.u... a a
z . .. ‘
{11.5%
.o X.

 

 

f‘ 00)
L:‘ j. '\

This is to certify that the

dissertation entitled
SURFACE-INITIATED LIVING POLYMERIZATIONS ON
GOLD: SYNTHESIS AND CHARACTERIZATION OF
NANO-THIN POLYMER FILMS

presented by

Jong-Bum Kim

has been accepted towards fulfillment
of the requirements for

 

Ph.D. degreein Chemistry

 

a“ ma

.7 Mm professor

 

Date 3 ”74’? 2002'

MSU is an Affirmative Action/Equal Opportunity Institution 0-12771

 

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

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 c:/CIRC/DateDue.p65-p.15

SURFACE-INITIATED LIVING POLYMERIZATIONS ON
GOLD: SYNTHESIS AND CHARACTERIZATION OF
NANOMETER THICK POLYMER FILMS
By

long-Bum Kim

A Dissertation

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
Department of Chemistry

2002

ABSTRACT

SURFACE-INITIATED LIVING POLYMERIZATION S ON GOLD: SYNTHESIS

AND CHARACTERIZATION OF NANOMETER THICK POLYMER FILMS
By

long-Bum Kim

We successfully synthesized polymer brushes on gold surfaces by surface-
initiated atom transfer radical polymerization (ATRP) of (meth)acrylates at room
temperature. The temperature used in this polymer brush synthesis is the lowest reported
for surface-initiated radical polymerizations, and we report the direct measurement of the
molecular weight of polymer brushes grown on a flat surface. The molecular weight of
the polymer brushes suggests a high grafting density of poly(methyl methacrylate)
(PMMA) on gold surfaces. Little polymer forms in solution, and a simple solvent rinse

can clean the polymer brushes on the surface.

We also elucidated the termination mechanism of surface-initiated ATRP. To
demonstrate the dependence of film growth on radical concentration, polymer brushes
were synthesized using different catalyst concentrations. The surface-initiated ATRP of
methyl methacrylate (MMA) with no addition of Cu(II)Br2 revealed a rapid initial
increase in thickness, but early termination of film growth. In contrast, polymerization
with initial addition of Cu(II)Br2 showed less deviation from ideal linear growth with
time. There exists an optimum catalyst concentration to yield maximum film thickness of

polymer after a certain period of reaction time.

ATRP from a surface is especially useful for synthesis and characterization of
triblock copolymer brushes. Surface-tethered triblock copolymers composed of
poly(methyl acrylate), PMAA, and poly(2-hydroxyethyl methacrylate) (PHEMA) were
grown from gold surfaces by a series of atom transfer radical polymerizations at ambient
temperature. GPC determinations of the molecular weights of model PMMA-b-PMMA—
b-PMMA tn'block polymer brushes demonstrate that this method can yield relatively
homogeneous polymer brushes (polydispersity around 1.5). Between the synthesis of
different blocks, films were exposed to a large excess of Cu(II)Br2 to quench

polymerization prior to washing and reinitiation.

We also examined surface-initiated ring-opening polymerization (ROP) of lactide
on PHEMA surfaces to synthesize biodegradable polylactide (PLA) thin films. Using the
dual living polymerization system of ATRP and ROP, we successfully synthesized
PHEMA-g-PLA on gold and controlled the thickness of PLA by varying polymerization
time. The growth of PLA on PHEMA followed lst order kinetics as evidenced by a linear
increase in PLA thickness with polymerization time. This suggests living polymerization.
Degradation of PLA film was tracked by surface reflectance FI‘IR. Changes in surface

morphology were also studied by optical microscopy during hydrolytic degradation.

To My Family

iv

ACKNOWLEDGMENTS

First, I would like to thank my advisor, Professor Gregory Baker, for his kind
guidance to help me to achieve my goals at Michigan State University. He has been an
excellent research advisor as well as a mentor for me. What I have learned from him will
be invaluable to my success at the next level. I wish to express my deep appreciation to
my committee members. Working with Professor Merlin Bruening was very good luck
for me because I learned his sincere approach to science with critical insights. I also
learned good organic chemistry and surface science from Professor Babak Borhan, Simon
Garrett and Gary Blanchard, which are crucial parts of my knowledge. I would like to
thank Professor Ned Jackson and Rob Maleczka for their guidance during my first year at
MSU.

I would like to give my thanks to all Baker group members: Sue, Cory, Yiyan,
Micah, Chun, Gao, Tara, Mao, Kirk, Erin, Fadi, Tianqi, Ying, Bao, quei, Feng, Leslie,
and Ping for making my life as a Baker Group member a memorable one. I also enjoyed
working with Wenxi, Milind, Matt, Skanth, and Anagi. We were such an excellent
research team. Special thanks to Lisa Dillingham for her being kind to me during my stay
in Chemistry Department. She has been always a big help for me.

Finally, I want to thank my family. They have always been supportive of pursuit
my dreams. My wife Seonmee and daughter Yerin, they deserve my biggest thanks and

love. My life would not have been the same without their love. Thank you for everything.

TABLE OF CONTENTS

Page
List of Abbreviations ....................................................................................................... x
List of Schemes ............................................................................................................... xii
List of Tables ................................................................................................................... xiv
List of Figures .................................................................................................................. xv
Chapter 1. Introduction ........................................................................................ 1
1. Atom Transfer Radical Polymerization (ATRP) ....................................................... 1
1-1. Comparison of ATRP with Conventional Radical polymerization .................. 1
1-2. Mechanism ........................................................................................................ 3
1-3. Kinetics ............................................................................................................. 5
1-4. Catalyst and Initiating Systems ......................................................................... 9
1-5. Ligands .............................................................................................................. l4
I-6. Solvents and Other Factors ............................................................................... 18
II. Surface—initiated Polymerizations ............................................................................. 20
11-1. Preparation of Polymer Brushes ....................................................................... 20
11-2. Surface-initiated Free Radical Polymerization ................................................. 22
11-3. Surface-initiated ATRP .................................................................................... 29
11-4. Surface-initiated Cationic Polymerizations ...................................................... 35
11-5. Surface-initiated Anionic Polymerizations ....................................................... 37
11-6. Surface-initiated Ring-opening Metathesis Polymerization ............................. 40
11-7. Electropolymerization and Plasma Polymerization on Surfaces ...................... 41
11-8. Application of Surface-initiated Polymerization .............................................. 42
111. References ................................................................................................................. 48

vi

Chapter 2. Surface-initiated ATRP on Gold at Ambient Temperature

.............. 55
I. Introduction .............................................................................................................. 55
II. Surface-initiated ATRP with MegTREN .................................................................. 58
111. Control Experiment .................................................................................................. 61
IV. Molecular Weight Determination ............................................................................. 63
V. Conclusion ................................................................................................................ 64
V1. References ................................................................................................................ 67

Chapter 3. Kinetic Study of Surface-initiated Atom Transfer Radical

fl
0

<25;

Polymerization from Flat Surfaces ............................................. 69
Introduction .............................................................................................................. 69
Dependence of Polymer Film Growth on Radical Concentration ............................ 72
Dependence of Polymer Film Growth on Stirring ................ ' ................................... 87
Conclusions .............................................................................................................. 89
References ................................................................................................................ 90

Chapter 4. Synthesis of Triblock Copolymer Brushes by Surface-

Initiated Atom Transfer Radical Polymerization ................. 92
I. Introduction .............................................................................................................. 92
H. Synthesis of Tethered PMA-b-PMMA-b—PHEMA Brushes on Gold ...................... 97
111. Molecular Weights of Triblock Homopolymer Brushes .......................................... 102
IV. Quantification of the Efficiency of Q& R ................................................................ 104
V. Comparison of Cu(11)Br2 Quenching with Simple Solvent Washing ...................... 110
VI. Conclusions ............................................................................................................. 112
VII. References ............................................................................................................... 115

vii

Chapter 5. Controlled Growth and Degradation of Biodegradable
Nanometer-Thick Films of PLA from PHEMA Surfaces .. 118

I. Introduction .............................................................................................................. 118
11. Synthesis of PHEMA-g-PLA ................................................................................... 120
III. Kinetics of Lactide Polymerization from PHEMA Surface ..................................... 124
IV. Control Experiments Without Surface -OH Groups ............................................... 124
V. Degradation of PLA Film ........................................................................................ 129
VI. Surface Morphology Study of PLA Film During Degradation ................................ 134
VII. Conclusions ............................................................................................................. 140
VII1.References ................................................................................................................ 141
Chapter 6. Experimental ...................................................................................... 144
1. General Details ......................................................................................................... 144
1-1. Materials ........................................................................................................... 144
I-2. Synthesis of MegTREN ..................................................................................... 144
L3. Synthesis of anbpy ......................................................................................... 145
1-4. Analytical Methods ........................................................................................... 146
II. Kinetic Studies of ATRP .......................................................................................... 147
II-l. General Procedures for Surface-initiated Polymerization ................................ 147
11-2. Preparation of Monomer/Catalyst Solution ...................................................... 147
11-3. Polymerization for Kinetic Study ..................................................................... 148
11-4. Solution Polymerization Procedures ................................................................ 149
11-5. Polymerization With or Without Stirring ......................................................... 150
11-6. Simulation of Polymer Brush Growth at Different Catalyst Concentration ..... 150
111. Synthesis of Triblock Copolymer Brushes .............................................................. 151

viii

111-1. Preparation of Substrate Immobilized with Initiator ....................................... 151

111-2. Procedures for Block Formation ..................................................................... 152
111-3. Quenching Procedure ...................................................................................... 153
111-4. Detachment of Polymer Brushes from Gold Surfaces ..................................... 153
V1. Synthesis of Biodegradable Nanometer-Thick Films of PLA ................................. 154
VI-l. Preparation of Monomer Solution and Substrates .......................................... 154
VI-2. Ring-opening Polymerization from PHEMA Surface .................................... 155
VI-3. Hydrolytic Degradation of PLA film .............................................................. 156
VI~4. Observation of PHEMA-g-PLA Films During Hydrolytic Degradation ........ 156
VII. References ............................................................................................................... 157

ix

AIBN

ATRP
2-BPB
BPMOA
bpy
depy
DMF
dNbpy
EXAFS

GPC
HEMA

HMTETA

LB
LCP
MA
MAA
uCP
MeHQ

MCfiTREN

MMA
Mn
MUD
Mw
NMR
OEGMA
OTf

PDI

PDMAEMA

PGMA
PHEMA
PMMA

PMDETA

PPEI
PS
PTBA
ROMP
RP

LIST OF ABBREVIATIONS

2,2’-Azobisisobutyronitrile

Atomic Force microscopy

Atom transfer radical polymerization
2-Bromopropionyl bromide
N,N-bis(2-pyn'dylmethyl)octylarnine
Bipyridine

4,4'-diheptyl-2,2’-bipyridine
N,N-dimethylformamide
4,4’-di(5-nonyl)-2,2'-bipyridine
Extended X-ray absorption fine structure
Fourier transform infrared

Gas chromatography

Gel permeation chromatography
2-hydroxyethyl methacrylate

1, 1 ,4,7, 10, lO-hexamethyltn'ethylenetetramine
Langmuir-Blodgett

Liquid crystalline polymer

Methyl acrylate

Methacrylic acid

Microcontact printing

Methyl hydroquinone
tris[2-(dimethylamino)ethyl]amine
Methyl methacrylate

Number average molecular weight
Mercaptoundecanol

Weight average molecular weight
Nuclear Magnetic Resonance
Oligo(ethylene glycol) methacrylate
Triflate

Polydispersity index calculated as Mw/Mn
Poly(dimethylaminoethyl methacrylate)
Poly(glycidyl methacrylate)
Poly(2-hydroxyethyl methacrylate)
Poly(methyl methacrylate)
N,N,N',N',N"-pentamethyldiethylenetriamine
Poly(N-propionylethylenimine)
Polystyrene

Poly(tert-butyl acrylate)

Ring-opening metathesis polymerization
Rate of polymerization

SAM
SEM
TEMPO
TMEDA
TPMA
T-triazine
UV

Xn

Self-assembled monolayer

Scanning electron microscopy
2,2,6,6-tetramethylpiperidinyloxy
Tetramethylethylenediarrfine
tris[(2-pyridyl)methyl]amine

1 ,3,5-tris[(dimethylamino)propyl]hexahydro-triazine
Ultraviolet

Degree of polymerization

xi

Scheme

Scheme 1.1.

Scheme 1.2.
Scheme 1.3.
Scheme 1.4.
Scheme 1.5.
Scheme 1.6.
Scheme 1.7.

Scheme 1.8.
Scheme 1.9.

Scheme 1.10.

Scheme 1.11.
Scheme 1.12.

Scheme 1.13.
Scheme 1.14.
Scheme 1.15.
Scheme 1.16.
Scheme 1.17.

LIST OF SCHENIES

Page
Radical generation steps in conventional and atom transfer radical
polymerization ......................................................................................... 2
Example of an atom transfer radical addition ......................................... 3
Mechanism of ATRP of methyl methacrylate ......................................... 4
Unit reactions in ATRP of styrene .......................................................... 7
ATRP initiators ....................................................................................... 1 1
Possible structures of CuX/dNbpy catalysts ............................................ 13
Proposed structures of the Cu(I)(bpy)2 and Cu(11)(bpy)2 catalysts in
ATRP ....................................................................................................... 14
Ligands used for copper-mediated ATRP ............................................... 16
Polymer brushes on a flat solid substrate ................................................ 21
Synthesis of polystyrene brushes on silica and cleavage of the
polymers from the surface ....................................................................... 25
Surface-immobilized initiators: (a) TEMPO-based and (b) iniferter ...... 29
Synthesis of a polystyrene—b-poly(methyl methacrylate) brush by
sequential carbocationic polymerization and ATRP ............................... 32
Cationic polymerization of hexyl isocyanate on gold nanoparticles ....... 37
Surface-initiated anionic polymerization of styrene on gold ................... 38
Surface-initiated anionic polymerization of styrene on clay ................... 39
Surface-initiated ROMP of norbomene-based chiral monomers ............ 40
Surface patterning by microcontact printing, surface-initiated
ATRP, and etching .................................................................................. 43

xii

Scheme 1.18. Surface-initiated ROMP from Si02/Si surface ........................................ 44
Scheme 2.1. Surface-initiated AT RP of MA with Me6TREN at RT ........................ 59
Scheme 3.1. Surface-initiated ATRP of (meth)acrylates on gold ................................ 73
Scheme 3.2. Kinetic scheme for ATRP: pathways for radical generation and

consumption ............................................................................................ 74
Scheme 4.1. Possible reaction pathways of a polymeric radical in ATRP of

IVIIVIIAI ....................................................................................................... £965
Scheme 4.2. Synthesis of PMA-b-PMMA-b-PI-IEMA triblock copolymer

brushes ..................................................................................................... 98
Scheme 5.1. Synthetic pathway for PHEMA-g-PLA ................................................... 122

xiii

LIST OF TABLES

Table Page

Table 2.1. Molecular weight of PMMA brushes ...................................................... 63
Table 4.1. GPC data for PMMA standards and polymers grown from surfaces
using ATRP ............................................................................................. 105

xiv

Figure

Figure 1.1.

Figure 1.2.

Figure 1.3.

Figure 1.4.

Figure 1.5.

Figure 1.6.

Figure 1.7.

Figure 1.8.

Figure 1.9.

Figure 2.1.

Figure 2.2.

LIST OF FIGURES

Page

Rate of methacrylic acid polymerization initiated by UV irradiation

of a xanthate system as a function of (a) fraction of xanthate in the
monolayer and (b) pH .............................................................................. 23
Time-dependent properties of polymer chains grown by surface-

initiated free radical polymerization of styrene: (a) molecular

weight M.., (b) grafting densities 6(PS), and (c) polydispersity of

the covalently attached polymers ............................................................ 27
Schematic illustration of ATRP initiator immobilization on Si by

the LB technique ...................................................................................... 31

AFM images of polyacrylamide films prepared by (a) surface-

initiated ATRP and (b) surface-initiated free radical polymerization ..... 34
Surface-initiated cationic polymerization of 2-oxazolines ...................... 36
Sequential surface derivatization of PET films ....................................... 41
Optical micrographs of patterned surfaces .............................................. 42
Surface patterning by surface-initiated ROMP ....................................... 45

Schematic depiction of an LC brush swollen in a low molecular

weight nematic ........................................................................................ 47
Differences in the radical generation mechanism for surface-

initiated free radical polymerization and surface-initiated ATRP ........... 56
Reflectance FTIR spectra of (a) a SAM of MUD (4), (b) initiator
immobilized on a MUD SAM (5), and (c) a grafted PMMA layer

on the initiator surface (6) ....................................................................... 6O

XV

Figure 2.3.

Figure 2.4.

Figure 2.5.

Figure 3.1.

Figure 3.2.

Figure 3.3.

Figure 3.4.

Figure 3.5.

Figure 3.6.

Figure 3.7.

Figure 3.8

Figure 3.9.

Reflectance FTIR spectra of (a) a SAM of MUD (4), (b) propionyl

group anchored MUD SAM (7), and (c) after surface

polymerization ......................................................................................... 62
Detachment of polymer brushes from gold by 12 treatment followed

by molecular weight determination by GPC ........................................... 65
Tapping mode AFM images of (a) a bare gold surface and (b)

PMMA brushes on gold .......................................................................... 66
Dependence of PMA film growth on the presence of Cu(11)Br2:

(a) without initial addition of Cu(II)Br2 and (b) with initial addition

of Cu(II)Br2 ............................................................................................. 75
Dependence of additional PMA film growth on the amounts of
remaining surface-initiating sites ............................................................ 78
Solution phase ATRP results: (a) first order kinetic plot of MMA
polymerization and (b) molecular weight and distribution of

polymers. ................................................................................................. 79
Growth of PMA layer thickness vs polymerization time ........................ 82
Optimum catalyst concentration that yields the thickest PMA film

after 4 hr of polymerization ..................................................................... 84
Simulation of surface-tethered radical concentrations as a function

of catalyst concentration .......................................................................... 85
Influence of radical concentrations on the probability of

recombination .......................................................................................... 86
Growth of PMA layer thickness vs polymerization time: (a)
polymerization without stin'ing; (b) polymerization with

Digital images of the glass reactor used for polymerization with

stirring ..................................................................................................... 89

xvi

Figure 4.1. Reflectance FTIR spectra of (a) a layer of initiator immobilized on
gold (1), (b) a grafted PMA layer on the initiator surface (2), (c) a
grafted PMA-b-PMMA diblock copolymer (3), and (d) a grafted

PMA-b-PMMA-b—PHEMA triblock copolymer (4) ................................ 100
Figure 4.2. Tapping mode AFM images of PMA(58A)-b-PMMA(111A)—b-

PHEMA(75A) ......................................................................................... 101
Figure 4.3. Ellipsometric thickness of PMA films vs polymerization time ........... 107
Figure 4.4. Thickness of multiblock PMMA films vs polymerization time .............. 113
Figure 4.5. Thickness of multiblock PMA films vs polymerization time ................. 114

Figure 5.1. Reflectance FT IR spectra of (a) 2: a grafted PHEMA layer (174

nm) on gold, (b) 3: a grafted PHEMA(17O nm)-g-PLA(29 nm)

layer after 90 min of lactide polymerization, (c) 3: a grafted

PHEMA(173 nm)-g-PLA( 127 nm) layer after 180 min of lactide

polymerization, (d) 3: a grafted PHEMA(178 nm)-g-PLA(179 nm)

layer after 240 min of lactide polymerization, and (e) 3: a grafted

PHEMA( 178 nm)-g-PLA(409 nm) layer after 480 min of lactide

polymerization ......................................................................................... 123
Figure 5.2. Plot of PLA thickness increase vs polymerization time .......................... 125
Figure 5.3. Reflectance FTIR spectra of (a) a surface grafted PHEMA layer on

gold, (b) after lactide polymerization from PHEMA, (c) a surface

grafted acetylated-PHEMA layer, and (d) after attempted lactide

polymerization from acetylated-PHEMA ................................................ 126
Figure 5.4. Reflectance FI‘IR spectra of (a) a surface grafted PMMA layer on

gold, (b) after attempted lactide polymerization from PMMA, (c) a

surface grafted PHEMA layer, and (d) after lactide polymerization

initiated from PHEMA ............................................................................ 127

xvii

Figure 5.5. Reflectance FTIR spectra of (a) PHEMA (174 run) before PLA

growth, (b) PHEMA (174 nm)-g-PLA(464 nm) before degradation,

and (CH1) PI-IEMA-g-PLA films at various times during

hydrolytic degradation ............................................................................. 131
Figure 5.6. Plot of PLA carbonyl peak area vs degradation time .............................. 133
Figure 5.7. Surface images of PHEMA(177 nm)-g-PLA(353 nm) obtained by

optical microscopy during hydrolytic degradation .................................. 135
Figure 5.8. Evolution of surface domains during degradation ................................... 137
Figure 5.9. Differences in the optical microscopic images of domains as a

function of changes in focal length ......................................................... 138
Figure 5.10. Evolution in surface morphology during the drying of water-

swollen films ........................................................................................... 139
Figure 6.1. Digital images of five stacked initiator-immobilized samples in a

reaction vial ............................................................................................. 149

xviii

Chapter 1. Introduction

1. Atom Transfer Radical Polymerization (ATRP)
I-l. Comparison of ATRP with Conventional Radical Polymerization

Radical polymerization is one of the most useful and practical synthetic methods
for forming polymers because of its applicability to a wide range of monomers, relatively
simple reaction conditions, and compatibility with solution, bulk, suspension, and
emulsion polymerization. Free radical polymerization, however, is limited by the fast
irreversible termination of growing radicals by coupling, disproportionation, and chain
transfer reactions. This termination leads to poor control over molecular weight and
polymer structure, and thus much effort has been devoted to developing controlled
radical polymerizations that produce well-defined polymers with narrow molecular
weight distributions. Since Matyjaszewskil and Sawamoto2 independently reported Atom
Transfer Radical Polymerization (ATRP) in 1995, ATRP has become one of prevailing

polymerization methods for minimizing termination reactions in radical polymerizations.3

ATRP is a versatile method for polymerizing acrylates, methacrylates, and
styrene. Using ATRP, the molecular weights of these polymers can be controlled easily
by varying monomer/initiator ratios and polymerization times. Molecular weight
distributions are narrower than for polymers prepared by conventional free radical
polymerizations. The improved control over polymerization afforded by ATRP can be
ascribed to a different mechanism for radical generation. In conventional free radical

polymerization, a propagating radical is irreversibly generated by photochemical or

thermal processes, while in ATRP, metal-catalyzed atom transfer leads to “reversible”
radical generation. This reversibility keeps the radical concentration low, which
minimizes radical termination by recombination. Scheme 1.1 shows the radical

generation steps in conventional and atom transfer radical polymerization.

