Zhao designed novel binary polymer brushes by combining
ATRP and NMP. As shown in Figure 1.19, a Y-shaped SAM was formed on the surface
that incorporated both ATRP and NMRP initiators. ATRP of MMA was followed by
removal of the terminal halogen using Bu3SnH. Styrene was then polymerized by NMP

to form binary polymer brushes that Show lateral phase separation.

-35-

PMMA
PS

PMMA

Br
I ,. ,3
O O O)?
N 0’N
O/ O O
O O o O
O O
O
O O

 

O o
NMRP
1) ATRP ________,
MMA, 75 °C Styrene
R 115 °c

2) Buaan

 

 

 

7////// /////////// /////////

M'xed PMMA/PS
Y-SAM PMMA Brushes Brlushes

Figure 1.19. Synthesis of mixed PMMA/PS brushes from an asymmetric di-functional
initiator-terminated SAM (Y-SAM) by combining ATRP and NMRP techniques.
Reprinted with permission from J. Am. Chem. Soc. 2004, 126, 6124-6134. Copyright ©
2004 American Chemical Society.

Controlled A T RP from surfaces of particles.

Surface-initiated ATRP is not confined to the formation of polymer brushes on
planar substrates. Polymer/inorganic hybrid nanoparticles have also been prepared by
surface-initiated ATRP of various monomers from colloidal initiators,6’ 80‘ 83'” aiming to
improve their stability in different solvents. A variety of inorganic substrates can be

modified by polymer brushes; the following discussion will emphasize polymer brushes

-36-

grown on the surfaces of silica nanoparticles. Using mild conditions (water, 20 °C),
Armes et a1.89 grafted poly(PEGMA) and poly((Z-N-morpholino)ethyl methacrylate))
(PMEMA) brushes on the surface of 300 nm silica particles prepared by the Stober
process. The PMEMA grafted silica particles Showed lower critical solution temperature
(LCST) behavior, and began to aggregate at 34 °C. The aggregates completely re-

dispersed on cooling below the LCST.

Bromoisobutyrate
-based siloxane

Ethanol 20°C

 

300nm

Figure 1.20. Poly(PEGMA) and PMEMA brushes grown from silica nanoparticles.

Zhao et (11.90

reported the first synthesis of comb-coil polymer brushes on the
surface of silica nanoparticles. A poly(HEMA) brush was grown from the nanoparticle by
surface-initiated ATRP. Then, concurrent polymerization of tert-butyl acrylate (tBA) and
lactide were initiated from the terminal halogen group by AGET ATRP, and from the OH

groups of the poly(HEMA) side chains by ROP, respectively.

-37-

 

HO 0 Br
1. ,CH3
OH (c,H,O),SiCH,CH,NH2 BI‘C‘CCB' Br
HO : CH3 : Br
OH Br
HO Br
ATRP HEMA
( PLA
LA B Br
r
L“ '\ PBA ° 0 e
— 6
O O
CuClzlbpy .1-
4—_____
Sn Oct Br
0 0
BA Br

Figure 1.21. Preparation of comb-coil polymer brushes on the surface of silica
nanoparticles.

Physicochemical properties of polymer brushes

Physical states of polymer brushes

The physical nature of surfaces (“hard” or “soft”) largely depends on the physical
states of polymer brushes (crystalline or amorphous). To date, most polymer brushes
have been amorphous, including glassy polymers such as PS or rubbery polymers such as
polydimethoxysilane (PDMS).

Liquid crystalline polymer (LCP) brushes have attracted substantial interest due to
the application of LC materials for displays and as semi-conducting materials. Peng et

al.91 synthesized and studied side chain LCP brushes comprised of phenylbenzoates

-33-

tethered to a polymethacrylate with a spacer. These films, up to 200 nm thick and
prepared by surface-initiated free-radical polymerization, were used as alignment layers
for LC systems. Alignment layers orient LCs at surfaces and provide the large area
domains needed for displays. As the brushes were heated or exposed to solvent, the LC
phase underwent a reversible nematic to isotropic LC phase transition. Recently, Huck et
al.92 reported the synthesis of side chain LCP brushes formed by surface-initiated ATRP
of active acrylate mesogens on silica and glass substrates (Figure 1.22). Glass sides with
polymer brushes were oriented face to face and separated by a spacer. A nematic LC, 4’-
pentyl-4-biphenyl carbonitrile, was inj ected into the gap between the slides and spread as
a result of capillary forces. Homeotropic alignment was observed, which was ascribed to
the interaction of LC side chain with the LC phase. Alignment was also observed on
nanopattemed substrates, particularly in regions where polymer brushes were present on
both the top and bottom substrates. This suggests potential applications for brushes in

displays and optoelectronics.
=>_O
o \—\_\—W
o_©_\<°H : O
O

 

 

O
O
\ ./\/\

O/S' OJYB' CuCl, CuClz, CuBr2, bpy
O DMF, acetone, 50 °C

0 O

\S./\/\ W
O /O
C6H12

Hit—CW

Figure 1.22. Synthesis of side chain LCP brushes by surface-initiated ATRP.

-39-

Although much progress has been made on amorphous polymer films and LCP
brushes, there are few examples of studies on the crystalline state of polymer brushes.
Studies of polymer brush crystallization should help explain polymer crystallization in
constrained geometries. Chapter II will elaborate on the crystallization behavior of

polymer brushes.

Stimuli-responsive polymer brushes

Stimuli-responsive polymer brushes are attractive for applications as “smart” or
responsive surfaces. Such brushes would alter their conformations upon changes in
temperature, solvents, pH, photo-illumination, or another external trigger, resulting in

changes in the surface physical properties.

Solvent induced polymer brush rearrangement.

For simple A-B block copolymers, phase separation of the A and B blocks
generally occurs when the heat of mixing is positive. Phase separation of A-B block
copolymers also occurs on surfaces, but being tethered to the surface, the polymer chains
are constrained to separate in nanometer-sized domains. Irnmersing block copolymer
brushes such as PMMA-b-PS in different solvents results in solvent-induced structuring

1.93"94 studied the rearrangement of PMMA-PS block polymer

of the surface. Brittain et a
brushes in CH2C12 (a good solvent for both blocks) and cyclohexane (a good solvent for
PS but not for PMMA). When the polymer brushes are in CH2Cl2, both blocks swell and

stretch away from the substrate, yielding a smooth surface. When the polymer brushes

-40-

are in a mixture of CH2C12 and cyclohexane, the PMMA blocks collapse to form

aggregates and avoid interaction with cyclohexane.

addition of
lmmersuon cyclohexane

in CHZClz

 

     

Figure 1.23. Speculative model for nanopattem formation in di-block copolymers.

A-B-A triblock copolymers showed similar phenomena. For PMMA-b-PS-b-
PMMA, the film showed extended conformations when exposed to the good solvent
(CH2C12), and a folded brush conformation when exposed to cyclohexane, which is a

good solvent only for the middle block.95

Good solvent for
middle block
—————>
)

‘—
Good solvent for
all blocks

   

Figure 1.24. Reversible response of triblock copolymer brushes to different solvents.

Polymer brushes can show different solvent induced changes in their surface
morphologies provided they have sufficient conformational freedom. Pyun et al.96
synthesized an A-B-C triblock copolymer (PDMS-b-PS-b-Poly((3—(dimethoxy

-41-

methylsilyl)propyl acrylate (DMSA) using anionic ring opening polymerization and
ATRP. The polymer was “grafted to” the silica surface resulting in polymer brushes with
low grafting densities. The polymer brushes rearranged to form glassy or rubbery
surfaces when immersed in different solvents (toluene and hexane) (Figure 1.25).

Various polymer brush architectures were constructed on surfaces to study their
properties. For example, Li et (11.97 grew mixed PtBA/PS brushes on silica particles by the
sequential ATRP of tBA and NMP of styrene from Y-shaped-initiators anchored on silica
particles. The block copolymer chains tethered to these particles reorganized in response
to changes in solvent, forming stable suspensions in CHC13 and MeOH, which are good

solvents for PS and poly (acrylic acid) (PAA, hydrolysis product of PtBA), respectively.

-42-

Swollen Brush in Toluene Collapsed Brush in Hexane

PDMS
p5 - PDMS excess hexane
—’ PS -
PDMSA excess toluene PDMSA
Rubbery
surface

 

Ultralight tapping
MA0 —> 1

     
 

5

Normal tapping
AIAo = 0.9

500

Figure 1.25. Tapping mode AF M height images of PDMS-b-PS-b-PDMSA brushes after
the following treatments: (left) after immersion in toluene and drying with nitrogen,
(right) after immersion in toluene, gradual addition of hexane and drying with nitrogen.
Reprinted with permission from Macromol. Chem. Phys. 2004, 205, 411. Copyright ©

2004 Wiley Intersciences.

-43-

 

 

 

(b)

 

 

 

ppm

Figure 1.26. ‘H NMR spectra of PAA/PS particles dispersed in (a) CDCl3, (b) DMF-d7,
and (c) CD30D. A drop of DMF-d7 was added into the particles prior to CDC13 and
CD3OD to increase the concentration of the dispersed nanoparticles. Reprinted with
permission from J. Am. Chem. Soc. 2005, 127, 6248-6256. Copyright © 2005 American
Chemical Society.

T emperature-responsive polymer brushes

Poly(N-isopropylacrylamide) (PNIPAAM) films are widely studied as
temperature responsive films since its LCST is ~32-33 °C, an easily accessible and
biologically relevant temperature. Jiang et al. prepared PNIPAAM films on a well-
controlled rough surface by surface-initiated ATRP.98 The films switch between super-
hydrophilicity (contact angle ~O °) and super-hydrophobicity (contact angle ~150 °)
within a 10° temperature range. At 25 °C (below the LCST of PNIPAAM), the polymer

-44-

adopts an extended brush architecture and the surface is hydrophilic due to H-bonding
between the polymer chains and water. At 40 °C (above the LCST), the polymer

structure is collapsed and the surface is hydrophobic due to H-bonding exclusively

between polymer chains.

Heafingup
——>

4_—
Codmgdown

 

 

 

>LCST

 

Figure 1.27. Diagram showing reversible formation of intermolecular hydrogen bonding
between PNIPAAM chains and water molecules (left) and intramolecular hydrogen
bonding between the C=O and N-H groups in PNIPAAM chains (right) below and above
the LCST. This mechanism is proposed to explain the thermally responsive wettability of
a PNIPAAM thin film.

-45-

a) 160 b) about0° 1493°

 

 

 

 

 

 

 

 

 

\40°C
120
. ‘I\‘\‘\
l 80
CAP” #/ Heating up
0, -Coo|ingdown
° °D§3n3° __. °° 25°C
C) 160, 29°C 4.000 d) 160 .III'Iljislololllllrll'l
. ‘ I ‘ l
' 120 HIiIlI‘IIIHI' li ill'il
Ill i ii I II II J l
I 30 [HIM ‘le l' ”I'IIHI' IIl'Il'iIi '
, ['IIIII IIIIIII “I III
I CAI“ III'II IIIIII'H'I'U' IIlI'Il
.. 0111...; Clll.l.l
I 2b 30 4o 50 o 4 32$L 12 16 20
TI°C —- Cycle _.

Figure 1.28. Surface-roughness-enhanced wettability of a PNIPAAM-modified surface.
a) The relationships between groove spacing (D) of rough surfaces and the water contact
angle at 25 °C (A), and at 40 °C (I). b) Water drop profile for thermally responsive
switching between super-hydrophilicity and super-hydrophobicity of a PNIPAAM-
modified rough surface with a groove spacing of about 6 pm, at 25 °C and 40 °C. The
water contact angles are ~0° and l49.3°, respectively. c) The temperature (T) dependence
of water contact angles for PNIPAAM thin films on a rough substrate with groove
spacing of about 6 mm (A) and on flat substrate (I). (1) Water contact angles at two
different temperatures for a PNIPAAM-modified rough substrate with a groove spacing
of 6 pm. Half cycles: 20 °C; and integral cycles: 50 °C. Reprinted with permission from
Angew. Chem, Int. Ed. 2004, 43, 357-360. Copyright © 2004 Wiley Intersciences.

-46-

Applications of polymer brushes

Polymer brushes can be applied in many different areas. For example, LCP
brushes, discussed previously, can be used as semiconducting polymers for LC displays.
With the development of nanotechniques, switchable surfaces have been used for
bioanalysis, microfluidic devices, and protein separations. Some representative examples

of polymer brush applications are described below.

Nanotechnology—Patterning of polymer brushes

One of the most important applications of polymer brushes is that they are well-
suited for the fabrication of nano- or micro-patterned arrays, which could be useful in cell
growth control, biomimetic microfabrication, drug delivery and microelectronics.
Polymer brushes can be patterned by traditional photolithography or chemical
amplification of patterned SAMS. Photolithography was first explored by Ruhe et al., 99
who exposed polymers to UV light through a mask. A good example of the complex
patterning that can be carried out on polymer brushes was described by Hawker and co-
workers.100 A PtBA brush was formed on a silica substrate, followed by spin-casting on
the surface a polystyrene film containing bis(tert-butylphenyl)iodonium triflate. After
exposing the surface to UV irradiation through a mask, the PtBA brushes were converted
to PAA brushes, resulting in a surface with distinct hydrophobic and hydrophilic areas.
SAM patterning was obtained by micro-contact printing (uCP), which was first

introduced by Whitesides et al.,101

and the pattern initiator monolayer was amplified from
the by surface initiated polymerization. Caster and co-workers 102 used scanning probe

lithography and surface-initiated ATRP to nanopattem PNIPAAM brushes.

-47-

   

1. immobilization

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ofinitiator
2.ATRPof
lePAAM
w) m)
65
nm
20 20
20
1° 10 ”m 10 10 "m
55‘ 55‘1 2 3 4 5 55‘
A 1 2 3 4 5 1 2 3 4 5
E ll )1
:7 JJ A A __ A A ii A
.1: 0 0 0
%
z
-55 -5s. . -55. . .
051015 0 51015 0 51015
Width (pm) Vlfidth (pm) width (um)

Figure 1.29. (top) Preparation of surface-confined PNIPAAM polymer brush
nanopattems by combining “nanoshaving” and surface-initiated ATRP using a surface-
tethered thiol initiator. (Bottom) Contact mode AFM height images (20 um X 20 um) and
corresponding typical height profiles of a PNIPAAM brush line nanopattem imaged at
room temperature in (a) air, (b) MQ-grade water, and (c) a mixture of MeOH/water (1 :1,
v:v). Reprinted with permission from Nano Lett. 2004, 4, 373-376. Copyright © 2004
American Chemical Society.

-43-

Biological application in protein resistance and cell adhesion

PEO (or PEG) is very important in biological applications due to its ability to
resist protein absorption and cell adhesion (non-specific adhesion resistance). PEO-
containing surfaces used for cmrent nanotechnology application are generated by self-
assembly of OEG alkane thiols on gold surfaces. PEO-containing polymer brushes are
attractive due to their synthetic robustness, and high surface grafiing densities. As shown
earlier (Figure 1.12), Andruzzi et al. compared the OEG tethered polystyrene brushes
with OEG SAMS, and found that polymer brushes were superior in their ability to inhibit
protein (less than 6%) and cell (less than 3%) adhesion when compared to surface
assemblies with the same ethylene glycol length.

Biocompatible and non-biofouling PEO containing polymer brushes patterned
onto gold or silica by various lithography techniques can be applied to “molecular
recognition”. The non-pattemed regions are backfilled with a molecule that has a specific
interaction with cell or other biomolecules. For example, Figure 1.31 shows a
nanopattemed gold surface modified with poly(PEGMA) brushes, '03 and a cell-adhesive
protein was absorbed on the surface after the features were backfilled with SAM 2. Two

different cells were patterned on the surface.

-49-

 

 

 

 

 

Figure 1.30. A) a PEGMA polymerized by ATRP from a micropattemed gold surface.
(B) NIH 3T3 fibroblast cells cultured on patterns of adsorbed fibronectin (20 um circles
and 40 um stripes). Reprinted with permission from Biomacromolecules, 2005, 6, 2427-
2448. Copyright © 2005 American Chemical Society.

-50-

 

 

 

Polymer brushes grafted membranes.

Ito et al. reported pH sensitive,104 photosensitive,'°5 and oxidation/reduction
sensitive106 polymer brushes covalently tethered to porous membranes, which can be used

to regulate liquids flowing through porous membranes.

.-’ solute solute

(a) Hydrogel membrane

 

 

 

   

 

 

 

or chemical
Immoblllzatlon of ___- _ __ .—
clonal censltlve

PO'VMOI‘

Signal
(b) Polymer brushes membrane Signal sensitive
polymer
Water so|ute Water SOlUte
Porous membrane 1 U
, Graft polymerization .1121); MM Signal n. Am
I m; f g d

 

 

 

 

 

 

 

Figure 1.31. Preparation of polymer brushes grafted to porous membranes and an
illustration of the mechanism for regulating water and solute permeation through a
polymer brush grafted membrane and a hydrogel membrane.

Polymer brushes grafted on membranes afford selective gas transport.
Balachandra et 01.107 grafted cross-linked poly(ethylene glycol dimethacrylate)
(Poly(EGDMA)) and poly(HEMA) brushes on the surface of porous aluminum
membrane supports. Gas permeation studies with poly(EGDMA) films showed a
COz/CH4 selectivity of 20, and a C02 permeability coefficient of 20 Barrer. The

COz/CH4 selectivity of poly(HEMA) films was only 0.9, however, fluorination of

-51..

poly(HEMA) increased the COz/CH4 selectivity to 9. Described in Chapter V are the

preparation of PEO-containing polymer brushes grafied from porous alumina membrane

supports, and the permeability of gasses through the membranes.

Multilayer Polyelectrolyte
Film (MPF)

   

4/
BrCOCl-I(CH,)Br
~(c,H,),/TH1=
o 'c

 

'c

   

.3 1:5 , CuCl/ Cqu bpy
c 9 . ‘ +—
;: .g .g- eooww water! DMF
>2 S. 5'- RT
4)‘ (D‘
"- ‘3
z 5

MPF Cross-linked polymer — " V

film Initiator studied to MPF

Figure 1.32. Schematic diagram showing polymerization of EGDMA from a
polyelectrolyte surface modified with an initiator. Reprinted with permission from J.

Membr. Sci. 2003, 227, 1-14.

-52-

References:

1.

10.

11.

12.

13.

14.

Zhao, B.; Brittain, W. J., Prog. Polym. Sci. 2000, 25, 677-710.

Milner, S. T., Science 1991, 251, 905-914.

Edmondson, S.; Osborne, V. L.; Huck, W. T. 3., Chem. Soc. Rev. 2004, 33, 14-22.

Jennings, G. K.; Brantley, E. L., Adv. Mater. 2004, 16, 1983-1994.

Husseman, M.; Malmstrom, E. E.; McNamara, M.; Mate, M.; Mecerreyes, D.;
Benoit, D. G.; Hedrick, J. L.; Mansky, P.; Huang, E.; Russell, T. P.; Hawker, C.
J ., Macromolecules 1999, 32, 1424-1431.

Huang, X.; Wirth, M. J ., Macromolecules 1999, 32, 1694-1696.

Kim, J. B.; Bruening, M. L.; Baker, G. L., J. Am. Chem. Soc. 2000, 122, 7616-
7617.

Jones, D. M.; Brown, A. A.; Huck, W. T. S., Langmuir 2002, 18, 1265-1269.

Pyun, J .; Kowalewski, T.; Matyjaszewski, K., Macromol. Rapid Commun. 2003,
24, 1043-1059.

Pyun, J .; Matyjaszewski, K., Chem. Mater. 2001, 13, 3436-3448.

Ionov, L.; Minko, S.; Stamm, M.; Gohy, J. F.; Jerome, K.; Scholl, A., J. Am.
Chem. Soc. 2003, 125, 8302-8306.

Currie, E. P. K.; Norde, W.; Stuart, M. A. C., Adv. Colloid. Interface Sci. 2003,
100, 205-265.

Luzinov, I.; Minko, S.; Tsukruk, V. V., Prog. Polym. Sci. 2004, 29, 635-698.
Huck, W. T. 8.; Yan, L.; Stroock, A.; Haag, R.; Whitesides, G. M., Langmuir

1999, 15, 6862-6867.
-53-

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

Zhou, F.; Jiang, L.; Liu, W. M.; Xue, Q. J ., Macromol. Rapid. Commun. 2004, 25,
1979-1983.

Zhou, F.; Liu, W. M.; Hao, J. C.; Xu, T.; Chen, M.; Xue, Q. J ., Adv. Funct. Mater.
2003, 13, 938-942.

Senaratne, W.; Andruzzi, L.; Ober, C. K., Biomacromolecules 2005, 6, 2427-
2448.

Rutenberg, I. M.; Scherman, O. A.; Grubbs, R. H.; Jiang, W. R.; Garfunkel, E.;
Bao, Z., J. Am. Chem. Soc. 2004, 126, 4062-4063.

Kamigaito, M.; Ando, T.; Sawarnoto, M., 2001, 101, 3689-3745.

Matyjaszewski, K.; Xia, J. H., Chem. Rev. 2001, 101, 2921-2990.

Patten, T. E.; Matyjaszewski, K., Adv. Mater. 1998, 10, 901.

Advincula, R.; Zhou, Q. G.; Park, M.; Wang, S. G.; Mays, J .; Sakellariou, G.;
Pispas, S.; Hadjichristidis, N., Langmuir 2002, 18, 8672-8684.

Advincula, R. C., J. Dispersion Sci. Tech. 2003, 24, 343-361.

Jordan, R.; Ulman, A.; Kang, J. F .; Rafailovich, M. H.; Sokolov, J ., J. Am. Chem.
Soc. 1999, 121, 1016-1022.

Zheng, L.; Xie, A. F .; Lean, J. T., Macromolecules 2004, 3 7, 9954-9962.