Scheme 1.1. Radical generation steps in conventional and atom transfer radical

polymerization

i) Conventional Free Radical Polymerization

THG $143 heat or UV $143
Nae—p—N=N—p—CEN ———> 2 Nae—p + sz

CHa CHa CH3
AIBN

ii) Atom Transfer Radical Polymerization
I? it'a i? P's
CHgCHgo—C—(p—Br + Cu(l)Br/2bpy —‘ CHsCHzo—C—?' + Cu(ll)Br2/ 2 bpy
CHa CHa

 

Ethyl 2-bromoisobutyrate

ATRP is considered to be a “living” polymerization, a chain-growth
polymerization without irreversible termination and chain transfer reactions. Thus, once
chains are initiated, they propagate until the supply of monomer is exhausted or until an
equilibrium is established between propagation and depropagation reactions. The chain

ends remain active until terminating agents are intentionally introduced. In practice a true

living radical polymerization is rarely achieved because free radicals are so reactive that
it is not possible to prevent some active chain ends from participating in termination and
chain transfer reactions. In certain cases, however, these side reactions occur so
infrequently that well-defined polymers can be prepared. These polymerizations are
termed “controlled” polymerizations.4 ATRP exhibits many characteristics of living
polymerizations, but most polymer chemists prefer to describe them as
“controlled”/living polymerizations because termination is unavoidable in radical

polymerization.

1-2. Mechanism

ATRP is a multistep repetition of atom transfer radical addition reactions
(Scheme 1.2).5’6 An organic radical is first generated by transferring a halogen atom from
an organic halide to a transition metal complex. Addition of the radical to an alkene

followed by transfer of the halogen atom to the resulting radical completes the reaction.

Scheme 1.2. Example of an atom transfer radical addition

 

Cu(l)Cl + ccu4 Cu(lI)Cl2 + ~ccna

 

~ccna + H20=CH—X ——> Hztf—OH—X

CCIg
|

OH—X + Cu(ll)Cl2 —: Pup—CH—x + Cu(I)CI
CC|3 CCla

In ATRP systems, reversible termination is used to reduce the steady state
concentration of growing radicals and suppress radical termination reactions. Once a
radical is generated from a dormant initiator, it can either add monomer or deactivate by
reacting with a metal complex to regenerate the dormant initiator and a metal in a lower
oxidation state. A series of similar activation/deactivation reactions occurs during
reinitiation and propagation. Scheme 1.3 shows the ATRP of methyl methacrylate
(MMA), including the reversible radical generation steps that occur after the addition of

each monomer.

Scheme 1.3. Mechanism of ATRP of methyl methacrylate

II I II I
R—c-p—Br + Cu(I)Br .._—_—-—_ R—c—p- Cu(II)Br2

CHa CHa

i0

00-h

i? I‘ 9‘3 9 in 9H.
R—c-p-CHg—p—Br + Cu(I)Br v—_— n—c-p—CHZ—p- Cu(ll)Br2
CHG COZCHa CHs cozcn3

io

OCH;

.. Rh 9'3 9 P's P‘s
R—C-(IBH—Cl-lg— CHz—tP—Br + Cu(I)Br ;_ R—C-CI-l—CI-lz—(f—CI-lz—(P Cu(II)Br2
Cl-Ia COcha cozcua cue 00201-13 cozcns

111 Further
Polymerization

Even though the steady state concentration of active radicals in ATRP has not yet
been directly measured by spectroscopy, solving the rate expressions for ATRP implies a
radical concentration of ~10’7 to 10'8 M. Since the radical concentration is low,
deactivation is several orders of magnitude faster than recombination between two
radicals, and thus irreversible termination is suppressed. However, some early
termination invariably occurs during the initial formation of Cu(II) species, which are
needed for control via the persistent radical effect.7’8 Electron paramagnetic resonance
studies of the bulk ATRP of styrene at 110 °C showed that after 20 min, the Cu(II)
concentration was nearly steady state, corresponding to 4—6 mol % of the initial Cu(I)
species.9 Since the polymerization used l-phenylethyl bromide and CuBr(dNbpy)2
(dNbpy = 4,4’-di(5-nonyl)-2,2’-bipyridine) as the initiating system with no added Cu(II)

species, the Cu(II) concentration built up in situ.

I-3. Kinetics

ATRP provides well-defined polymers with low polydispersities (1.05 < MWIMn <
1.5). The number average degree of polymerization (X..) is determined by the ratio of the
concentrations of consumed monomer to initial initiator: Xn = A[M]/[I]o.10 As shown in
equation 1, ATRP usually follows first-order kinetics with respect to monomer, initiator,
and catalyst (Cu(I)) concentration, while the rate of polymerization is inversely
proportional to the deactivator (Cu(11)). Equation 2 describes the evolution of the

polydispersity (MW/Mn) with conversion (p).ll

 

 

k [Cu‘]
=k _..._ M I .1
RP ”kml H ]°[XCu"] eq
MW =1+[ kplllo n IE_1] eq. 2
M, kaXCu 1 p

The two equations show the trade-off between polymerization rate and control
over molecular weight. A high kw/kdmt ratio leads to a high radical concentration and a
corresponding high rate of polymerization. At the same time, a relatively low kw,“ leads
to an increase in the polydispersity, probably through irreversible termination. In the
limit where kw >> km“ the polymerization rate is very high but the reaction yields a
polydisperse polymer. This case is best described as a conventional redox-initiated
radical polymerization process. The equations show that adding Cu(II) favors the
deactivation of radicals, slows the polymerization rate, and provides a lower
polydispersity. Equation 2 also indicates that MW/Mn decreases with conversion, a
characteristic of living polymerizations. Thus, an ATRP process with a low kp/kdcw ratio
and a high concentration of deactivator will provide a narrow molecular weight

distribution at high degrees of polymerization.11

ATRP successfully polymerizes styrene, acrylates, and methacrylates.
Matyjaszewski and coworkers investigated the ATRP kinetics for methyl acrylate (MA),
MA, and styrene. For the homogeneous polymerization of MA using CuBr(dNbpy)2,12
plots of the evolution of molecular weight vs conversion and ln([M]o/[M]) vs time were
linear, confirming lst order kinetics. The equilibrium constant for reversible radical

formation at 90 °C, Kgq = 1.2 x 109, is much smaller than in the ATRP of styrene or

MMA. For comparison, Keq for the ATRP of MMA with CuCl at 90 °C is 7.0 x 107, and
Keq for the ATRP of styrene with CuBr is 2 x 10's. The differences in Kcq are probably
due to different radical stabilities or the differences in the C—halogen bond strength.

ATRP of styrene at 110 °C (1 mol % initiator and catalyst) has been studied in

detail.”

The characteristic time intervals for the various steps in ATRP can be estimated
from the reciprocal of the product of the bimolecular rate constants and the reagent

concentration (1: = l/k2[R]). Scheme 1.4 shows the characteristic times for the individual

steps in the ATRP of styrene, activation, deactivation, propagation, and termination.

Scheme 1.4. Unit reactions in ATRP of styrene

i) activation (Ta = 22 s); deactivation (1,, = 0.018 ms)

 

 

activation
Pn—Br + Cu(I)Br / 2 bipyr L He + Cu(ll)Br2 / 2 bipyr
deactivation
ii) propagation (1:p = 0.12 3)
Hr + Styrene -—-> F...-
iii) termination (t. = 0.13)
P". + Pm. ———> Pn_ Pm

* Pr; and P"; represent polystyrene radicals: H—-CHz— I-l-—CH2— .

O n-1 or

 

 

The calculated periodicities shown in Scheme 1.4 lead to a plausible kinetic
scenario for styrene polymerization.” Early in the polymerization, a high concentration
of radicals is generated and their rapid recombination leads to ~5 % loss of chain ends
and the in situ production of ~5 % of Cu(II) species from the Cu(I) catalyst. The Cu(II)
species deactivate growing radicals, decrease their concentration, and suppress
terminatiOn. At 30 % conversion, an average chain is activated to a radical every 22 sec,
and then is deactivated after 0.018 msec. Since the deactivation rate is seven times faster
than propagation (t = 12 msec), addition of one monomer to a polymer chain requires
seven activation/deactivation cycles (2.5 min), and 4 h to prepare a polymer chain with a
degree of polymerization of 100. Termination is 830 times slower than propagation (t =
12 msec), leading to good control over the polydispersity. For the ATRP of MA, similar
characteristic time estimates predict that 14 monomers are added to a growing radical
during each activation/deactivation cycle. The number of monomer units (n) added

during each activation step can be expressed as
n = kp[M]/kdew[D] eq. 3
where [M] and [D] are the concentration of monomer and deactivator respectively.

The above kinetic description is simplified because it fails to take into account
chain transfer reactions and termination by processes other than bimolecular coupling or
disproportionation. ATRP kinetics vary with the kind and concentration of catalyst,
ligand, monomer, solvent, and temperature. In the next sections, these factors and their

influence on ATRP are explored in more detail.

1-4. Catalyst and Initiating Systems

In ATRP, the catalyst plays an important role in determining the rate of the
polymerization and the molecular weight distribution of the resulting polymers. The
catalyst activity can be adjusted by changing the ligand, counterion, or transition metal

ion. A variety of metal catalysts have been developed to initiate and control ATRP,

14,15 2.1 I 17.18 . 19,20
’

including copper, ruthenium, 6 nicke Iron, and rhodium complexes.” The
copper halide systems are the most popular because of their versatility with vinyl and
(meth)acrylate monomers.22 Since copper-based ATRP was used for this research, this

discussion emphasizes copper catalysts.

The halogen atoms in the radical generation step play important roles in defining
the position of the equilibrium.23 The C-Cl bond energy in benzyl chloride is ~ 284.7
kJ/mol while the C-Br bond energy in benzyl bromide is 213.5 kJ/mol.24 Thus, initiation
of C-Br is faster than C-Cl, reflecting the difference in the carbon-halogen bond energy.
The partial ionic character of the Cu-X bonds in a ligated complex such as Cu(bpy)2"X‘
may also be a factor in determining the equilibrium position. Scheme 1.5 shows various
initiators used in ATRP. Besides those listed in the scheme, Ot-haloketones also have been
used as ATRP intiators. Radicals generated a to functional groups are stabilized in the
order CN > C(=O)R > C(=O)OR > Ph > C1 > Me. Tertiary alkyl halides are better
initiators than secondary and primary alkyl halides because of the higher stability of
substituted radicals. Alkyl chlorides should be the least efficient initiators because the
order of bond strength in alkyl halides is R-Cl > R—Br > R—I. Chlorosulfonylphenyl
groups (-Ph—SOzCI) are also good ATRP initiators because the Cl atom can be easily

abstracted by copper catalysts, producing —Ph—SOz radicals.

Control experiments on halide-free catalysts show that the ions coordinating with
copper have little effect on the deactivation step. The halides of Cu bipyridine complexes
were replaced by triflates (14Cu(OTt)2, or LzCu(OTt), L = a substituted bipyridine) and
the complexes were added to model free radical polymerizations initiated by AIBN.
14Cu(OTf)2 had no effect on MA and styrene polymerizations, while 14Cu(OTf) reduced
the polymerization rate of MA, but had no effect on styrene polymerization. Thus, the
control seen in ATRP does not come from the interactions of growing radicals with
copper complexes but in the reversible halogen atom transfer.25

Mixed halide systems (R-Br/CuCl) slow the polymerization rate and reduce the

26

rate of termination. As a result, control over polymerization is improved, and

polydispersity is reduced. To better understand halogen exchange between initiator and

CuCl, model studieszns

were run at 90 °C using several initiators (R-Br = benzyl
bromide, l-phenylethyl bromide, methyl-2-bromopropionate, and ethyl-2-
bromoisobutyrate) with CuCl. NMR and GC results show that within 30 min, 80—90
mole% of the R-X groups are R-Cl formed by the back transfer of Cl from CuBrCl to
R-F" For R-Cl/CuBr systems, R-Cl is also the dominant alkyl halide at steady state. Thus
for mixed halide systems, the rate of halogen exchange is rapid and alkyl chlorides are

preferentially formed over alkyl bromides. These results also suggest that the mixed

halide system should give fast initiation.

10

Scheme 1.5. ATRP initiators

i) halogenated alkanes and benzylic halides

i i'

RCCla n— —x R— —Cl X
R = H
.. 0
Br
Ph
CFa R = H. CH3 R = H, Ph x = Br, CI
COzCHa X = 31', 0'

ii) a-bromoesters

AJXR “@9103 M 8'
Br B CHg

R = H, CH3

PMS: P v 3'
. 0 MAY”

0
iii) sulfonyl chlorides R
9—8 Cl +8 Cl 0 8 Cl 0 ff
IT— I I— C,
0 O O
R O
R = CCI3, CH3 R = OCHa, CH3, F, COZH R = CI, OCH3 R = H, N(CH3)2

kid ‘0 im >MKE4.

To have good control over the polymerization kinetics and polymer molecular
weight distribution in ATRP, a few major criteria should be fulfilled. First, fast initiation
is required to generate significant numbers of polymer chains and ensure that all chain
ends start to grow polymer at the same time. Second, the equilibrium position in the
radical generation step should strongly favor the dormant species. Fast deactivation of
active radicals to dormant chain ends is indispensable for minimizing chain termination
reactions that broaden the polymer molecular weight distribution. Third, all initiators
must grow polymers for the molecular weight of the final polymer to be defined by the
[monomer]/ [initiator] ratio. In terms of the above three requirements, the mixed halogen
initiation system, R-Br/CuCl is possibly more advantageous than the R-Br/CuBr or R-
Cl/CuCl because R-Br/CuCl allows faster initiation than the R-Cl/CuCl system and faster
deactivation than R-Br/CuBr. The experimental data show that mixed halide systems
provide better control over ATRP. For example, the ATRP of MMA using a R-Br/CuCl

system showed lower polydispersity than polymerization with R-Br/CuBr.28

To date, the exact structure of the catalysts in solution is unclear, and even the
structure of the widely studied Cu/bpy system is uncertain. 1H NMR studies suggest that
the Cu(I) and Cu(II) complexes are labile with fast exchange between coordinated and
free bpy.29 Temperature-dependent color changes of catalyst solutions have been
interpreted as a structural change from a monomeric Cu(I) species to a dimeric halogen
bridged species.26 Based on literature data on the coordination chemistry of copper
complexes in polar solvents, Matyjaszewski proposed the structures shown in Scheme
1.6 as possible forms of CuX/dNbpy during the polymerization.3 There are three types of

complexes that can have a 1:1 Cu(I)X to bidentate ligand ratio, the halogen bridged dimer

12

1, complexes 2 and 5 which have two bidentate ligands coordinated to Cu(I) and either a

halide or dihalocuprate counter-anion,”3 1

and complexes 3 and 4, where monomer acts
as a ligand. The structure of the CuX/bpy complex depends on the solvent polarity,32

the monomeric form 5 should dominate in polar solvents, while bridged dimer 1 could
exist in less polar media. More complex structures also may be possible. The maximum
rates of polymerization for styrene and MA were observed when the ratio of dNbpy to
CuX was 1:2,12’20 suggesting 5 as the active catalyst. X-ray and EXAFS data point to a
cationic bipyramidal structure [X-Cu(11)(bpy)2]+ for the solid catalyst isolated from
nonpolar media, however, a neutral distorted square planar structure (X2Cu(II)/bpy) may
be preferred in solution.33 It was also proposed that the active Cu/bpy complexes

involved in the ATRP are tetrahedral Cu(I)(bpy)2 and trigonal bipyramidal X-

CU(H)(bPY)2 Complexes (Scheme 1.7).34'35

Scheme 1.6. Possible structures of CuX/dNbpy catalysts.

N\ /x\ / N\ / _|+ - N\ /M :\C /-""-]+
C /C"’\ j/C‘KD =<N/:\:> CUX’Z ‘— (N /C"\x ‘_ /Cx‘\

C::I;::.':: \ \ g /3/

5

l3

Scheme 1.7. Proposed structures of the Cu(I)(bpy)2 and Cu(II)(bpy)2 catalysts in

 

 

 

ATRP.
1%
\ / \ / k. ~1M"s"
PK / activation
Pn—Br + in; ‘
N ' deactivation
/ \ / \ [wow-1.4
/ kp ~103M"s"

1.5. Ligands

While the metal catalyst controls the polymerization reaction by mediating the
atom transfer equilibrium, the ligands play critical roles in solubilizing the catalyst in
organic media and tuning the rates of the activation and deactivation steps. Extensive
research has focused on finding new ligands that are readily available and whose
structures can easily be varied for tuning ATRP. Common ligands for copper halides are

12,20,37-39 and infines.40’4l Recently,

multidentate aliphatic amines,36 aromatic amines,
Matyjaszewski et a1. investigated the structure-activity relationship for Cu catalysts with

the multidentate ligands shown in Scheme 1.8.11

According to kinetic and electrochemical studies, catalyst activity increases as the
ligand stabilizes the Cu(II) state of the catalyst. In general, the activity of ligands
decreases in the order: alkyl amine z pyridine > alkyl imine >> aryl imine > aryl amine.
A sufficiently high deactivation rate is required to obtain good control over

polymerization because the molecular weight distribution is influenced by the rate of

14

deactivation. In terms of the rates of the activation and deactivation, N,N,N',N',N"-
pentamethyldiethylenetriamine (PMDETA) and N,N-bis(2-pyridylmethyl)octylamine
(BPMOA) are suitable ligands for Cu-based ATRP of a wide range of acrylates,

methacrylates, and styrenic monomers.

Initially bpy was used as the ligand in copper-mediated ATRP of styrene and
(meth)acrylate monomers."37 The Cu/bpy catalyst has limited solubility, and a poorly
soluble Cu(II) complex will suppress the deactivation step and compromise control over
the polymerization. To improve the solubility of the catalyst, bpy was replaced with 4,4'-
diheptyl-2,2'-bipyridine (depy) and dNbpy and ATRP was better controlled.2‘)""2
Besides controlling solubility, ligands also lead to steric effects in ATRP. ATRP with
4,4'-di-tert-butyl-2,2'-bipyridine produced polymers with a narrow molecular weight
distribution, while ATRP using 6,6'-disubstituted 2,2'—bipyridine as the ligand failed to
yield polymers.43 The bulky substituted groups presumably hindered coordination of

ligands to the copper catalyst.

15

Scheme 1.8. Ligands used for copper-mediated ATRP

i) bidentate ligands

O—QWCQD W

dNbpy, when R: 5-nonyl
depy, when Fl: n-heptyl CHIN—03th

ii) tridentate ligands

eHn

>~SS< O5 OSSO OSSO

\C(E>\O/RIIH17Ca/NBLCBH17 H17Ce/N-IHKC8HW

iii) tetradentate ligands

 

 

 

MeeTREN

l6

Catalysts containing multidentate amines such as PMDETA and l,1,4,7,10,10-
hexamethyltriethylenetetramine (HMTETA) have faster polymerization rates than those
using bpy as the ligand.36 One possible explanation is that the copper-amine complexes
have lower redox potentials than copper-bpy complexes,44 shifting the equilibrium from
dormant species towards radicals, which results in a higher polymerization rate. In some
cases, polymerizations using copper catalysts with multidentate amine ligands showed
significant deviations from first-order kinetics. For example, first order kinetic plots of
the ATRP of MMA are curved when PMDETA is used as the ligand, and styrene
polymerizations using tetrarnethylethylenediamine (TMEDA) significantly deviate from
first-order kinetics at high conversions. Both results are consistent with either an increase

in the concentration of growing radicals or loss of initiating sites during polymerization.36

The increased activity of catalysts with multidentate ligands makes ATRP
possible at room temperature. ATRP of MA proceeded at ambient temperature with
tris[2-(dimethylamino)ethyl]amine (MegTREN) as a ligand,38 while the use of dNbpy
required 90 "C.'2 Currently, the most popular ligands for ATRP are 1,3,5-
tris[(dimethylamino)propyl]hexahydro—triazine (T-tr‘iazine),23 MegTREN,38 N,N-bis(2-
pyridylmethyl)octylamine (BPMOA),45 tris[(2-pyridyl)methyl]amine (TPMA),“? N-(n-

pentyl)-2-pyridylmethanimine,4° PMDETA,36 and HMTETA.36

In conclusion, ligands serve two purposes. They fine-tune the equilibrium
constant that governs atom transfer between active radicals and dormant initiators, and
they control the solubility and stability of catalytic complexes in various monomers and

solvents at different temperatures.

17

1-6. Solvents and Other Factors

In addition to the catalyst concentration and activity, several other experimental
parameters such as temperature, use of zero-valent metals, and the solubility of the
catalyst affect the rates of polymerization and termination. For some catalysts, the Cu(II)
complexes have limited solubility and have been observed to precipitate from nonpolar
solvents.23 Loss of Cu(II) often leads to non-first-order kinetics and a loss of control over
polydispersity. ATRP of n-butyl acrylate with CuBr/bpy was poorly controlled in
benzene (MW/Mn = 2.4), while the same polymerization using the more soluble
CuBr/dNbpy afforded well-defined polymer with the expected molecular weight and a
low polydispersity (MW/Mn = 1.1).10 In ethylene carbonate, the same polymerizations
catalyzed by CuBr/bpy provided good control over molecular weight, a low PDI, and a
rate of polymerization that was faster than a bulk polymerization of n-butyl acrylate.lo
The authors speculated that the rate increase is due to the formation of a monomeric

copper species with dissociated anions.

In general, the use of polar solvents improves the solubility of Cu(II) complexes
and leads to homogeneous, well-controlled ATRP. Polar solvents reported to be
compatible with ATRP include y-butyrolactone,23 DMF,lo carbonates (ethylene,
dimethyl, and propylene carbonate),10 and ethers (anisole, diphenyl ether, 1,4-dioxane,
1,4-dimethoxybenzene, and TI-IF).10 Mixed solvent systems have also been investigated,
such as the ATRP of 2—hydroxyethyl methacrylate (HEMA)10 in a mixture of methyl
ethyl ketone and l-propanol, and in methanol/water mixtures. Recently, a room
temperature ionic liquid, l-butyl-3-methylimidazolium hexafluorophosphate was shown

to be a good solvent for copper-mediated ATRP of MMA.46

18

While most ionic and coordination polymerizations cannot be carried out in the
presence of water, radical polymerizations are compatible with water and carbonyl
groups. Recently, Armes reported ATRP of methoxy-capped Oligo(ethylene glycol)

methacrylate (OEGMA) in aqueous media using a Cu(D/bpy catalyst.47 Both OEGMA
and the catalyst are water soluble. The rate of aqueous ATRP at 20 °C is faster than

ATRP run in organic solvents at elevated temperatures. Due to its polarity, water is
believed to promote the formation of [Cu(bpy)2]+ a very active mononuclear catalyst.
Although compatible with water, ATRP is like normal free radical polymerizations in
terms of its sensitivity to oxygen, which must be excluded to limit formation of
unreactive peroxy radicals.48

Monomer structure is one of the major factors that determine the propagation rate
(kp) of polymerization since it defines the stability of the radical formed after its addition
to growing chain ends. The polymerization rate of acrylate and methacrylate is different
because the propagating radical of MMA is more stable and thus chains grow more

slowly.10

Because of its tolerance to many organic functional groups, ATRP is useful for
the synthesis of polymers with complex architectures. Using ATRP, polyhedral
oligomeric silsesquioxane acrylates were polymerized to give hybrid polymers with
previously unattainable compositions and low polydispersities.49 In addition, ATRP and
a second polymerization mechanism can be combined to yield well-defined block
copolymers. For example, ATRP of MMA and the ring-opening metathesis
polymerization (ROMP) of 1,5-cyclooctadiene were initiated simultaneously from a

single ruthenium complex.50

19

Even though ATRP normally gives well-defined final polymers, there are several
problems that must be overcome before it gains widespread applicability in industry. In a
typical ATRP, the catalyst concentration is 01—1 mol % relative to monomer, and the
residual catalyst colors the final polymer. Extra steps must be taken to remove catalyst
after polymerization. Potential solutions include the use of supported catalysts51 and
promoters. ATRP of MMA was reported for a CuBr-HMTETA catalyst system supported
on silica gel,”53 and both visible light and methyl hydroquinone (MeHQ) were shown to
accelerate ATRP54 and improve its living character at low catalyst levels. In the case of
MeHQ, the activation energy for MMA polymerization was reduced from 60.3 to 44.9
ltJ/mol.4o This suggests that MeHQ may act as a coordinating ligand for copper and thus

affect the position of the Cu(D/Cu(II) equilibrium.

II. Surface-initiated Polymerizations
II-l. Preparation of Polymer Brushes

The preparation of polymer brushes on solid surfaces has been of great interest
because of their potential as sensing layers, anti-corrosion layers, for controlling the
wetting of surfaces, and for nanostructuring surfaces. More than two decades ago,
DeGennes developed a model for understanding polymer brush systems and suggested
the synthesis of polymer brushes from surface-immobilized initiators.55 According to
Milner’s definition, polymer brushes are long-chain polymer molecules attached by one
end to a surface or interface by some means, with a density of attachment points high
enough so that the chains are obliged to stretch away from the interface, sometimes much

farther than the typical unstretched size of a chain.56

20

Scheme 1.9. Polymer brushes on a flat solid substrate

1 (.951.

The physical properties of polymer brushes largely depend on the chain length,

 

grafting density, backbone flexibility, and excluded volume. Due to their confinement to
a surface, polymer brushes respond to stimuli in a collective and uniform way, and
usually these responses trigger changes in surface properties. Since these polymer chains
are covalently bound to the surface, adhesion between the polymers and the substrate is
much stronger than can be achieved from simple physisorption. However, due to limited
access of catalyst and monomer to areas close to the interface, polymers grown from

surfaces may be less uniform than those grown under homogeneous conditions.

To obtain homogeneous polymer brushes, the grafting density must be high and
uniform, the PDI of the polymers should be near 1,57 and all chains must be terminally
grafted onto surfaces. Strategies for attaching polymer brushes to surfaces include the
"grafting to" technique,58 tethering preformed polymer chains from solution to a surface,
and the "grafting from" technique pioneered by Sogah,59 polymerization from surface-
anchored initiators. The former results in a lower density of polymer brushes than the
“grafting from” technique because steric hindrance eventually prevents incoming
polymer chains from diffusing through the film to surface reaction sites. This section will

focus on the “grafting from” technique, i.e. surface-initiated polymerizations.

21

II-2. Surface-initiated Free Radical Polymerization

Niwa used UV light to photoinitiate the polymerization of methacrylic acid
(MAA) from a self-assembled monolayer (SAM) of xanthate derivatives on goldfs'0 Both
the formation of the xanthate SAM and the photopolymerization process were monitored
using a quartz crystal microbalance. The SAM was a mixed monolayer prepared from
double-chained xanthate 1, and the xanthate-free analog 2. The polymerization rate was
strongly dependent on the mole fraction of 1 (fl) in the SAM and reached a maximum at
f; = 0.2. The reaction also was pH-dependent, with almost no polymerization at pH 7,
and a rapid increase as the pH was lowered to 2.3, the lowest pH tested (Figure 1.1). The
reason for this dependence is unclear, although one possibility is that increasing
electrostatic repulsion at high pH between incoming MAA and PMAA decreases the
monomer diffusion rate and tums off the polymerization. The surface properties of the
PMAA brushes were characterized by their interaction with cytochrome c, a typical redox
protein with a surface covered with positively charged lysine residues at neutral pH. The
rapid adsorption of cytochrome c on PMAA was monitored using a quartz crystal

microbalance.
V5_$\/\/\/\/\/\I_
O O\—\
O

Vii—Mo
o \—\
/\3_5/\/\/\/\/\)‘°/—/

N—E—CHz-S-iO-CHcha 1

N—C—CHg-CI 2
8

22

(a) (b)

 

 

 

 

 

 

 

 

 

12‘-
10‘-
.:'~ 70)
'm 2 8-
.2. J“
«- '2
.53 x 5‘
x 0.
tr“ 0.-
4h
2-
2 4 6 8 10
'q D”

Figure 1.1. Rate of methacrylic acid polymerization initiated by UV irradiation of a
xanthate system as a function of (a) fraction of xanthate in the monolayer and (b) pH.
Reprinted with permission from Macromolecules 1996, 29, 3681-3685. Copyright 1996
American Chemical Society.

Prucker and Riihe reported surface-initiated free radical polymerization of styrene
from a self-assembled monolayer of azo initiators covalently bound to high surface area
silica gels.“ Scheme 1.10 shows the polymerization reaction and the scheme for
detachment of the polymer chains from the silica gel. The initiators have three important
functionalities: (1) an azo group that generates free radicals upon heating or UV
irradiation, (2) a chlorosilane that allows the initiator to be anchored to the surface
through reaction with silanol groups of the silica substrate, and (3) an ester that can be

hydrolyzed to detach the polymer brushes from the surface. After polymerization of

23

styrene, the ester bonds that connected the polystyrene (PS) to the surface were cleaved
and the molecular weights of the polymers were determined to allow comparison between
free radical polymerization in solution and at a surface. Irradiation of selected areas of
surfaces allowed preparation of rrricrostructured thin polymer layers by surface-initiated
free radical polymerization.62 Velten et al. attached cation-bearing peroxides to mica
surfaces via ion exchange. Initiation of styrene polymerization from the initiators gave
the surface-bound polymers.63 When treated with a poor solvent (methanol), the

polystyrene chains appeared as distinct molecular droplets in SEM images.

Prucker and Rtihe also investigated the kinetics and mechanism of surface-
initiated free radical polymerization from a monolayer of azo initiators attached to the
surface of silica particles.64 The rate of decomposition of the surface-immobilized
initiator was monitored by DSC and by a volumetric method that quantifies the amount of
nitrogen generated during decomposition. Dilatometry was used to follow the kinetics of
polymer brush formation. The polymer chains were detached from the surface and the
molecular weights of the degrafted polymers and their distribution were studied as a

function of the reaction parameters during polymerization.

24

Scheme 1.10. Synthesis of polystyrene brushes on silica and cleavage of the polymers
from the surface

CN
Me Me

+ Cl—é|i—/\/O M «Ni/Me
Me 0 e N

 

 

CN
Me O MeMe
O—Si—/\/ if
I Me
Me N
toluene/styrene
heat, -Nz
CN
Me
SW0 +
. I Me
Me 0 norbonded
L potystyrene
n
p-TsOH
MeOH/toluene
CN
Me O
OH Me/
0 §|i——/\/ + 0 Me

Me

 

25

Wittmer et al. predicted strong differences between polymer brushes grown from
surfaces and polymers generated in solution.65 They assumed that long chains are more
efficient at adding monomers than short chains because they are more mobile and easily
accessible to monomers, and thus polymer brushes formed at the surface should have a
higher polydispersity compared to the same reaction occurring in solution. In fact, the
PDI of the detached polymer brushes prepared by Prucker and Riihe ranged from 1.5 to 2,
which is close to the PDI of free radical polymerizations in solution. Consequently, it can
be concluded that surface immobilization does not cause excessive broadening of the
molecular weight distribution. Figure 1.2 shows the molecular weights and

polydispersities of the detached polymer brushes.

Minko et al. used in situ ellipsometry to study the kinetics of the surface-initiated
free radical polymerization of styrene from a silica surface with either immobilized azo-
initiators or a physisorbed hydroperoxide macroinitiator.66 The kinetics of the grafting
process could be described by clasSical treatments of free radical polymerization, taking
into account the termination of growing chain ends by disproportionation and coupling,
or by chain transfer reactions. Such behavior resulted in a linear dependence of the
polymerization rate on the surface concentration of the initiator and an inverse square
root dependence on the initiator concentration in solution. The amount of grafted polymer
approached a constant value with extended polymerization time, with the saturation value
dependent on the polymerization rate. Using the same polymerization technique, they

prepared grafted polymer coatings of styrene and MMA on titanium dioxide powder.67

26

(a)

 

 

 

 

 

 

 

 

 

 

 

 

175 .
L
:- 150 r
1—
g ’ 1
a 125 ~ .
fl 1 1
$2
._:: 100 r 1
2 I ‘
75 .
50 r r . r a r I r A r A r a r
0 2 4 6 8 10 12 14
time[h]
(b) (C)
10 .fifi O O 4 2.0~ U
’61 ‘ I 2I: ’ D .
Q 8 _ 1 1.9” d
n , g- , .
2 3 P J Q 1.81- d
(0 fl .
O. ’ a i
,2 4 - .9 1.7LL
as 1 '0
>~
8.. 2 . - E 1.6-
I 1
I 1
(Q 0 P ‘ 1.5
OTTA4 6‘810A12‘14 o 2 4 6 8101214
time [11) time [11]

Figure 1.2. Time-dependent properties of polymer chains grown by surface-initiated free
radical polymerization of styrene: (a) molecular weight Mn, (b) grafting densities 8(PS),
and (c) polydispersity of the covalently attached polymers. Reprinted with permission
from Macromolecules 1998, 31, 602-613. Copyright 1998 American Chemical Society.

27

Prucker and Riihe’s approach was extended to the formation of block copolymers.
A poly(e-caprolactone) macroinitiator containing azo groups was physisorbed on a
silicon oxide surface. Initiation of polymerization from the macroinitiator gave the block
copolymer.68 The principal advantage of this approach is that a simple physisorbed

macroinitiator system allows the creation of hydrophobic layers on hydrophilic surfaces.

Despite the success of Prucker and Riihe’s approach, traditional free radical
polymerization cannot provide well-defined polymer structures or block copolymer
brushes. Hawker et al. used 2,2,6,6-tetramethylpiperidinyloxy (TEMPO) (Scheme 1.11a)
as a persistent radical to achieve living free radical polymerization from a surface. The
initiator contains a benzyl radical trapped with TEMPO and a site for linking the initiator
to SiOz surfaces. Hawker was able to grow well-defined polymer brushes with a PDI of
1.14 and controlled molecular weight.‘59 However, it was necessary to add predetermined
amounts of free alkoxyamine initiator to the reaction mixture to control polymer grth
from the surface bound initiators. The free initiator establishes a concentration of
nitroxide radicals in the polymerization mixture that is high enough to control chain
growth from initiators on the surface and in solution. The calculated grafting density was

~200 A2 per polymer chain for molecular weights > 20,000 g/mol.

Besides nitroxide-mediated polymerization, ATRP and polymerization with
iniferters have been used to carry out “living” radical polymerizations from surfaces. The
term, iniferter originates from initiator, chain transfer agent, and termination species
because iniferters act in all three roles. Using the photoiniferter technique, de Boer et al.
synthesized polymer films grafted from iniferter monolayers on glass or silicon wafers

(Scheme 1.11b).70 The ability to control the thickness of the grafted polymer and prepare

28

well-defined diblock copolymer brushes affirmed the living character of the process. The

use of ATRP for surface-initiated polymerization is described in detail in the next section.

Scheme 1.11. Surface-immobilized initiators: (a) TEMPO-based and (b) iniferter

 

 

 

 

_ i— ——Si-—
2. a
I l l / /(B /A
I 8102 surface I [ Glass or silicon-water I
(a) (b)

II-3. Surface-initiated ATRP

Ejaz et al. deposited Langmuir-Blodgett (LB) films of ATRP initiators onto
oxidized silicon substrates (Figure 1.3)." The initiator, 2-(4-chlorosulfonylphenyl)ethyl
trimethoxysilane (CT S) was heated to attach the initiator to the substrate, and then the
polymerization was started by the addition of MMA and a Cu(I) catalyst. The
polymerization was not controlled, but the addition of p-toluenesulfonyl chloride as a free

initiator resulted in a well-controlled polymerization. The role of the free initiator is to

29

increase the Cu(II) concentration, thus increasing the rate of the deactivation process.
AFM showed the formation of a homogeneous polymer layer on the substrate. von Weme
and Patten immobilized initiators on silica nanoparticles and showed that structurally
well-defined polymer-nanoparticle hybrids could be prepared by surface-initiated
ATRP.72 The addition of a small amount of free initiator to the polymerization gave
control over the molecular weight, but the addition of 5-15 mol% of deactivator (a Cu(II)
complex) failed to give controlled polymerization. This suggests that significant amounts

of a Cu(II) complex are needed to establish equilibrium in the radical generation step.

Zhao reported the first synthesis of diblock copolymer brushes, PS-b-PMMA, by
sequential carbocationic polymerization and ATRP.73 Scheme 1.12 shows the synthetic
pathway for growing grafted diblock copolymer from silicate surfaces. The authors stated
that the addition of free initiator during ATRP was necessary to ensure a sufficient
concentration of deactivating Cu(II) species, otherwise the polymerization was not

controlled.

30

' Initiating group for A'I'RP

0:
S read H drol sis
1-12 ,8' on the water surface 1
I‘
at "term "11.1w
H co
fined TOCH‘

CTS Monolayer formation

Compression

11111111111111"m

| Condensation

 

/ LB Trough

Deposition

Fixation
”-

 

   

Figure 1.3. Schematic illustration of ATRP initiator immobilization on Si by the LB
technique. Reprinted with permission from Macromolecules 1998, 31 , 5934-5936.
Copyright 1998 American Chemical Society.

31

Scheme 1.12. Synthesis of a polystyrene-b-poly(methyl methacrylate) brush by
sequential carbocationic polymerization and ATRP (Reprinted with permission from J.

Am. Chem. Soc. 1999, 121, 3557-3558. Copyright 1999 American Chemical Society)

   

E; 1'
g V ; initiator 9 CH,
2 ' anchoring O- S.‘ 0CD3
g 0 (1‘13
8 .L
styrene
T104
-78°C. c1120,
DtBP, 50 min

CH:
“CM—immio
CH3 G"

 

 

N I methyl methacrylate
‘ N’V x/‘ N CuBr, anisole
' I 90°C, 6 11
CH, ('33:
'O-I-Iwz-wi-ICH2-C ..
n m
“‘3 c0201.

In another study, styrene and methyl acrylate were polymerized from an initiator
layer of 2-bromoisobutyrate immobilized on silicon wafers (3). The polymerizations
showed a linear increase in the polymer thickness with reaction time.74 Controlled
growth was achieved by the addition of a deactivating Cu(H) species, but more
importantly, control was achieved without a sacrificial initiator in solution. Shah et al.

also used surface-initiated ATRP of acrylates and methacrylates to prepare a variety of

polymer brushes.75
iii ll THa
1; O—Si—(CHzm—O—C—O—Br 3
'0') Cl-b

ATRP is reported to be very tolerant of functional groups, impurities, and water.
Water accelerates ATRP and allows polymerization to take place at room temperature.
Huang et al. synthesized crosslinked polymer films on gold surface76 and Jones et al.
grew poly(glycidyl methacrylate) (PGMA) brushes and PMMA-b-PHEMA copolymer
brushes from Si surfaces.77 Both groups used aqueous ATRP to obtain controlled growth
of polymer films from surfaces. Wirth reported that the surface topography of
polyacrylamide films grown by ATRP is quite different from films grown from surface-
irmnobilized azo initiator by free radical polymerization.78 Figure 1.4 shows AFM
images of the polymer surface. The surface roughness was 0.5 nm for the film prepared
by ATRP and 3.1 nm for the film formed by conventional radical polymerization

methods.

33

    

Ourn Ourn

Figure 1.4. AFM images of polyacrylamide films prepared by (a) surface-initiated ATRP
and (b) surface-initiated free radical polymerization. Reprinted with permission from
Anal. Chem. 1998, 70, 4023-4029. Copyright 1998 American Chemical Society.

Surface initiated ATRP can be applied to curved as well as flat surfaces. Recently,
Ejaz et al. anchored 2 (4 L‘ ‘” ,lri.---,l) groups to a porous glass filter and
polymerized MMA from the surface.79 The polymer brushes were cleaved from the glass
by treatment with HF, and GPC measurements showed that Mn increased linearly with
monomer conversion, in good agreement with the theoretical value. The polydispersity

of the polymer brushes was low (1.1 S PDI S 1.2), confirming that graft polymerization

from curved surfaces also is controlled.

Surface-initiated ATRP provides several advantages over conventional radical
polymerizations including good control over the molecular weight of the polymers and
hence the thickness of polymer brushes, the ability to tailor surface properties through the
polymerization of a wide spectrum of commercially available vinyl monomers, tolerance
to water and impurities, and the option of carrying out polymerizations at relatively low

temperatures with active catalytic systems.