Quirk, R. P.; Mathers, R. T.; Cregger, T.; Foster, M. D., Macromolecules 2002,
35, 9964-9974.

Sakellariou, G.; Park, M.; Advincula, R.; Mays, J. W.; Hadjichristidis, N., J.
Polym. Sci. A: Polym. Chem. 2006, 44, 769-782.

Jordan, R.; Ulman, A., J. Am. Chem. Soc. 1998, 120, 243-247.

-54-

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

41.

42.

Jordan, R.; West, N.; Ulman, A.; Chou, Y. M.; Nuyken, 0., Macromolecules
2001, 34, 1606-1611.

Zhao, B.; Brittain, W. J., J. Am. Chem. Soc. 1999, 121, 3557-3558.

Prucker, 0.; Ruhe, J ., Macromolecules 1998, 31, 592-601.

Prucker, 0.; Ruhe, J ., Macromolecules 1998, 31 , 602-613.

Kim, N. Y.; Jeon, N. L.; Choi, I. S.; Takami, S.; Harada, Y.; Finnie, K. R.;
Girolami, G. S.; Nuzzo, R. G.; Whitesides, G. M.; Laibinis, P. E.,
Macromolecules 2000, 33 , 2793-2795.

Moon, J. H.; Swager, T. M., Macromolecules 2002, 35, 6086-6089.

Juang, A.; Scherman, 0. A.; Grubbs, R. H.; Lewis, N. S., Langmuir 2001, 17,
1321-1323.

Liu, X. G.; Guo, S. W.; Mirkin, C. A., Angew. Chem, Int. Ed. 2003, 42, 4785-
4789.

Husemann, M.; Mecerreyes, D.; Hawker, C. J .; Hedrick, J. L.; Shah, R.; Abbott,
N. L., Angew. Chem, Int. Ed. 1999, 38, 647-649.

Choi, I. S.; Langer, R., Macromolecules 2001, 34, 5361-5363.

Joubert, M.; Delaite, C.; Bourgeat-Lami, E.; Dumas, P., Macromol. Rapid
Commun. 2005, 26, 602-607.

Zhao, B.; Brittain, W. J ., Macromolecules 2000, 33 , 342-348.

Kim, J. B.; Huang, W.; Wang, C.; Bruening, M. L.; Baker, G. L., Bottle Brush
Brushes: Ring-Opening Polymerization of Lactide from Poly(hydroxy
methacrylate) Surfaces. Wiley-VCH VerlagGmbH&Co. KGaA: Weinheim, 2004.

Baum, M.; Brittain, W. J ., Macromolecules 2002, 35 , 610-615.

-55-

43.

44.

45.

46.

47.

48.

49.

50.

51.

52.

53.

54.

55.

56.

Matyjaszewski, K., Controlled Radical Polymerization. Washington, DC, 1998.

Andruzzi, L.; Senaratne, W.; Hexemer, A.; Sheets, E. D.; Ilic, B.; Kramer, E. J .;
Baird, B.; Ober, C. K., Langmuir 2005, 21, 2495-2504.

Matyjaszewski, K., Macromol. Symp. 2002, 183, 71-81.

Connal, L. A.; Gurr, P. A.; Qiao, G. G.; Solomon, D. H., J. Mater. Chem. 2005,
15, 1286-1292.

Wang, A. R.; Zhu, S. P., Macromolecules 2002, 35, 9926-9933.

Brandts, J. A. M.; van de Geijn, P.; van Faassen, E. E.; Boersma, J .; van Koten,
G., J Organometallic Chem. 1999, 584, 246-253.

Kotani, Y.; Kamigaito, M.; Sawamoto, M., Macromolecules 1999, 32, 2420-2424.

Kato, M.; Kamigaito, M.; Sawamoto, M.; Higashimura, T., Macromolecules
1995, 28, 1721-1723.

Bielawski, C. W.; Louie, J .; Grubbs, R. H., J. Am. Chem. Soc. 2000, 122, 12872-
12873.

Wang, J. S.; Matyjaszewski, K., J. Am. Chem. Soc. 1995, 117, 5614-5615.

Matyjaszewski, K.; Wei, M. L.; Xia, J. H.; McDermott, N. E., Macromolecules
1997, 30, 8161-8164.

Muhlebach, A.; Gaynor, S. G.; Matyjaszewski, K., Macromolecules 1998, 31 ,
6046-6052.

Percec, V.; Barboiu, B.; Neumann, A.; Ronda, J. C.; Zhao, M. Y.,
Macromolecules 1996, 29, 3665-3668.

Uegaki, H.; Kotani, Y.; Kamigaito, M.; Sawamoto, M., Macromolecules 1998,
31 , 6756-6761.

-56-

57.

58.

59.

60.

61.

62.

63.

64.

65.

66.

67.

68.

69.

70.

Moineau, G.; Minet, M.; Dubois, P.; Teyssie, P.; Senninger, T.; Jerome, R.,
Macromolecules 1999, 32, 27-35.

Lecomte, P.; Drapier, 1.; Dubois, P.; Teyssie, P.; Jerome, R., Macromolecules
1997, 30, 7631-7633.

Patten, T. E.; Matyjaszewski, K., Acc. Chem. Res. 1999, 32, 895-903.

Pintauer, T.; Matyjaszewski, K., Coordin. Chem. Rev. 2005, 249, 1155-1184.

Coca, S.; Jasieczek, C. B.; Beers, K. L.; Matyjaszewski, K., J. Polym. Sci. Polym.
Chem. 1998, 36, 1417-1424.

Wang, X. S.; Lascelles, S. F .; Jackson, R. A.; Armes, S. P., Chem. Commun.
1999,1817-1818.

Wang, X. S.; Armes, S. P., Macromolecules 2000, 33, 6640-6647.

Huang, W. X.; Kim, J. B.; Bruening, M. L.; Baker, G. L., Macromolecules 2002,
35, 1175-1179.

Perrier, S.; Haddleton, D. M., Macromol. Symp. 2002, 261-272.

Coullerez, G.; Carlmark, A.; Malmstrom, E.; Jonsson, M., J. Phys. Chem. A 2004,
108, 7129-7131.

Lee, S. B.; Russell, A. J .; Matyjaszewski, K., Biomacromolecules 2003, 4, 1386-
1393.

Ali, M. M.; Stover, H. D. H., Macromolecules 2004, 37, 5219-5227.

Liu, S. Y.; Weaver, J. V. M.; Tang, Y. Q.; Billingham, N. C.; Armes, S. P.; Tribe,
K., Macromolecules 2002, 35, 6121-6131.

Cai, Y. L.; Armes, S. P., Macromolecules 2004, 3 7, 7116-7122.

-57-

71.

72.

Fujii, S.; Cai, Y. L.; Weaver, J. V. M.; Armes, S. P., J, Am. Chem. Soc. 2005, 127,
7304-7305.

Dufresne, M. H.; Gauthier, M. A.; Leroux, J. C ., Bioconjugate Chem.

2005, 16, 1027-1033.

73.

74.

75.

76.

77.

78.

79.

80.

81.

82.

83.

Haddleton, D. M.; Perrier, S.; Bon, S. A. F ., Macromolecules 2000, 33, 8246-
8251.

Nomura, E.; Ito, K.; Kajiwara, A.; Kamachi, M., Macromolecules 1997, 30, 2811-
2817.

0h, J. K.; Min, K.; Matyjaszewski, K., 2006, 39, 3161-3167.

Kim, J. B.; Huang, W. X.; Miller, M. D.; Baker, G. L.; Bruening, M. L., J. Polym.
Sci. A: Polym. Chem. 2003, 41, 386-394.

Lemieux, M.; Usov, D.; Minko, S.; Stamm, M.; Shulha, H.; Tsukruk, V. V.,
Macromolecules 2003, 3 6, 7244-7255.

Minko, S.; Patil, S.; Datsyuk, V.; Simon, F.; Eichhom, K. J .; Motomov, M.;
Usov, D.; Tokarev, 1.; Stamm, M., Langmuir 2002, 18, 289-296.

Mori, H.; Boker, A.; Krausch, G.; Muller, A. H. E., Macromolecules 2001, 34,
6871-6882.

Mori, H.; Seng, D. C.; Zhang, M. F.; Muller, A. H. E., Langmuir 2002, 18, 3682-
3693.

Huang, W. X.; Baker, G. L.; Bruening, M. L., Angew. Chem, Int. Ed. 2001, 40,
1510.

Zhao, B.; Haasch, R. T.; MacLaren, S., J. Am. Chem. Soc. 2004, 126, 6124-6134.

Pyun, J .; Jia, S. J .; Kowalewski, T.; Patterson, G. D.; Matyjaszewski, K.,
Macromolecules 2003, 36, 5094-5104.

-53-

84.

85.

86.
87.
88.

89.
90.
91.
92.

93.

94.

95.
96.

97.

98.

von Weme, T.; Patten, T. E., J. Am. Chem. Soc. 2001, 123, 7497-7505.

Bontempo, D.; Tirelli, N.; Masci, G.; Crescenzi, V.; Hubbell, J. A., Macromol.
Rapid Commun. 2002, 23, 418-422.

von Weme, T.; Patten, T. E., J. Am. Chem. Soc. 1999, 121, 7409-7410.

Mandal, T. K.; Fleming, M. 8.; Walt, D. R., Chem. Mater. 2000, 12, 3481-3487.

Xiao, D. Q.; Wirth, M. J ., Macromolecules 2002, 35, 2919-2925.

Perruchot, C.; Khan, M. A.; Kamitsi, A.; Armes, S. F.; von Weme, T.; Patten, T.
E., Langmuir 2001, 1 7, 4479-4481.

Zhao, H. Y.; Kang, X. L.; Liu, L., Macromolecules 2005, 38, 10619-10622.

Peng, B.; Johannsmann, D.; Ruhe, J ., Macromolecules 1999, 32, 6759-6766.

Hamelinck, P. J.; Huck, W. T. S., J. Mater. Chem. 2005, 15, 381-385.

Zhao, B.; Brittain, W. J .; Zhou, W. S.; Cheng, S. Z. D., J. Am. Chem. Soc. 2000,
122, 2407-2408.

Zhao, B.; Brittain, W. J ., Macromolecules 2000, 33, 8813-8820.

Boyes, S. G.; Brittain, W. J .; Weng, X.; Cheng, S. Z. D., Macromolecules 2002,
35, 4960-4967.

Pyun, J .; Jia, S. J .; Kowalewski, T.; Matyjaszewski, K., Macromol. Chem. Phys.
2004, 205, 411-417.

Li, D. J .; Sheng, X.; Zhao, B., J. Am. Chem. Soc. 2005, 12 7, 6248-6256.

Sun, T. L.; Wang, G. J.; Feng, L.; Liu, B. Q.; Ma, Y. M.; Jiang, L.; Zhu, D. B.,
Angew. Chem, Int. Ed. 2004, 43, 357-360.

-59-

99.

100.

101.

102.

103.

104.

105.

106.

107.

Prucker, 0.; Schimmel, M.; Tovar, G.; Knoll, W.; Ruhe, J., Adv. Mater. 1998, 10,
1073.

Husemann, M.; Morrison, M.; Benoit, D.; Frommer, K. J .; Mate, C. M.; Hinsberg,
W. D.; Hedrick, J. L.; Hawker, C. J., J. Am. Chem. Soc. 2000, 122, 1844-1845.

Xia, Y. N.; Whitesides, G. M., Angew. Chem, Int. Ed. 1998, 37, 551-575.

Kaholek, M.; Lee, W. K.; LaMattina, B.; Caster, K. C.; Zauscher, S., Nano Lett.
2004, 4, 373-376.

Ma, H. W.; Hyun, J. H.; Stiller, P.; Chilkoti, A., Adv. Mater. 2004, 16, 338-+.

Ito, Y.; Park, Y. S.; Imanishi, Y., J. Am. Chem. Soc. 1997, 119, 2739-2740.

Ito, Y.; Ochiai, Y.; Park, Y. S.; Imanishi, Y., J. Am. Chem. Soc. 1997, 119, 1619-
1623.

Ito, Y.; Nishi, S.; Park, Y. S.; Imanishi, Y., Macromolecules 1997, 30, 5856-5859.

Balachandra, A. M.; Baker, G. L.; Bruening, M. L., J. Membr. Sci. 2003, 227, 1-
14.

-60.-

Chapter II

Crystallization of Polymer Brushes

This chapter describes the crystallization of comb-like polymer brushes formed by
surface-initiated atom transfer radical polymerization (ATRP) of PEG-methacrylates
(PEGMA) from gold surfaces in aqueous media. Optical microscopy and atomic force
microscopy (AF M) show that the ethylene oxide side chains of poly(PEGMA) crystallize
to form two-dimensional spherulites. The crystal growth rates are film-thickness
dependent. AFM and reflectance FT-IR measurements on < 100 nm fihns are consistent
with formation of crystalline lamellae predominantly oriented parallel to the surface, but
thicker films are mainly comprised of perpendicularly-oriented lamellae. The
crystallization of polymer brushes occurs by heterogeneous nucleation and one-

dimensional growth.

Introduction

Thin and ultra-thin films are attractive for tailoring surface properties because of
their wide range of physicochemical properties. Studies on polymer films, primarily
physically absorbed films deposited on substrates by spin-coating, show that the
interactions between the substrate and polymer layers usually lead to physical properties
that are different than those of the bulk polymer. For example, the glass transition
temperature (T 8) may be enhanced or depressed depending on the film thickness and
interfacial interactions,1 and restriction of chain mobility greatly affects the

10.11

Crystallization process. Polymer morphology, 2'9 degree of crystallinity, preferred

-61-

2, 7,12 3,10,11,13-18

polymer chain orientation, and crystal growth rates may vary with the
film thickness.

Although physical adsorption of films on substrates is a convenient and versatile
method for surface modification, the stability of physisorbed coatings is limited in the
presence of good solvents. For this reason, films covalently bonded to substrates, such as
polymer brushes, are being rapidly adopted for the study of thin films. Some properties of
polymer brushes are unique to their architecture and have not been explored in detail.
For example, the effect that the polymer brush architecture may have on crystallization is
unclear since the extended chains of densely grafted brushes may pre-organize and favor
crystallization, while chain entanglements and the restricted mobility of the polymer
chains in these brushes may retard the crystallization process. Luzinov et al. 19’ 20
studied the surface morphology of PEG brushes formed by reacting the end groups of
activated PEGs of different molecular masses with a surface primed with poly(glycidyl
methacrylate). After PEG grafting, the surface was either partially or totally covered by
finger-like crystalline domains, depending on the molecular mass of the constituent PEGs.
Since the thicknesses of the films were similar, the variations in crystalline morphology
must be related to the grafting density and the molecular masses of the PEGs.

To further understand the effects of brush architecture on crystallization, this
work examines the crystallization of comb-like polymer brushes obtained fiom ATRP of
poly(ethylene glycol) methyl ether methacrylate (PEGMA) from initiators immobilized
on Au surfaces. In the case of PEGMA macromonomers, polymerization from surfaces

yields methacrylate comb polymer brushes with hydrophilic poly(ethylene oxide) (PEO)

side chains as shown in Figure 2.1. The side chains of poly(PEGMA) and other comb

-62-

polymers crystallize when they are sufficiently long, irrespective of the tacticity of the

polymer backbone.”23

9H2
H3C-C-ICI-O(CH2CH20>CH3
+ 0 22-23

 

Figure 2.1. Architecture of comb-like poly(PEGMA) brushes grown from a surface.

The melting temperature (T,,.) of the PEO segment in comb polymers depends on
the length of side chains and ranges from below room temperature to ~65 °C, the T m of
PEO. This range is compatible with thermally unstable substrates such as self-assembled
monolayers on Au, enabling detailed IR analysis of the structure and orientation of the
crystalline PEO side chains. In addition to their relevance for understanding
crystallization of thin films, poly(PEGMA) brushes have important potential applications
as gas separation membranes and materials for limiting non-specific protein adsorption
on surfaces. Studies of chain crystallization in poly(PEGMA) brushes complement
extensive research on crystallization of bulk PEO,” 25 PEO films on surfaces,“ 3‘ ‘4’ ‘5
PEO segments in block copolymers,” 27 as well as films of PMMA/PEO blends.7' 28' 29

Crystalline PEO comb-polymers are usually arranged in a “sandwich” structure,
with layers of amorphous backbones alternating with lamellae of crystalline side chains.
For example, Inomata et al. suggested a crystalline lamellae packing for PEO side chains

of poly(MMA-stat-PEGMA), with amorphous backbone connecting lamellae.23

-63-

Neugebauer et al. proposed a model of alternating disordered and crystalline domains for
polymethacrylates with PEG and alkyl side chains. 2‘

Although the chain conformations in adsorbed films and brushes are different,
polymer crystallization studies of physisorbed films provide a useful starting point for
understanding the crystallization of brushes. As film thickness decreases to nm
dimensions in the physisorbed polymer films, the spherulitic morphology observed in

6’7’ 30 or finger-like patterns for

bulk polymers evolves into sheaf-like aggregates,
monolayer films.9’ 3" 32 AFM makes it possible to observe individual lamellae in
spherulites. For most ultra-thin polymer films studied to date, crystalline lamellae
preferentially orient parallel to the substrate, with the polymer chain axis oriented normal
to the surface. The orientation of crystalline lamellae may change with the fihn thickness,
and lamellae oriented parallel to the surface may give way to perpendicular lamellae as
the film thickness increases. For example, Schonherr et al.2 studied PEO films and
reported crystalline lamellae oriented parallel to the substrate when the films were < 300
nm thick, and perpendicular lamellae for film thickness > 1 am. This trend in lamellae
orientation was also found in crystalline low density polyethylene (LDPE) films7 and
poly(di-n-hexylsilane) films.12 The accepted explanation for these thickness effects is
that perpendicular orientation minimizes the energy for the primary nucleation step, and
therefore, the same orientation is favored for secondary nucleation and crystal growth,
resulting in the favored perpendicular lamellae orientation. In ultra-thin films, lamellae

lying flat on the surface reduce the surface energy more effectively and are favored over

perpendicular lamellae.

-64-

Frank et al.11 reported that the crystallization rate and crystallinity of poly(di-n-
hexylsilane) films decreased significantly when the fihn thickness was < 50 nm. Since
then, several research groups confirmed a reduction in crystal grth rates as film
thickness decreased, but conflicting results remain regarding the correlation between film
thickness and crystal grth rate. Taguchi et al.16 studied the grth rate of isotactic
polystyrene films and claimed that the fihn growth rate (G) can be related to film
thickness (d) using a simple equation: G(d) = C(00) (1- a/d), where C(00) refers to the
growth rate of bulk polymer, and the constant a is ~ 6 nm. In contrast to the above results,
Schonherr et al.2 studied isothermal crystallization of PEO films on silicon substrates,
and concluded from an Avrami analysis of FT-IR data that the crystallite growth rate and
film thickness have an exponential relationship, i.e, G(d) = C(00) exp (1- constant/d).

l. '4’ 18 studied isothermal crystallization of PEO films ranging from 13

Dalnoki-Veress et a
nm to 2 pm, and reported a non-monotonic decrease in crystal grth rate for film
thickness > 40 nm, which they related to morphology changes.

This chapter describes the synthesis and characterization of crystalline polymer
brushes with an emphasis on the effect of film thickness on morphology, chain

orientation in crystallites and the crystallization rate. The crystallization kinetics for

brushes is interpreted in the context of the Avrami-Evans theory.

-65-

Experimental Section

Materials. Poly(ethylene glycol) methyl ether methacrylate (PEGMA, Mn=.1,100
g/mol), Cu(I)Br (99.999%), Cu(II)Br2 (Aldrich, 99.999%), and ll-mercapto-undecanol
[HS(CH2)I 10H] were used as received. 2,2’-Bipyridine (bpy) was recrystallized from hot
hexane and sublimed under vacuum at 60 °C. The disulfide initiator
((CH3)2CBrC00(CH2)IIS)2) was synthesized using a slightly modified version of a
literature procedure.33 Deionized water (Milli-Q, 18 MQcm) was used as the
polymerization solvent.

Synthesis of initiator [Br-C(CH3)2-C00(CHz)uS)2] 33 CH2C12 (150 mL), 10%
aqueous potassium bicarbonate (20 mL) and ll-mercaptoundecanol (4.0 g) were added to
a 250 mL Schlenk flask. Bromine (0.51 mL) was slowly syringed into the well-stirred
mixture, and the initial orange-red color of bromine quickly disappeared. After continued
stirring for 2 h, the organic and water layers were separated, and the water layer was
extracted with 50 mL of CH2C12. The combined organic layers were dried over MgSO4,
filtered, and evaporated to dryness to yield 2.95 g (75%) of H0(CH2)1 IS-S(CH2)110H (1)
as a white powder. 1H-NMR (300 MHz, CDC13): 5 1.23 (br, 32H, CH2), 1.52 (m, 4H,
CH2), 1.62 (m, 4H, CH2), 2.62 (t, 4H, SCHz), 3.61(t, 4H, 0CH2). Under a blanket of Ar,
2-bromoisobutyryl bromide (1.46 mL) was added drop-wise to a 0 °C solution of
disulfide [1] (2.0 g, 5 mmol) and triethyl amine (3.4 mL, 0.07 mol) in 120 mL CH2C12.
After stirring at 0 °C for 1 h, and at 25 °C for 2 h, the solution was filtered to remove
triethylamine hydrobromide and then extracted with 2N sodium carbonate solution
saturated with aqueous ammonium chloride (2 x 50 mL). The dichloromethane was

removed by rotary evaporation and the crude product was purified by column

-66-

chromatography using 10:1 hexane/ethyl acetate as the eluent. Removal of the solvent
provided 1.5 g of the initiator (43%) as a pale yellow viscous liquid. 1H-NMR (300 MHz,
CDC13): 5 1.31 (br, 28H), 1.7 (br, 8H), 1.9 (s, 12H), 2.7(t, 4H, J = 7.3), 4.18 (t, 4H, J =
6.6). 13C-NMR: 5 172 (C=0), 66.4 (0CH2), 56.2 (C), 39.4 (SCHz), 31.0(CH3), 29.7(CH2),
29.5(CH2), 29.4(CH2), 28.7(CH2), 28.6(CH2), 26.0(CH2). Anal. Calcd. for

C30H56Br204S2: C, 51.13; H, 8.01. Found C, 51.27; H, 7.69.