34

II-4. Surface-initiated Cationic Polymerizations

Ulman et al. reported surface-initiated cationic polymerization of 2-oxazolines as
a novel and efficient approach to immobilization of poly(2-ethyl-2-oxazolines).80 Figure
1.5 outlines the reaction pathway for the preparation of these poly(N-
propionylethylenimine) (PPEI) brushes. The polymer chains were grown from triflates
immobilized on SAM, and after several days, the chain ends were functionalized with a
dialkyl amine to terminate the polymerization and form amphiphilic polymer brushes.
The resulting PPEI film was uniform in thickness and was stable to exhaustive Soxhlet
extraction with a good solvent. The formation of the polymer layers was confirmed by

ellipsometry, contact angle measurements, and external reflectance FI‘IR.

Even though controlled radical polymerization methods can provide a broad range
of polymers, this approach is restricted to polyolefins that typically have flexible
backbones. Seery et al. used surface-initiated cationic polymerization of isocyanates at
room temperature to prepare polymer brushes on silica or gold nanoparticles (Scheme
1.13).81 The use of a nanoparticle system allows for characterization of brushes by NMR
spectroscopy and transmission FTIR because the particles have a high surface area to
volume ratio and are easily suspended in solution. The kinetics of the polymerization of
hexyl isocyanate from the surface of gold nanoparticles was studied by real time IR
spectroscopy.82 Surprisingly, the kinetic plots were consistent with a second-order
process after a short induction period, while analogous homogeneous polymerizations
showed the expected first-order dependence on monomer concentration.83 The exact
interpretation of the kinetic results and the observation of an induction time are not

currently understood.

35

GI OHOHOHOH
€sz
0
so CF N’ ‘0 €sz 6:]
+Au 0(502-CF3)2 6 2 3 I [ OTTK
———->
EtOH;r.t.;16 h 1 day; 0" C—) reflux
c= 10 W vaporphase
.a.s.s.§. R R
HUT A"
E E
o d j .
ten-nine!
car‘s 02* "749] _ kt) functionality
I on N Mum OTf H
l1 L._i czHg LAM“- C2Hs>j E polymer
—>
—> —>
7days ; 1 day; Au
CHCI,; reflux CHCI,; r. t.
R R

Figure 1.5. Surface-initiated cationic polymerization of 2-oxazolines. Reprinted with
permission from J. Am. Chem. Soc. 1998, 120, 243-247. Copyright 1998 American
Chemical Society.

Scheme 1.13. Cationic polymerization of hexyl isocyanate on gold nanoparticles

l CH3(CH2)rN=C=O

I'O 'ICI

.—S\/\/\/\/\/\o/ NAVii:

Cl

Cl

 

 

II-S. Surface-initiated Anionic Polymerizations

Jordan et al. reported surface-initiated anionic polymerization of styrene from a
SAM of 4’-bromo-4-mercaptobiphenyl on a gold surface.84 The bromo-functionalized
SAM was metalated with sec-BuLi to generate 4’-lithio-4-mercaptobiphenyl, the initiator
for the anionic polymerization (Scheme 1.14). The grafting density for this polystyrene
brush was 320—360 Azlchain. Based on ~20 A2 per organic molecule in the SAM, one out
of 16—18 biphenyl thiols initiates the growth of a polymer chain from the surface.
Polarized infrared spectroscopy was used to study the conformation of the polymer

chains. Since only vertical components of vibrating dipoles are detectable, reflection

37

absorption IR spectroscopy is especially sensitive to the chain orientation with respect to
the reflecting surface. The spectrum of the polystyrene brush was compared to that of
bulk film, and differences in the ratios of vCH2(as)/VCH2(s) in the spectra of bulk films
and polymer brushes indicate that the average orientation of the polymer brushes is
different from that of chains in bulk films, and that the polymer brushes are considerably

elongated due to the high grafting density.

Qingye et al. also reported styrene polymerization from surfaces. They prepared
1,1-diphenylethylene (DPE) derivatives with quartemary ammonium tethers, and then
immobilized the initiator by cation exchange (Scheme 1.15).85 There was a linear
relationship between the monomer concentration and the Mn of the cleaved polymers,

which is consistent with a living anionic polymerization mechanism.

Scheme 1.14. Surface-initiated anionic polymerization of styrene on gold

ooo ooo ooo

L Si02 $102 $102

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

38

Scheme 15. Surface-initiated anionic polymerization of styrene on clay

(CHa)12—0

 

11-6. Surface-initiated Ring-opening Metathesis Polymerization

As shown in Scheme 1.16, Buchmeiser et al. prepared surface-grafted polymer
supports by surface-initiated ring-opening metathesis polymerization (ROMP) of
norbomene-based chiral monomers containing L-valine and L—phenylalanine.86 The
chiral monomers could be polymerized at room temperature using molybdenum and
ruthenium initiators, and the thickness of the films could be varied over a wide range (up
to l urn).87 The film thickness increased with monomer concentration, but the polymer
grth stopped after ~30 min, presumably due to deactivation of the immobilized

catalyst.

Scheme 1.16. Surface-initiated ROMP of norbomene-based chiral monomers

1. initiator

Woman
2.

 

 
 

Mb

initiator: MO(N'2,6'M62'CQH3)(CHCM92Ph)(OCM6(CF3)2)2
Cl2Ru(=CHPh-p-F)(PCy3)2

 

40

11-7. Electropolymerization and Plasma Polymerization on Surfaces

Crispin et al. grafted poly(acrylonitrile) and poly(ethyl acrylate) on transition

88 Because this method yields a very stable

metal surfaces by electropolymerization.
polymer/metal interface, the deposited polymer film adheres to the metal surface even
after washing with solvent. The strong adhesion was attributed to the formation of metal—
carbon chemical bonds. Cohn et al. used the plasma polymerization of
tetrafluoroethylene to generate a highly cross-linked fluorinated polymer layer on
poly(ethylene terephthalate) (PET).89 Exposing the fluoropolymer film to an ammonia

plasma introduced amines on the fluoropolymer surface, which were then used as

anchoring sites for further derivatization via diisocyanate spacers. (Figure 1.6).

Plasma “0' PEG

1
II
It
I

I
-——N— III—(CH2)g—N=C=O
H H

Substrate
TF E Plasma

o o
l I
—N—d—N—(CH2)g—N—tlr—o—(CHz—CHz—O)fiH

H H H
_L_J

HDI: Hexamethylene diisocyanate

 

 

 

Figure 1.6. Sequential surface derivatization of PET films.

41

II-8. Applications of Surface-initiated Polymerization

One of the most common applications of surface-initiated polymerizations is the
formation of nano-pattemed surfaces by soft lithography techniques that combine
microcontact printing (uCP) and graft polymerization. Hawker et al. combined
photolithography with nitroxide-mediated “living" free radical polymerization to yield
patterned polymer brushes with well-defined hydrophobic and hydrophilic domains
(Figure 1.7).“)0 They extended this concept to surface-initiated polymerization of tert-
butyl acrylate (PTBA). Patterned polymer layers have also been prepared by aqueous

ATRP.77

(a) (b)

200 um features Water droplet

   
     

PAA brush PTBA brush

 

Figure 1.7. Optical micrographs of patterned surfaces: (a) 10-p.m features in a continuous
polymer brush showing regions of poly(tert-butyl acrylate) (dark) and poly(acrylic acid)
(light) and (b) interaction of a water droplet with ZOO-um features showing an unusual
wetting profile and preferential interaction with poly(acrylic acid) brush domains.
Reprinted with permission from J. Am. Chem. Soc. 2000, 122, 1844-1845. Copyright
2000 American Chemical Society.

42

Shah et al. reported the use of surface-initiated ATRP to amplify patterned
initiator layers on gold films. PMMA, PHEMA, PTBA, and poly(dimethylaminoethyl
methacrylate) (PDMAEMA) were grown from spatially patterned initiators and then the
pattern was transferred into the substrates by using the brushes as barriers to wet
chemical etching of gold (Scheme 1.17).75 The same group also demonstrated that
patterned polymer brushes can be prepared by the surface-initiated ROP of e-

caprolactone from nricrocontact printed gold surfaces.75

Scheme 1.17. Surface patterning by microcontact printing, surface-initiated ATRP, and
etching (Reprinted with permission from Macromolecules 2000, 33, 597-605. Copyright

2000 American Chemical Society)

 

 

 

 

 

 

 

 

CH3(CH2)153H l Polymerization by ATRP at 60 °C
Au (100 nm)
TI (8 nm);‘r:—_._g?ss_—Wfi pMMA
l pCP

Selective Etching of Gold
Y: O I j
l (BrC(CH3)2COO(CH2)1oS)2 M
Chemisorbed at 60 °C

 

 

 

43

Surface-initiated ring-opening metathesis polymerization also is capable of
generating patterned polymer films on silicon both in the plane and normal to the surface
(Scheme 1.18). The combination of uCP and surface-initiated ROMP enables one to
control polymer composition and thickness in both lateral and vertical directions at a
molecular level (Figure 1.8), and offers new possibilities for fabricating surface features
in microelectronic devices.”91 The lateral resolution of the surface features was 2 pm,
which is limited by the uCP technique. The thickness of the polymer fihns was controlled

by monomer concentration and reaction time, and ranged from 5—100 nm.

Scheme 1.18. Surface-initiated ROMP from SiOZISi surface

.. tbs... *6 2:530

g 038' Icy. g
.1.

0— S SI(OE1)3 O—IS
OI [Ru] : O
“i” ' “I“ i(OEl)a n

 

 

 

Polyme OTS
v r v (e)

(a) I'_—_I::.'°’

 

 

l a. b. and o
> 21th
10.11-m A I.— _
Wllnl _ 2 m
d
1031:: i _
Polymer 0T8 M00 nun
v

 

 

III 013 e [Ruleetalyst
! nun-sols {- MM

 

 

 

 

Figure 1.8. Surface patterning by surface-initiated ROMP: (a) optical micrograph of 2-
um-wide lines of a patterned poly[norbomenyl-Si(OEt)3] film generated by
polymerization from the surface before (thickness ~10 nm), (b) optical micrograph of the
same sample after reactive ion etching (RIE) with SF6 for 3 min at 30 W in a parallel-
plate etcher, (c) SEM micrograph of the sample before RIB [same sample as shown in
(a)], (d) SEM micrograph of the sample after RIE, and (e) schematic representation of
patterned polymer formation by uCP, activation, ring-opening metathesis polymerization
on the surface of Si02/Si, and RIE. Reprinted with permission from Appl. Phys. Lett.
1999, 75, 4201-4203. Copyright 1999 American Institute of Physics.

45

Peng et al. synthesized poly(methacrylate) brushes with side chains of
phenylbenzoate mesogens on silicon oxide surfaces and the polymer backbone and the
mesogen units are connected through a flexible spacer. The thickness of these side-chain
liquid crystalline polymer (LCP) brushes (up to 200 nm) could be controlled by monomer
concentration and reaction time.” The optical texture of segregated polymer brushes
exhibits a strong memory effect. Heating the brush or exposing it to solvent renders it
isotropic, and upon cooling or removal of solvent, the identical texture reappears as it
returns to the nematic state. Utilizing the LCP brushes, it was possible to induce a certain
orientation of the nematic director in an adjacent bulk LC phase. This orientation could
be tilted because there may be a competition between the orienting action of the stretched

polymer chains and the uncovered fraction of the substrate (Figure 1.9).

Wirth et al. used surface-initiated ATRP to grow 10 nm thick films of
polyacrylamide on the surface of porous silica gel (86 nm pore size, 36 m2/g surface
area).78’93 Size-exclusion chromatography of proteins using the modified silica as the
stationary phase confirmed that the silica pores remained open after polymerization; four
proteins, thyroglobulin (669 kDa), ovalbunrin (44 kDa), ribonuclease A (13.7 kDa) and
aprotinin (6.5 kDa) eluted in order of decreasing size, in accordance with a size-exclusion
mechanism. Electrophoresis capillaries were coated with linear or cross-linked
polyacrylamide by surface-initiated ATRP.94 Using free radical polymerization to coat
capillaries with cross-linked polyacrylamide films is impractical because of concurrent
polymerization of acrylarnide in solution. The modified capillaries were used to separate
a mixture of three proteins, cytochrome c, lysozyme, and ribonuclease A by

electrophoresis, and could be used for more than 100 runs without clogging.

46

 

 

Figure 1.9. Schematic depiction of an LC brush swollen in a low molecular weight
nematic. Competing orienting actions from the brush and the bare surface may result in a
tilted alignment. The chemical structure of the mesogen used in this study is indicated in
the inset. Reprinted with permission from Macromolecules 1999, 32, 6759-6766.
Copyright 1999 American Chemical Society.

47

There are examples of similar grafting techniques applied to different substrates.
For example, Tsubokawa et al. used a variety of surface grafting techniques to tailor the
surface properties of carbon thin films using radical, cationic, and anionic graft
polymerization,” and Piletsky et al. utilized surface photograft polymerization and

molecular imprinting techniques to introduce specific binding sites into porous

membranes.96

111. References
(1) Wang, J .-S.; Matyjaszewski, K. J. Am. Chem. Soc. 1995, 117, 5614-5615.

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

(3) Matyjaszewski, K.; Xia, J. Chem. Rev. 2001, 101, 2921-2990.

(4) Matyjaszewski, K.; Mueller, A. H. E. Polym. Prepr. (Am. Chem. Soc, Div.
Polym. Chem.) 1997, 38, 6-9.

(5) Curran, D. P. Synthesis 1988, 489-513.

(6) Bellus, D. Pure Appl. Chem. 1985, 57, 1827-1838.

(7) Fischer, H. Macromolecules 1997 , 30, 5666-5672.

(8) Fischer, H. J. Am. Chem. Soc. 1986, 108, 3925-3927.

(9) Matyjaszewski, K.; Kajiwara, A. Macromolecules 1998, 31, 548-550.

(10) Matyjaszewski, K.; Nakagawa, Y.; Jasieczek, C. B. Macromolecules 1998, 31,
1535-1541.

48

(11)

(12)

(13)

(14)

(15)

(16)

(17)

(18)

(19)

(20)

(21)

(22)

(23)

(24)

(25)

Matyjaszewski, K.; Goebelt, B.; Paik, H.-j.; Horwitz, C. P. Macromolecules
2001, 34, 430-440.

Davis, K. A.; Paik, H.-j.; Matyjaszewski, K. Macromolecules 1999, 32, 1767-
1776.

Matyjaszewski, K. Macromolecules 1999, 32, 9051-9053.
Percec, V.; Barboiu, B. Macromolecules 1995, 28, 7970-7972.

Haddleton, D. M.; Jasieczek, C. B.; Hannon, M. 1.; Shooter, A. J.
Macromolecules 1997 , 30, 2190-2193.

Ando, T.; Kato, M.; Karnigaito, M.; Sawamoto, M. Macromolecules 1996, 29,
1070-1072.

Granel, C.; Dubois, P.; Jerome, K.; Teyssie, P. Macromolecules 1996, 29, 8576-
8582.

Uegaki, H.; Kotani, Y.; Karnigaito, M.; Sawamoto, M. Macromolecules 1997,
30, 2249-2253.

Ando, T.; Kamigaito, M.; Sawamoto, M. Macromolecules 1997, 30, 4507-4510.
Matyjaszewski, K.; Patten, T. E.; Xia, J. J. Am. Chem. Soc. 1997, 119, 674-680.

Moineau, G.; Granel, C.; Dubois, P.; Jerome, R.; Teyssie, P. Macromolecules
1998, 31, 542-544.

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

Shen, Y.; Zhu, S.; Zeng, F.; Pelton, R. H. Macromol. Chem. Phys. 2000, 201,
1169-1175.

Kerr, J. A. Chem. Rev. 1966, 66, 465-500.

Matyjaszewski, K.; Woodworth, B. E. Macromolecules 1998, 31 , 4718-4723.

49

(26)

(27)

(28)

(29)

(30)

(31)

(32)

(33)

(34)

(35)

(36)

(37)

(38)

(39)

Kitagawa, S.; Munakata, M. Inorg. Chem. 1981, 20, 2261-2267.

Shipp, D. A.; Wang, J.-L.; Grimaud, T.; Matyjaszewski, K. Polym. Prepr. (Am.
Chem. Soc., Div. Polym. Chem.) 1998, 39, 504-505.

Matyjaszewski, K.; Shipp, D. A.; Wang, J .-L.; Grimaud, T.; Patten, T. E.
Macromolecules 1998, 31 , 6836-6840.

Matyjaszewski, K. ACS Symp. Ser. 1998, 685, 258-283.

Healy, P. C.; Pakawatchai, C.; White, A. H. J. Chem. Soc., Dalton Trans. 1985,
253 1-2539.

Pallenberg, A. J .; Koenig, K. S.; Bamhart, D. M. Inorg. Chem. 1995, 34, 2833-
2840.

Munakata, M.; Kitagawa, S.; Asahara, A.; Masuda, H. Bull. Chem. Soc. Jpn.
1987, 60, 1927-1929.

Pintauer, T.; Qiu, J.; Kickelbick, G.; Matyjaszewski, K. Inorg. Chem. 2001, 40,
2818-2824.

Goto, A.; Fukuda, T. Macromol. Rapid Commun. 1999, 20, 633-636.

Matyjaszewski, K.; Paik, H.-j.; Zhou, R; Diamanti, S. J. Macromolecules 2001,
34, 5125-5131.

Xia, J .; Matyjaszewski, K. Macromolecules 1997, 30, 7697-7700.
Wang, J .-S.; Matyjaszewski, K. Macromolecules 1995, 28, 7901-7910.
Xia, J .; Gaynor, S. G.; Matyjaszewski, K. Macromolecules 1998, 31 , 5958-5959.

Wang, J.-L.; Grimaud, T.; Matyjaszewski, K. Macromolecules 1997, 30, 6507-
6512.

50

(40) Haddleton, D. M.; Kukulj, D.; Duncalf, D. J .; Heming, A. M.; Shooter, A. J.
Macromolecules 1998, 31, 5201-5205.

(41) Haddleton, D. M.; Crossman, M. C.; Dana, B. H.; Duncalf, D. J .; Heming, A. M.;
Kukulj, D.; Shooter, A. J. Macromolecules 1999, 32, 2110-2119.

(42) Patten, T. E.; Xia, J .; Abernathy, T.; Matyjaszewski, K. Science 1996, 272, 866-
868.

(43) Patten, T. E.; Xia, J.; Abernathy, T.; Matyjaszewski, K. Polym. Prepr. (Am.
Chem. Soc., Div. Polym. Chem.) 1996, 37, 575-576.

(44) Bernhardt, P. V. J. Am. Chem. Soc. 1997, 119, 771-774.
(45) Xia, J.; Matyjaszewski, K. Macromolecules 1999, 32, 2434-2437.

(46) Carmichael, A. J .; Haddleton, D. M.; Bon, S. A. F.; Seddon, K. R. Chem.
Commun. 2000, 1237-1238.

(47) Wang, X. S.; Armes, S. P. Macromolecules 2000, 33, 6640-6647.
(48) Matyjaszewski, K. Macromolecules 1998, 31 , 4710-4717.
(49) Pyun, J .; Matyjaszewski, K. Macromolecules 2000, 33, 217-220.

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

(51) Kickelbick, G.; Paik, H.-j.; Matyjaszewski, K. Macromolecules 1999, 32, 2941-
2947.

(52) Shen, Y.; Zhu, S.; Zeng, F.; Pelton, R. Macromol. Chem. Phys. 2000, 201 , 1387-
1394.

(53) Shen, Y.; Zhu, S.; Pelton, R. Macromol. Rapid Commun. 2000, 21, 956-959.

(54) Guan, 2.; Smart, B. Macromolecules 2000, 33, 6904-6906.

51

(55)

(56)

(57)

(58)

(59)

1268.

(60)

(61)

(62)

(63)

(64)

(65)

(66)

(67)

(68)

(69)

De Gennes, P. G. Macromolecules 1980, 13, 1069-1075.
Milner, S. T. Science 1991, 251, 905-914.

Dan, N.; Tirrell, M. Macromolecules 1993, 26, 6467-6473.
Bridger, K.; Vincent, B. Eur. Polym. J. 1980, 16, 1017-1021.

Hertler, W. R.; Sogah, D. Y.; Boettcher, F. P. Macromolecules 1990, 23, 1264-

Niwa, M.; Date, M.; Higashi, N. Macromolecules 1996, 29, 3681-3685.
Prucker, 0.; Riihe, J. Macromolecules 1998, 31, 592-601.

Prucker, 0.; Schimmel, M.; Tovar, G.; Knoll, W.; Riihe, J. Adv. Mater.
1998, 10, 1073-1077.

Velten, U.; Tossati, S.; Shelden, R. A.; Caseri, W. R.; Suter, U. W.; Hermann, R.;
Mueller, M. Langmuir 1999, 15, 6940-6945.

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

Wittmer, J. P.; Cates, M. E.; Johner, A.; Turner, M. S. Europhys. Lett. 1996, 33,
397-402.

Minko, S.; Sidorenko, A.; Starnm, M.; Gafijchuk, G.; Senkovsky, V.; Voronov,
S. Macromolecules 1999, 32, 4532-4538.

Sidorenko, A.; Minko, S.; Gafijchuk, G.; Voronov, S. Macromolecules 1999, 32,
4539-4543.

Stohr, T.; Ruhe, J. Macromolecules 2000, 33 , 4501-4511.

Husseman, M.; Malmstrcim, E. E.; McNamara, M.; Mate, M.; Mecerreyes, D.;
Benoit, D. G.; Hedrick, J. L.; Mansky, P.; Huang, E.; Russell, T. P.; Hawker, C.

52

(70)

(71)

(72)

(73)

(74)

(75)

(76)

(77)

(78)

(79)

(30)

(81)

(82)

(83)

J. Macromolecules 1999, 32, 1424-1431.

de Boer, B.; Simon, H. K.; Werts, M. P. L.; van der Vegte, E. W.; Hadziioannou,
G. Macromolecules 2000, 33, 349-356.

Ejaz, M.; Yamamoto, S.; Ohno, K.; Tsujii, Y.; Fukuda, T. Macromolecules 1998,
31, 5934-5936.

von Weme, T.; Patten, T. E. J. Am. Chem. Soc. 2001, 123, 7497-7505.
Zhao, B.; Brittain, W. J. J. Am. Chem. Soc. 1999, 121, 3557-3558.

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

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

Huang, W.; Baker, G. L.; Bruening, M. L. Angew. Chem, Int. Ed. 2001, 40,
1510-15 12.

Jones, D. M.; Huck, W. T. S. Adv. Mater. 2001, 13, 1256-1259.
Huang, K.; Doneski, L. J .; Wirth, M. J. CHEMTECH 1998, 28, 19-25.
Ejaz, M.; Tsujii, Y.; Fukuda, T. Polymer 2001, 42, 6811-6815.
Jordan, R.; Ulman, A. J. Am. Chem. Soc. 1998, 120, 243-247.

Seery, T. A. P.; Dhar, R; Huber, D. L.; Vatansever, F. Polym. Prepr. (Am. Chem.
Soc., Div. Polym. Chem.) 1999, 40, 148-149.

Huber, D. L.; Seery, T. A. P. Polym. Prepr. (Am. Chem. Soc., Div. Polym.
Chem.) 1999, 40, 709-710.

Patten, T. E.; Novak, B. M. Macromolecules 1993, 26, 436-439.

53

(84)

(85)

(86)

(37)

(38)

(89)

(90)

(91)

(92)

(93)

(94)

(95)

(96)

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

Zhou, Q.; Fan, X.; Xia, C.; Mays, J .; Advincula, R. Chem. Mater. 2001, 13,
2465-2467.

Buchmeiser, M. R.; Sinner, F.; Mupa, M.; Wurst, K. Macromolecules 2000, 33,
32-39.

Kim, N. Y.; Jeon, N. L.; Choi,1. S.; Takami, 8.; Harada, Y.; Finnie, K. R.;
Girolami, G. S.; Nuzzo, R. G.; Whitesides, G. M.; Laibinis, P. E.
Macromolecules 2000, 33, 2793-2795

Crispin, X.; Lazzaroni, R.; Geskin, V.; Baute, N.; Dubois, P.; Jerome, R.; Bredas,
J. L. J. Am. Chem. Soc. 1999, 121, 176-187.

Cohn, D.; Stern, T. Macromolecules 2000, 33, 137-142.

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

Jeon, N. L.; Choi, I. S.; Whitesides, G. M.; Kim, N. Y.; Laibinis, P. E.; Harada,

Y.; Finnie, K. R.; Girolami, G. S.; Nuzzo, R. G. Appl. Phys. Lett. 1999, 75, 4201-

4203.

Peng, B.; Johannsmann, D.; Riihe, J. Macromolecules 1999, 32, 6759-6766.
Huang, X.; Wirth, M. J. Macromolecules 1999, 32, 1694-1696.

Huang, X.; Doneski, L. J .; Wirth, M. J. Anal. Chem. 1998, 70, 4023-4029.
Tsubokawa, N.; Ueno, H. J. Appl. Polym. Sci. 1995, 58, 1221-1227.

Piletsky, S. A.; Matuschewski, H.; Schedler, U.; Wilpert, A.; Piletska, E. V.;
Thiele, T. A.; Ulbricht, M. Macromolecules 2000, 33, 3092-3098.

54

Chapter 2

Surface-Initiated ATRP on Gold at Ambient Temperature

1. Introduction

This chapter describes the surface-initiated ATRP of MMA on gold surfaces at
room temperature. The polymerization temperature in this polymer brush synthesis is the
lowest reported for surface-initiated radical polymerizations, and it is the first report of
the direct measurement of the molecular weight of polymer brushes grown on a flat
surface. The molecular weight of the polymer brushes suggests a high grafting density of
PMA on gold surfaces. No polymer forms in solution, and a simple solvent rinse can

clean the polymer brushes on the surface.

Growth of polymer brushes is an attractive method to modify and control
interface properties."5 Strategies for attaching polymer brushes to surfaces include the

" grafting to" technique}5

tethering preformed polymer chains from solution onto a
surface, and the "grafting from" technique}8 polymerizing from surface-anchored
initiators. The latter results in a higher density of polymer brushes on a surface because
the “grafting to” technique eventually faces serious steric hindrance that prevents

incoming polymer chains from diffusing through the film to surface reaction sites.

Several research groups used radical,9 cationic,10 anionic,ll ring-opening,12 ring-
opening metathesisl3 and ATRP7'14 to grow polymer chains from a surface. Of these
approaches, ATRP is especially attractive because it is a living process that gives a low

polydispersity index (PDI). To prepare well-organized polymer brushes, the PDI (MW/Mn)

55

should be close to 1. Block copolymers can also be prepared15 "6 with ATRP by changing
monomer solutions and re-initiating at dormant sites. Utilizing ATRP for surface
polymerization is a very convenient way to avoid the occurrence of polymerization in
solution. This feature reduces purification of the polymer brushes to a simple washing
process. As shown in Figure 2.1, surface-initiated ATRP generates only surface-bound
radicals, while surface-initiated free radical polymerization produces surface-bound
radicals and radicals in solution. Polymerization in solution is problematic since the
polymer surface is easily contaminated by physisorbed polymer from solution. This
problem is magnified for cross-linkable monomers, as solution polymerization can lead to

gelation.

(a) Surface-initiated free radical polymerization

N2
_ hv
N—N‘\__ . e .
or A \—
(b) Surface-initiated ATRP
C-Br + Cu1-Br 4—_’ C' + Cu'I-Brz

 

Figure 2.1. Differences in the radical generation mechanism for surface-initiated free

radical polymerization and surface-initiated ATRP.

56

Building polymer layers on gold is interesting because gold surfaces are
chemically homogeneous, virtually free of contamination, easy to clean, and applicable to
a wide variety of analytical techniques for thin film characterization. Ill addition, SAMs
of (1)-functionalized thiols enable easy preparation of a variety of well-ordered initiator
monolayers. Research on surface-initiated polymerization has focused on silicon

12.17
(1

substrates, and studies on go] are rare. Since the Au-S bond is somewhat unstable

above 60 °C'7'20 or under UV irradiation,2| growing well-ordered polymer brushes from

thiols on gold is a challenge. Shah et al. showed that the thermal stability of the SAM
affects the overall thickness of polymer brushes.l7 To create a more thermally stable
SAM at high polymerization temperatures, they intentionally prepared the SAM at the
polymerization temperature. Polymerization at room temperature (RT) will thus be
beneficial for preparing grafted polymer layers from thiol monolayers on gold. Until
recently, controlled radical polymerizations at RT were not possible. However,
Matyjaszewski described the synthesis of polyacrylates by ATRP at RT using a copper
catalyst derived from CuBr and MegTREN.22 This highly active catalyst system was

adapted to the synthesis of polymer brushes at room temperature.

One challenge in surface-initiated polymerization is the characterization of
polymer chains grown on surfaces.”"6 Two research groups published direct
measurements of the molecular weight of polymer chains detached from a curved
surface."'9 To have measurable amounts of detached polymers, they grew polymer chains
on silica gel with a large surface area. As Jordan et al. pointed out, polymer growth on a
flat surface might be quite different from growing polymers on a curved surface, because

these sterically different systems might have different kinetics and different polymer

57

brush conformations.ll This chapter describes the application of RT ATRP to the growth
of polymer brushes from thiol monolayers on gold, and the direct measurement of the

molecular weight of polymer chains grown from a flat surface.

11. Surface-initiated ATRP with MegTREN

Scheme 2.1 outlines the synthetic pathway for the preparation of grafted PMMA
chains. 4 was prepared by immersing a gold-coated Si substrate in a 1 mM solution of 1—
mercaptoundecanol (MUD) for one day. The ellipsometric thickness of the MUD layer
was 12:1:1 A. To produce the anchored initiator (5), 4 was treated with 2-bromopropionyl
bromide (2-BPB) in the presence of triethylarnine. A concentrated 2-BPB solution (0.1
M) was used to achieve near-quantitative initiator immobilization. Since a thiol SAM
could be unstable in the presence of acid bromides, 4 was dipped in the acid bromide
solution for only 2 min. This step was carried out in a dry box because acid bromides are

moisture sensitive, especially in the presence of an organic base.

Initiator inunobilization is apparent from the appearance of a carbonyl peak at
1743 cm'1 in the reflectance FI'IR spectrum (Figure 2.2, spectrum b). To perform ATRP,

5 was immersed in MMA containing 0.1 mol % CuBr and 0.1 mol % MegTREN23 at 25
°C for 12 hours. Since ligated Cu(I) is sensitive to 02, all polymerizations were carried

out in a dry box filled with helium. After polymerization, the newly formed film was
washed with dry THE A large increase in the carbonyl peak, at 1745 cm"1 in the external
reflection FTIR spectrum confirmed the formation of PMMA brushes (Figure 2.2,

spectrum c). The ellipsometric thickness of the polymer film measured at 5 different

58

spots on a 75 cm2 wafer was 370:5 A showing that the polymer film is remarkably

uniform.

Scheme 2.1. Surface-initiated ATRP of MMA with MegTREN at RT

S\/\/\/\/\/\/OH 4
0
EtaN 11
dry THF 3’09“)”
Br
9 i1
S—(CHzltt—O-C-F—Br 5
C113
MMA
CtBr/MOGTREN
RT

R 9113 l
s—(CHaln—o-C-p ore—c Br 6
out cozcrta n

,/‘
\ N'—
OCH3 ’N
MMA MesTREN

59

 

0.05

 

 

 

 

 

 

 

0)
0
C
CB
9
8
.o b
< xes ( )
xes (a)
3500 3000 2500 2000 1 500 1 000

Wavenumber (cm")

Figure 2.2. Reflectance FTIR spectra of (a) a SAM of MUD (4), (b) initiator
immobilized on a MUD SAM (5), and (c) a grafted PMMA layer on the initiator surface

(6).

111. Control Experiment

To show that MMA polymerization only occurs on surfaces with immobilized

initiators, 7 was prepared by the immersion of 4 in a solution of 0.1 M propionyl chloride

and pyridine for 2 min. Thus, the only difference between the initiator-immobilized

surface (5) and the control surface (7) is the existence of the or-bromocarbonyl

functionality that acts as a radical initiator for ATRP.

re

Au

0
—s—(CHa)n—o—i':—CH.20tu 7

 

 

 

The evidence for the synthesis of 7 is again based on the appearance of an ester

carbonyl peak at 1746 cm'1 (Figure 2.3, spectrum b). A control polymerization was

canied out on 7 under the same polymerization conditions used with 5. Ellipsometry and

FTIR confirm that

no polymerization occurs on this surface. There was no change in the

thickness of the organic layer or in the 1R intensity of the carbonyl peak after the

polymerization (Figure 2.3, spectrum c).

61

 

 

 

 

 

 

W
I 0.001
8
g - L W (b)
‘9
O
(D
.0
<
Mix“ “(a) ‘ AWAW
3500 3000 2500 2000 1 500 1000

Wavenumber (cm")

Figure 2.3. Reflectance FTIR spectra of (a) a SAM of MUD (4), (b) propionyl group
anchored MUD SAM (7), and (c) after surface polymerization.

62

IV. Molecular Weight Determination

To determine the molecular weight and polydispersity of the grafted brushes, the
PMMA chains were detached from the surface using 12, and characterized by GPC
(Figure 2.4). For the 370 A-thick film, GPC yields an Mn of 44,500 and a PDI of 1.3. As

shown in Table 1, these polymerizations are very repeatable.

Table 2.1. Molecular weight of PMA brushes

 

 

 

 

PMMA film thickness (A) M, PDI (MW/Mn)
331 33,100 1.3
370 44,500 1.3
402 68,900 1.5

 

 

 

 

 

* The thickness of PMA films was measured by ellipsometry, and the
number average molecular weights (Mn) of PMA chains are relative
to PMMA standard samples.

Since the detached PMMA brushes should exist as disulfides,24 the S-S bonds
were cleaved with NaBH425 and the GPC measurements were repeated. The same Mn
value and PDI were obtained, which implies that the disulfides (RIS-SRZ), if formed, are
likely to be unsymmetrical disulfides formed from the PMMA chain and a MUD
molecule that was buried by the growing PMMA brushes during polymerization. Thus,
Mn only depends on the length of the PMMA brush since the molecular weight of the
short chain thiolate is negligible compared to that of the PMMA brush. Based on the film

thickness and Mn, ~10% of the surface-bound initiators actually initiate PMMA chains,

63

and thus each PMMA brush is surrounded by 9 short chain thiolates. Thus it is natural

that a polymer brush desorbs as an unsymmetrical disulfide.

Using the Mn value from GPC analysis, the average cross-sectional area was
calculated. For the calculation, Ax=M/ptNA was used, where Ax is the average cross-
sectional area of polymer chain, M is the molecular weight of the chain, p is the density
of PMA (1.1 g/cm3), t is the polymer thickness, and NA is Avogadro’s number. The
average cross-sectional area calculated by this method is ~18O A2. This result suggests a

high grafting density, and is consistent with results of Shah et al.‘7

Using the polymer
that formed in solution as a proxy, they used Mn to calculate an average cross-sectional
area of ~200 Azlchain. AFM and ellipsometry show that these films are both
macroscopically and microscopically uniform. After grafting of PMA, the surface
roughness decreased from 1.92 nm (bare gold) to 0.54 nm (gold with grafted PMMA) as

shown in Figure 2.5.

V. Conclusions

There are several advantages in using room-temperature ATRP for “grafting
from” surfaces. First, the process is compatible with substrates that are sensitive to high
polymerization temperatures. Second, a simple washing step without Soxhlet extraction
gives clean polymeric thin films because polymer chains grow only on the surface. There
should be no polymerization in solution as long as there is no chain transfer from surface
propagating sites to solution. Thus, thermal polymerization will be negligible at RT.

NMR analysis showed no signs of polymerization in solution. This implies that the

monomer could be recovered from solution and reused. Perhaps the biggest advantage of
this technique is the high grafting density of polymer brushes, which results in films that

are uniform at the A level.

In conclusion, ambient temperature ATRP using MegTREN allows successful
synthesis of dense chemically bound PMMA brushes on gold surfaces. The polymer
films are uniform over large areas, and grow only on the surface without parallel

formation of polymer in solution.

B: short alkyl chains from MUD
or MMA oligomers

 

 

 

 

Gold Water
I 12/CH2CI2
__ _ 1st GPC
F PMMA I—s s R Analysis
I NaBH4

 

_ 2nd GPC
L PMMA I—SH + HS R Analysis

 

Figure 2.4. Detachment of polymer brushes from gold by 12 treatment followed by

molecular weight determination by GPC.

65

X/Y 0.2 um/div
Z 25.0 nm/div

E
C
0.
O
In
X/Y 0.2 um/div
2 25.0 nm/div

 

Figure 2.5. Tapping mode AFM images of (a) a bare gold surface and (b) PMMA

brushes on gold. Images in this dissertation are presented in color.

(1)

(2)

(3)

(4)

(5)

(6)

(7)

(8)

(9)

(10)

(11)

(12)

References
Milner, S. T. Science 1991, 251 , 905-914.
Huang, X.; Doneski, L. J.; Wirth, M. J. CHEMTECH 1998, 28, 19-25.

Bruening, M. L.; Zhou, Y.; Aguilar, G.; Agee, R.; Bergbreiter, D. E.; Crooks, R.
M. Langmuir 1997, 13, 770-778.

Sidorenko, A.; Minko, S.; Schenk-Meuser, K.; Duschner, H.; Starnm, M.
Langmuir 1999, 15, 8349-8355.

Prucker, O.; Naumann, C. A.; Riihe, J .; Knoll, W.; Frank, C. W. J. Am. Chem.
Soc. 1999, 121 , 8766-8770.

Bridger, K.; Vincent, B. Eur. Polym. J. 1980, 16, 1017-1021.

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.

Hemnann, W. A.; Stumpt, A. W.; Prierrneier, T.; Bogdanovic, S.; Dufaud, V.;
Basset, J .-M. Angew. Chem, Int. Ed. Engl. 1997, 35, 2803-2805.

Prucker, 0.; Riihe, J. Macromolecules 1998, 31 , 592-601.
Jordan, R.; Ulman, A. J. Am. Chem. Soc. 1998, 120, 243-247.

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

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

67

(13)

(14)

(15)

(16)

(17)

(18)

(19)

(20)

(21)

(22)

(23)

(24)

(25)

Jeon, N. L.; Choi, I. S.; Whitesides, G. M.; Kim, N. Y.; Laibinis, P. E.; Harada,
Y.; Finnie, K. R.; Girolami, G. S.; Nuzzo, R. G. Appl. Phys. Lett. 1999, 75, 4201-
4203.

Ejaz, M.; Yamamoto, S.; Ohno, K.; Tsujii, Y.; Fukuda, T. Macromolecules 1998,
31 , 5934-5936.

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

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

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

Bain, C. D.; Troughton, E. B.; Tao, Y. T.; Evall, J.; Whitesides, G. M.; Nuzzo, R.
G. J. Am. Chem. Soc. 1989, 111, 321-335.

Schlenoff, J. B.; Li, M.; Ly, H. J. Am. Chem. Soc. 1995, 117, 12528-12536.

Carnillone, N., 111; Chidsey, C. E. D.; Liu, G. Y.; Scoles, G. J. Chem. Phys. 1993,
98, 3503-3511.

de Boer, B.; Simon, H. K.; Werts, M. P. L.; van der Vegte, E. W.; Hadziioannou,
G. Macromolecules 2000, 33, 349-356.

Xia, J.; Gaynor, S. G.; Matyjaszewski, K. Macromolecules 1998, 31 , 5958-5959.
Ciarnpolini, M.; Nardi, N. Inorg. Chem. 1966, 5, 41-44.

Templeton, A. C.; Hostetler, M. J .; Kraft, C. T.; Murray, R. W. J. Am. Chem.
Soc. 1998, 120, 1906-1911.

Stahl, C. R.; Siggia, S. Anal. Chem. 1957, 29, 154-155.

68

Chapter 3

Kinetic Study of Surface-initiated Atom Transfer Radical

Polymerization from Flat Surfaces

I. Introduction

This chapter describes elucidation of the termination mechanism of surface-
initiated atom transfer radical polymerization (ATRP). To demonstrate the dependence of
kinetics on radical concentration, polymer brushes were synthesized at different catalyst
concentrations. The surface-initiated ATRP of methyl methacrylate (MMA) with no
addition of Cu(11)Br2 showed rapid initial growth followed by early termination, whereas
polymerization with initial addition of Cu(II)Br2 and use of a CuCl catalyst showed less
deviation from ideal, constant growth. For a given reaction time, there is a specific
catalyst concentration that yields maximum film thickness. Reduction in film thickness
when stirring the solution suggests that stirring increases chain mobility and enhance
termination by recombination. Solution phase ATRP of MA was also tested to study

early stages of polymerization kinetics and the evolution of molecular weight.

Because Atom Transfer Radical Polymerization (ATRP) normally gives polymers
with narrow molecular weight distributions, it has become one of the most popular
polymerization methods for vinyl monomers such as acrylates, methacrylates, and
styrene.1 ATRP also offers additional advantages when used to grow polymers from
surfaces.24 The thickness of grafted polymer brushes is easily controlled by varying

reaction time, and in the absence of added free initiator and chain transfer, polymer

69

chains grow only from the surface. This reduces the purification procedure to a simple
rinsing of the polymer surface.5 In some cases, surface-initiated ATRP can proceed even
at ambient temperature in both organic6 or aqueous media.7'8 ATRP is a controlled/living
process, and thus a surface-confined polymerization can be quenched and re-initiated to
synthesize multi-block copolymer brushes.9 Polymer thin films tethered through covalent
bonds can be prepared by surface-initiated polymerizations, and the resulting thin
polymer films can be used to modify and control surface properties.10 Tethered polymer
films have many potential applications including anticorrosion layers, chemical
separations, control of surface wetting and morphology, and as a route to nanostructured

surfaces.l "16

In surface-initiated radical polymerizations, there have been some reports of
radical termination. The kinetics and mechanism of surface-initiated free radical
polymerization were investigated by Prucker and Rtihe.18 They confirmed that the
immobilization of the azo initiator does not alter initiator properties significantly because
the observed initial rates of polymerization are in close agreement with those in the
polymerization of bulk styrene with 2,2'-azobisisobutyronitrile (AIBN). They also noted
that the main differences between surface and solution polymerizations are mostly due to
changes in termination reactions. The concentration of growing radicals on a surface
increases with elevated temperature, resulting in an increased probability of termination
of growing radicals by recombination or disproportionation because the radicals on a
surface are located in close vicinity. For the same polymerization system, Minko et al.

predicted that termination between surface free radicals can occur when initiation is

70

rapid.19 Very soon after the start of polymerization, however, termination will decrease

greatly as the concentration of surface free radicals decreases.