Scheme 2.1. Synthetic procedure for preparation of initiator 2

2 113-6011217011

l Br2

1 S+CH2)-OH
11

s—(CH2)—0H
11

0
E13N I Br/U\<Br

1? it's
s-(CHzl—o-c—c-Br
11 CH3

9H3
s-(CHZIH—o-p-c—Br
o CH3
Preparation of immobilized initiators on Au surfaces. Gold-coated wafers (200

nm of gold sputter-coated on 20 nm of Cr on Si (100) wafers) were UV/03-cleaned for 15

min before use. After washing with water and ethanol, the slides were immersed in a 1

-67-

mM solution of disulfide initiator 2 in ethanol for 24 h to form the self-assembled
monolayer. The films were washed with ethanol, sonicated in ethanol for 1 min, again
washed with ethanol, and then dried under a stream of N2. The substrates were
characterized by ellipsometry and by reflectance F T-IR.

Surface-initiated polymerization of PE GMA. PEGMA (35 g, 0.032 mol) and 20
mL of Milli Q water were added to a Schlenk flask and stirred until the PEGMA
completely dissolved. The flask was connected to a vacuum line and the solution was
degassed by three freeze-pump-thaw cycles, and finally backfilled with Ar. Under a
blanket of Ar, CuBr (0.024 g), CuBr2 (0.011 g), and bpy (0.077 g) were added to the
solution, and the mixture was stirred for at least 1 h to form a homogeneous dark brown
solution. The solution of monomer and catalyst ([CuBr] = 1 mM, [CuBr2] = 0.3 mM, and
[bpy] = 3 mM) was transferred into a N2-filled glovebag and added to a series of vials,
each containing one initiator-anchored slide. Slides were removed from vials at
predetermined times, and the polymer films grown from the gold surfaces were rinsed
sequentially with water, THF, and anhydrous ethanol. The films were then dried under a
stream of N2,

Solution ATRP of PE GMA. PEGMA (20 g, 0.018 mol) and 15 mL of water were
added to a Schlenk flask. The flask was connected to a vacuum line and degassed via
three freeze-pump-thaw cycles. CuBr (0.065 g, 0.45 mmol), CuBr2 (0.034 g, 0.15 mmol)
and bpy (0.14 g, 0.9 mmol) were added to the mixture under Ar, and the solution was
stirred until a homogeneous dark brown solution was formed. Degassed ethyl 2-
bromoisobutyrate (0.18 g, 0.9 mmol) was syringed into the reaction mixture to initiate the

polymerization. The solution became increasingly viscous and after 2 h, the

-68-

polymerization was quenched by exposing the solution to air. After dilution with water
or THF, the resulting blue solution was filtered through a short silica column to remove
the copper catalyst. Removal of the solvent yielded 5 g of poly(PEGMA) as a white solid.
IH-NMR indicated > 99 % conversion of PEGMA (see Figure 2.15). M11 = 60,000 g/mol.

Characterization methods. Reflectance FT-IR was performed using a Nicolet
Magma-IR 560 spectrometer with a PIKE grazing angle of 80 °C. Reflectance FT-IR
experiments used p-polarized light. Since only transition moment components
perpendicular to the surface provide absorption peaks in the spectra, the molecular
conformation of samples were inferred from IR absorption spectra using the previously
reported PEG assignments. Film thicknesses were measured by a rotating analyzer
ellipsometer (M-44; J .A. Woollarn) at an incident angle of 75°. The data were analyzed
using WVASE32 software, and the thickness and refractive index determinations were
performed on at least three spots on each substrate. The refractive index was fitted with
the film thickness, and the measured refractive index was around 1.5. AF M images were
measured using the close-contact mode of a nano-RTM instrument, and both height and
phase images were recorded simultaneously. The silicon tip was used with a spring
constant of 36 N/m, with tip curvature of 10-20 nm, and a resonance frequency of 286-
339 KHz. The scan rate was set to 0.85 Hz and the set-point was adjusted to optimize
image quality.

Crystallization studies. The morphology of crystalline polymer brushes were
imaged using Nikon 0ptiphot2-POL polarizing optical microscope equipped with a
Mettler FP 82 hot-stage and a CCD camera. Samples for isothermal crystallization

experiments were heated at 50 °C for 8 h in a vacuum oven to erase the thermal history of

-69-

the film. The films were immediately transferred to the hot stage, and held at 23 i 2 °C
under a flow of N2. In non-isothermal crystallization experiments, the samples were
loaded into the hot stage and heated to 45 °C (above the melting temperature), and then
cooled at 2 °C /min to ambient temperature, and further cooled at 1 °C /min to ~5 °C by
purging the hot-stage with the boil off from liquid N2. For other experiments, the slides

were cooled to ambient temperature in a desiccator under reduced pressure or under N2.

Polymer solutions for spin-coating were prepared as 1-10 wt % solutions in
solvent CH2Cl2 and filtered through 0.02 pm PTFE filters. All films were deposited at
3000 rpm, with the film thickness controlled by varying the concentration of the polymer
solution. After spin-coating, the films were dried under vacuum and annealed at 30 °C.

The films were characterized by external reflectance FT-[R and by ellipsometry.

Results and discussion

Preparation of polymer brushes. Scheme 2.1 shows the synthetic route used to
grow poly(PEGMA) from gold. Methacrylates containing 4-5, 8-9, and 22-23 ethylene
oxide repeat units were polymerized, but only poly(PEGMA) with 22-23 PEO units
crystallized at room temperature. Research on bulk poly(PEGMA) shows that the atactic
backbone of poly(PEGMA) and a portion of side chains adjacent to the backbone
constitute an amorphous phase, and the ethylene oxide chain must exceed a minimum
chain length, typically 9-10 ethylene oxide units to crystallize. In the remainder of this
Chapter, PEGMA refers to the methacrylate with 22-23 PEO units, unless otherwise

Specified. Polymerization studies of other PEG methacrylates appear in Chapter III.

-70-

Immobilization of the initiator was confirmed by the appearance of a carbonyl
peak in the reflectance FT-IR spectrum (Figure 2.2a) as well as formation of a 17 i 2 A-
thick film, as measured by ellipsometry. The PEGMA macromonomer was polymerized
from the surface using aqueous solution containing the CuBr/CuBr2/bpy ATRP catalyst.
During the first 8 h of the polymerization, the film thickness increased linearly with
polymerization time, consistent with a controlled polymerization (see Figure 2.3), but at
longer times, the film growth rate decreased, which may reflect reaction of active chain

ends with adventitious oxygen in the glovebag.

IR spectra of the films (Figure 2.2b) show characteristic bands for poly(PEGMA):
C-H stretching (2950-2800 cm'l), C=0 stretching (1740 cm”), CH2 bending (1460 cm"),
CH2 wagging (1350 cm"), overlapping CH2 twisting (~1300-1240 cm'l), and a broad C-
0-C stretching band (1150 cm'l). The width and position of the IR peaks are consistent
with an amorphous polymer brush film. As described below, the evolution of the IR

bands indicates that these films crystallize at room temperature.

-71-

 

o CH

s(c1+2+o— -C+Br

Initiator-anchored A CH3
substrate 0 CH3
s-(0H2—)-o— —c+ar

CH3

0 CH3

11 CH3

 

0 CH3
+sfcnfi—o— -CI:+Br
0 01-13

+sfc112—)-o- c—éer

CH3

CH3

H3C 22-23

 

io/VOTEF)

 

Macromonomer PEGMA

0

22-23

 

 

 

CuBr, Cu(||)Br2.bpy / \
“N

H20, r.t.

sfcrufi—o-c— c—(—CH2—c’C—O)—Br

Scheme 2.2. Growth of poly(PEGMA) films from gold.

-72-

\

/

 

0.04

1L

c: Crystalline polymer film

JL_

b: Amorphous polymer film

“Ju- - m;-

a: Initiator x 30

 

Absorbance

 

 

 

 

W

3600 3200 2800 2400 2000 1600 1200 800

Wavenumbers (cm'l)

Figure 2.2. Reflectance FTIR spectra of (a) an initiator monolayer, and (b) an amorphous
114 nm thick poly(PEGMA) film. Spectrum b was taken immediately after its synthesis.
Trace (c) shows a poly(PEGMA) film of the same fihn thickness after crystallization at

room temperature.

-73-

 

 

 

 

500
400 - D
E
- 300 4 ' '3
in .
U) .'
2
f, 200 - F
.E
l-
100 - P
13'
111
o I I I I I
o 5 1o 15 20 25 30
Time (h)

Figure 2.3. Evolution of the ellipsometric fihn thickness with time for the ATRP of
PEGMA at room temperature from initiators anchored to gold surfaces. Conditions:

[CuBr] = 1 mM, [CuBr2] = 0.3 mM, [bpy] = 3 mM, and [M]:[H20] = 2:1 (v/v).

-74-

Morphology of crystalline polymer brushes. The optical microscopy experiments
described here provide a qualitative picture of the crystallization behavior as a function of
temperature and film thickness. While stored at room temperature, the films crystallized
and the mirror-like finish of as-prepared fihns developed surface features that were
visible to the naked eye. When observed between crossed polarizers, the as-prepared
polymer films initially appeared dark (amorphous), but eventually the Maltese cross
patterns characteristic of spherulites filled the field of view. The images shown in Figure
2.4 are from samples annealed at 45 °C and then cooled to the desired temperature. Since
the Tm of poly(PEGMA) films are ~35-40 °C, the super-cooling (AT = T c-Tm) at ~5 °C
supports a sufficiently high crystallization rate to enable crystallites to be observed in a
short period of time.

Panels a and b in Figure 2.4 show the evolution in polymer morphology for a 240
nm-thick fihn. At 10 °C, large and well-defined “spherulites” cover most of the field of
view. The domains have similar radii, consistent with nucleation of each domain at
comparable times. At 5 °C, the crystalline domains grow to cover the entire surface, as
shown in panel b. These disc-like domains are large, ~100 nm in diameter but only 240
nm thick, resulting in an aspect ratio comparable to digital video discs (DVDs). Given
their symmetry and disc-like geometry, the domains may be viewed as a 2-dimensional
slice through the center of a spherulite. Crystallization under isothermal conditions at

room temperature under N2 also produced films with spherulitic morphologies.

Panels b, c and (1 show the effects of film thickness on crystallization. All three
samples were imaged at 5 °C under comparable conditions. While the 240 nm film was

fully crystallized, some areas of the 148 and 76 nm films remained amorphous. The 76

-75-

run film appeared as sheaf-like aggregates, which may be typical of the early stages of
spherulite development. These observations show that the crystallization behavior
poly(PEGMA) films is similar to thin PEO films, and that both the crystalline

morphology and crystallization rate depend on the poly(PEGMA) film thickness.

 

Figure 2.4. Optical micrographs of poly(PEGMA) brushes grown from gold surfaces and
viewed through crossed polarizers: (a) and (b) show the crystallization of a 240 nm at 10
°C and at 5 °C, respectively; (c) and (d) show the crystallization of 148 nm and 76 nm,

films, respectively at 5 °C. Images in this dissertation are presented in color.

-76-

AFM measurements provided a more detailed view of the crystalline polymer
brush morphology. Figure 2.5 shows flat and 3-D height images obtained for a 108 nm
film. Panel a shows a clearly resolved boundary (1 ) between two spherulitic domains and
well-defined perpendicular lamellae (2) that emanate from the center of the “spherulite”.
Panel b shows the apparent overlap of two spherulitic domains, possibly at the site of a
film defect, resulting in twisted branches and a more irregular boundary between the two

domains.

The evolution of film morphology with thickness was examined using 20-300 nm
films that were crystallized under identical conditions (isothermal crystallization). Figure
2.6 shows AFM images for 122 and 310 nm films. The image of the 310 nm film shows
thick, closely packed perpendicular lamellae. The observed branching of lamellae in the
122 nm film is the expected consequence of the radial growth of “spherulites”.

The morphologies of films < 100 nm are more complicated and are thickness-
dependent. Panel a of Figure 2.7 Compared to thicker films, the lamellar morphology
seen in the height image is less resolved, but more clearly shown in the phase image. The
morphology of the 70 nm film is similar to that of the 92 nm film. A switch from
perpendicularly oriented lamellae to lamellae presumably lying parallel to the surface
came from the image of a 55 nm film, which clearly shows a worm-like film morphology
that resembles aggregates of lamellae parallel to the surface (Figure 2.8). As the film
thickness decreased to 35 nm, however, the AFM image of the film appeared
homogeneous and nearly featureless, suggesting either an amorphous film, or a

morphology dominated by parallel lamellae.

-77-

(a)

25 nm

Onm

 

(b)

30nm

iOnm

       

0 pm 2.5 pm 5 pm

Figure 2.5. AFM images of a 108 nm crystalline polymer brush film. At left are the
height images and at right are the corresponding 3-D views. In panel a, the scan distance
is 4 x 4 pm, with a 2 distance of 25 nm. In panel b, the scan distance is 5 x 5 pm, and the

2 distance is 30 nm. Images in this dissertation are presented in color.

-73-

(a) 310nm

 

20 nm
0 nm
0 pm 2.5 pm 5 pm
(b) 122nm
. 40 nm
0 nm

 

Figure 2.6. AFM images of crystalline polymer brushes. (a) shows data for a 310 nm
thick film and (b) shows data for a the 122 nm film. Both image are 5 x 5 um, and the
heights of the 310 and 122 nm films are 20 and 40 nm, respectively. Images in this

dissertation are presented in color.

-79-

      
    

 

 

   

   

(a) 92 nm
24 nm f. - 475 mV
i’ ' { ' 2.,
it >- . ’25:?" y”:- '-.' #2-
r: {3,15}? ' . gut-c; ./ 3 In?“ ' '1
14:13:- e‘ - '1?-
“‘1‘: "'- " ': -
Onm --».p; _ OmV
0pm 25pm 5pm 0pm 25pm 5pm
(b) 70 nm
31 "m 919 mV
Onm V
Ont
0 pm 2.5 pm 5 pm 0 11m 2-5 pm 5 pm
(c) 35 nm
11 nm 561 mV
Onm OmV
0pm 25pm 5pm 0pm 25pm 5pm

-30-

Figure 2.7. (Left) AFM height images (5 x 5 pm) of crystalline polymer brushes with
thickness of: (a) 92 nm, 2 distance is 24 nm; (b) 70 nm, 2 distance is 31 nm; (e) 35 nm, 2
distance is 10 nm. (Right) Phase images of crystalline polymer brushes with thickness of:
(a) 92 nm, 2 distance is 475 mV; (b) 70nm, 2 distance is 919 mV; (c) 35 nm, 2 distance is
561 mV. Images in this dissertation are presented in color.

15 nm

 

Figure 2.8. AFM height images of a crystalline film with thickness of 55 nm. Scan
distance of the left image is 5 x 5 pm with a 2 distance of 15 nm. The scan distance in the
right image is l x 1 pm with a 2 distance of 15 nm. Images in this dissertation are
presented in color.

~81-

Spectroscopic characterization

Reflectance FT-IR measurements can provide spectroscopic confirmation of
crystallization and information on the orientation of the poly(PEGMA) side chains
relative to the surface. As seen in Figure 2.2c, crystallization causes several
poly(PEGMA) IR bands to sharpen. Notable changes consistent with crystallization
include the C-O-C band at 1150 cm'1 splitting into strong peaks at 1120 and 1148 cm],
and the CH2 wagging peak at ~1350 cm'l evolving into a sharp peak at 1360 cm'1 with a
small shoulder at 1345 cm'l. Both indicate a high degree of crystallinity for the PEO
chain.

PEO usually crystallizes as a 72 helix, which contains seven CH2CH2O repeat
units and two turns of the helix in the 19.3 A between equivalent sites along the helix.
The PEO chain in the helix is a succession of nearly trans (CCOC), trans (COCC), and
gauche (OCCO) conformations along the chain.” 25 This structure, shown in Figure 2.9,
has been confirmed by X-ray measurement as well as extensive spectroscopic studies.
PEO adopts a planar zig-zag conformation under tension,34 but when the stress is
removed, the zig-zag conformation disappears rapidly.

The symmetry of the IR bands for the helical PEO have been assigned to two
species, one with transition moment parallel to the helical axis, and the other with

transition moment perpendicular to the helical axis (See Table 2.1).

-32-

Top view

 

Side view

 

 

Figure 2.9. Crystalline structure of PEO with top view and side view.

Polymer brushes with different film thicknesses were characterized by reflectance
FT -IR spectroscopy (Figure 2.10). All films were annealed at 50 °C to erase their thermal
history, and then held at room temperature until crystallization was complete. Trace a is
the IR spectrum of a 13 nm-thick film, and the width and position of the peaks indicate
that this film was amorphous. All fihns of comparable thickness (< 20 nm) proved to be
amorphous, possibly because the restricted conformations of ultrathin films prevent
efficient packing of the PEO side chains into crystalline lamellae. Another possibility is
that films < 20 nm thick have melting points at or below room temperature. However, the
the latter seems unlikely, since a 300 nm fihn melts at 38 i 1 °C, and the melting point

only decreases to 34 :l: 1 °C for a 80 nm film.

-83-

Table 2.1. Absorption band assignments for crystalline PEO

 

Ban (1 assignment 3,1, Transition Wavenumber Wavenumber (cm'l)

 

moments ° (cm") a (this study)

5(CH2) J- 1473,1466 1467
6((CH2) 11 1461,1453 d

w (CH2) 4 1358 1360

w (CH2) 11 1342 1344

1(an ) J- 1278 1280

1(an ) 11 1240 1242

v(C0) J- 1147 1148

0(C0) -L 1116 1120
0(CO) 11 1103 e

v(C0)+r(CH2)+u(CC) J- 1060 1062

r (CH2)+ 1) (CH2 ) 11 958 964

r (CH2) + 0(CC) J- 947 947

r(CH2)+u(CO) J- 844 843

 

3 Ref 24,29 b 5 = scissor; w = wag; t = twist; u = stretch; r = rock. c J- and II represent
polarization direction perpendicular and parallel to the helical axis of PEO, d overlapped.

e not observed.

-84-

The IR spectrum of a 35 nm thick fihn (spectrum b) shows only four peaks
between 1400 and 900 em‘1 (1344, 1242, 1120, 964 cm"). The CH2 wagging (1360, 1344
cm"), CH2 rocking (964, 947 cm'l) and combination bands (843 cm") are sensitive to the
conformation order.24 Absorption peaks at 1344, 1242, and 964 cm"1 in trace b
correspond to vibrations with transition moments parallel to the PEO helical axis.
Moreover, since the reflectance FT-IR measurements used p-polarized light with the
active transition moment normal to the surface, the helical axis of ethylene oxide units
was oriented perpendicular to the substrate.

The FT-IR spectra evolved as the film thickness increased. Spectrum c shows a
small band at 1148 cm], which appears as a shoulder of the 1120 cm'1 band. Splitting of
this C-O-C absorption confirms crystallization of the ethylene oxide segments. In
addition to the 1344 and 1242 cm'1 bands seen in thinner films, there are new absorptions
at 1360 and 1280 cm}. These bands are more obvious in trace (1. The 1360, 1280, 1148,
947 and 843 cm'1 bands correspond to vibrations with transition moments perpendicular
to the helical axis. Thus, there are two populations of helical PEO in 92 nm thick film,
one with its axis perpendicular to the surface and the other parallel to the surface. In
thicker films (traces e and f), the 1360, 1280, 947 and 843 cm‘1 bands dominate indicating
an increasing proportion of PEO segments oriented parallel to the surface.

The FT-IR data confirm the previous interpretations of the AFM images. The
morphology of films >100 nm thick appears to be dominated by perpendicularly oriented
lamellae, since polarized FT-[R measurements identify the dominant helix orientation in
these films as parallel to the surface, and hence consistent with lamellae oriented normal

to the surface.

-85-

 

1148 1120
1467 1360 134412801242

I ||l|

843

I 0.1 964 947
1062

   
 
 

   

 

f) 240 nm

e) 177 nm

(1) 92 nm '1 A j \
c) 70 nm
b) 35 nm

a)13 nm (intensity x 5) A

-
A

Absorbance

 

 

 

 

 

 

1 600 1500 1400 1 300 1200 1 100 1000 900 800

Wavenumber (cm'1)

Figure 2.10. Reflectance FTIR spectra of poly(PEGMA) brushes of various thicknesses;
a) 13 nm; b) 35 nm; c) 70 nm; (1) 92 nm; e) 177 nm; 1) 240 nm. The brushes were initially

annealed at 50 °C and then allowed to crystallize at room temperature in N2.

-86-

The morphologies of films <100 nm thick are more complicated, with AFM
showing a progressive loss of perpendicular lamellae as films become thinner. For very
thin fihns such as the 35 nm thick film shown in Figure 2.7, the AFM image was
featureless, but FT-IR detected relatively sharp bands at 1342 and 1242, 1120, and 964
cm'1 indicating some order, and by inference, crystallinity. In contrast, the 55 nm fihn of
Figure 2.8 shows signs of structure in the AFM image, different than seen in thicker fihns,
suggesting lamellae oriented parallel to the surface. The FT-IR spectra of thin films are
consistent with such a conclusion.

IR measurements indicate that the PEO chains in the 92 and 70 nm-thick fihns are
orientated both parallel and perpendicular to the surface. AFM only showed irregular
perpendicular lamellae, possibly due to the overgrth of perpendicular lamellae on the
top of parallel lamellae. In addition, AF M usually detects the top layer morphology of the
film, while vibrational bands shown in the IR spectra reflect mean orientations of
polymer chains.

The two orientations for crystalline lamellae in films are shown in Figure 2.11.
Crystallization of thin polymer brushes in lamellae parallel to the surface would reduce
the polymer and surface interaction, which is favorable. Rafailovich et al.7 claimed that
for the crystallization of thick LDPE films, the energy of the primary nucleation step is
minimal when the chains are oriented parallel to the surface. However, as described in
later sections of this chapter, crystallization of poly(PEGMA) occurs by heterogeneous
nucleation. Formation of perpendicular lamellae also may be driven by comb-like
structure of poly(PEGMA), which favors effective side chain interactions when the side

chains are oriented parallel to the surface. Bulk comb polymer samples typically

-37-

crystallize in layers with the crystalline side chains oriented roughly perpendicular to the
backbone. Densely packed comb polymer brushes favor an extended polymer backbone
when the surface density is high enough, with the polymer backbones roughly oriented
perpendicular to the surface and the side chains parallel to the surface. Crystal grth
follows the same orientation, leading to the perpendicular lamellae. A similar side chain

orientation was described recently by Gabriel et al.35

In addition to identifying chain orientation, IR spectroscopy suggests that films
must be >20 nm to exhibit detectable crystallinity. Several groups studied spin-coated
PEO films and reported different results regarding the critical film thickness for
crystallization. Schdnherr et al.2 did not detect crystallinity in fihns < 15 nm. However,

Reiter et al. 9' 32

reported crystallization in monolayer-thick films (~ 5 nm). The
poly(PEGMA) case is structurally different from spin-coated PEO. First, the backbones
of comb-like structures usually interfere with crystallization side chain segments adjacent
to the polymer backbone, which should lead to a higher critical thickness for
crystallization than linear PEO films. (We note that the melting point of the
poly(PEGMA) brushes are significantly lower than linear PEO, which suggests a lower
thermodynamic stability). In addition, polymer brushes with high graft densities may be
viewed as pre-organized for crystallization, but more restricted in terms of available

chain conformations. More detailed studies of the crystallization rates will illuminate the

importance of these issues.

-33-

 

 

 

Figure 2.11. Lamellae orientations in the films of poly(PEGMA): (a) Chain axis of PEO
side chains is oriented perpendicular to the substrate, lamellae parallel to the surface; (b)

Chain axis of PEO side chains is parallel to the substrate, perpendicular lamellae.

-89-

Crystallization kinetics of polymer brushes.

The classical method used to interpret the grth of polymer spherulites is the

application of the Avrami-Evans theory.3"‘”39 The Avrami equation is

v/v =1 - exp(-Kt ") Eq.].

where v/v is the volume percent crystallinity, t is the crystallization time, K is a
crystallization rate constant, and n is the Avrami exponent, a parameter that reflects the
dimensionality of crystal growth. For atherrnal nucleation (homogeneous), all nucleation
events occur at t = 0, and n can have values of 1, 2, and 3, corresponding to one, two, and
three dimensional growth, respectively. For thermal nucleation (heterogeneous),
nucleation occurs as a function of time), and n can have values of 2, 3, and 4, which
correspond to one, two and three dimensional growth.” 41 The Avrami exponent is
obtained by measuring crystallinity as a function of time and analyzing the experimental
data using equation 2, a logarithmic version of equation 2.1.

log [- In (1- v/v)] = n logt + log K Eq. 2.
where n and K are extracted from the slope of a plot of log [- In (1- v/v)] vs. log t and the
intercept, respectively.

The crystallization of thin polymer brush films is conveniently monitored by
optical microscopy, which allows real-time imaging of the growth of individual
spherhulites. The volume percent crystallinity in thin films can be approximated as:

vc/v~A,/A ~fc7rr2/7rR2
where A, and A are the crystalline area and the total area observed, respectively. For the
case of a single spherulite, the radius r is tracked over time and related to R, the radius of

the firlly grown spherulite with a degree of crystallinityfi.
_90-

The crystallization rates for a series of poly(PEGMA) thin films of various
thickness were measured using polarized optical microscopy. Prior to crystallization, the
films were annealed at 50 °C for 12 h to erase their thermal history. The fihns were
immediately inserted into a hot stage purged with N2, quenched to 23 i 2 °C, and the
evolution of the film morphology was captured by a video camera. Figure 2.12 shows the
growth of spherulites in a 240 nm thick poly(PEGMA) film. The use of the term
“spherulite” is technically correct only at the initial stages of crystallization, when the
radius of the spherulite is less than the film thickness. This stage lasts only a few seconds
and the radii of the spherulites quickly exceed the film thickness. With radii > 50 um,
these disk-like objects correspond to a slice through the center of a “spherulite”, a

cylindrical crystallite.

The “spherulite” growth rates were measured for films of different thicknesses
and plotted as a function of time (Figure 2.13a). For each sample, a constant growth rate
(G: dR/dt) of 0.02 to 0.07 um/s was observed, which increased with fihn thickness. The
average spherulite grth rate is plotted in Figure 2.13b. The growth rate increases with
film thickness (regime I), but then appears to level off for fihns >300 nm (regime II). The
increase in crystal grth rate is likely due to an increase in the average chain mobility of
longer chains, enabling side chains interact with each other and incorporate into the
lamellar structure more easily. At film thickness > 300 nm, the effect of the surface-
anchoring on chain mobility has a minor influence on crystal growth, resulting in grth

rates independent of fihn thickness.

-91-

 

Figure 2.12. Optical micrographs showing the growth of spherulites in a 240 nm thick
film at 23 t 2 °C. The samples are viewed using crossed polarizers. The scale bar

represents 100 um. Images in this dissertation are presented in color.

-92-

 

 

  
  
  
  
    

 

 

  

 

 

 

 

 

500
450 _ O 370nm
I 310nm
400 - A 250nm
E I 210nm
3 35° ‘ o 122nm A
0
g 300 . O 113nm
E A 80nm
g 250 -
in
'6
m 200 -
.2
E 150 -
100 -
50 -
0 . . .
0 50 100 150 200
Time (min)
8
7 _
Regime | {1‘ I. 161

8-4 1 l

+ Regime ll

Growth rate (x10'2um/s)
#

 

 

 

o I l l l
O 100 200 300 400 500
Thickness (nm)

Figure 2.13 a) The evolution of spherulite radii with time for the crystallization of
poly(PEGMA) brush films of various thickness at room temperature. b) Spherulite
growth rates measured as a function of the poly(PEGMA) film thickness.

-93-

The linear fit of the Avrami equation to the experimental data is shown in Figure
2.14. For all data sets, the Avrami exponent extracted from the slopes was ~2, which is
consistent with either homogeneous nucleation (atherrnal) and two-dimensional growth
or heterogeneous nucleation (thermal), and one-dimensional growth (see Table 2.2.). For
homogeneous nucleation, the distribution of nuclei in the film should be random, and
therefore, the nucleation pattern in a given film should be different each time it is
crystallized. For heterogeneous nucleation, secondary nucleation may occur at defect
sites, and assuming the sites are not lost, the nucleation pattern should be conserved in a

subsequent crystallization.

Table 2.2. Avrami exponent n determined from 2-D spherulite grth monitored by in

situ polarized optical microscopy.

 

 

Film thickness (nm) Growth rate (><10'2 um/s) Avrami Exponent n

76 1.7 :t 0.3 2.1
113 2.4 :t 0.2 2.1
210 4 at 1.4 2.2
260 4.7 i 0.7 2.4
290 6.7 :t 0.3 —
310 6.2 i 0.2 2.2
370 6.6 i 0.2 2

 

-94-

To identify the nucleation mode, we heated crystalline films above their melting
temperature to erase their thermal history and let the samples cool to room temperature
and crystallize. Alter crystallization was complete, we repeated the experiment three
times, and followed the crystallization process using polarized optical microscopy. By
comparing the images for each melt/crystallization cycle, we found that the distribution
of nucleation sites were the same for each cycle, indicating heterogeneous nucleation and

1-D crystallization (radial growth).

 

 

 

 

 

 

 

 

0.5
-l- 370nm
0 _ <> 310nm + 0 A I
o 250nm +0 ~ '
I 240nm ++°
-o 5 ~ 0 38
: 1:1 210nm + . g
> '12, A
\o X 12201“ +
? .1 - 0 I
- .2,
c o 80nm + 0
E. -1.5 ~ .
X
C) O
0 ° I,
" . Cl
2 +
-2.5 ~ 0
'3 1 r I
2.5 3 3.5 4 4.5

log t

Figure 2.14. Linear fit of the Avrami equation to the-experimental data for crystalline

polymer brushes with different film thickness.

-95-

F rank et al. used Avrami analysis to interpret the dimensionality of crystal growth

in poly(di-n-hexylsilane)ll and PEO films.2 For the poly(di-n-hexylsilane) film, polymer
chains are oriented parallel to the surface, and high T c and low film thickness favor one
dimensional growth. For PEO films, the polymer chains were oriented perpendicular to
the surface, and the crystal growth was 3-dimensional. The dimensionality of crystal
grth is related to the polymer and interface interactions.
Several studies have reported the crystallization rates of PEO thin films.2’ 3’ 14’ 18
The results are dependent on the crystallization conditions (T c, supercooling), the
substrate, and the molecular weight of PEO. However, linear PEO is a poor model for the
crystallization of polymer brushes because of the different architectures of linear and
comb polymers as well as the processes involved in main-chain and side-chain
crystallization. Therefore, we studied the crystallization behavior of spin-coated
poly(PEGMA) thin films for comparison.

Poly(PEGMA) was prepared by ATRP of the PEGMA macromonomer in water
using a CuBr/CuBr2/bpy catalyst. After a 2 h polymerization at room temperature, the
polymerization was exposed to air to terminate the reaction. Purification provided
poly(PEGMA) as a white solid, with M" = 60,000g/mol 1H-NMR and IR spectra (see
below) confirmed the chemical structure. Thin films were spin-coated onto gold coated
silicon wafer substrate using CH2Cl2 as the solvent. After drying under vacuum and
annealing at 30 °C, the samples were allowed to crystallize at room temperature.

Optical microscopy of the spin-coated film showed a spherulitic morphology with
extinction of banded rings, which is possibly due to the twisting and dislocation of

lamellae. Like the poly(PEGMA) brush, the orientation of the crystalline lamellae in

-96-

spin-coated fihns also depended on the film thickness. For a 60 nm thick film, the
preferred lamellae orientation is parallel to the surface with the PEO side-chains oriented
perpendicular to the surface. FT-IR data from the thick film (290 nm) is consistent with
the PEO side-chains parallel to the surface, suggesting an perpendicular lamellae

orientation.

d+e+f

 

 

 

 

 

Figure 2.15.1H-NMR spectrum of poly(PEGMA).

-97-

 

DA

A_

Absorbance

 

 

 

 

 

3500 3000 2500 2000 1500 1000 500

Wavenun'ber (cm'1)

Figure 2.16. Reflectance FT-IR spectra of 56 and 290 nm thick poly(PEGMA) films that

were spin-coated on the gold-coated silicon wafer.

-93-

The “spherulite” growth rate measured for a spin-coated 300 nm thick
poly(PEGMA) film is shown in Figure 2.17. The spin-coated film crystallized twice as
fast (0.12 urn/s) as a polymer brush (0.06 prn/s) of comparable thickness. The slower
crystalline growth rate for polymer brushes indicates that the decreased mobility imposed
by the brush architecture trumps any benefits of chain alignment. Alignment effects may
prove to be important only when the density of polymer chains on the surface approaches

full coverage.

 

 

   
    

 

 

 

700

o Spin-coated film -300 nm
600 ~

I Polymer brushes-300 nm
500 ~

400 - G = 0.12 Irm/s

300 -

Radius (um)

200 -

100 i

 

 

 

0 1000 2000 3000 4000 5000 6000
Time (s)

Figure 2.17. Crystal grth rates measured for spin-coated poly(PEGMA) and
poly(PEGMA) brushes of comparable film thickness (~300 nm).

-99-

Conclusions.

Surface-initiated ATRP of PEGMA macromonomer from gold substrates lead to
the nanometer-thck films of comb-like polymer brushes. Polymers with side chains of 22-
23 ethylene oxide units align and crystallize. The film morphology, orientation of
crystalline lamellae, and crystal grth rates all depend on the fihn thickness. Polarized
optical microscopy and AFM show that the polymer brush films crystallized into
spherulitic domains. In addition, both AFM and reflectance FT-IR spectroscopy are
consistent with crystalline lamellae preferentially oriented perpendicular to the surface
when the film thickness exceeds 100 nm. For thinner films, the favored orientation
shifted to crystalline lamellae parallel to the surface, with the PEO side-chains oriented
perpendicular to the surface. Crystal growth rates decreased with fihn thickness, and an
Avrami analysis of the data indicated heterogeneous nucleation and one-dimensional
crystal growth. Polymer brushes crystallized slower than physically absorbed films, likely

due to mobility restrictions imposed by the brush architecture.

-100-

References:

10.

11.

12.

13.

14.

Prucker, 0.; Christian, S.; Bock, H.; Ruhe, J .; Frank, C. W.; Knoll, W., Glass
transition in ultrathin polymer films. In Organic Thin Films, 1998; Vol. 695, pp
233-249.

Schonherr, H.; Frank, C. W., Macromolecules 2003, 36, 1188-1198.

Schonherr, H.; Frank, C. W., Macromolecules 2003, 36, 1199-1208.

Pearce, R.; Vancso, G J ., Macromolecules 1997, 30, 5843-5848.

Pearce, R.; Vancso, G J ., Polymer 1998, 39, 1237-1242.

Mellbring, 0.; 0iseth, S. K.; Krozer, A.; Lausmaa, J .; Hjertberg, T.,
Macromolecules, 2001, 34, 7496-7503.

Wang, Y.; Ge, S.; Rafailovich, M.; Sokolov, J .; Zou, Y.; Ade, H.; Luning, J .;
Lustiger, A.; Maron, G, Macromolecules 2004, 37, 3319-3327.

Sakai, Y.; Imai, M.; Kaji, K.; Tsuji, M., Macromolecules 1996, 29, 8830-8834.

Reiter, G; Sommer, J. U., J. Chem. Phys. 2000, 112, 4376-4383.

Despotopoulou, M. M.; Frank, C. W.; Miller, R. D.; Rabolt, J. F., Macromolecules
1995, 28, 6687-6688.

Despotopoulou, M. M.; Frank, C. W.; Miller, R. D.; Rabolt, J. F., Macromolecules
1996, 29, 5797-5804.

Hu, Z. J .; Huang, H. Y.; Zhang, F. J .; Du, B. Y.; He, T. B., Langmuir 2004, 20,
3271-3277.

Mareau, V. H.; Prud'homme, R. E., Macromolecules 2005, 38, 398-408.

Dalnoki-Veress, K.; Forrest, J. A.; Massa, M. V.; Pratt, A.; Williams, A., J. Polym.
Sci. B: Polym. Phys. 2001, 39, 2615-2621.

-101-

15.

l6.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

Schultz, J. M.; Miles, M. J ., J. Polym. Sci. B; Polym. Phys. 1998, 36, 2311-2325.

Taguchi, K.; Miyaji, H.; Izumi, K.; Hoshino, A.; Miyamoto, Y.; Kokawa, R.,
Polymer 2001, 42, 7443-7447.

Marentette, J. M.; Brown, G R., Polymer 1998, 39, 1405-1414.

Massa, M. V.; Dalnoki-Veress, K.; Forrest, J. A., Eur. Phys. J. E 2003, 11, 191-198.

Zdyrko, B.; Klep, V.; Luzinov, 1., Langmuir 2003, 19, 10179-10187.

Zdyrko, B.; Varshney, S. K.; Luzinov, 1., Langmuir 2004, 20, 6727-6735.

Neugebauer, D.; Theis, M.; Pakula, T.; Wegner, G; Matyjaszewski, K.,
Macronolecules 2006, 39, 584-593.

Xu, J. T.; J i, J ., Polymer 2003, 44, 6379-6385.

Inomata, K.; Nakanishi, E.; Sakane, Y.; Koike, M.; Nose, T., J. Polym. Sci. B:
Polym. Phys. 2005, 43, 79-86.

Yoshihara, T.; Murahashi, S.; Tadokoro, H., J. Chem. Phys. 1964, 41 , 2902.

Miyazawa, T.; Ideguchi, Y.; Fukushima, K., J. Chem. Phys. 1962, 37, 2764-&.

Sun, X.; Zhang, H.; Zhang, L.; Wang, X.; Zhou, Q.-F., Polym. J. 2005, 3 7, 102-
108.

Richardson, P. H.; Richards, R. W.; Blundell, D. J .; MacDonald, W. A.; Mills, P.,
Polymer 1995, 36, 3059-69.

Wang, M. T.; Braun, H. G; Meyer, E., Polymer 2003, 44, 5015-5021.

Hoffmann, C. L.; Rabolt, J. F., Macromolecules 1996, 29, 2543-2547.

Bartczak, Z.; Argon, A. S.; Cohen, R. E.; Kowalewski, T., Polymer 1999, 40,
2367-2380.

-102-

31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

41.

Reiter, G; Castelein, G; Hoemer, P.; Riess, G; Sommer, J. U.; Floudas, G, Eur.
Phys. J. E. 2000, 2, 319-334.

Reiter, G; Sommer, J. U., Phys. Rev. Lett. 1998, 80, 3771-3774.

Shah, R. R.; Merreceyes, D.; Husemann, M.; Rees, 1.; Abbott, N. L.; Hawker, C.
J .; Hedrick, J. L., Macromolecules 2000, 33, 597-605.

Takahash.Y; Sumita, 1.; Tadokoro, H., J. Polym. Sci. B; Polym. Phys. 1973, 11,
2113-2122.

Gabriel, S.; Dubruel, P.; Schacht, E.; Jonas, A. M.; Gilbert, B.; Jerome, R.; Jerome,
C., Angew. Chem. Int. Ed. 2005, 44, 5505-5509.

Avrami, M., J. Chem. Phys. 1939, 7, 1103.

Avrami, M., J. Chem. Phys. 1940, 8, 212.

Avrami, M., J. Chem. Phys. 1941, 9, 177.

Evans, U. R., Trans. Faraday. Soc. 1945, 41, 365.

Gedde, U., Polymer Physics, 1 st Edition. Kluwer Academic Publishers: Dordrecht,

1995.

Wunderlich, B., Macromolecular Physics. Academic Press: New York, 1973,
1976; Vol. 1, 2.

-103-

Chapter 111

Synthesis of Comb-like Polymer Brushes on Planar and

Nanoparticulate Surfaces

This chapter describes the synthesis and characterization of comb-like polymer
brushes on planar and non-planar surfaces using surface-initiated ATRP. Comb-like
polymer brushes were formed on gold and silica surfaces by polymerization of PEGMA
macromonomers in aqueous media. Polymerization of macromonomers is challenging
because their high molecular weight results in low monomer concentrations. Studies of
the kinetics of surface-initiated ATRP on planar substrates were used to identify optimum
conditions for the growth of PEGMA films. In the second part of this chapter, these
polymerizations were used to grow comb-like polymer brushes on the surface of silica

nanoparticles.

Introduction

Poly(ethylene glycol) (PEG) coated surfaces are important in biological and
medical applications due to PEG’s ability to resist non-specific protein adsorption and
cell adhesion.1 A variety of chemical methods have been developed to deposit strongly
adhering PEG coatings on surfaces including self-assembly of PEG-containing thiols on
gold,2 coupling reactive groups on PEG with silicon substrates} 4 or the use of priming
layers such as poly(glycidyl methacrylate) (PGMA). However, these “grafting to”
methods usually result in a low PEG grafting density and thin films. Alternatively, PEG
brushes can be grafted from initiators anchored on surfaces through the polymerization of

—104-

monomers that incorporate PEG segments. The advantage of using comb-like polymer
brushes is that dense arrays of PEG side chains are formed in a one-step polymerization.
A preference for well-defined brush structures implies their synthesis by a controlled
polymerization technique such as ATRP. Zhu et al. 5 used ATRP to grow
poly(PEGMA(4-5)) (the numbers refer to the average number of oxyethylene repeat units
in the poly(ethylene glycol) methyl ether chain) from silica by ATRP and obtained films
< 30 nm thick. Recently, Ma et al. 6 used ATRP to synthesize poly(PEGMA(8-9)) from a
gold coated surface, however, the limiting film thickness was only 100 nm after a 16 h
polymerization. Obviously, faster polymerizations of such monomers are needed to
provide ready access to materials designed to exploit PEG-functionalized surfaces.

In the previous chapter, we described the polymerization of PEGMA from gold
and studied the crystallization behavior of polymer brushes. In this chapter, we describe
optimization of the polymerization conditions for the synthesis of polymer brushes from
PEG-substituted methacrylates (Scheme 3.1), with an emphasis on generating thick films

on surfaces using short polymerization times.

 

,CH3
H20=C\
,c=o
(Lt/\oit’nCHa
PEGMA
0 CH3 1 m=4-5
ll '
stcrlanIo-c—Qer 2. m=8-9 i? 9H3 [CH3
CH3 3. "1:22.23 siCHzio-C-g—PCHz—C—i-Br
(1? CH3 c e c IIB L1 (13' H CH3 >=0n
s(CH2-)-o—c+8r U r‘ ”( ) '2' “a" O
11 CH3 H20,l‘.i. ?
o
+-

3

H3O

Scheme 3.1. Surface-initiated ATRP of PEGMA from gold surfaces.