Matyjaszewski et al. simulated the kinetics of surface-initiated polymerization,S
and concluded that initiator coverage (0) is a major factor in defining whether the growth
in layer thickness depends linearly on reaction time. High densities of surface-
immobilized initiators (0 > 0.5) lead to a nonlinear dependence of conversion on time and
a much broader distribution in chain length. They predicted that distributions might
become even broader for systems with additional bimolecular activation/ deactivation
equilibria (ATRP, for example) because the bimolecular processes require diffusion of
catalysts and monomer to the chain ends. Niwa et al. also observed that surface
polymerization depends on initiator coverage.20 Polymerization hardly proceeded when
the surface coverage of a photoactive xanthate was high (0 z 1). Rp showed the smallest
value at 0 = l, and with decreasing 0, Rp gradually increased and gave a maximum value
at 0 = 0.2. These results are good indications that a high radical concentration on two-

dimensional surfaces results in extensive termination.

Plots of polymer film thickness as a function of polymerization time are similar
for surface-initiated free radical polymerization (FRP) and ATRP.21 The similarity
between ATRP and FRP kinetics suggests that termination of growing chain ends. Since
the presence of two growing radicals next to each other or in close proximity is a matter
of probability, it might be impossible to avoid termination by recombination of two
radicals. Interestingly, Prucker and Rtihe showed that in thermally induced surface
radical polymerization the film thickness decreases as polymerization temperature

increases. To understand these observations, the authors noted that the film thickness

71

depends on both the number of polymer chains attached to surface and the molecular
weight of these chains. They suggested that too high concentration of radicals on surface
at high temperature can not yield maximum thickness because termination reaction

becomes serious during polymerization.

To date, most research on surface-initiated ATRP focused on the synthesis,
characterization, and application of grafted polymer layers, while systematic kinetic
studies are rare. Because polymerization kinetics depend on the concentrations of
catalytic system, temperature, solvent polarity, and mixing/diffusion effects,
polymerization from a two-dimensional surface and solution polymerizations could show
quite different kinetic behavior. Even though there have been suggestions in the literature
that termination reactions significantly impact on polymer brush growth, the termination
mechanism of surface polymerization is not well understood. Here, we report kinetic data
for surface-initiated ATRP of methyl methacrylate (MMA) and methyl acrylate (MA) to

help elucidate the dominant termination mechanism in surface polymerizations.

II. Dependence of Polymer Film Growth on Radical Concentration

Scheme 3.1 shows the preparation of PMA brushes grafted on gold. The self-
assembled monolayer initiator, 1 was prepared using a literature procedure.6 To perform
ATRP, l was immersed in a solution of MMA or MA at 23 °C containing the copper
catalyst, prepared in situ from the appropriate copper halides and ligands. The formation

of polymer brushes is apparent from the increase in the ellipsometric thickness of the

72

organic layer and the appearance of large carbonyl peaks at 1743—1745 cm’1 in the

reflectance FI‘IR spectrum.

Scheme 3.1. Surface-initiated ATRP of (meth)acrylates on gold

 

o
Au —S—(CHZ)1,—o—t'i-CH—er 1

Ha

 

 

 

MA or MMA
CuCl / CuBrg / 2Me5TREN
THF-Cl—bCN / 23°C

0 R
Au —s—(CI-12)n—o—<':'—CH orig—d Br 2
. this dozen-l, n

PMMA, when R = CH;
PMA, when R = H

 

 

 

 

ATRP (Scheme 3.2) is described as a “living/controlled” radical polymerization
because the irreversible termination reactions that consume radicals are suppressed. In
the absence of termination, one should observe a linear growth in the thickness of
polymer films during surface-initiated ATRP. Shown in Figure 3.1a is the evolution of
thickness during the surface-initiated ATRP of MMA using CuBr/MegTREN as the

catalyst and without added CuBrz.

73

Scheme 3.2. Kinetic scheme for ATRP: pathways for radical generation and consumption

 

 

kp
ka +M
activation
Pn—Br + Cu(I)X - Pn' + Cu(l|)X2
deactivation
k9 Pm' termination

kt

Pn— Pm

Contrary to expectation, the data show a rapid initial growth followed by a decline
in the grth rate and cessation of film growth by 10 hr. Similar kinetic data were
reported by Ejaz et al. for the surface-initiated ATRP of MMA from SAMs of
immobilized p-toluenesulfonyl chloride,22 and Jones and Huck reported that surface-
initiated ATRP of acrylates and methacrylates slowed down and lost their controlled
character at long reaction times.‘6 In their study of the surface-initiated ATRP of HEMA
in water, they also observed rapid growth at early stages of the polymerization and
termination after about 60 min. This behavior can be explained by the loss of active
catalyst, or through the loss of surface initiating sites, most likely through the

recombination of two growing radicals.

74

400

 

 
 
 
 

‘2’

Thickness (A)
8

 

   

o with copper(ll) bromide
A without copper(ll) bromide

 

 

 

 

 

I l l I l I I I I I l I I

0 60 120 180 240 300 360 420 480 540 600 660 720
Reaction Time (min)

Figure 3.1. Dependence of PMA film growth on the presence of Cu(II)Br2: (a) without
initial addition of Cu(II)Br2 and (b) with initial addition of Cu(II)Br2

Two publications related to surface-initiated ATRP display plots of grafted
polymer thickness vs polymerization time. ATRP of a sugar-carrying methacrylate using
CuBr/4,4'-di-n-heptyl-2,2'-bipyridine catalyst showed curvature2 similar to that of Figure
3.1, while ATRP of methyl acrylate using CuBr-CuBr;/N,N,N',N",N"-
pentarnethyldiethylenetriarnine (PMDETA) catalysts showed a linear growth of polymer
thickness until ~6 hr.5 The curvature seen in the first case is consistent with some portion
of surface-initiators being lost during polymerization by termination reactions, while the
linear growth observed in the second example may reflect choices of monomer structure,

catalytic system, and reaction media, that minimize the radical concentration and

75

recombination reactions. According to Wittmer’s suggestion for surface-initiated
polymerization, longer chains compete more effectively for access to monomers than
shorter chains, resulting in a wider molecular weight distribution, compared to the same
polymerization in solution.23 However, the polydispersities of polymer chains detached
from surfaces,3’6’9’l7 did not show the expected wide molecular weight distribution. Thus,
termination of growing radicals seems to be the major factor that determines kinetics of

surface polymerization and molecular weight distribution of final polymers.

To confirm the loss of surface initiating sites and rule out loss of catalyst activity,
5 samples were collected at various stages during a surface polymerization. The samples
were rinsed with THF, dried under a stream of nitrogen, and the film thickness for each
sample was measured by ellipsometry (filled triangles, Figure 3.2). Each sample was
then immersed in a solution of monomer with fresh catalyst. For this “regrowth” process
the concentration of the catalyst was reduced and CuBr; was added to promote thicker
film growth. After 4 hours, the samples were removed from the reaction vials and
immediately washed with THF. As shown in Figure 3.2a' (hollow triangles), the
thickness added (equal to the length of the arrows) during the regrowth process was
largest for the 30 min sample, and decreased as the contact time with the catalyst
increased. Since the change in film thickness is a proxy for the number of surface-bound
initiators, the data are consistent with a continuous loss of surface-bound initiators

through radical disproportionation or coupling during ATRP.

Termination by disproportionation generates a saturated chain end and a polymer
chain end that contains a double bond. Disproportionation was excluded as a termination

mechanism since no hands were detected by surface reflectance FT 1R near 1620—1670

76

cm'l that could be attributed to C=C stretching. The loss of the double bonds via
polymerization also was discounted since that process should lead to cross-linking, and
only soluble products were obtained. As described in Chapter 2, we successfully used
iodine to detach polymer brushes from similar surfaces, which demonstrates that there is
no significant cross-linking between polymer chains. However, iodine treatment failed to
detach cross-linked polymer films of poly(ethyleneglycol dimethacrylate) and PHEMA

layers. Thus the data are consistent with termination by coupling.

To test whether the observed kinetic behavior is specific to surface
polymerization, we carried out a comparable solution phase ATRP at room temperature
as a control. As shown in the first order kinetic plot (Figure 3.3a) the solution
polymerization using Cu(I)Br/MegTREN as the catalyst shows qualitatively the same
trends seen in the surface polymerizations, indicating that termination also is a problem in
solution polymerization. This kinetic plot resembles simulations of solution ATRP that

include termination,27 but the experimental plot shows less curvature.

77

120

 

 

 

 

 

 

 

 

f A additional polymerization for 4 hr at LG
100 - A polymerization at HC
- A
'3 80 1
U)
9 s
.5 60
o
E A
" 40 - i T (a’)
A A 4 (a)
A
20 -
O I I I I I I I I

 

 

0 1 2 3 4 5 6 7 8 9
Reaction Time (hr)

Figure 3.2. Dependence of additional PMA film growth on the amounts of remaining
surface-initiating sites: (a) represents the same polymerization shown in Figure 3.4 and
(a’) the reused samples after polymerization for 4 hr at [MA] 2 M, [CuCl/MegTREN] 40
mM, [CuBrZ/Me6TREN] 12 mM. The thickness of the PMA film was measured by
ellipsometry at two different spots on a sample and averaged. The thickness deviations

for the samples are within 12 A

78

 

0.20
0.18 - (a) .
0.16 -
0.14 -

H .

0.12 -

0/[M

- 0.10 - .

In([M

0.08 ~
0.06 1 o
0.04 -
0.02 -

 

 

 

0.00 F r .
0 100 200 300 400 500
Reaction Time (min)

I r

 

12000 1 .35

(D)
10000 . 1 1.30

r 1.25

A A - 1.20
' A——>

- 1.15

D
PDI (WM)

3
I

~1.10

Molecular Weight (Mn)
0,
>
O

i

- 1.05

 

 

 

O T r T T W T I 1 .00
2 4 6 8 10 12 14 16 18
Conversion (°/o)

Figure 3.3. Solution phase ATRP results: (a) first order kinetic plot of MMA
polymerization and (b) molecular weight and distribution of polymers.

79

The data in Figure 3.31) show a near-linear increase in molecular weight and a
PDI of ~ 1.20. This suggestion of control must be tempered with the non-zero y-intercept
for the Mn data, which is consistent with poor control in the 0—6 % conversion range.
The higher than expected molecular weight must reflect an inefficient deactivation step,
and substantial termination. The high activity of the Cu(I)Br/MegTREN system24 enables
room temperature ATRP, but complete control over the polymerization has not yet been
achieved. Since avoiding termination might not be possible, minimizing termination

could be a more realistic goal.

In ATRP, control over the polymer architecture depends on the fast reversible
generation of radicals. Thus, a suitable concentration of a soluble Cu(II)Br2 deactivator
is essential. Most homogeneous ATRPs do not require the addition of Cu(II)Br2 to the
polymerization since the concentration of the deactivator automatically builds up at early
stages in the polymerization. In surface-initiated ATRP, however, the absolute amount of
initiator on the surface is small, and thus the concentration of Cu(Il)Br2 formed in
solution will be negligible. Once radicals are generated on the surface, they cannot easily
revert to their dormant state and eventually they will terminate by coupling.
Exacerbating coupling is the confinement of the surface initiators to a layer, which
ensures that radicals on the surface will be in close proximity. Thus addition of Cu(II)Br2
at the beginning of the reaction is indispensable for reversible radical generation, and

suppression of radical termination reactions.

To try to achieve a more linear grth rate, Cu(I)Br was replaced by Cu(1)Cl and
Cu(II)Br2 was added at the beginning of the polymerization. Mixed halide systems (R-

Br/CuCl) effectively slow the polymerization rate and reduce termination25 by

80

maintaining a lower concentration of growing radicals. The resulting polymerization
kinetics showed less deviation from linear growth (Figure 3.1b), but the curvature

indicated that termination was still significant.

In principal, a sufficiently high km“ should reduce the steady state concentration
of radicals to the point where bimolecular termination processes are unlikely. At the
same time, however, the rate of polymerization may be reduced to impractical levels.
Thus, for a given polymerization time, there should be an optimum set of reaction
conditions that lead to the thickest polymer brush. Shown in Figures 3.4 and 3.5 are the
results from a series of surface initiated ATRPs of MA run at different catalyst
concentrations. In each polymerization, the ratio of the Cu(I) catalytst and the Cu(II)C12
deactivator was held constant. At the highest catalyst concentration (plot a in Figure
3.4), there was almost no growth of PMA, but the thickness of the polymer film increased
with decreases in catalyst concentration. At even lower catalyst concentrations, (plots (1
and e), the rate of film growth slowed considerably but showed relatively linear growth
over an 8 hour period. These data are consistent with chain termination via coupling
playing an important role in defining the film growth profile. Given the two-dimensional
nature of surface-bound initiators, one can envision that high concentrations of catalyst
lead to a spatially dense concentration of surface-tethered radicals that recombine easily.
Thus, [R0] is generally not constant and polymer film growth fails to follow first order

kinetics.

81

40

 

35-

30-

25-

Thickness (A)

20-

15

 

10

(d)
/: (a)

(e)

 

600

I I I

2 3 4

Reaction Time (hr)

T

5

6

 

500-

.1;
8

Thickness (A)
Q)
8

 

  

 

 

(C)

(b)

(a)

 

N- Ir
(D
#- If

82

5
Reaction Time (hr)

at...)

 

 

Figure 3.4. Growth of PMA layer thickness vs polymerization time: (a) [MA] 2 M,
[CUCI/MCéTREN] 40 mM, [CuBrleegTREN] 12 mM; (1)) [MA] 2 M,
[CuCl/MegTREN] 2 mM, [CuBr2Me6TREN] 0.6 mM; (c) [MA] 2 M, [CuCl/MegTREN]
0.1 mM, [CuBrz/MegTREN] 0.03 mM; ((1) [MA] 2 M, [CuCl/MegTREN] 5x10"3 mM,
[CuBr2Me6TREN] 1.5x10’3 mM; (e) [MA] 2 M, [CuCl/MegTREN] 2.5x10'4 mM,
[CuBr2/Me6TREN] 7.5x10’5 mM. Surface-initiated ATRP was done without stirring in a
(1:1) mixture of CH3CN-THF at 23 °C, and in drybox. The thickness of the PMA film
was measured by ellipsometry at two different spots on a sample and averaged. The

thickness deviations for the samples are within :l:2 A.

To validate the idea that polymerizations from surface-immobilized initiators can
be optimized to yield maximum grth of polymer film for a specific reaction time, we
plot in Figure 3.5 data from polymerizations run for 4 hours at various catalyst
concentrations. The observance of an optimum catalyst concentration is reminiscent of
data for the UV-initiated polymerization of styrene from surface-bound azo initiators.21
In that case, the optimum light intensity represents a balance between propagation rate

and the irreversible termination of radicals.

A computer simulation was designed to model the temporal variation in the
surface radical concentration and the observed polymer brush growth rate. A suitable
model would enable accurate prediction of the Cu (1) and Cu (H) concentrations needed
to attain the desired rate of polymerization and the final polymer brush thickness. The
growth process was divided into short time intervals and the concentration of active chain

ends, [R0] was calculated for that interval. We assumed that the cumulative concentration

83

of active chain ends, [R0] is proportional to the final thickness of the polymer brushes, i.e.

thC monomer concentration IS constant.

 

 
 
    
    
 

 

 

 

 

500
(g) conc. of CuCl
e (a): 40 mM
400 ‘ (b): 2 mM
(c): 0.1 mM
.3? (d): 5x10-3 mM
7:300 - (e): 2.5x10-4mM
8
C
6
1?: 200 - b)
'—
100 -
(d
(e) Curve: for eye guide a)
O r l r l
0 1 2 3 4 5
Log [CuCl] (umol)

Figure 3.5. Optimum catalyst concentration that yields the thickest PMA film after 4 hr
of polymerization. (a)—(e) 30 mol% of CuBrz relative to CuCl was used for each

polymerization.

Based on the ATRP kinetic scheme depicted in Scheme 3.2, the change in the
radical concentration with time, d[R°]/dt (equation 3) can be described by three terms, the
rate of radical generation via the reaction of Cu(I) with initiator, and the loss of radicals

through deactivation and bimolecular termination.

84

d[R-]

dt = ka[RBr][CuX] — kd [R-][Cux,] - k, [R-]2 (3)

Different values of ka, kg, and k126 were tested in the simulation, with k8 = 0.8
Umol-sec, kd = 1x105 Umol-sec, and k, = 1x109 Umol-sec giving the best fit to the
experimental data. The Cu(I)/Cu(11) concentrations used in the simulation match those
used in the experiments (runs a-e of Figure 3.4). The simulation results are plotted in

Figure 3.6.

 

1 .2e-7
(c)
1.0e-7 -
8.0e-8 .

6.06-8 1

4.08-8 ‘ (b)

((1)
206-8 1 /
(a)

V __ (e)
0.0 . - - - . . .

0 1 2 3 4 5 6 7 8 9
Reaction Time (hr)

 

Normalized Cumulative Radical Conc.

 

 

Figure 3.6. Simulation of surface-tethered radical concentrations as a function of catalyst

concentration: the same values of copper concentrations as in Figure 3.4 were used for

(a). (b), (C), (d) and (e)-

85

The simulations capture the general trends seen in the experimental data, although
they show some deviation from the experimental data. For example, runs a, b, and c all
have the same initial growth rate, while the simulation of run c yields a lower initial
slope. Also, run (1 crosses run a earlier in the simulation than is seen in the experimental
data. The most likely explanation is that the simulation fails to capture all of the details of
the surface polymerization. In particular, the model does not address the 2-dimensional
nature of the surface-bound initiator layer at the beginning of polymer growth, and its
evolution toward 3-dimensionality during polymerization. As shown in Figure 3.7, the
chain ends are in close proximity in surface polymerizations, enhancing the likelihood of
termination. In addition, the rate constants for radicals tethered to surfaces may deviate
from those measured in solution. The simulations also show that the linear growth seen
for runs d and 6 does not correspond to a controlled polymerization. These

polymerizations are better described as steady state systems where [R'] is constant.

 

 

High concentration of Low concentration of
radicals radicals
coupling

 

Figure 3.7. Influence of radical concentrations on the probability of recombination.

86

III. Dependence of Polymer Film Growth on Stirring

In solution ATRP, a homogeneous solution of the deactivating complex is critical
for control over the polymerization, and the use of a polar co-solvent to dissolve Cu(H)
complexes and/or stirring is commonly used to ensure homogeneity during the
polymerization. In contrast, polymerizations from surfaces are inherently heterogeneous
since polymer growth is limited to a thin layer near the surface. It is plausible that access
to monomer, catalyst, or deactivating agent may be diffusion-limited and the kinetics of
surface-initiated polymerizations may be different from homogeneous reactions in
solution. Figure 3.8 shows two polymerizations designed to test for such effects. The
evolution of the thickness of PMA films with time are shown for polymerizations run
without (plot a) and with stirring (plot b). The polymerization run with stirring reached a
limiting thickness of ~200 A before 400 min, but a parallel reaction run in the absence of
stirring (Figure 3.8b) grew to nearly twice the thickness in the same time. For the stirring

experiments, we used a glass reactor shown in Figure 3.9.

The simulation results would suggest that the limiting thickness for the stirred
reaction reflects more termination by coupling than the non-stirred reaction. One could
argue that stirring increases the mobility of chain ends and improves the probability that
two radicals to be located near each other, which results in early termination by
recombination. An alternative mechanism is the formation of a Cu concentration
gradient. If the Cu(II) formed by the activation process remains localized near the
radical, then the rate of the deactivation process would be enhanced and the probability

for termination by coupling would decrease.

87

 

 

 

 

400
. (a)
300 1
A O
'S
8
8 200 1 O o O O (b)
x O
.2 O
.c O O
l— O
O o O
100 1 0°
0
O
O t I l I I T
0 100 200 300 400 500 600

Reaction Time (min)

Figure 3.8. Growth of PMMA layer thickness vs polymerization time: (a) polymerization
without stirring; (b) polymerization with stirring. Polymerization conditions for surface-
initiated ATRP: [MMA] 2 M, [CuBr/MegTREN] 2 mM, [CuBr2/2anbpy] 0.6 mM, in
solvent mixture of CH3CN-THF (1:1), at 23 °C, and in drybox. The thickness of the
PMMA film was measured by ellipsometry at three different spots on a sample and

averaged. The thickness deviations for the samples without stirring and with stirring are

$4 A and :9 A respectively.