-105-

Experimental

Materials. Unless otherwise noted, all chemicals were obtained from Aldrich.
PEGMA(4-5), (Mn = 300 g/mol) and PEGMA(8-9), (Mn = 875 g/mol) were eluted
through a basic alumina column to remove inhibitor. PEGMA(22-23) (Mn = 1,100 g/mol),
Cu'Br (99.999%), Cu'Cl (9999904101013:2 (99.999%), 1,1,4,7,10,10-hexamethyltriethyl
enetetramine (HMTETA), tetramethyl-1,4,8,1l-tetraazacyclotetradecane (Me4Cyclam),
4,4’-di(n-nonyl)-2,2’-bipyridine (anbpy) 10-undecen-1-ol, 2-bromoisobutyryl bromide,
pyridine, chlorodimethylsilane (97%) and H2PtC16 were all used as received. Milli-Q
water (18 M9) and N,N-dimethylformamide (DMF) (HPLC grade, inhibitdr free) were
used as polymerization solvents. Fumed silica (FS) (Aerosil A200), with a primary
particle size approximately 12 nm and an average surface area of 200 mZ/g was kindly

supplied by Degussa.

Characterization methods. Film thicknesses were measured by a rotating
analyzer spectroscopy ellipsometer (M-44; J .A. Woollam) at a 75° angle of incidence.
The refractive index and film thickness were calculated simultaneously from the values
of A and (1). Thermal gravimetric analyses (TGA) of fumed silicas were obtained in dry
air using a Perkin Elmer TGA 7 instrument at a heating rate of 10 °C/min. Each sample
was held at 110 °C for 30 min to remove adsorbed water prior to initiating the run.
Differential scanning calorimetry (DSC) analyses of polymers were obtained using a
Perkin Elmer DSC 7. Samples were run in sealed aluminum pans under an N2
atmosphere. The coolant for the system was liquid nitrogen. Each sample was heated

from room temperature to 120 °C at a rate of 10 °C/min, held at 120 °C for 30 min, cooled

-106-

to -90 0C at a rate of -50 °C /min, and then heated to 120 °C at a rate of 10 °C/min. The

melting point was taken as the maximum of the melting endotherrn.

Polymerization using the CuBr/CuBr/HM TE T A catalyst complex. A catalyst
solution was prepared in an oxygen-free drybox by adding CuBr (0.057 g, 20 mmol),
CuBr2 (0.045 g, 6 mmol) and HMTETA (0.13 g, 60 mmol) to 20 mL of DMF, and then
transferred to a N2-filled glove bag. A second Schlenk flask was loaded with PEGMA(4-
5) (30 g, 0.09 mol) and H20 (15 mL). The flask was connected to a vacuum line and the
solution was degassed using three freeze-pump-thaw cycles, back-filled with N2, and then
transferred to the N2-filled glove bag. Catalyst solution (5 mL) was added to the second
flask, and the resulting blue-green solution was stirred until homogeneous. The final
catalyst concentrations were [CuBr] = 2 mM, [CuBr2] = 1 mM and [HMTETA] = 6 mM.
The solution was added to a series of vials, each containing one initiator-anchored slide.
Slides were removed from vials at predetermined times, ranging from 5 min to 1 h.
(Polymerizations longer than 1 h generally became viscous with some gel formation.)
The polymer films were rinsed sequentially with water, DMF, and anhydrous ethanol,

and finally dried under a stream of N2.

Polymerization using CuBrMe4Cyclam/CuBr2(anbpy) 2 as the catalyst. In a
drybox, CuBr2 and anbpy (1:2 molzmol) were mixed and the resulting green complex
was stored for future use. A catalyst solution was prepared in an oxygen-free drybox by
adding CuBr (0.057 g, 20 mmol), CuBr2(anbpy)2 (0.21 g, 10 mmol) and Me4Cyclam
(0.1 g, 20 mmol) to 10 mL of DMF. The solution was stirred until homogeneous (~ 20
min). A second Schlenk flask was loaded with PEGMA(4-5) (18 g, 0.05 mol) and H20 (9

mL). The flask was connected to a vacuum line and the solution was degassed using three

-107-

freeze-pump-thaw cycles, and finally back-filled with N2. In a N2-filled glovebag, a
portion of the catalyst solution (3 mL) was added into the flask. The final concentrations
were [CuBr] = 2 mM, [CuBr2 (anbpy)2] = 1 mM, and [Me4Cyclam] = 2 mM. The
catalyst monomer solution solutions were combined, causing the formation of a brown
solution and deposition of a precipitate. The solution was immediately transferred into
the vials, each containing an initiator-anchored gold slide. The polymer fihns were taken
out from the solution at predetermined times, ranging from 5 min to 1 h, rinsed
sequentially with water, DMF, and anhydrous ethanol, and finally dried under a stream of

N2.

Synthesis of undecen-I-yl, 2-bromo-2-methyl propionate (I).7 A 500 mL round
bottom flask was loaded with 10-undecene-l-ol (42.78 g, 0.25 mol), anhydrous THF (250
mL), and 25 mL of pyridine. The flask was cooled to 0 0C, and 2-bromoisobutyryl
bromide (24.8 mL) was added drop-wise to the flask. After stirring the reaction mixture
at room temperature for 12 h, the solution was filtered to remove pyridinium
hydrobromide, and THF was removed under reduced pressure. Hexane (100 mL) was
added to the mixture, and the solution was washed with 2N HCl and water (2x). The
organic phase was collected and dried over MgS04, filtered, and the solvent was removed
by rotary evaporation. Purification by column chromatography on silica using 10/1 (v/v)
hexane/ethyl acetate as the eluent provided 1 (Rf = 0.6) as a transparent liquid in 82.4%
yield. 1H-NMR: 5 l.2-1.45 (br, 12H), 1.64 (m, 2H), 1.94 (s, 6H), 2.05 (m, 2H), 4.17 (t,
2H), 4.95 (d, 2H), 5.8 (m,1H) ppm. 13C-NMR: 8 171.6 (C=0), 139.1 (CH), 114.1 (CH2),
66.2 (OCH2), 56.01 (C), 32.97 (SCH2), 30.81 (CH3), 29.4 (CH2), 29.3 (CH2), 29.24 (CH2)

29.18 (CH2), 28.4 (CH2), 25.8 (CH2).

-108-

Synthesis of (11-(2-bromo-2-methyl)propionyloxy)undecyldimethylchlorosilane
(2). H2PtC16 (5 mg) and undecen-l-yl, 2-bromo-2-methyl propionate (l) (3.00 g, 9.43
mmol) were added to a dry flask. Chlorodimethylsilane (10.43 mL, 94.3 mmol) was
syringed into the mixture and the reaction mixture was stirred at room temperature under
N2. The reaction was monitored by 1H-NMR, and alter the reaction was complete, the
yellowish solution was diluted with anhydrous toluene, and quickly passed through a
column of activated carbon and silica. Removal of the solvent provided (1 l-(2-bromo-2-
methyl)propionyloxy)undecyldimethylchlorosilane (2) as a colorless liquid in 60% yield.
lH-NMR: 5 0.4 (s, 6H), 0.9 (t, 2H) 1.22-1.42 (br, 16H), 1.8 (m, 2H), 1.93 (s, 6H), 4.16 (t,
2H). 13C-NMR: 5 171.7 (C=0), 66.2 (OCH2), 56.01 (C), 32.97 (SCH2), 30.81 (CH3),
29.4 (CH2), 29.3 (CH2), 29.24 (CH2), 29.18 (CH2), 28.4 (CH2), 25.8 (CH2), 22.1 (CH2),

19.0 (CH2), 1.69 (CH3).
0

_ ‘(CHaie-OH + 131/kg,

Pyridine
THF

0
1 ‘(CH2)9-o—5-<Br

HSiMe2C|
H2PtC|5

\ 0
ll
2 Cl-si—(CH2)II—o-c~<er

 

Scheme 3.2. Synthetic route to (l1-(2-bromo-2-methyl)propionyloxy)

undecyldimethylchlorosilane.

-109-

Surface-initiated AT RP of PRGMA from Si wafers. Silicon (100) wafers were
cleaned in a toluene ultrasound bath for 5 min. The wafers were then rinsed with toluene,
acetone and anhydrous ethanol, and dried with a stream of N2. The samples were
oxidized in a UV/ozone chamber for 15 min. The thickness of the S102 layer measured by
ellipsometry was 16-22 A. The growth of initiators was done in a drybox or glove bag
purged with N2. To attach the initiator to the surface, the silica wafers were immersed in
an initiator/toluene solution for 24 h. The fihn was washed with a large amount of toluene
and ethanol, and dried under a stream of N2. The increase in film thickness measured by
ellipsometry, 16 :1: 2A, confirmed the presence of the initiator.

A typical polymerization procedure is described. A Schlenk flask was loaded
with PEGMA(4-5) (42 g) and MilliQ water (21 mL). The solution was degassed by three
freeze-pump-thaw cycles, and finally filled with argon. Cu(I)Br (0.018 g), Cu(II)Br2
(0.008 g), and bpy (0.06 g) were added to the solution under argon. The mixture was
stirred until a homogeneous dark brown solution formed (at least 1 h). The final catalyst
concentrations were [CuBr] = 2 mM, [CuBr2] = 0.6 mM, and [bpy] = 6 mM. The solution
was transferred into a glove bag and added to a series of vials, each containing one
initiator-anchored gold slide. Slides were removed from vials at predetermined times.
Polymer films grown from silica wafers were rinsed with a large amount of water,
followed by sonication in toluene, washed with ethanol, and finally dried under a stream
of N2.

Grafling of polymer brushes on the surface of fumed silica. Fumed silica was
placed under vacuum at 120 °C for 12 h to remove adsorbed moisture. Fumed silica (0.5

g) was added to a Schlenk flask and the flask was placed under vacuum and back-filled

-110-

with N2. Initiator (5 1.1L) and anhydrous toluene (10 mL) was syringed into the flask under
N2, and the reaction mixture was stirred for 24 h. After the reaction was complete, the
mixture was filtered through a 40 um pore-size filter under reduced pressure. The fumed
silica was exhaustively washed with toluene under Soxhlet conditions, and dried under
vacuum to yield ~2 g of white particles. TGA analysis and IR spectra (Figure 3.11.)
confirm attachment of the initiator (C=O peak at 1725 cm”).

For the polymerization of PEGMA from the silica particles, fumed silica (with
initiator attached) was added to a Schlenk flask, evacuated, and back-filled with argon.
Degassed solvent (H2O or H2O/MeOH) and the monomer mixture was added and the
solution was stirred to ensure complete dispersion of the silica in the solution. The
catalyst solution (CuBr/CuBr2/bpy complex) was added to the monomer solution to start
the polymerization. After 2 h, polymerization was terminated by exposure of solution to
the air. Water was added to the mixture, causing the solution to turn blue. The solution
was sonicated for 2 min, and then the solids were collected by centrifugation. After
decanting the solvent, an aqueous solution of saturated ethylenediaminetetraacetic acid
(EDTA) was added to the mixture followed by sonication and centrifugation. After
repeating this procedure three times, the solution was colorless. The solvent was decanted,
and the remaining solvent was removed under vacuum. The poly(PEGMA)-modified
silica was isolated as a white powder. The same procedure applied to PEGMA(8-9)
yielded a white, slightly sticky solid.

Control experiments. A200 fumed silica was added to a Schlenk flask, and then
evacuated and backfilled with Ar to establish an inert atmosphere. A degassed mixture of

monomer and solvent (H2O or H2O/MeOH) was added to the silica and stirred to disperse

-lll-

the silica. Then a mixture of CuBr/CuBr2/bpy dissolved in an aqueous solution of
monomer (see Table 3.1) was added to the solution and stirring was continued for 2 h.
The work-up was started by adding water to the mixture, causing the solution to turn blue.
After sonication for 2 min, the solids were isolated by centrifugation. The solvent was
decanted and saturated aqueous EDTA was added to the mixture followed by sonication
and centrifirgation. The resulting colorless solution was decanted and the residual solvent
was removed under vacuum. The powdery white solid was subjected to TGA analysis.
The procedure for the other control experiment was similar, except the silica used had an

attached initiator layer, and no CuBr2 was added to the catalyst mixture.

Results and Discussion

Surface-initiated ATRP of PEGMA on Au. 1n surface-initiated ATRP, the absolute
amount of initiator on the surface is small, and thus the concentration of Cu"Br due to the
reaction of catalyst with initiator is negligible. Once radicals are generated on the surface,
they cannot easily be converted to their dormant state, which leads to termination through
bimolecular radical coupling and disproportionation. Thus, the addition of Cu"Br at the
beginning of the reaction is necessary for controlled polymerizations from surfaces.
Mixed halide systems (CuBr2/CuCl) were reported to effectively slow the polymerization
rate and reduce termination by maintaining a lower concentration of growing radicals.8
Therefore, our initial experiments on the surface-initiated ATRP of PEGMA used a
CuCl/CuBr2/bpy catalyst system. The mixed halide system provided good control over
polymerization as expected, but the polymerization rate, ~2 nm/h and 80 nm after 48 h,
was too slow (Figure 3.1). The likely reasons for the low polymerization rate were a low

Keq (kac,<<kde,c,), the intrinsic low concentration of polymerizable groups in the

-112-

macromonomer, and possible rate reductions due to reduced access of the
macromonomer to the growing chains on the surface. Switching from CuCl to CuBr, a
more reactive catalyst, provided faster grth rates at the same catalyst concentration
(Figure 3.1), however, the limiting film thickness was < 100 nm and curvature in the fihn
thickness-time relationship indicates significant termination. Similar results were
obtained for the polymerization of PEGMA(8-9) catalyzed by CuBr and CuCl. Again, we
found that mixed halide system provided good polymerization control, but thin films. As

shown in Figure 3.2, the difference between the two catalyst systems is dramatic.

80

 

Thickness (nm)

 

 

 

 

0 . i 4. i ‘e
0 5 10 15 20 25
Time(h)

Figure 3.]. Evolution of the ellipsometric film thickness with time for the grth
of poly(PEGMA(22-23)) from Au. Conditions: [M]:[H20] = 3:2 (v/v). (a) [CuBr] =10
mM, [CuBr2] = 5 mM, [bpy]= 30 mM. (b) [CuCl] = 10 mM, [CuBr2] = 5 mM, [bpy] = 30
mM. The curves are guides to the eye.

-ll3-

500
450 -
400 ~
350 -
300 ~
250 -
200 a
150 -
100 -

 

Film thickness (nm)

50-

 

 

 

O T 1 1 1
0 10 20 30 40

Time (h)

Figure 3.2. Evolution of the ellipsometric film thickness with time for the grth
of poly(PEGMA(8-9)) from Au. Conditions: (a) [CuBr] = 28.5 mM, [CuBr2] = 9.6 mM,
[bpy] = 85.5 mM; (b) [CuCl] = 28.5 mM, [CuBr2] = 9.6 mM, [bpy] = 85.5 mM. The

curves are guides to the eye.

-114-

Previous results from Kim et al. showed that high catalyst concentrations led to a
spatially dense concentration of surface-bound radicals that easily combine. Thus, the
radical concentration is not constant and polymer film grth fails to follow first order
kinetics in monomer. Reducing the catalyst concentration can increase the film grth
rate and provide improved control over the polymerization. Figure 3.3 shows the effects
of catalyst concentration on the film grth rate and film thickness in the ATRP of
PEGMA(22-23). The early stages of the polymerizations are similar, but the highest
growth rate and highest limiting film thickness (400 nm) correspond to the lowest
catalyst concentration. Near-linear grth occurred during the initial 8 h of
polymerization, and deviation from linear grth at later stages suggests some
termination. Polymerizations using high catalyst concentrations were plagued by
significant termination and ceased after 5 h, yielding a maximum film thickness of ~50
nm.

Applying the same approach used for the polymerization of PEGMA(8-9) gave
qualitatively similar results (Figure 3.4). Two catalyst concentrations were tested, [CuBr]
= 28.5 mM, and [CuBr] = 3 mM. Again the higher catalyst concentration provided the
thinner film. Interestingly, the film grown with [CuBr] = 3 mM is roughly comparable to
data set b in Figure 3.3, but the film derived from the lower molecular weight monomer
was significantly thicker. Increasing the monomer concentration leads to thicker films
(Figure 3.5.) as expected. Compared to the case of small monomers, a higher monomer
concentration may help overcome the slow diffusion of the macromonomers to the

reactive surface sites.

-115-

450

 

400 -

(a)
350 1

300 i
250 -

(b)
200 -

Thickness (nm)

150 —
(c)
100 a

 

 

50 - .
/ (d)

 

0 I T T T l I I
0 5 10 15 20 25 30 35

 

Time (h)

Figure 3.3. Evolution of the ellipsometric film thickness with time for the growth
of poly(PEGMA(22-23)) films from Au. [M]:[H20] = 3:2 (v/v). (a) [CuBr] = 0.7 mM,
[CuBr2] = 0.21 mM, [bpy] = 2.1 mM; (b) [CuBr] = 2 mM, [CuBr2] = 0.6 mM, [bpy] = 6
mM; (0) [CuBr] = 5 mM, [CuBr2] = 1.5 mM, [bpy] = 15 mM; ((1) [CuBr] = 10 mM,
[CuBr2] = 3 mM, [bpy] = 30 mM.

-1l6-

 

600

500 i

400 -

300 r

200 .

Film thickness (nm)

 

100

 

 

 

o I l I r 1
0 1 0 20 30 40 50

Time (h)

Figure 3.4. Evolution of the ellipsometric film thickness as a function of time for
the polymerization of poly(PEGMA(8-9)). [M]:[[H2O] = 1:1. (a) [CuBr] = 28.5 mM,
[CuBr2] = 9.6 mM, [bpy] = 85.5 mM. (b) [CuBr] = 3 mM, [CuBr2] = 1 mM, [bpy] = 9

mM. The curves are guides to the eye.

-ll7-

 

 

Film thickness (nm)

 

 

 

o 1 1 1 1 F
0 10 20 30 40 50

Time (h)

Figure 3.5. Evolution of the ellipsometric film thickness with time for the
polymerization of PEGMA(8-9) at different monomer concentrations. Conditions: [CuBr]
= 28.5 mM, [CuBr2] = 9.6 mM, [bpy] = 85.5 mM. (a) [M]:[H20] = 2:1 and (b)
[M]:[H2O]=1 :1. The curves are guides to the eye.

-118-

Rapid growth of comb-like polymer brushes on the surface. As described earlier,
our motivation for optimizing the grth conditions for poly(PEGMA) brushes is the
potential utility of thick brushes as membranes, electrolytes and in other applications.
Because of the large surface area required in gas separations, processes that take hours to
deposit films are not economically viable. There are alternatives to surface initiated
ATRP, such as plasma polymerization or UV-induced graft polymerization, for
generating polymer films on membranes. However, plasma polymerization is
incompatible with non-volatile monomers, both methods can lead to polymer degradation,
and neither affords significant control. As shown earlier, controlled grth of 100 nm
films via ATRP can takes several hours, but practical syntheses require controlled growth
with rates >100 nm/h. Our goal is the synthesis of polymer membranes in minutes using
surface-initiated polymerization. Dense films are preferable since they may enhance
membrane selectivity.

Our earlier results suggest that a more rapid polymerization system requires a low
concentration of a highly active catalyst in a concentrated monomer solution. Copper
based ATRP catalysts with tetradentate ligands such as HMTETA are more reactive than
catalysts with bpy and other bidentate ligands.9 Recently, Bao et al. reported the ultrafast
AT RP of tert-butyl acrylate (20 nm/min) from Au and silica surfaces using a
CuBr/Me4Cyclam/CuBr2/(anbpy)2 catalyst in DMF/anisole. The same catalyst system
might also be useful for the polymerization of PEGMA. Since the
CuBr/Me4Cyclam/CuBr2/(anbpy)2 catalyst is sparingly soluble in water, we used DMF
as a co-solvent. Despite the formation of a precipitate when a DMF solution of the

catalyst was added to the monomer/water mixture, the PEGMA polymerization rate was

-119-

 

300

250 - 0

200 - O

150 i 9

Thickness (nm)

100 —

50‘

 

 

 

0 r 1 1 1 1 r
0 10 20 30 40 50 60 70

Time (min)

Figure 3.6. Evolution of the ellipsometric film thickness with time for the
polymerization of PEGMA(4-5) using Me4Cyclam as the ligand. Conditions:
[M]:[H2O+DMF] = 1:1; [CuBr] = 2 mM, [CuBr2/(anbpy)2] = 0.6 mM, [MeaCyclam] =
6 mM.

-120-

 

500
450 - 0
400 - .
350 -
E 300 — .
3 250 i
Q)
E l
g 200
.C
l- 150 ~ 5
100 -
50 ~ °
0 1 T

 

 

 

0 10 20 30 40 50 60 70

Time (min)

Figure 3.7. Evolution of the ellipsometric film thickness with time for the
polymerization of PEGMA(4-5) using HMTETA as the ligand. Conditions:
[M]/[H2O+DMF] = 1:1, [CuBr] = 2 mM, [CuBr2] = 0.6 mM, [HMTETA] = 6 mM. The

data points are the average of two independent polymerizations.

-121-

very fast, with 100 nm thick films formed in 5 min (Figure 3.6). Using a related ligand,
HMTETA, a 5 min polymerization of PEGMA(4-5) using water as a solvent yielded a 60
nm thick film. Remarkably, the linear growth in film thickness with time implies that the
polymerization is well-controlled.

Figure 3.8. compares PEGMA polymerizations using the HMTETA and
MeaCyclam ligands. Both systems rapidly polymerize PEGMA(4-5) (~10 nm/min),
however, each system has limitations. Polymerizations using Me4Cyclam were much less
controlled, and the rate of polymerization visibly decreased alter 15 min. Polymerization
using HMTETA became viscous after 1 h polymerization, which suggests substantial
chain transfer. Significant chain transfer has been observed in free-radical10 and
photoiniferter mediated11 polymerization of PEGMA. The coefficient for chain transfer to
PEG 12 is an order of magnitude higher than MMA (1.2 x 10'5 ),13 and is believed to be
dominated by of hydrogen abstraction from the ethylene glycol units. Although increases
in solution viscosity, and presumably chain transfer, were occasionally seen in the
CuBr/bpy system, it is unclear why chain transfer is favored for the HMTETA catalyst.
A complication of the Me4Cyclam system is the marginal solubility of the catalyst. The
reaction solution is inhomogeneous and a portion of the CuBr catalyst complex
precipitates. When pure DMF is used as the solvent, the HMTETA system fails to
generate thick films. Another problem of heterogeneous polymerization is that they

provide less uniform films.

-122-

 

500
450 7 9
400 —
350 -
300 2
250 1
200 ~

Thickness (nm)

150- I Q
100— '
50-

 

 

 

0 1 1 l 1 1 1
0 10 20 30 40 50 60 70

Time (min)
Figure 3.8. Evolution of the ellipsometric film thickness with time for the

polymerization of PEGMA(4-5) using HMTETA(O) and Me4Cyclam (I) as ligands. The
data are re-plotted from Figures 3.6 and 3.7.

-123-

Synthesis of PE G-functionalized nanoparticles. Particles with PEO-modified
surfaces have applications ranging from stationary phases for chromatographic
separations to polymer electrolytes in lithium ion batteries.14 Initially, comb-like polymer
brushes were grown on the surface of silicon wafers since this substrate facilitates fihn
characterization by ellipsometry. Later, PEG-functionalized nanoparticles were prepared
by the polymerization of PEGMA from the surface of fumed silica in aqueous media at
room temperature.

Growth of polymer brushes on silica wafers. Scheme 3.3 shows the synthetic
steps used to grow polymer brushes from silicon wafer surfaces. Traditionally, people use
trichlorosilanes to anchor initiators and other organic species on SiOz surfaces, but
polymerization of trichlorosilanes competes with surface anchoring, complicating
quantification of the anchoring step. Therefore, initiators were attached to silicon wafers
and silica nanoparticles as monochlorosilanes. To generate initiator layers on $02, clean
silicon wafers were placed in an initiator/toluene solution for 24 h. Ellipsometric
measurements showed an increase in fihn thickness (20 i 3A) confirming successful
attachment of the initiator. Et3N was occasionally used to accelerate the initiator
anchoring reaction.

Surface-initiated ATRP was catalyzed by a mixture of CuBr/CuBrZ/bpy in
aqueous media at room temperature. Figure 3.9 shows the polymerization of PEGMA(4-
5). Thick films (>300 nm) were obtained in 8 h by using a relatively low catalyst
concentration ([CuBr] = 2 mM, [CuBr2] = 0.6 mM, [bpy] = 6 mM). The film thickness-
time relationship is linear, indicating a controlled polymerization, and a lower catalyst

concentration yielded thinner films. Both are consistent with minimal termination.

-124-

\ i?
Cl—S/i—(CH2)11-O—C—<Br

Toluene, 24h

i?
i-(CH2)11-O—C—<Br

i?
i-(CH2)11-O—C—<Br

 

 

CH3
CuBr, CuBr2, bpy H20=C:
H20, rt. O,C=0 CH
v wo”; 3
\ 9 9H3 9H3
o—Si—(CH2)11-o—c-c—(CH2—c)—Br
/ CH3 o=c "

 

6fCH2CH20)CH3
m

Scheme 3.3. Growth of comb-like PEO polymer brushes from SiOz surfaces.

-125-

 

 

 

 

350
F
300 -
,I
250 - "
E 200 ,9
7,; ‘ (a)
8 ," , I
c 150 ~ .I x6
x .' ,’
.2 i' , ’
.C ,’
+- 100 q ,. ,9 (b)
50 * l'lf,‘
o ' . . . .
o 2 4 6 8 10
Time (h)

Figure 3.9. Evolution of the ellipsometric film thickness with time for the
polymerization of PEGMA(4-5). Conditions: [M]:[H20] = 2:1, (a) [CuBr] = 2 mM,
[CuBr2] = 0.6 mM, [bpy] = 6 mM; (b) [CuBr] = 1 mM, [CuBr2] = 0.3 mM, [bpy] =3mM.

-126-

Synthesis of comb-like polymers on the surface of fumed silica. Growth of
PEGMA from nanoparticles (Scheme 3.4) follows the same procedure used for flat SiOz
surfaces. Activated fumed silica particles were dispersed in a toluene solution of initiator
and the reaction mixture was stirred for 24 h. White particles were isolated after filtration,
exhaustive Soxhlet extraction, and drying under vacuum. Initiator attachment to fumed
silica was confirmed from the OH stretching (2900 cm'l) and C=O stretching (1725 cm")
bands in the transmission IR spectrum (Figure 3.4). Considering the 8-9 wt% loss seen in
TGA analyses (Figure 3.5), the surface area of A200 (200 mZ/g), and the areal density of
silanols on SiOz surfaces (1 mmol OH/g), we estimate that ~% of the OH groups on the

surface reacted with the monochlorosilane initiators.

 

 

     
 

 

 

 

OH
OH
Me O CH Br
i u 3
OH Cl-S'i-(CH2)11—O-C-C:-Br
Me CH3
Fumed silica OH ‘ Br
particle 7
0H Toluene, r.t. 20h Br
OH
OH
i0 Ethylene oxide units
0 as side chain
PEGMA(22-23) ? *
PEGMA(4-5) O Fumed silica /
#5 particle

 

ATRP

Scheme 3.4. Modification of silica particles by surface-initiated ATRP of PEG-
methacrylates

-127-

PEGMA(4-5) and PEGMA(22-23) were used as macromonomers for
polymerization. Fumed silica with surface anchored initiators was dispersed in a
PEGMA/water solution, degassed, and filled with Ar. The catalyst mixture (CuBr, CuBr2,
and bpy) was then added to initiate the polymerization. The macromonomer and catalyst
concentrations used in the polymerizations are listed in Table 3.1. After 2 h, the
polymerization was terminated by exposure to air. The particles were purified, dried, and
characterized as before. As shown in Figure 3.11, IR spectra show a large increase in the
C=O stretching and CH2 wagging bands. However, the C-O-C stretching bands of the

oxyethylene units overlapped with a strong Si-O-Si peak and were not observed.

 

 

 

 

 

0.05
0
0
5 (b)
'E
O
in
.0
<
M
4000 3200 2400 1 600 800

Wavenumber(cm'1)

Figure 3.10. FT-IR spectra of (a) initiator anchored fumed silica particles; (b)
poly(PEGMA(22-23)) grown from fumed silica particles.

-128-

Table 3.1. Reaction conditions for the polymerization of PEGMA(22-23) and
PEGMA(4-5) from the surface of fumed silica particles.

 

 

Entry Monomers [M] [M]:[I]:[CuBr]:[CuBr2]:[pr] Time (h)
1 PEGMA(22-23) 0.5 300: l : 1.1 :0.34 : 3.3 2
2 PEGMA(4-5) 0.5 167 :1 : 0.75 : 0.25 :2.5 1.5
3 PEGMA(4-5) 0.8 500: 1: 1.1 : 0.34 :3.3 0.5

 

TGA analysis indicated an 80% weight loss, as shown in Figure 3.6. To prove that
the polymer films are covalently attached to the fumed silica surfaces, two control
experiments were performed. First, fumed silica, the macromonomer, and catalyst
solutions were mixed under the same condition used for ATRP. After isolating the silica
using the washing and drying protocol described earlier, the particles were analyzed by
TGA. The weight loss was ~5% (Table 3.2). In the second control experiment, initiator-
anchored fumed silica was combined with a monomer/water solution containing CuBr2,
but no CuBr catalyst. After stirring for 2 h, the particles were isolated and analyzed by
TGA. The 8% weight loss supports covalent anchoring of poly(PEGMA) to the surface,

rather than simple physisorption of the macromonomer.

-129—

 

 

 

120 Weight
retention
(a)
100 . _ 93.8%
"——‘ 8849
A 80 ~ /°
25 (b)
3 60 ~
0
.l
j; 4M)—
o
3: 20 «
--- 7;r%
° ‘ (C)
-20 r

 

 

 

0 100 200 300 400 500 600 700
Temperature (°C)

Figure 3.11. TGA analysis of: (a) bare fumed silica; (b) initiator-anchored fumed silica;
(c) poly(PEGMA(22-23)) grown from fumed silica.

Table 3.2. Results from TGA control experiments.

 

Control Experiment TGA weight loss

Mixture of fumed silica and monomer/catalyst solution 5-6%

Mixture of initiator attached fumed silica with monomer and 8°/
0
CuBr2 solution.

 

-130-

Poly(PEGMA)(22-23) will crystallize at room temperature, but it has been
proposed that when PEO chains interact with polar groups such as surface silanols, the
dipolar and hydrogen bonding interactions in these groups would inhibit crystallization.
Because of the high surface area of A200, it is plausible that poly(PEGMA(22-23)) may
not crystallize. DSC results indicate that the attached PEGMA(22-23) chains are
crystalline with T m ~ 40 °C. Therefore, shorter macromonomers must be employed for
applications where cystallization is deleterious.

Poly(PEGMA(4-5)) modified silica particles were prepared as described for
poly(PEGMA(22-23)). These polymerizations are unremarkable, as the two monomers
have nearly identical structures. The monomer and catalyst concentrations used for two
polymerizations, targeting XI, = ~170 and 500, respectively, are shown in Table 3.1. TGA

analysis indicated a 90% weight loss for these materials.

Table 3.3. DSC results of poly(PEGMA) silica particle composites.

 

Entry TGA analysis DSC results

 

Weight Loss T.,, (°C) A Hm (J/g) Tc (°C) A He (J/g)

 

1 87% 40.04 76.44 -0.83 63.70

2 87% 40.39 71.95 -4.76 58.10

 

 

 

 

 

-131-

 

 

120 Weight
(a) retention
10° ‘ _ 93.8%
— O
.. 80 . 884/0
55 (b)
3 so -
O
.J
S _
13’ 4o
3 20 ~
— 7.2%
O i (C)
-20

 

 

 

0 100 200 300 400 500 600 700
Temperature (°C)

Figure 3.12. TGA analysis of (a) bare fumed silica; (b) initiator-anchored fumed silica; c)
poly(PEGMA(4-5)) grown from the firmed silica surface.

Summary and Outlook

Comb-like polymer brushes were successfully grown from the surface of planar
substrates (gold and silicon) and silica nanoparticles. Surface-initiated ATRP of PEGMA
macromonomers yielded thick films under mild conditions. TGA analyses shows that the
grafted polymer comprises ~ 90% of the particle weight. Poly(PEGMA(22-23)) grafted
to silica particles is crystalline, however, poly(PEGMA(4-5)) grafted to similar silica
particles is amorphous. The fast growth of polymer brushes from a variety of surfaces

suggests possible applications in high throughput membrane formation.

-132-

References

1.

10.

ll.

12.

13.

Lee, J. H.; Lee, H. B.; Andrade, J. D., Prog. Polym. Sci. 1995, 20, 1043-1079.

Mrksich, M.; Whitesides, G. M., ACS Symposium Series 1997, 680, 361-373.

Zhu, X. Y.; Jun, Y.; Staarup, D. R.; Major, R. C.; Danielson, S.; Boiadjiev, V.;
Gladfelter, W. L.; Bunker, B. C.; Guo, A., Langmuir 2001, 17, 7798-7803.

Sofia, S. J .; Premnath, V.; Merrill, E. W., Macromolecules 1998, 31 , 5059-5070.

Feng, W.; Chen, R.; Brash, J. L.; Zhu, S., Macromol. Rapid. Commun. 2005, 26,
1383-1388.

Ma, H. W.; Wells, M.; Beebe, T. P.; Chilkoti, A., Adv. F unc. Mater. 2006, 16,
640-648.

Matyjaszewski, K.; Miller, P. J .; Shukla, N.; Irnmarapom, B.; Gelman, A.;
Luokala, B. B.; Siclovan, T. M.; Kickelbick, G.; Vallant, T.; Hoffrnann, H.;
Pakula, T., Macromolecules 1999, 32, 8716-8724.

Xia, J. H.; Matyjaszewski, K., Macromolecules 1999, 32, 2434-2437.

Johnson, R. M.; Ng, C.; Samson, C. C. M.; Fraser, C. L., Macromolecules 2000,
33, 8618-8628.

Bo, G.; Wesslen, B.; Wesslen, K. B., J. Polym. Sci. A: Polym. Chem. 1992, 30,
1799-1808.

Luo, N.; Hutchison, J. B.; Anseth, K. S.; Bowman, C. N., Macromolecules 2002,
35, 2487-2493.

Okamura, S.; Katagiri, K.; Motoyama, T., J. Polym. Sci. 1960, 43, 509-516.

Brandrup, J. 1., EH, Polymer Handbook. Wiley Interscience: New York, 1975.

~133-

14. MacCallum, J. R., Vincent, C.A., Polymer Electrolyte Reviews-I . Elsevier
Science Publishing Co. Inc.: New York, 1987.

-134-

Chapter IV

Crosslinked Polymer Brushes Containing PEO Segments

This chapter describes the synthesis and characterization of crosslinked polymer
brushes with poly(ethylene oxide) (PEO) segments. Crosslinked brushes were grown
from gold substrates by copolymerization of a macromonomer, poly(ethylene glycol)
methyl ether methacrylate (PEGMA), and a crosslinker, bisphenol A ethoxylate
dimethacrylate (BisA-EDMA) using surface-initiated ATRP. The crosslink density of the
film was estimated from FT-IR measurements and the aqueous swelling of the films, as

measured by in-situ ellipsometry.
Introduction

PEG coated surfaces are very effective in reducing protein adsorption and cell
adhesion, and crosslinked PEG films should be particular well-suited for biological
applications," 2 because of the enhanced dimensional stability and mechanical properties
of crosslinked polymer films. Moreover, crosslinking inhibits crystallization, and fihns
derived from PEGMA and BisA-EDMA may prove usefirl for membrane-based gas
separations.3 In addition, since crosslinking seldom goes to 100% conversion, the residual
vinyl groups may serve as sites for attaching additional functional groups.4

Currently, most crosslinked polymer films are formed by photo-initiated
polymerization or conventional free radical polymerization. However, these methods

afford limited control over the polymer architecture and often suffer from side reactions.

-135-

Surface-initiated ATRP can generate polymer brushes with a range of architectures under
mild conditions, but only a few studies address crosslinked polymer brushes. Huang et al.
previously reported the controlled synthesis of crosslinked polymer films by ATRP of
ethylene glycol dimethacrylate (EGDMA) at room temperature. In this study, we extend
that work to the copolymerization of a macromonomer (PEGMA) and a crosslinker
(BisA-EDMA), and study the relationship between crosslinking density and film
properties. The polymers incorporate oligomeric PEO units, which suggests potential

medical applications and useful materials for membrane-based separations.

Experimental

Materials. Unless otherwise noted, all materials were obtained from Aldrich.
BisA-EDMA, Mn = 1,700 g/mol, was passed through a column packed with a commercial
inhibitor removal agent (Aldrich). PEGMA (MIn =1,100g/mol), Cu(I)Br (99.999%), and
Cu(II)Br2 (99.999%) were used as received. 2,2’-Bipyridine (bpy) was recrystallized
from hot hexane, and then sublimed under vacuum at 60 °C to afford white crystals.
Milli-Q water (18 MD), and N,N-dimethylformamide (DMF) (HPLC grade, inhibitor free)
were used as polymerization solvents.

Characterization methods. Reflectance FT-IR was performed using a Nicolet
Magma-IR 560 spectrometer with a PIKE grazing angle of 80 °C. Film thicknesses were
measured by a rotating analyzer spectroscopy ellipsometer (M-44; J .A. Woollam) at a 75°
angle of incidence. The refractive index and film thickness were calculated
simultaneously by the values of A and (1). Film thickness measured under N2 and water

were obtained using a home-built cell with glass windows. After measuring the refractive

-136-

index and film thickness of the polymer film in Nz-filled cell, the cell was filled with
deionized water. The optical constant of water-swollen film is between 1.33 (the

refractive constant of water) and 1.5 (polymer film refractive index).

Synthesis of crosslinked polymer brushes. A stock catalyst solution was prepared
in a drybox by adding CuBr (0.043 g, 10 mM), CuBr2 (0.02 g, 3 mM), and bpy (0.14 g,
30 mM) to 30 mL of DMF. The solution was sealed in a flask and transferred to a N2
filled glovebag. Monomer (PEGMA), crosslinker (BisA-EDMA) and water were added
to a Schlenk flask and stirred until homogeneous. The overall concentration of monomer
and crosslinker was fixed at 0.4 M, and the monomer to crosslinker ratio was varied to
obtain the desired polymer brush composition. For a polymerization using 30 mol%
crosslinker, 15.4 g of PEGMA (0.014 mol), 10.8 g of BisA-EDMA (6 mmol), and water
(19 mL) were added to a Schlenk flask. The solution was degassed via three freeze-
pump-thaw cycles, and back-filled with N2. A portion of the stock catalyst solution (5 mL)
was added to the mixture under N2, resulting in final catalyst concentrations of [CuBr] =
1 mM, [CuBr2] = 0.3 mM, and [bpy] = 3 mM. The solution was stirred until
homogeneous and then was added to a series of vials, each containing one initiator-
anchored gold slide. Slides were removed from vials at predetermined times. Each slide
was rinsed with a copious amount of water, ultrasonically cleaned in water and ethanol

baths, rinsed with ethanol, and dried under a stream of N2.

-137-

Determination of cross-link density of polymer brushes.

(a) The composition of copolymer films.

The composition of the crosslinked films was evaluated by infrared spectroscopy.
The C=O stretching band at 1739 cm’1 provides information about PEGMA and BisA-
EDMA while the band for aromatic C=C stretching at 1608 cm’1 is an exclusive marker
for BisA-EDMA. The composition of the polymer film is x mol% crosslinker (BisA-
EDMA), and (1- x) mol% PEGMA. The ratio of two absorption peaks (aromatic C=C
stretch and C=C peaks) in the IR spectra of a 100% BisA-EDMA film is:
AarC-C /Ac=o =3 (4.1)
and for the copolymer films:
A' arC-C /Ac.=o = b (4-2)
Since, the C=O absorption of peaks for BisA-EDMA and PEGMA (at the same thickness)
are comparable, combining equations (4.1) and (4.2) provides x, the mole fraction of the
crosslinker in the film.
x = b/a
The same calculations were made for a series of films polymerized at various crosslinker
feeds, and the calculated mole fraction of cross-linker in the polymer, F2 was plotted as a
function of the mole fraction of cross-linker in the feed solution, f2.
(b) Crosslink density determination.
The cross-link density of polymer films can be estimated from the residual vinyl
groups in the polymer film. As PEGMA has only one double bond, any residual double
bonds in films must be due to unpolymerized C=C double bonds in the BisA-EDMA

crosslinker. As shown in Figure 4.1, a crosslink forms when both vinyl groups of the

-138-

crosslinker react; a branch forms when only one vinyl group reacts. The first situation
includes cases of “intrarnolecular” crosslinking, formation of a loop, which is commonly
seen in copolymerization of vinyl and divinyl monomers.5 While loops do not have the
same consequences as a regular crosslink, in this study, loops will be counted as

crosslinks. Thus, the calculated crosslink density will overestimate the true value.

M

I II

Figure 4.1 Crosslinking and branching in films containing BisA-EDMA

If the probability of crosslinking is y, the probability of branching is l-y. From the IR

spectrum of a BisA-EDMA homopolymer film:

A C=C /A arC-C = C1 . (4.3.)
In the copolymer films:
A' C=C /A arC-C = C2 (4.4.)

Combining equations (4.3) and (4.4), the probability of cross-linking (case I) is
y =1- (2C2/C.)
Taking into account of the actual mole fraction of crosslinker in films (defined as X), we

obtain the crosslinking density in the polymer film, which is defined as:

Crosslink density = x x y

-139-

Results and Discussion

Synthesis of crosslinked polymer brushes. Scheme 4.1 outlines the synthetic
route used to prepare crosslinked polymer brushes. PEGMA and the BisA-EDMA
crosslinker were copolymerized from initiators anchored gold surfaces by aqueous
surface-initiated ATRP. The preparation of the initiator-anchored gold substrates was
described in previous chapters. For surface polymerization, the slides were immersed in
an aqueous solution of monomer, crosslinker and the CuBr/CuBrZ/bpy catalyst mixture.
In each polymerization, the total concentration of monomer and crosslinker is held
constant, and the mole ratio of monomer to crosslinker were varied to obtain polymer
brushes with different crosslink densities. In some UV-induced polymerizations, it was
found that the crosslink density in the polymer increased as the crosslinker concentration
increased in the pre-polyrner solutionf"8 Significant chain transfer and the use of
sacrificial initiator are incompatible with these polymerizations due to the formation of
gels in the solution, which are very difficult to remove from the substrate. After
polymerization, polymer films were rinsed with large amounts of water, sonicated in
DMF, rinsed with ethanol, and finally dried in a stream of N2.

F T-IR spectra of the substrates showed a large increase in the ester carbonyl band
at 1730 cm], appearance of broad C-O-C peaks at 1150 cm", and aromatic C=C bands at
1608 cm'1 and 1530 cm"1 (compare Figure 4.2 trace a & b), confirming the growth of a
copolymer layer from the surface. In addition, the C=C stretching band at 1640 cm'1
indicates unreacted BisA-EDMA vinyl groups in the copolymer film, which may be used
for further derivatization. From the width and position of the absorption peaks, we

conclude that the crosslinked polymer film is amorphous.