88

.-

--.A. ‘mfl—q
r
1

(a) Top View (b) Side View
2.5 cm . ' .
— m

  

Figure 3.9. Digital images of the glass reactor used for polymerization with stirring.

Images in this dissertation are presented in color.

IV. Conclusions

MMA polymerizations with stirring terminated after 6 hr, while the
polymerizations without stirring showed faster growth during the same time period. The
results might imply that stirring increases the mobility of chain ends and increases the
chance that two radicals near each other meet and recombine, leading to termination.
Polymerizations using high catalyst concentration (40 mM) showed almost no growth of
PMA, while the polymerization using low catalyst concentration (0.1 mM) resulted in a
thick PMA film. The results from polymerizations at different catalyst concentration
indicate that the longer the samples are kept at high concentration of catalyst, the more
initiators on surface are lost. Finally, it can be concluded that the major termination

mechanism of surface-initiated ATRP is recombination.

89

V.

(1)

(2)

(3)

(4)

(5)

(6)

(7)

(8)

(9)

(10)

(11)

(12)

References
Matyjaszewski, K.; Xia, J. Chem. Rev. 2001, 101, 2921-2990.
Ejaz, M.; Ohno, K.; Tsujii, Y.; Fukuda, T. Macromolecules 2000, 33, 2870-2874.

Husseman, M.; Malmstroem, 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. F.

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

 

v—W ‘ d- 0.. ‘I‘EL".

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

Huang, W. X.; Baker, G. L.; Bruening, M. L. Angew. Chem, Int. Ed. 2001, 40,
1510-1512.

Huang, W.; Kim, J.-B.; Bruening, M. L.; Baker, G. L. Macromolecules 2002, 35,
1 175-1 179.

Kim, J .-B.; Huang, W.; Bruening, M. L.; Baker, G. L. Macromolecules
Submitted.

Zhao, B.; Brittain, W. J. Prog. Polym. Sci. 2000, 25, 677-710.
Huang, X.; Doneski, L. J.; Wirth, M. J. CHEMTECH 1998, 28, 19-25.

Sedjo, R. A.; Mirous, B. K.; Brittain, W. J. Macromolecules 2000, 33, 1492-
1493.

90

(13)

(14)

(15)

(16)

(17)

(18)

(19)

(20)

(21)

(22)

(23)

(24)

(25)

(26)

Zhao, B.; Brittain, W. J .; Zhou, W.; Cheng, S. 2. D. Macromolecules 2000, 33,
8821-8827. ’

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

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

Jones, D. M.; Huck, W. T. S. Adv. Mater. 2001, 13, 1256-1259.
Prucker, 0.; Ruehe, J. Macromolecules 1998, 31 , 592-601.
Prucker, 0.; Ruehe, J. Macromolecules 1998, 31 , 602-613.

Minko, S.; Sidorenko, A.; Starnm, M.; Gafijchuk, G.; Senkovsky, V.; Voronov,
S. Macromolecules 1999, 32, 4532-4538.

Niwa, M.; Date, M.; Higashi, N. Macromolecules 1996, 29, 3681-3685.

Prucker, O.; Schimmel, M.; Tovar, G.; Knoll, W.; Ruehe, J. Adv. Mater. 1998,
10, 1073-1077.

Ejaz, M.; Yamamoto, S.; Ohno, K.; Tsujii, Y.; Fukuda, T. Macromolecules 1998,
31, 5934-5936.

Wittmer, J. P.; Cates, M. E.; Johner, A.; Turner, M. S. Europhys. Lett. 1996, 33,
397-402.

Queffelec, J .; Gaynor, S. G.; Matyjaszewski, K. Macromolecules 2000, 33, 8629-
8639.

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

Matyjaszewski, K.; Paik, H.-j.; Zhou, R; Diamanti, S. J. Macromolecules 2001,
34, 5125-5131.

91

 

“ ant-2.13m

 

 

Chapter 4

Synthesis of Triblock Copolymer Brushes by Surface—Initiated

Atom Transfer Radical Polymerization

I. Introduction

This chapter describes the synthesis and characterization of triblock copolymer
brushes. Surface-tethered triblock copolymers composed of poly(methyl acrylate),
poly(methyl methacrylate), and poly(2-hydroxyethyl methacrylate) were grown from
gold surfaces by a series of atom transfer radical polymerizations at ambient temperature.
GPC determination of the molecular weights of model triblock polymer brushes
demonstrate that this method can yield relatively homogeneous polymer brushes. After
the synthesis of each block, films were either exposed to a large excess of Cu(II)Br2 to
quench polymerization prior to washing and reinitiation, or were simply rinsed with
solvent. Comparison of the thicknesses of multiblock homopolymer films with those
prepared using a single initiation step shows that in the Cu(II) quenching approach, >95
% of the active chains support growth of an additional block. However, when simple
solvent rinsing was used after each block, only 85-90% of active chains were preserved

during the quenching step.

Block copolymers are especially attractive materials because of their ability to
phase separate and self assemble into spherical, rod-like, and lamellar geometries. When
deposited on surfaces, thin films of block copolymers can phase-segregate by the

selective adsorption of one block of the copolymer"2 or orient upon application of

92

 

 

external fields.3 Each of these techniques leads to formation of a pattern that can be used
as a template for further materials deposition. Recent examples of materials patterned in
this way include vertically oriented nanowires,4 SiOz posts on surfaces,5 and surface-

induced nanopattems of diblock copolymer films on rrrica.6

Tethering multiblock copolymers to surfaces is particularly interesting because it
provides responsive, controllable interfaces7 with nanoscale features.8 Brittain et al.
showed that copolymer brushes reversibly self-organize on exposure to different

solvents.9’l l

The topology of multiblock copolymer brushes also suggests their use in
fornring multilayered materials where the layers can be arranged in a predetermined
order. Growth of such films from tethered initiators complements other approaches to the
formation of layered interfaces, such as deposition of Langmuir-Blodgett films,‘2

13-15

zirconium phosphonates, and layered polyelectrolytes.16

The most common approaches for surface-tethered block copolymer synthesis are
(1) the preparation of polymeric macroinitiators that can be isolated and used later to
initiate the polymerization of a second block, and (2) the sequential addition of two or
more monomers to a polymerization. The first approach often employs different
polymerization mechanisms to form each block, while the second approach is restricted
to monomers that can be polymerized sequentially. Examples of the macroinitiator
approach include the synthesis of surface-tethered polystyrene-b-poly(methyl
methacrylate) (PS-b-PMMA), polystyrene-b-poly(methyl acrylate) (PS-b-PMA) and PS-
b-poly(N,N-dimethylaminoethyl methacrylate) by sequential carbocationic
polymerization and atom transfer radical polymerization (ATRP).9’”’18 PS-b-PMMA was

also prepared from surface-immobilized azo initiators by successive reverse-ATRP and

93

“I

 

 

ATRP.ll Using the second approach, surface-tethered PS-b-PMA and PS-b-poly(tert-
butyl acrylate) were synthesized by sequential ATRP,19 and surface-immobilized
alkoxyanrines were used for the preparation of PS-b-poly(styrene-co-methyl

methacrylate) brushes.20

This study extends the surface-initiated synthesis of block copolymers from
diblock to triblock systems using low-temperature ATRP. The major focus of this work is
determining the efficiency with which additional blocks can be added to chains growing
from a surface. When working with polymers composed of many blocks, this efficiency
becomes especially important as shown by significant thickness differences between
heptablock films prepared using two different strategies for radical quenching between

the deposition of blocks.

ATRP is very attractive for formation of block copolymers because radical
generation is “reversible” due to an equilibrium between the active
(radical/Cu(mBrzlligand) and dormant (initiator/Cu(I)Br/ligand) states. Scheme 4.1
shows the reversible generation and quenching of a radical in the ATRP of methyl
methacrylate (MMA). Ignoring chain transfer reactions, polymer radical I can either
undergo further polymerization by adding more monomer, convert to the dormant
macroinitiator II, or terminate through bimolecular coupling or disproportionation
reactions. Since the radical concentration in ATRP is very low,21 termination is usually
the least common scenario. One can intentionally favor the formation of the dormant
macroinitiator by adding a high concentration of Cu(II)Br2 to capture the radicals and

shift the equilibrium to the dormant state.11 This quenching should effectively stop

94

polymerization and keep growing chain ends alive for polymerization of subsequent

blocks.

This quenching and re-initiation (Q & R) approach is particularly powerful for the
preparation of surface-tethered block copolymers because the polymeric radical is
confined to the substrate, and the quenching step simply involves transferring substrates
to a concentrated Cu(II)Br2 solution that does not contain Cu(I)Br. Thus the length of the
polymer chain is controlled by the elapsed polymerization time before exposure to the
Cu(II)Br2 solution. After rinsing to remove residual Cu(11)Br2, the polymerization can be
re-initiated to extend the chains, or multiblock copolymers can be synthesized by simply
switching to a different monomer after each cycle. The main advantages of this approach
are the facile synthetic procedure, easy control over the thickness of each polymeric
layer, restriction of polymerization to surface-bound chains, and a low polymerization

temperature that is compatible with temperature-sensitive substrates.

95

—_

._-mr Lh‘ q

Scheme 4.1. Possible reaction pathways for a polymeric radical in ATRP of MMA

 

 

 

 

 

n 0 ,
<1 '12 9 F12 on, i I
R1—C—C—Br + Cu(I)Br __. Rl-c—c. 00001312 : I 1
Ha éHs '.
o R2 CH, CH3
1:
22: QIkSflSHS 1 = R‘-('i-' 0142—0 0142—0. Cu(l|)Br2
011. (502th m $020113
> Father Polymerization
> Termination
/ -Cu(l)Br 2
Reversible Quenching 9 l? 01"3 CHs
I .__ - II = Rl-c— orig—0 CHz—C—Br
Cu(I)Br (in, dozens M 002614,
Irreversible Quenching Re-initiation

> Block Copolymer Brushes

 

 

with excess Conner; Cu(I)Br

96

11. Synthesis of Tethered PMA-b-PMMA-b-PIIEMA Brushes on Gold

The synthetic route to tethered PMA-b-PMMA-b-PHEMA brushes is shown in
Scheme 4.2 and synthetic details appear in the Experimental section. To perform ATRP,
a self-assembled monolayer of initiator, 1, was immersed in a MA solution containing
Cu(I)Br/MegTREN and Cu(II)Br2/2(anbPY) at 25 °C. Although Cu(I)Br/N1e6TREN
complexes are active and allow ATRP to proceed at room temperature, the presence of
Cu(II)Br2 is indispensable for controlling polymerization. Because of the poor solubility22
of the Cu(II)Br2 complex in neat monomer, a mixture of CH3CN and TI-IF (1:1 vzv) was
used as a polar medium to obtain a homogeneous reaction system. These polar solvent
mixtures should not affect the polymerization because ATRP is very tolerant toward
functional groups such as ethers and nitriles.23 After 20 nrin, the MA polymerization was

quenched by transferring the substrate to a solution of Cu(II)Br2/2(anbpy) in CH3CN-

THF (1:1). Matyjaszewski er al. reported that the Cu(II)Br2 complex with anbpy'

successfully controls surface-initiated ATRP, acting as a deactivator and removing the

necessity for sacrificial initiators.19

The formation of PMA on the surface is apparent from the appearance of a
carbonyl peak at 1743 cm“1 in the reflectance FTR spectrum of these films (Figure 4.1,
spectrum a). After polymerization of the initial block and quenching, 2 was washed with
CHgCN-TI-IF, and polymerization was re-initiated at 25 °C by transferring the substrate
to a MMA solution containing fresh copper catalyst. Following the MMA polymerization
and quenching, the reflectance FTIR spectrum of the film contained a more intense
methyl peak at 2953 cm’1 and an increased carbonyl peak, confirming the formation of

the PMMA block (Figure 4.1, spectrum b).

97

no. run-r

97"”—

Scheme 4.2. Synthesis of PMA-b-PMMA-b-PHEMA triblock copolymer brushes

 

Au -—S—(CHQ)11—O—0—ICH—Br 1
Cl'b

1i) Initiation In MA
iii) Quenching

Au —S—(CH2)11—O-C—OC+ Hz—C—+r ‘2

1i) Re- initiation in MMA

 

 

 

 

 

 

 

 

 

iii) Quenching
.. I I 91
Au -—R CHz—C Cl-lg—C r 3
$020th ( I coacrh m

 

 

 

l i) Re-initiation in HEMA

 

 

 

 

 

 

 

ii) Quenching
.. I I a. I I 11 I
Au -—-R GHQ—C CHz—C Cl-Ig—C Br 4
002cm t L COZCHGJ m L COch-IECHZOHJ n
PMA PMMA PHEMA

o
a = —S——(CH2)11—O-C—Cl+—
Cl'b

98

.e: ”mart

Polymerization from 3 was then initiated in a HEMA solution at 40 °C. After the
HEMA polymerization, the substrate was simply rinsed with a series of solvents
including CH3CN-THF, DMF, EtOAc, EtOH, and deionized water. Since little, if any,
polymer formed in solution,24 a simple solvent rinse was sufficient to clean the
copolymer brushes on the surface. Evidence for the formation of the third block,
PHEMA, is the appearance of a broad hydroxy (OH) peak at 3200—3600 cm‘1 and a
further increase in the intensity of the carbonyl peak (Figure 4.1, spectrum c). The
ellipsometric thicknesses of the film after addition of each layer were: 2: 65:1 A, 3:
133i1 A, and 4: 27113 A. The thicknesses were measured at three different spots on the
same sample. Each succeeding polymer layer was purposely doubled in thickness by
controlling the polymerization time to provide clear changes in the IR spectrum after
each polymerization cycle. Formation of the last block required longer polymerization
times because fewer initiators were left on the surface (vide infra). Ellipsometry and
PT IR spectroscopy showed that no polymerization occurred from a propionylated MUD
SAM (no a-bromocarbonyl functionality) under the same polymerization and washing
conditions used for the synthesis of each block of the triblock copolymer brushes. This

result demonstrates that polymer grows only from surfaces with immobilized initiators.

99

 

0.02

 

 

[N (d)

 

 

k (b)

X4 , . (a)
I l i T

4000 3500 3000 2500 2000 1500 1000
Wavenumber (cm'1)

 

 

?

 

Figure 4.1. Reflectance FT IR spectra of (a) a layer of initiator immobilized on gold (1),
(b) a grafted PMA layer on the initiator surface (2), (c) a grafted PMA-b-PMMA diblock
copolymer (3), and (d) a grafted PMA-b-PMMA-b-PHEMA triblock copolymer (4).

100

As mentioned earlier, one of the most interesting properties of multiblock
copolymers is their ability to form patterned structures. AFM images of surface-tethered
PMA-b-PMMA-b—PHEMA films suggest formation of nanodomains whose properties
depend on their environment. Upon exposure to a nonpolar solvent, hexanes, these
nanodomains form separate islands (Figure 4.28). Presumably, chains in the hydrophilic
PHEMA block strongly self-assemble in a nonpolar medium to minimize surface energy,
while the relatively hydrophobic PMMA block moves to the subsurface (valley) region to
form discrete islands. Treatment with deionized water leads to relatively smooth surfaces
with pseudo spherical domains in contact with each other (Figure 4.2b). This probably

occurs because water is a better solvent for PHEMA than for PMMA.

A A 100.0 nm
- 100.0 nm

   

Figure 4.2. Tapping mode AFM images of PMA(58A)-b-PNMA(111A)-b—
PHEMA(75A) (a) after immersing in hexanes at ~60 °C for 30 min, and (b) after
immersing in deionized water at ~70 °C for 30 min. Images in this dissertation are

presented in color.

101

111. Molecular Weights of Triblock Homopolymer Brushes

To determine the molecular weight of polymer brushes, one must detach the
polymer chains from the substrate. Figure 2.4 shows the detachment of PMA brushes
from a gold surface. As previously shown for gold colloids, 12 is capable of breaking
Au—S bonds.25 After the iodine treatment, detachment of the polymer chains was
confirmed by reflectance FTIR and NMR. The FTIR spectrum of the Iz-treated surface
did not show any significant polymer peaks and the NMR of the collected polymer chains
from solution was identified as PMMA. Since detached polymer brushes should exist as
disulfides,25 we treated detached PMMA homopolymer brushes with NaBH4 to cleave the
S—S bonds.26 GPC determination of the Mn and PDI values for the PMMA homopolymer
brushes were the same before and after reduction with NaBH4,24 suggesting that the
disulfides are asymmetric (PMMA-S—S-R) and that the molecular weights of R groups
(MUD initiator fragments or oligomers) are negligible compared to those of the PMMA

chains.

Due to the insolubility of PHEMA?"28 PMMA-b-PMMA-b—PMMA (triblock
PMMA, which is in fact a homopolymer) was synthesized by the Q & R strategy to
confirm the formation of reasonably well-defined triblock chains. Comparison of the Mn
and molecular weight distribution of PMMA-b—PMMA-b—PMMA with those for a
PMMA brush of comparable thickness grown using a single initiation step provides
information about the efficiency of the Q & R cycles. If a significant fraction of initiating
sites were lost during each Q & R step, the molecular weight distribution of the triblock
brush would be broadened, resulting in a larger polydispersity index (PDI = Mw/Mn). The

average thickness of the three separate PMMA-b-PMMA-b-PMMA films was 381:3 A,

102

including 16 A for the initiator monolayer, and the thickness of each layer was about 120
A. After detaching the PMMA brushes from the surface using iodine?“25 the molecular
weight distribution was determined by GPC. The results (Mn = 37,400, PDI = 1.48) agree
reasonably well with the values for PMMA homopolymers grown by a single initiation
step without Q & R. The reported Mn and PDI values for a 370 A thick PMMA film
grown by a single initiation step are 44,500 and 1.30, respectively.24 The slightly lower
Mn and higher PDI of PMMA-b-PMMA-b-PMMA may suggest, however, that some
termination reactions occur during Q & R cycles. A second triblock polymer brush,
PMA-b-PMA-b-PMA, synthesized by the Q & R strategy and characterized in the same
way yielded Mn = 41,600 and PDI = 1.32 for 389i2 A thick films. The MD and low PDI
values from these experiments provide further confirmation that Q & R steps are capable

of forming triblock copolymers.

Table 4.1 shows Mn and PDI values as well as portions of the GPC
chromatograms for brushes detached from gold substrates. The table also includes two
control experiments using commercial PMMA standards: runs a and b confirm that
exposure of polymer chains to iodine should not affect Mn and PDI, while run c shows
that the detection limit for PMMA in the GPC system (RI detector, THF as solvent) is
<0.2 mg/mL. When the detached polymer brushes are dissolved in 150 uL THF, their
calculated concentration would be about 0.2 mg/mL, assuming that there is no loss during
the work-up procedure. The fact that the signal to noise ratio is a little higher for the
control sample than for detached brushes indicates that some portion of the polymer
brushes was probably lost during the ug-scale work-up and GPC sample injection. At

these low concentrations of polymer, lower molecular weight species may be difficult to

103

detect due to their minimal weight fraction, which might exclude some low molecular
weight polymer chains. Additionally, the baseline at longer retention times is difficult to
determine. Thus, we note that PDI values for the polymer brushes are probably
artificially low. Approximate baselines used for the GPC determinations are shown as
dotted lines in Table 4.1. Triblock PMMA (run (1) shows same baseline drift, while
triblock PMA (run e) has a relatively stable baseline. The use of polymerization
conditions that minimize termination (see below) may reduce the polydispersity of

polymer brushes.

IV. Quantification of the Efficiency of Q & R

To quantify the efficiency of the Q & R process, the thicknesses of PMA films
prepared using several quenching (Cu(mBrz) and re-initiation cycles were compared with
values for films prepared with a single initiation step. The hollow circles in Figure 4.3
show the evolution of PMMA film thickness as a function of polymerization time when
using a single initiation step. Films prepared with 1-3 Q & R cycles (triangles and
diamond) have nearly the same thickness/time relationship, suggesting that most of the

active polymerization sites were preserved during the quenching process.

Quantification of the quenching process requires a model of film growth because
the rate of termination of growing chains is significant compared to the rate of
polymerization from the surface. Loss of active chain ends during surface
polymerization is implied by the nonlinear relationship between film thickness and time
shown in Figure 4.3. This is consistent with solution ATRP using MeGTREN, which also

shows termination of growing chains.22

104

Table 4.1. GPC data for PMMA standards and polymers grown from surfaces using

 

 

 

 

 

 

 

 

ATRP
measured
polymer sample GPC peak Mn & PDI
PMMA standard
sample, (a)—(c)
(a) before iodine (a) (a) Mn = 59,200
treatment PDI = 1.07
(b) after iodine (b) Mn = 59,000
treatment PDI = 1.07
(c) at low conc. (c) Mn = 62,700
0.2 mg/mL PDI = 1.05
Minutes
((1) Triblock PMA ((1) Mn = 37,400
(381 A thick film) PDI = 1.48
(e) Triblock PMA (e) Mn = 41 ,600
(389 A thick film) PDI = 1.32
”.55 . .133 ,. 55.
Minutes
(f) PMMA by single (f) Mn = 44,500
initiation step ' ‘ PDI = 1.30
(370 A thick film) 11.40' iaeo' ‘ i580
Minutes

 

105

 

‘93- .‘.' v

*Footnotos for Table 4.1

“H” The manufacturer-specified Mn and PDI values of the PMMAstandard sample were
60,200 and 1.10, respectively. The standard was purchased from Polysciences, Inc.