-140-

=< ,0 CH3
ii CH3 V0 Monomer (PEGMA)
3(CH2—)1—10—c—{—Br

0 22-23
CH3

0 CH3

+ wax—W.;.-

CuBr, CuBr2, bpy

 

Cross-linker
H20, rt.
(BisA-EDMA)
V
9H3 W3 9H3
A www—(CHz—c—+(—CH2—c—+(-CH2—c—)—Br
I " o=c x o=c v
C=C? l i
A
9H2
9H2
0
\|’22-23
CH3 ' 9
o=c o=c
l I
-(-CH2—C'3-)-m c=CH2
CH3 CH3

—= M O 0 wt.

Scheme 4.1. Synthetic route to crosslinked copolymer films grown from gold substrates.

-141-

.voom 05 E coo—£388 $38 on 55 SE SSS—once coo—E—mmos a 3v EH SE Agwmmvba a A3 “oontsm
=< Baa—n 38:28 .8825 5 80¢ :38» “WEE 8:58 we 88on ENE oocfloocom .Né 95w;

 

 

 

 

 

 

 

 

 

AwfiurenEacgu;
can 22:. can? coca comm coon comm gov
i< ANV
V
a.
S
o
m
B
U
com? 82 m
Ono
0-0 2680?. to

 

 

 

 

 

 

-142—

Figure 4.3 shows the evolution of film thickness as a function of reaction time for
different feed solution compositions. As the mole fraction of crosslinker added to the
polymerization increased, the polymerization rate and film thickness decreased.
Homopolymerization of PEGMA yielded a 90 nm thick film in a 4 h polymerization, but
the film thickness dropped to 50 run when the feed composition 50 mol% crosslinker. In
addition, the polymerization rate decreased to half of the homopolymer rate. Addition of
the BisA—EDMA crosslinker has a negative affect on polymerization, however, the reason
is not yet resolved. One possible explanation is that as more crosslinker is incorporated
into the film, the polymer chains are more entangled, which makes the terminal Br less
accessible to the copper catalyst. Therefore, both polymerization rate and polymer film
thickness decrease.

Characterization of crosslinked polymer films. The composition of polymer
brushes was estimated by integration of characteristic absorption peaks shown in FT-IR
spectra of crosslinked polymer brushes. The mole fraction of crosslinker in the polymer
film was inferred from the absorbance of aromatic C=C peak, relative to that of the C=O
band (see the experimental section for details). The results from fihrrs prepared with
various feed compositions are shown in Figure 4.4. The data shows that the feed and film
compositions are nearly linearly related, indicating that despite their size and structural
differences, incorporation of PEGMA and BisA-EDMA is differences in reactivity or

diffusion to the reactive sites surface on the substrate.9

-143-

 

 

 

 

 

 

 

 

1000
900 - i
800 — i
700 - i
5 600 - i f
g 5
c 500 -
1‘» ° 1
5 400 - 1 i 2 -0°/
300 - 9 Q O 15%
200 ~ , A 050%
A80%
100 “ 1 (>100%
0 l I I T
o 1 2 3 4 5
Time (h)

Figure 4.3. Evolution of film thickness with time for the copolymerization of
PEGMA and BisA-EDMA. The overall concentration of monomer and crosslinker for all
runs is 0.4 M. Catalyst: [CuBr] = 1mM; [CuBr2] = 0.3mM; [bpy] = 3mM. The inset

shows the mole % composition of BisA-EDMA in the monomer feed solution.

-144-

 

 

 

 

 

 

1
_ 0.9- .. ,x’
o 1- ,
g 0.8 ~
9.
o 0.7 s 0 ,/ 9
0 ,’
.5 0'6 7 j x// .ii.
2
"6 0.5 — f /
g g
.3 0.4 ‘ i ,’/
“E 0.3 ~ .
'3 i x
E 02 — ,/
i:
0.1 - ,/
o I I I U I
o 02 0.4 0.6 0.8 1

f2, Mole fraction of M2 in comonomer feed

Figure 4.4. The relationship between the feed composition and the measured film

composition. Each point is the average of at least three independent samples.

We are interested in quantifying the residual vinyl groups in the film since they
may be used for further elaboration of the membranes. As detailed in the experimental
section, comparison of the vinyl C=C and the aromatic C=C stretching bands provide
information on the fraction of BisA-EDMA double bonds that were not polymerized The
data show that the fraction of BisA-EDMA vinyl bonds polymerized is approximately
80%, and independent of the BisA-EDMA composition in the feed. There was, however,

significant run to run variation in films prepared with high crosslinker feed ratios.

-145-

 

 

 

 

 

 

 

1.2
2
.0
3
8 1.0- i .. ..
g8 o.a~ ii i T
. a i
0 w
E S. 0,-
7) 3 ' JL
in O.
2 in '
u U
*5 g 0.4 — r
c 9 .
O
'3
g 0.2 ~
9
§

0.0 I I I l

o 0.2 0.4 0.6 0.8 1

1‘.

f2, Mole fraction M2 in comonomer feed

Figure 4.5. The fraction of crosslinkers in which both double bonds participate in

the polymerization. Each point is the average of at least three independent samples.

The crosslink density can be defined from the data above:

M(x - link) . M(BisA — EDMA)
M(BisA — EDMA) M(BisA — EDMA) + M(PEGMA)

 

Density(X — link) =

M (x - link)

Density(X — link) =
M (BisA — EDMA) + M (PE GMA)

(4.5)

 

The near linear relationship between the feed and measured compositions, and the
nearly constant crosslink efficiency (~80%) imply that the crosslink density should be
linearly related to the mole fraction of the crosslinker in the feed, as shown in Figure 4.6.
Except for the 80 mol% point, the general trend is a linear relationship, which may reflect

some variability in membrane synthesis or the limitations of the IR data.

-146-

 

 

100

l
80 r
70 ~
60 r
50 —
40 r
30 a

20. ,‘s i

10-

Crosslinking density (%)

 

 

 

0 20 40 60 80 100
Feed Crossiinker mo|e%

Figure 4.6. Crosslink density vs. the mole percentage of crosslinker in the feed.

-147-

In addition to estimating the crosslink density, FT-IR spectra also provide
information on the morphology of the films. The IR data show that films copolymerized
from feed solutions with less than <15 mol% BisA-EDMA are crystalline, but all BisA-
EDMA mole fractions >15 mol% produced amorphous films.

Swelling behavior of crosslinked copolymer films. Figure 4.7. shows dry and
water swollen thicknesses for polymer films prepared at different crosslinker feed ratios.
The dry films ranged from 20-30 mm and correspond to compositions of 0% to 100%
poly(BisA-EDMA). In Figure 4.7b, the normalized swelling is plotted as a function of
crosslinking density to better illustrate the relationship between film structure and

swelling behavior. The swelling ratio is defined as

d(Wet) ‘ d(d’y) X 100% (4.6)
d (dry)

Swelling ratio =

Where d is the film thickness.
As shown in Figure 4.7b, the swelling ratio decreased with cross-link density and
pl ateaued at >40% crosslink density. The former indicates increasing dimensional
Stability with crosslinking, which is characteristic of crosslinked films, while 160%
SVVelling may represent the inherent swelling due to the long PEO segments in the two

I11¢Z>nomers. Qiao et al. 10 observed the increased swelling of PEG dimethacrylate network

as the length of PEG segments increased.

-148-

 

2000

 

I in water
1600 0 dry state

 

 

 

1200 *

 

800 1 I

Film thickness (A)

400 3

 

 

0 20 40 60 80 100
Crosslinker feed percentage ( mo|°/o)

 

300 ,
280 l
260 -
24o -
220 -
200 -
130 -
160 - ,
14o -
120 -

1 00 T . . .
0 20 4O 60 8O 1 00

Crosslinking density (%)

Swelling percentage (%)
99
o
e

 

 

 

Figure 4.7. a) (Top) thickness of dry and water-swollen films as a function of the
mole percentages of crosslinker in the monomers feed solution; b) (bottom) the

1‘EIationship between the normalized swelling of films and the film crosslink density.

-149-

Conclusion and Outlook:

Crosslinked polymer brushes containing PEO segments were grown from the
surface of gold substrates using surface-initiated ATRP. The addition of the cross-linker
to PEGMA slowed the polymerization rate. Using the aromatic ring of BisA-EDMA as a
reference, the film structure and crosslink density were measured by an analysis of the IR
spectra of crosslinked films. Aqueous swelling of copolymer films in water decreases
with increasing crosslink density, and is nearly constant for crosslink densities >40%.
The gas permeation characteristics of crosslinked-polymer membranes grafted on porous

alumina supports are described in Chapter V.

-150-

References:

Nolan, C. M.; Reyes, C. D.; Debord, J. D.; Garcia, A. J .; Lyon, L. A.,
Biomacromolecules 2005, 6, 2032-2039.

Groll, J .; Amirgoulova, E. V.; Ameringer, T.; Heyes, C. D.; Roecker, C.;
Nienhaus, G. U.; Moeller, M., J. Am. Chem. Soc. 2004, 126, 4234-4239.

Lin, H.; Wagner, E. V.; Freeman, B. D.; Toy, L. G.; Gupta, R., Science 2006, 311,
639-642.

Heggli, M.; Tirelli, N.; Zisch, A.; Hubbell, J. A., Bioconjugate Chem. 2003, 14,
967-973.

Matsumoto, A.; Ohashi, T.; 0e, 11.; Aota, H.; Ikeda, J. I., J. Polym. Sci. A: Polym.
Chem. 2000, 38, 4396-4402.

Baker, J. P.; Hong, L. H.; Blanch, H. W.; Prausnitz, J. M., Macromolecules 1994,
27, 1446-1454.

Elliott, J. E.; Macdonald, M.; Nie, J .; Bowman, C. N., Polymer 2004, 45, 1503-
1510.

Kizilay, M. Y.; Okay, O., Macromolecules 2003, 36, 6856-6862.

Odian, G., Principles of Polymerization. Wiley-Interscience: New York, 2004.

Qiao, C.; Jiang, S.; Dong, D.; Ji, X.; An, L.; Jiang, B., Macromol. Rapid Commun.
2004, 25, 659-663.

-151-

Chapter V

Polymer Membranes for Gas Separation

This chapter describes the development of polymeric membrane materials and
studies of gas permeation through membranes, emphasizing the separation of CO2 from
light gases such as H2 and N2. Comb and crosslinked polymer brushes containing PEO
units were grafted to the surface of porous alumina membrane supports by surface-
initiated ATRP. Due to the interactions between CO2 and the ether oxygens of the
ethylene oxide units, the modified membranes are highly permeable to CO2 and
selectively permeate CO2. For poly(PEGMA) films, the measured permeability of CO2 is
~300 Barrer with a C02 /H2 selectivity of 14. CO2 permeability and selectivity decreased
as the PEO side chain of the comb polymer was shortened. CO2 and H2 permeability, and

the CO2/H2 permselectivity decreased as more crosslinker was incorporated in the film.

Background

During the last decade, gas separation by polymeric membranes has been widely
studied as an alternative to traditional gas separation technology. The emphasis of this
research has focused on the recovery of H2 from petrochemical process streams, removal
of CO2 from hydrocarbon mixtures, and removal of acidic gases from natural gases. 1

The steady state permeability of gas A, through a film with thickness of l, is
defined as:2

-152-

where FA is the steady state flux of gas through the film (cm3 (STP)/cm2 s), l is the film
thickness and Ap is trans-membrane pressure (cmHg). Permeability coefficients are
commonly expressed in units of Barrers, where 1 Barrer = 10'10 cm3 (STP) cm/(cm2 s
cmHg). When the upstream pressure (p) is much higher than the downstream pressure
(p2), Ap can be approximated as pi. The permeability of gas A through a polymer matrix
can be expressed as the product of the average effective diffusivity, DA, and the apparent
solubility of gas A in the polymer, S A?
PA: DA X 3A (2)

According to this diffusion/solubility model, a penetrant gas dissolves in the high
pressure side of the film, diffuses through the film, and then desorbs at the low pressure
side of the film. The ideal membrane selectivity for gas A over gas B is expressed as:

on (NR) = PA/PB = (DA/DB) x (SA/ SB) = a r) x or s (3)
where or D is the diffusivity selectivity, and a s is the solubility selectivity.

The gas diffusivity increases with decreasing penetrant (gas) size (lower critical
volume or kinetic diameter of the penetrant), increasing polymer chain flexibility
(increased free volume), and decreasing polymer chain and penetrant interactions. The
penetrant solubility increases with increasing condensability (higher boiling point or
critical temperature) and stronger gas-polymer interactions.3’ 4 As shown in Table 5.1, the
critical volume and kinetic diameter (dk) are used as a metric for penetrant size.
Considering only molecular size, H2 is smaller and has a higher diffusion coefficient than
CO2. However, CO2 has a higher boiling point and critical temperature than H2, and

therefore, it is more condensable than H2.

-153-

Many commercially available membranes are based on glassy polymers, and are
designed to maximize the gas diffusivity selectivity. Those membranes are more
permeable to H2 than to CO2, since an favors H2 much more strongly than as favors CO2.
One way to improve the permeability of glassy membranes is to increase chain mobility
by adding a plasticizer. However, plasticization of glassy polymers would be detrimental
to membrane performance. When the polymer absorbs a high concentration of a
condensable gas (CO2), the polymer swells and softens. The diffusion coefficient of large
molecule (CO2) increases more rapidly than small molecule (H2), resulting in the loss of

diffusion selectivity.

Table 5.1. Physical properties of H2, C02, and N25, 6

 

 

 

Size Condensibility S°lub1my
paramenter
Critical Boilin oint Critical
Volume (1 1, (nm) (13p temperature 8 (MPa 0'5)
(cm3/mol) (K)
H2 65.1 0.289 20.4 33.2 6.6
N2 89.8 0.364 77.4 126.2 5.3
CO2 93.9 0.330 195 304.2 12.2

Since the diffusivity ratio OLD [=D(CO2)/D(H2)] is unfavorable, a common strategy
for separating CO2 and H2 is to design rubbery membranes that have a high solubility
ratio as [= (S(CO2)/ S(H2)]. A high solubility for CO2 relative to light gases also is very
useful for CO2/N 2 separation, since their sizes and critical volumes are similar (see Table
5.1.). The solubility of a gas depends on the physical properties of the gas, such as boiling

point and critical temperature, but more importantly, solubility is related to polymer/gas

-154-

interactions. PEO is perhaps the best material for the separation of CO2 due to favorable
interactions between the ether oxygens and CO2. Membranes more permeable to CO2
than H2 are called “reverse selective” membranes. Plasticization, which is detrimental to
the glassy polymers, actually benefits PEO-containing reverse selective membranes,
since it increases D(CO2)/D(H2), yielding a higher CO2/H2 selectivity.

High molecular weight PEO is semi-crystalline. The formation of a crystalline
phase would reduce the CO2 gas permeability because only ethylene oxide segments in
the amorphous phase could interact with CO2 and solubilize CO2. There are several
strategies for incorporating PEO segments into polymers while avoiding PEO
crystallization including the use of low molecular weight PEO, block copolymers with
short PEO blocks, graft and comb polymers with short PEO segments as side chains, and
cross-linked PEO networks. PEO-containing monomers commonly used to synthesize
such structures are shown in Scheme 5.1. Table 5.2 lists CO2 permeabilities and CO2/H2
selectivities for representative membrane materials. More CO2/N 2 separation data appear

in reference 10.

-155-

_ ,CH3
Hzc—q
lc=o

CH
Oi~/\0’)j1 3

1. n=4-5
PEGMA 2.n=88

3. n = 22-23

Wofiok:

6. n=4

7. n=14

trio-ick-

4. n=4

PEGDMA
5. n = 14

O

TYVO ”O O 0 30949:

8. BisA-EDMA

Scheme 5.1. Structures of representative PEG (meth)acrylates and PEG di(meth)acrylates
used in this study and reported in the literature for membrane separations.

-156-

Table 5.2. Representative examples of pure gas CO2 permeability and CO2/H2

selectivity in PEO-containing polymers.

 

 

 

Permeability and Selectivity P(CO2) 01(C02/H2) ° Reference
b
Polymer (Barrer)
Semi-crystalline PEO 13 (35 °C) 6.7 (1 atm) Lin7
Liquid PEG blend with solid PDMS N/A 7 .5(25 °C) Lin 8

 

PEG soft block and polyimide hard block 66 (35 °C) 7.8 (35 °C) Bonderg' ‘0

120 (35 °C) 9.8 (35 °C)

 

 

 

Crosslinked 50 wt% 2/ 50 wt% 7 a 260 (35 °C) 10 Freeman 8
PEO films
70 wt% 2 / 30 wt% 7 320 (35 °C) 11 (35 °C)
40 (-20 °C)
99 wt% 2 570 (35 °C) 12
Crosslinked PEG di-acrylate (7) 70 (23 °C) 9.5 Patel ”"3
ol ethers
p y 50/50 w/w 100 5
PEG di-acrylate/PPG diacrylate

 

 

 

 

 

 

a The number refers to the structures in Scheme 5.1.
b Permeability coefficient: lBarrer =10-10 cm3 (STP) cm/(cm2 s cmHg).
° Selectivity 0i (CO2/H2) = P (CO2)/ P(H2).

-157-

 

Our strategy is to develop composite polymer membranes in which a thin polymer
skin is deposited on the surface of a porous aluminum membrane support. Solution
casting,14 and solution coating,15 plasma polymerization 16‘ ‘7 and UV-induced grafting,'8’
19 are the most common ways to prepare composite membranes. The principal challenge
in these methods is the difficulty in controlling film thickness. Surface-initiated ATRP
allows controlled growth of polymer brushes on the surface, and it may provide access to
membranes with well-defined polymer brush skins. In this chapter, we describe the
grafting of PEO-containing combs and crosslinked polymer brushes from the surface of
porous alumina membrane supports, and examine the CO2, H2 and N2 gases permeability

through these membranes.
Experimental

Materials. Poly(ethylene glycol) methyl ether methacrylate (PEGMA, M,, = 300
g/mol and 1,100 g/mol), Cu(I)Br (99.999%), Cu(II)Br2 (99.999%), tetramethyl-1,4,8,l 1-
tetraazacyclotetradecane (Me4Cyclam), 4,4’-di-(n-nonyl)-2, 2’-bipyridine (anbpy) were
used as received from Aldrich. Bisphenol A ethoxylate dimethacrylate (BisA-EDMA, M,,
= 1,700 g/mol) was passed through a column packed with a commercial inhibitor removal
agent (Aldrich) to remove hydroquinone inhibitor. Milli-Q water (18 M0), N,N-
dimethylformamide (DMF) (HPLC grade, inhibitor free) were used as polymerization
solvents.

Characterization methods. Field-emission scanning electron microscopy
(FESEM, Hitachi s-47OOII, acceleration voltage of 15 kV) was used to characterize the

film grth on the porous aluminum membrane support. Membranes were freeze-

-158-

fractured using liquid N2 prior to sputter coating with 5 nm gold for cross-sectional image
analysis.

Anchoring initiators on the surface of porous alumina membrane support.
Porous alumina membranes (0.02 pm) were cleaned in an UV/O3 chamber for 15 min,
washed with anhydrous ethanol and dried with N2. The membranes were sputter-coated
with 5 nm gold, and then were cleaned in the UV/O3 chamber for an additional 10 min,
washed with ethanol and dried under a stream of N2. A self-assembled initiator
monolayer was formed by immersing the gold-coated membrane in a 1 mM ethanolic
solution of [Br-C(CH3)2-COO(CH2)ioS]2 for 24 h. The films were washed with a large
amount of ethanol, sonicated in ethanol for 1 min, washed again with ethanol, and finally
dried under a stream of N2. To directly attach initiators to the alumina membrane, the
membrane was immersed in a DMF solution of the trichlorosilane initiator (see chapter
111) for 24 h, and then the membrane was washed with DMF and anhydrous ethanol, and

dried under a stream of N2 stream. The synthetic scheme is illustrated in Figure 5.2.

Surface-initiated A T RP of poly(PE GMA ) films using CuBr/CuBr/bpy catalyst.
A catalyst solution was prepared in an oxygen-free drybox by adding CuBr (0.057 g, 20
mM), CuBr2 (0.045 g, 6 mM) and bpy (0.18 g, 60 mM) to 20 mL of DMF, and then
transferring the solution to a N2-filled glove bag. PEGMA (15 g) and Milli Q water (14
mL) were added to a Schlenk flask and stirred until homogeneous. The flask was
connected to a vacuum line and the solution was degassed using three freeze-pump-thaw
cycles, back-filled with N2, and then transferred to the N2-filled glove bag. A portion of
the catalyst solution (1.5 mL) was added to the flask, resulting in catalyst concentrations

of [CuBr] = 1 mM, [CuBr2] = 0.3 mM, and [bpy] = 3 mM. The solution was transferred

-159-

into a container holding an initiator-anchored membrane. After 12 h, the membrane was
removed from polymerization solution, rinsed with large amount of water and DMF,
cleaned in an ultrasonic bath with DMF, washed with anhydrous ethanol, and finally

dried under a stream of N2.

Polymerization of PE GMA using a C uBr/lile4C yclam/CuBr2(anbpy) 2 catalyst.
In a drybox, CuBr2 and anbpy (1:2, molzmol) were mixed together to form a green
complex and stored for future use. A catalyst solution was prepared in an oxygen-free
drybox by adding CuBr (0.057 g, 20 mM), CuBr2(anbpy)2 (0.21 g, 10 mM) and
Me4Cyclam (0.1 g, 20 mM) to 10 mL of DMF. The solution was stirred until
homogeneous (~ 20 min). A second Schlenk flask was loaded with PEGMA(4-5) (18 g,
0.05 mol) and H20 (9 mL). The flask was connected to a vacuum line and the solution
was degassed using three freeze-pump-thaw cycles, and finally back-filled with N2. In a
N2-filled glovebag, a portion of the catalyst solution (3 mL) was added into the flask. The
final catalysts concentrations were: [CuBr] = 2 mM, [CuBr2] = [(anbpy)2] = 1 mM, and
[Me4Cyclam] = 2 mM. The catalyst solution was added to the monomer solution, causing
it to turn brown with the formation of a precipitate. The solution was immediately
transferred into a vial containing the initiator-anchored membrane. The membrane was
taken out at a predetermined times, ranging from 3 min to 10 min, to control the film
thickness. The membrane was rinsed with a large amount of water and DMF, cleaned in
an ultrasonic bath with DMF, washed with anhydrous ethanol, and finally dried under a
stream of N2.

Growth of crosslinked copolymerfilm. A stock catalyst solution was prepared in

a drybox by adding CuBr (0.043 g, 10 mM), CuBr2 (0.02 g, 3 mM), and bpy (0.14 g, 30

-l60-

mM) to 30 mL of DMF. The solution was sealed in a flask and transferred to a N2 filled
glovebag. Monomer (PEGMA), crosslinker (BisA-EDMA) and water were added to a
Schlenk flask and stirred until homogeneous. The total concentration of monomer and
crosslinker was fixed at 0.4 M, and the monomer to crosslinker ratio was varied to obtain
the desired polymer brush composition. For a polymerization with 30 mol% crosslinker,
15.4 g of PEGMA (0.014 mol), 10.8 g of BisA-EDMA (6 mmol), and 19 mL water were
added to a Schlenk flask. The solution was degassed via three freeze-pump-thaw cycles,
and back-filled with N2. A portion of the stock catalyst solution (5 mL) was added to the
mixture under N2, resulting in final catalyst concentrations of [CuBr] = 1 mM, [CuBr2] =
0.3 mM, and [bpy] = 3 mM. The solution was stirred until homogeneous and then was
transferred into vials containing the initiator-anchored membranes. After a 12 h
polymerization, membranes were removed from the solution, rinsed with a copious
amount of water, ultrasonically cleaned in a DMF bath, washed with ethanol, and finally
dried under a stream of N2.

Gas permeation measurements. The permeation of gases (CO2, H2, and N2)
through the modified membranes was measured using a permeation cell with a pressure
relief valve. The permeate flux was measured as a function of gas inlet pressure (5-50
psig) using a soap-bubble flow meter. Measurements were performed for each gas
separately in the order H2, N2, and CO2. For each gas, the permeation cell was purged
with the gas several times to remove traces of other gases and to establish a stable gas
flux. The flowmeter readout was converted to the flux using Eq 4:

F=v/A (4)

-l61-

where F is the flux of the gas (cm3 (STP)/cm2 s), v is the flow rate of the gas (lemin),
and A is the area of the membrane (cmz). Permeability coefficients were calculated

according to Eq. 1.

 

N, Q

@ Erliii

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

co2 H2
Gas controller
12 . 34
Pressure
gauge
permeation cell
'— i—1
L
Pressure 7" *4 ‘4
relief valve
Digital Flow Meter
Polymer
membrane

Figure 5.1. Apparatus for gas permeation measurements.

-162-

Results and Discussion
Synthesis of and characterization of polymer films

Polymer brushes were grafted from the surface of alumina membrane supports
(0.02 pm pore size) using two methods. In the first, the membrane was sputter coated
with 5 nm of gold, and then the thiol terminated initiator was attached by forming a self-
assembled monolayer. Alternatively, trichlorosilane initiator precursors were directly
coupled with OH groups to form Si-O-Si bonds. Using surface-initiated ATRP, PEGMA
macromonomers were polymerized from the initiator layers to form comb polymer
brushes with PEO side chains. Including a cross-linker (BisA-EDMA, 30-100 mol%) in
the monomer pool allowed the growth of cross-linked films with various crosslink
densities on the surface of porous alumina membranes.
Figure 5.2 shows transmission FT-IR spectra of polymethacrylates grafted on the
surface of a porous alumina membrane support. Characteristic peaks observed in the
spectrum, include C-H stretching at 2885 cm"1 and ester C=O stretching at 1729 cm".
The lengths of the PEO side chains in the two polymers are different, hence the different
intensity ratios seen for C-H stretching and ester C=O bands. The ethylene oxide C-O-C
stretching band is obscured by the Al-O-Al stretching band, located at ~1110 cm".
Because of the high pore density in the alumina substrates, ellipsometric measurements
Were not possible on these surfaces, and thus film thicknesses were estimated from the
Cross-sectional FESEM images. To demonstrate the potential of these materials in gas
Separations, we examined CO2, N2 and H2 fluxes through these films grown on the

Surface of porous alumina membranes.

-l63-

 

Porous alumina U U U U U U
membrane support

Cl\ 9
CI—Si—(CH2)11—O-C—<Br
Cl

 

 

 

Sputter coated
with 5 nm gold

0J00001
s-(CHzto—ii—ég;

000000 snnra:

H3 CH3

 

 

 

 

 

c
HZC=C:

,c=o

O‘COFCHS U U U U L
Yimflwfif

 

 

 

 

 

 

Path A Path B

Figure 5.2. Synthesis of composite membranes with polymer skins.

-164-

 

Poly[PEGMA(4-5)]
grafted membrane

 

Absorbance

Poly[(PEGMA)(22-23)]
grafted membrane

 

 

 

4000 3500 3000 2500 2000 1500 1000

Wavenumber (cm'l)

Figure 5.3. Transmittance FT-IR spectrum of poly(PEGMA(22-23)) and
poly(PEGMA(4-5)) grafted to porous alumina membrane supports.

Pure gas permeation of poly(PE GMA ) brushes.

The PEO side chain of poly(PEGMA(22-23)) contains 22-23 oxyethylene units
per monomer repeat, which should provide a sufficient fraction of PEO to give high
selectivity. A poly(PEGMA) film prepared by a 12 h surface-initiated ATRP was
subjected to gas permeation measurements. As shown in Figure 5.4, this membrane
showed a remarkable CO2/H2 selectivity of 14. The film thickness estimated from

FESEM was ~1000 nm (Figure 5.5), and therefore the film has a C02 permeability of

-165-

~300 Barrers. This result is very promising when compared to prior work. This
selectivity is slightly higher than the value of ~10 for typical crosslinked PEG-acrylates
or PEG methacrylates, but the permeability is 4-fold greater.“ 20

Although the above results are very promising, this membrane was synthesized
from a PEGMA monomer with long PEG side chains. As described in Chapter 2, these
chains will eventually crystallize, degrading the membrane permeability. Indeed, later

measurements with the same membrane, showed decreased gas flux and a drop in CO2/H2

selectivity.

on
O

 

s cm")
a, N
O O

I

 

Flux x1000(em3 (STP) '1
c: 3 B 8 8 8

 

 

O
.3
O

20 30 4O 50
Pressure (psig)

Figure 5.4. Single-gas fluxes of CO2 and H2 through films prepared by polymerization
of PEGMA(22-23) from porous alumina as a firnction of trans-membrane pressure drop.

The outlet pressure was 1 atm and the measurements were performed at room
temperature.

-166-

Polymer Film , -
‘V

“-71 “N V-Lr 1,. ' ‘
xv”, “2 , 1.

4. " .w
2‘ ~ 3.»

'5

 

Figure 5.5. FESEM image of the filtrate side of a porous alumina after growth of a
poly(PEGMA(22-23)) film from the surface.

The CO2/H2 selectivity can be recovered by eliminating film crystallinity. Since
PEO is water-soluble, crystalline poly(PEGMA) fihns should be rendered amorphous by
soaking with water. As shown in Figure 5.5 a crystalline film treated with water had a
CO2/H2 selectivity of ~ 8, and a C02 permeability of ~ 75.7 Barrer. These data suggest
that the film was partially crystalline. Either the length of time that the film was in water
was too short to fully dissolve the crystals, or a small fraction of the PEO side chains
quickly crystallized after the film was dried. The results are summarized in Table 5.3.

We also considered the possibility of operating gas permeation above the melting
point of poly(PEGMA) (35-40 °C) to achieve high permeability and CO2/H2 selectivity,
or storing membranes in CO2 to plasticize the polymer matrix and inhibit crystallization.
However, our current lab setup does not permit high temperature measurements and long
storage in CO2. An alternative approach to minimize crystallization is to utilize
monomers with fewer ethylene glycol units, although this may result in a decrease in the

film selectivity as the fraction of PEO in the film will be lower.

-167-

 

 

 

 

 

 

 

 

100
90. OHydrogen I
e, lCarbondioxide
g 80‘ ANitrogen -
‘in 70 1 -
g 60‘ '
(I)
'2' 5°“ '
0 40* I
o
g 30- I
E?
x 20 7
'2 10 o O
m -i
e 9 ° . .
o , ' I A—‘——L—
0 10 20 30 40 50
Pressure(psig)

Figure 5.6. Single-gas fluxes of CO2 and H2 through a poly(PEGMA(22-23)) grafted
membrane as a function of trans-membrane pressure drop. The initially crystalline film

was soaked in water, and then measurements were taken after the film was dried.

Two additional macromonomers, PEGMA(4-5) and PEGMA(8-9), were
polymerized. As shown in Chapter 2, these poly(PEGMA) films will not crystallize due
to their relatively short PEO side chains. A catalyst prepared from CuBr, CuBr2 and
Me4Cyclam was used for the polymerization of PEGMA(8-9). After a 10 min reaction,
the membrane was removed from the reaction mixture, washed and dried. Figure 5.7
shows gas flux through a poly(PEGMA(8-9)) membrane. The H2 gas flux increased
linearly with the trans-membrane pressure, while CO2 increased non-linearly, which
suggests membrane plasticization. Considering the film thickness estimated from FESEM

is 700 nm (see Figure 5.8), the CO2 and H2 permeabilities are ~35 and 4 Barrer,

-l68-

respectively. The CO2/H2 selectivity of ~8 is good, but lower than measured for the
previous polyPEGMA films. The decrease of CO2 gas permeability is likely due to its
decreased solubility in the film, since there are fewer ethylene oxide units in the polymer
matrix. The H2 gas permeability also decreased, and in this case, the decreased
permeability is due to a decrease in gas diffusivity. Shorter PEO side chains results in
more tightly packed polymer backbones, which affects the permeation of H2. Decreased
CO2 permeability and CO2/H2 and CO2/N2 selectivity also were observed in

poly(PEGMA(4-5)) films. The results are summarized in Table 5.3.

 

 

 

 

16
£147 .
U
7m12' .
A 4 CO
E10 2I
CD
«7' 8‘ '
E I
3' 6' I
8 I
:- 4‘
X
3
E 2‘ H2 .
O Q O O O .
o i i i
15 25 35 45 55

Pressure (psig)

Figure 5.7. Single-gas fluxes of CO2 and H2 as a function of trans-membrane pressure
drop for permeation through poly(PEGMA(8-9)) films prepared by polymerization from

porous alumina.

-169-

(a)

poly(PEGMA)(9)
film

(1))

 

Figure 5.8. FESEM image of the filtrate side of porous alumina after grth of (a)
poly(PEGMA(8-9)) and (b) poly(PEGMA(4-5)).

~170-

 

.mEE AEOMAKEQ 5:25 cosmozcoa mew mo $85.5 .m.m 933.

353808 EEobE 92 Eob 33:28 32? owfiog 05 08 823823. 62.59: 2:. 2

2055508 “measure 93 .8.“ 35308

$56568 5:53.83 05 he owed?“ 05 8a mos—g com: 2E. .Awm So m NEoV\ Eo Ehmv mEo 2-3 H Sham _

{<2 3
ow 2
<\Z ~ H w
wd um aw 9N
NENOU ~E~OU
.. £2.83

98 A03 was C .3 cornice 3 @6808 3818—8 803 356868 bzzmogom 4

«<2 v.9... omN
N m.w on
<\Z v N H om
Vault.— _.m 4600.2
NZ ~E NOD

4 Eugene 329503 3:35:59—

ooo g

com

com

of

3.5
82592:.

MN-NN
mm-NN
9w

mé

as... Angew.—Ea
5 3:... DH .«e .5952

-l7l-

Pure gas permeation through crosslinked polymer films (poly(PEGMA -co-BisA-

EDMA))

Another strategy for inhibiting crystallization is to use a crosslinked polymer
network to frustrate crystallization. Therefore, PEGMA(22-23) was copolymerized with
the BisA-EDMA crosslinker, with 30-100 mol% crosslinker in feed solution. The C02,
H2, and N2 fluxes through different fihns were measured as a function of crosslinking,
and are shown in Figure 5.9. FESEM images of a series of crosslinked films are shown

in Figure 5.10.

-172-

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

120
5‘ 18 -« I
.. 30 mol% - E 50 mol%
,n ‘°°* cross-linker . ,° 16 - cross-linker '
a
5 A 14« I
A 301 I a
g . L”. 122 I
as m - Hg 10 I
g I ‘5’ 8 d I
" ‘0‘ 8 6 ‘ e
x I
5 9 ° 2 4 « l O .
E 20‘ ' o ’ x I '
e e ’
O A A 3 2 ~
0 -_.u r Y T o , T , .
o 10 P 20 30 40 50 o 10 20 30 40 so
mu" (“9) Pressure (psig)
so 70
f; 50 80 mol% «at; 60 q 100 mol% .
'5 cross-lunker ' '0 cross-linker -
i I 'u 5°J
n. 40 2 a: I o
'- t-
;9. 9. 4o 2 I ’
E 30 ' e ”E o
in 0 g 30 - 0
g 20 « I . A 8 I 9
O
.. A .. 20- o
X e ‘ x 3 A ‘
x 102 . x 2 ‘
a ‘ 2 10 U 2
lb 2 u. A ‘
o v v v v I v v VTrv v v v u v v v v v v o 1 ‘ T v
0 10 2° 3° 4° 5° 0 1o 20 30 40 so
Pressure (psig) Pressure (pslg)

Figure 5.9. Single-gas fluxes of CO2 (I), H2 (6) and N2 (A) through cross-linked co-
polymer fihns prepared by polymerization of PEGMA(22-23) with a cross-linker (BisA-
EDMA) from porous alumina as a fimction of trans-membrane pressure drop. For the

film with 50 mol% crosslinker, the flow rate of N2 is too low to be measured.

-173-

{25' Crosslinked
2 film layer
. .,

“W ‘t a. 32-1.:- --
‘ a». ‘1 ‘1 "F i .

.i

 

(C) ((1)

Figure 5.10. FESEM images (cross-sectional) of porous alumina (0.02 pm surface pore
diameter) afier growth of crosslinked films from BisA—EDMA and PEGMA(22-23). The
monomer solution containing (a) 30 mol% crosslinker; (b) 50 mol% crosslinker; (c) 80

mol% crosslinker; and (d) 100 mol% crosslinker (a poly(BisA-EDMA) film).

-l74-

The CO2/H2 and CO2/N2 selectivity were calculated fi'om the gas permeability, the
gas fluxes, and the film thickness estimated from FESEM images. The results,
summarized in Table 5.4, show that the CO2 permeability decreased from ~ 60 Barrer for
the 30 mol% crosslinker film, to 6 Barrer for a 100 mol% BisA-EDMA film. The H2 and
N2 gas permeabilities show the same trend. BisA-EDMA is approximately 78% ethylene
oxide units by weight compared to 91 wt% ethylene oxide units in PEGMA(?-?).
Therefore, increasing crosslinking decreases the ethylene oxide content in the polymer
films, which decreases the CO2 solubility in the polymer matrix. In addition, the gas
permeability decreased as the crosslink density increased for films containing 30 — 100%
BisA-EDMA. These data are somewhat surprising given the aqueous swelling
experiments described in chapter N, which showed that polymer swelling was
independent of crosslink density when the BisA-EDMA was > 40%. Freeman and co-
workers have pointed out that changes in gas diffusion and permeability in crosslinked
polymer systems are more related to changes in the glass transition temperature (T S) than
crosslink density changes.20 In the BisA-EDMA/PEGMA system, introduction of the
crosslinker increased the polymer T 3 due to the rigid benzene rings of BisA-EDMA,
which likely reduced the free volume of the polymer, and hence affected the gas diffusion

through the membrane.

-175-

 

agape? A<2Qm-<em-8-<20mm58 @3538; é
$33828 owmuoiw can 8.865000 3:53:qu 30 .v.m 2an

.mocmSEoE E9396 9.3 60¢ coon—:28 2:9 05 mo owfiog 2: 03 333828 cocoa»: BE 0
.mocmfiEoE Eouotmv 95 com 35308 $56508 3:58:on 05 mo owns?“ 05 an 335» @085 .3350 m
~80: Eu Ehmv mEu 2-2 H Shem ~ 93 .33...» 2.3 :oumsco 8 manages 33:28 203 356508 bzfiaogm .1

18-82205
FEB :OuuNCoEboméo 5 won: nous—cm poem 58958 2: E omficooba 208 Aéaméxmwmv yogi—$90 a

 

 

 

 

 

 

 

 

 

 

Wm M; w; wé #6 mm o\c_oE o2
Nd H Wm md H WA m6 H WV #4 H NR md H md ow .XLOE ow
(CZ NM (£2 NS #N COM o\c_oE Om
Nd H 9w 5.0 H N.m ed H 5.0 v H —N w H mm com o\c_oE Om
NZ\NOU NENOU NZ 5 NOU
« u veg
Eéuuozfiik
baaofiuh . «$933 5 suHEEuEU
u . . uENNomfiwoo btsgiswm new 385539

 

 

 

 

 

-176-

The CO2/H2 selectivity also decreased as more crosslinker was incorporated in the
film, from 3.2 for the polymer film containing 30 mol% crosslinker to about 1.3 for the
BisA-EDMA film, which suggests a decrease in CO2 solubility. The decreased film
thickness may contribute to the decreased selectivity. Hirayama et al.” used plasma
irradiation to prepare cross-linked polymer films from PEGMA and BisA-EDMA, and
reported that CO2 permeability decreased as the cross-linker content increased, but the
selectivity of CO2/N2 remains unchanged. A polymer containing 70 wt% PEGMA
exhibited a C02 permeability of 260 Barrers, and CO2/N2 selectivity of about 48, while
poly(BisA-EDMA had a C02 permeability of 128 Barrer and a CO2/N2 selectivity of 49.
However, the films were thick (um scale), and FESEM images show that our crosslinked
films (80 and 100 mol% crosslinker) were 25-50 nm. In these thin films, there may be

defects or pinholes cause the decreased selectivity.

Summary

Comb and crosslinked PEO containing polymer brushes were grafted on the
surface of porous alumina membrane support using surface-initiated ATRP, and used for
the separation of CO2 from light gases (H2 and N2). Preliminary results show that
poly(PEGMA) films are promising for separation of CO2 from light gases, with CO2/H2
selectivities as high as 14 and CO2/N2 selectivities about 40. Decreased CO2 permeability
and CO2/H2 selectivity were observed for membranes prepared from PEGMAs with
shorter PEO side chains. The CO2 permeability and CO2/H2 selectivity of crosslinked
films prepared by copolymerization of PEGMA with a crosslinker (BisA-EDMA)

decreased as the film crosslinking increased. The lower fraction of ethylene oxide units

~177-

in the films and an increase in the glass transition temperature with BisA-EDMA content

may account for decreased permeability.

-l78—

References:

1.

10.

11.

12.

l3.

14.

Baker, R. W., Membrane Technology and Applications, 2nd Ed. J. Wiley: New
York, 2004.

Ghosal, K.; Freeman, B. D., Polym. Adv. Tech. 1994, 5, 673-97.

Stern, S. A., J. Membr. Sci. 1994, 94, 1-65.

Petropoulos, J. H., Polym. Gas Sep. Membr. 1994, 17-81.

Reid, R. C., Prausnitz, G. K., Poling, B.E., The Properties 0d Gases and Liquids.
McGraw-Hill: New York, 1987.

Stannett, V. T., Diffusion in Polymers. Academic Press: New York, 1968.

Lin, H.; Freeman, B. D., J. Membr. Sci. 2004, 239, 105-117.

Lin, H. Q.; Freeman, B. D., J. Molecul. Struc. 2005, 739, 57-74.

Bondar, V. 1.; Freeman, B. D.; Pinnau, I., J. Polym. Sci. B: Polym. Phys. 2000,
38, 2051-2062.

Bondar, V. I.; Freeman, B. D.; Pinnau, I., J. Polym. Sci. B; Polym. Phys. 1999,
37, 2463-2475.

Patel, N. P.; Zielinski, J. M.; Samseth, J .; Spontak, R. J ., Macromol. Chem. Phys.
2004, 205, 2409-2419.

Patel, N. P.; Miller, A. C.; Spontak, R. J ., Adv. F unc. Mater. 2004, 14, 699-707.

Patel, N. P.; Miller, A. C.; Spontak, R. J ., Adv. Mater. 2003, 15, 729-733.

Staudt-Bickel, C.; Koros, W. J ., J. Membr. Sci. 1999, 155, 145-154.

-179-

 

15.

16.

17.

18.

19.

20.

21.

Yanagishita, H.; Kitamoto, D.; Haraya, K.; Nakane, T.; Okada, T.; Matsuda, H.;
Idemoto, Y.; Koura, N., J. Membr. Sci. 2001, 188, 165-172.

Hayakawa, Y.; Terasawa, N.; Hayashi, E.; Abe, T., J. Applied. Polym. Sci. 1996,
62, 951-954.

Ulbricht, M.; Belfort, G., J. Membr. Sci. 1996, 111, 193-215.

Kim, J. H.; Ha, S. Y.; Nam, S. Y.; Rhim, J. W.; Baek, K. H.; Lee, Y. M., J.
Membr. Sci. 2001, 186, 97-107.

Pieracci, J .; Crivello, J. V.; Belfort, G., Chem. Mater. 2002, 14, 256-265.

Lin, H. Q.; Freeman, B. D., Macromolecules 2005, 38, 8394-8407.

Hirayama, Y.; Kase, Y.; Tanihara, R.; Sumiyama, Y.; Kusuki, Y.; Haraya, K., J.
Membr. Sci. 1999, 160, 87-99.

-180-