0’) The PMMA standard sample was dissolved in 12/CH2C12 solution and kept at room
temperature for 15 hr. After removal of CHZCIZ and 12, the resulting PMMA was injected
into the GPC instrument.

(“M“) Triblock polymer brushes are, in fact, homopolymers synthesized by three cycles of
quenching and re-initiation. For GPC analysis, these were collected from three 2.7 cm2
substrates (total area: ~8 cmz).

“H” The low signal to noise ratio is due to the low concentration of polymer. These
peaks are magnified relative to a and b.

(0 The elution time is faster than other runs because two PLgel 100. mixed-B columns
were used without an additional PLgel 311 mixed-E column in series. The polymer
brushes were collected from ~75 cm2 surface area, yielding a relatively large amounts of
polymer chains and a higher signal to noise ratio than analysis of the other polymer

brushes.

Several other publications related to surface-initiated ATRP display plots of
grafted polymer thickness versus polymerization time. A similar plot for the ATRP of a
sugar-carrying methacrylate using a CuBr/4,4'-di-n-heptyl-2,2'-bipyridine catalyst
showed curvature29 similar to that in Figure 4.3, while ATRP of methyl acrylate using
CuBr/N,N,N',N",N"-pentamethyldiethylenetriamine (PMDETA) and CuBr2/PMDETA
showed a linear growth of polymer thickness with time.19 The linear growth observed in
the second example may reflect an optimization of monomer structure, catalytic system,
and reaction medium that minimizes termination by recombination. ATRP of ethylene
glycol dimethacrylate in aqueous media also showed a reasonably linear growth in

polymer thickness with time.30

106

400

 

 

     

 

 

 

 

 

 

 

300 4
3
8
8 200 -
x
.9
l5 0 without Q & R
A with 1 cycle of Q & R
100 a v with 2 cycles of Q & R
0 with 3 cycles of Q & Ft
a fit using equation 3
O I I T I l I I
O 50 1 00 1 50 200 250 300 350

Reaction Time (min)

Figure 4.3. Ellipsometric thickness of PMA films vs polymerization time. Thicknesses
obtained at three different spots on a sample were averaged, and the standard deviation of
these measurements was :l:2 A. Polymerizations were initially quenched at 15 min or 15
min and 75 min in the case of Cu(H) quenching. The line represents a fit to the thickness
of films prepared in a single reaction step (open circles) using the function, thickness =

16 + A ln(1+Bt) derived from equation 3.

107

To account for the loss of growing chains during polymerization from a surface, a
simple kinetic model for film growth was utilized. The loss of active chains, A, is
assumed to be second order with respect to surface concentration as represented by
equation 1. Note that A includes chain ends in both the dormant and radical states. The
increase in film thickness with time, dI/dt, should be first order with respect to the
number of active chains as shown in equation 2. Integration of equation 1 yields an
expression for the surface concentration of active chains at any given time, and insertion

of this expression into equation 2 followed by integration yields an expression for film

thickness as a function of time (equation 3), where [A0] is the initial concentration of
active initiator functionalities. To fit the data, 16 was added to the right side of equation
3 because the polymerization was initiated from a 16 A-thick surface-immobilized
initiator monolayer. The fit of the data in Figure 4.3 and 4.4 (open circles) to this

equation is excellent, showing that the model is reasonable.

dlA] 2

—=—k A

dt 1[ I (1)
dl

—=k A

dt 2[ l (2)
l=:—21n(l+k1[A0]I) (3)

1

108

To account for the loss of chain ends during a quenching and re-initiation step, a
quenching efficiency, Q, the ratio of active chain ends after and before quenching is
introduced. Thus if the first quenching step occurs at time t1, the number of active sites
after the quenching step will be given by equation 4. For multiple quenching steps, the
concentration of active sites after quenching step n (which occurs at time tn) is given by
equation 5, where n is the number of quenching steps and the 1,, values represent the times

at which each of the quenching steps occurred. The growth in film thickness, Al between

the (n-1)th and nth quenching steps for n22 will thus be given by equation 6. This

equation is obtained by substituting values of [A(t,,-1)] for [A0] and values of (tn-tn-1) for t
in equation 3. Equation 3 is still valid for film growth prior to the first quenching step.

The total film thickness is simply the sum of the A1,, values.

[A]: [A IQ (4)

1+k1[A0 1:,

 

[AolQn (5)

[A(tn)]= 0 0 2 o n-l o
l+k1[A ]t1+Qk1[A ](12-t1)+Q k1[A ](t3-t2)+...+Q k1[A ](t,,—t,,_1)

o n—l
.842... i. huge he.-.) _2 (6)
k] 1+kl[A0]11+le[Ao](12 -Il)+Q kl[A0](f3 -12)+...+Qn k1[Ao](In_1—tn_2)

109

-0 'Liluvfi

.—.____,_._.
r .-
I

 

The efficiency of quenching, Q can be estimated from equation 6 and the data in
Figure 4.3. These calculations show that Q > 0.95 as would be expected from the similar
thicknesses of films prepared in a single step and films prepared with quenching and
reinitiation. Equation 6 also shows that the effect of quenching will be more pronounced
as termination of active chains is reduced (lower k1 values) or as the number of
quenching steps increases. In the next section we compare thicknesses of films prepared

using up to six quenching steps and two different quenching procedures.

V. Comparison of Cu(II)Br2 Quenching with Simple Solvent Washing

Considering the short lifetime of radicals in ATRP,31 their concentration at any
given time should be very low. The combination of this low radical concentration and the
inertness of the dormant chain ends might allow replacement of the Cu(II)Br2 quenching
with simple solvent washing to remove unreacted monomer after each block is formed.
Matyjaszewski” and Dadmun32 reported recently that multiblock copolymers could be
prepared by ATRP in solution without a quenching step. However, in some cases,
polydispersity slightly increased with the number of blocks, implying that some
termination of polymer chains possibly happens during separation and re-initiation
procedures.32 This trend led us to use the Cu(II) quenching strategy to shift the
equilibrium to inert dormant chains during separation and re-initiation procedures. Cu(II)
quenching might be especially important for the preparation of copolymer brushes.
because chain ends growing from surfaces are in close proximity and may recombine

more readily than similar chains in solution.

110

To see whether the Cu(II) quenching step is necessary, polymerizations were
either quenched with a Cu(II)Br2 solution or simply washed with solvent after formation
of each block. FTIR spectroscopy and ellipsometry show that both approaches lead to the
successful formation of multiblock polymer brushes. However, the films prepared using
the solvent washing protocol repeatedly showed 2-4 % less increase (per cycle) in the
subsequent block thickness than those prepared by the Q&R strategy. To determine
whether this difference was significant, heptablock PMMA brush ([PMMA-b-kPMMA)
was prepared with and without Cu(II)Br2 quenching steps. Polymer films were either
quenched or solvent-washed after 15, 25, 40, 60, 90, and 150 min of reaction time and re-
initiated to prepare the next block. Figure 4.4 shows that as the number of Q & R cycles
increased, the difference in the thicknesses of films synthesized by the two methods also
increased. The results for heptablock PMA films ([PMA-b-kPMA) prepared using both
Cu(II)Br2 quenching and solvent washing are shown in Figure 4.5. Again, the higher
efficiency of the Cu(II)Br2 quenching strategy becomes evident as the number of
quenching steps increases. The termination of growing chain ends by recombination may
be unavoidable, but Cu(II)Br2 quenching can preserve more surface initiating species and
lead to thicker polymer layers. After heptablock formation, the accumulated difference in
the thicknesses of the samples prepared by the two methods were ~21 % and ~13 % for

[PMMA-b-kPMMA and [PMA-b-kPMA, respectively.

Assuming that the Cu(II)Br2 quenching step had an efficiency of 100%, equation

3 was used to fit the data for PMMA film growth (open circles, Figure 4.4). This

allowed us to obtain values for kzlkl and k1[A0]. Using these values and equation 6, Q for

solvent washing is estimated to be around 87%. Thus the Cu(II) quenching method is

111

 

 

considerably more efficient than solvent washing. In the case of PMA, the fit of the
thickness data for Cu(H) quenching to equation 3 was not sufficiently good to estimate
the quenching efficiency for the solvent washing procedure. However, the data clearly

show the lower efficiency of solvent washing compared to Cu(II) quenching.

VI. Conclusions

Surface-initiated ATRP was successfully quenched by a Cu(H)Br2/2(anbpy)
solution, and dormant initiating sites were preserved for re-initiation of multiblock
synthesis. Direct GPC measurement of the molecular weights of polymer brushes
detached from the surface demonstrates that this Q & R method is capable of producing
relatively homogeneous brushes. Reflectance FTIR spectroscopy, ellipsometry, and
molecular weight data confirm the formation of triblock copolymer brushes on gold
surfaces. After heptablock formation, the accumulated difference in the thicknesses of the
samples prepared by Cu(H)Br2 quenching and solvent washing became significant: ~21
% and ~13 % for [PMMA-b—kPMMA and [PMA-b-kPMA, respectively. The efficiency
of a Cu(II)Br2 Q & R cycle determined from the growth rate of PMA films was >95 %,
while the same efficiency for solvent washing was 85—90%. The Cu(II)Br2 Q & R
approach allows preparation of multiblock copolymers with control over the thickness of

each block at ambient temperature.

112

 

 

400

300 *

  
 

Thickness (A)
8

 

 

 

 

 

 

 

100 ‘ 0 Cu(ll) quenching
O solvent washing
a fit using equation 3
O I I I I I I I

 

0 50 100 150 200 250 300 350
Reaction Time (min)

Figure 4.4. Thickness of multiblock PMMA films vs polymerization time. Each
subsequent point represents one more cycle of polymerization followed by Cu(II)
quenching or solvent washing, e.g., 300 min data were obtained after 7 polymerization
steps with quenching or washing between each step. Error bars represent the standard
deviations of thickness determinations on at least three different spots on two different
samples. The line represents a fit to the thickness of films prepared with Cu(II)Br2
quenching (open circles) using the function, thickness = 16 + A ln(l+Bt) derived from

equation 3.

113

 

 

 

400

 

 

 

 

 

 

 

 

 

300 -
8:
(I)
513
c 200 ~
3:
.2
.c
'—
100 - o Cu(ll) quenching
O solvent washing
a fit using equation 3
O I I I I I I I
O 50 1 00 1 50 200 250 300 350

Reaction Time (min)

Figure 4.5. Thickness of multiblock PMA films vs polymerization time. Each
subsequent point represents one more cycle of polymerization followed by Cu(II)
quenching or solvent washing. Error bars represent the standard deviations of thickness
determinations on at least three different spots on two different samples. The line
represents a fit to the thickness of films prepared with Cu(II)Br2 quenching (open circles)

using the function, thickness = 16 + A ln(l+Bt) derived from equation 3.

114

 

VII. References
(1) Dan, N.; Tirrell, M. Macromolecules 1993, 26, 4310-4315.

(2) Hadziioannou, G.; Patel, 8.; Granick, S.; Tirrell, M. J. Am. Chem. Soc. 1986, 108,
2869-2876.

(3) Morkved, T. L.; Lu, M.; Urbas, A. M.; Ehrichs, E. E.; Jaeger, H. M.; Mansky, P.;
Russell, T. P. Science 1996, 273, 931-933.

(4) Thum-Albrecht, T.; Schotter, J.; Kastle, G. A.; Emley, N.; Shibauchi, T.; Krusin-
Elbaum, L.; Guarini, K.; Black, C. T.; Tuominen, M. T.; Russell, T. P. Science
2000, 290, 2126-2129.

(5) Kim, H.-C.; Jia, X.; Stafford, C. M.; Kim, D. H.; McCarthy, T. J.; Tuominen, M.;
Hawker, C. J.; Russell, T. P. Adv. Mater. 2001, 13, 795-797.

(6) Spatz, J. P.; Eibeck, P.; Moessmer, 8.; Moeller, M.; Kramarenko, E. Y.; Khalatur,
P. G.; Potemkin, I. 1.; Khokhlov, A. R.; Winkler, R. G.; Reineker, P.
Macromolecules 2000, 33, 150-157.

(7) Mansky, F.; Liu, Y.; Huang, E.; Russell, T. P.; Hawker, C. Science 1997, 275,
1458-1460.

(8) Zhao, B.; Brittain, W. J. Prog. Polym. Sci. 2000, 25, 677-710.
(9) Zhao, B.; Brittain, W. J. J. Am. Chem. Soc. 1999, 121 , 3557-3558.

(10) Zhao, B.; Brittain, W. J .; Zhou, W.; Cheng, S. Z. D. Macromolecules 2000, 33,
8821-8827.

(1 l) Sedjo, R. A.; Mirous, B. K.; Brittain, W. J. Macromolecules 2000, 33, 1492-
1493.

(12) Corkery, R. W. Langmuir 1997, I 3, 3591-3594.

115

(13) Alberti, G.; Casciola, M.; Costantino, U.; Vivani, R. Adv. Mater. 1996, 8, 291-
303.

(14) Katz, H. E. Chem. Mater. 1994, 6, 2227-32.

(15) Kaschak, D. M.; Mallouk, T. E. J. Am. Chem. Soc. 1996, 118, 4222-4223.
(16) Decher, G. Science 1997, 277, 1232-1237.

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

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

 

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

(20) Husseman, M.; Malmstrtim, 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.

(21) Davis, K. A.; Matyjaszewski, K. Macromolecules 2001, 34, 2101-2107.

(22) Queffelec, J .; Gaynor, S. G.; Matyjaszewski, K. Macromolecules 2000, 33, 8629-
8639.

(23) Matyjaszewski, K.; Patten, T. B.; Xia, J. J. Am. Chem. Soc. 1997, 119, 674-680.

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

(25) Templeton, A. C.; Hostetler, M. J .; Kraft, C. T.; Murray, R. W. J. Am. Chem.
Soc. 1998, 120, 1906-1911.

(26) Stahl, c. R.; Siggia, s. Anal. Chem. 1957,29, 154-155.

116

(27)

(28)

(29)

(30)

(31)

(32)

Robinson, K. L.; Khan, M. A.; de Banez, M. V.; Wang, X. S.; Armes, S. P.
Macromolecules 2001, 34, 3155-3158.

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

Ejaz, M.; Ohno, K.; Tsujii, Y.; Fukuda, T. Macromolecules 2000, 33, 2870-2874.

Huang, W.; Baker, G. L.; Bruening, M. L. Angew. Chem, Int. Ed. 2001, 40,
1510-1512.

_ by”

Matyjaszewski, K. Macromolecules 1999, 32, 9051-9053.

II. ‘ - 'It‘
l
is

Eastwood, E. A.; Dadmun, M. D. Macromolecules 2001, 34, 740-747.

117

 

Chapter 5

Controlled Growth and Degradation of Biodegradable N anometer-
Thick Films of PLA from PHEMA Surfaces

I. Introduction

This chapter describes the surface-initiated ring-opening polymerization (ROP) of
lactide on poly(2-hydroxyethyl methacrylate) (PHEMA) surfaces as a route to the
controlled growth of biodegradable thin films of polylactide (PLA). Using a combination
of atom transfer radical polymerization (ATRP) and ROP, we successfully synthesized
PHEMA-g-PLA on gold and used the polymerization time to control the growth of PLA.
The growth of PLA on PHEMA followed lst order kinetics with a linear increase in PLA

thickness versus polymerization time, indicating a polymerization with living character.

Polymers based on renewable resources are increasingly attractive alternatives to
those derived from petroleum. Poly(lactic acid) or poly(lactide) is environmentally
degradable and the monomer, lactic acid, is available from renewable resources such as
corn. PLA’s biodegradability and biocompatibility have led to applications as surgical
sutures, tissue scaffolds, and drug delivery systems, and now PLA is being introduced as
a commodity polymer for fibers and packaging. A number of block copolymers have
been synthesized to gain better control over PLA’s physical properties.HS

In addition to bulk forms of PLA, thin films of PLA are especially important for
8.9

applications in drug delivery systems7 and as drug-eluting coatings on medical devices.

The thickness of a coating and its degradation profile are important factors that must be

118

7h

 

considered in designing materials for controlled drug release. Therefore, precise control
over the thickness of the PLA layer is indispensable. Various techniques have been
developed to prepare thin-film coatings on solid substrates. Physically adsorbed layers
are often thermally unstable due to weak interactions between the polymer and solid
substrate and lack resistance to solvents and friction. Covalently attached films should

offer improved thermal stability and mechanical properties.

Surface-initiated ROP of lactide provides a unique method for the synthesis of
PLA layers covalently linked to hydroxy-terminated surfaces. While surface-initiated
polymerization has attracted much attention, there are a limited number of reports
devoted to surface-initiated ring opening polymerization. Husemann et al. used surface-
initiated ring opening polymerization to graft poly(e-caprolactone) from hydroxy-
terminated self-assembled monolayers (SAMs) as a step in a soft lithography patterning
scheme.lo Two recent examples of the surface-initiated ROP of lactides were reported by
Choi11 and Mijller.12 In both cases, PLA was initiated from —OH or -NH2 groups at the
termini of long chain alkyl thiolates on gold or silicate surfaces. However, neither study
reports a kinetic study of PLA growth, which is necessary for control over the thickness
of biodegradable films. Here we describe the synthesis and the kinetics of lactide

polymerization from PHEMA grafted on gold surfaces.

If we think of PHEMA as functioning in the same way as a SAM, one should
expect a relatively low grafting density. A comparison of the contact angles for water on
PHEMA,“ 63°, and on an (o-hydroxyalkyl thiolate SAM,25 34°, suggests that the surface
density of PHEMA -OH groups is less than full surface coverage. However,

participation of subsurface —OH groups would increase the grafting efficiency and

119

produce a comb-like structure. To prepare the surface anchored PLA film, a layer of
PHEMA on gold was synthesized by ATRP, and then ROP was used to grow PLA from
the PHEMA layer. Since both polymerizations are living processes, the thickness of each
layer is easily controlled by the polymerization time. The thickness of the PLA film could

be controlled from several run up to 450 nm.

The PHEMA-g-PLA system on gold has several important features. Polymer
layers on gold are chemically homogeneous and are compatible with a wide variety of
thin film characterization techniques. In addition, gold is biocompatible with potential
applications in drug delivery devices.l3 Compared to thiolate on gold SAMs, PHEMA
provides the thermal stability needed to support lactide polymerization. The Au—S bond

OCH-16

is unstable above ~60 and organic monolayers often detach from surfaces.

Finally, the PHEMA-g-PLA system is a biocompatible hydrogell7 that can be used as a

drug delivery system.'8‘2'

11. Synthesis of PHEMA-g-PLA

Scheme 5.1 outlines the synthesis of PHEMA-g-PLA tethered to a gold surface.
The detailed procedure for the preparation of a SAM of initiator l is described in the
Experimental section. ATRP of HEMA (2-hydroxyethyl methacrylate) was initiated from
the surface-bound initiator layer. The catalyst system was a mixture of Cu(I)Br/tris[2-
(dimethylamino)ethyl]amine (Me6TREN) (0.1 mol % based on monomer) and
Cu(H)Br2/2 equivalents of 4,4'-di-n-nonyl-2,2'-bipyridine (anbpy). The Cu(II) complex
(30 mol %, relative to Cu(I)), ensures the deactivation of active radicals and provides

some control over the polymerization. The dormant a-bromocarbonyl initiator was

120

 

activated by immersing substrate 1 in a 1:1 (v/v) CH3CN-THF solution of HEMA and the
copper catalyst. Polymerizations of HEMA were run at room temperature with the
thickness of the PHEMA film controlled by the reaction time. Detailed conditions for the
ATRP of HEMA can be found in Experimental section. After HEMA polymerization, the
substrate was washed sequentially with DMF, EtOH, EtOAc, EtOH and deionized water,
and was then dried under a stream of nitrogen. The formation of PHEMA was apparent
from the appearance of a carbonyl peak at 1733 cm‘1 and a broad hydroxy peak at
3200—3600 cm‘1 in the reflectance FTIR spectrum (Figure 5.1, spectrum a). A number

of PHEMA substrates were prepared to support kinetic experiments of the ROP of lactide
from PHEMA. The PHEMA films on these substrates all had similar ellipsometric

thicknesses, 174:2 nm.

Using the hydroxy groups of PHEMA side chains as initiators, rac-lactide was
polymerized in toluene at 95 °C with Sn(II) octanoate, (Sn(0ct)2) as the catalyst. The
polymerizations were run in a drybox to avoid contamination from water, which can act
as a competing initiator. At regular intervals, samples were removed for IR and film
thickness measurements. The IR spectra in Figure 5.1 show the growth of PLA on
PHEMA. The initial spectrum (a) showed a single carbonyl peak at 1733 cm”l from
PHEMA, but after 90 min of lactide polymerization, the carbonyl peak broadened

(Figure 5.1, spectrum b) and shifted to higher wavenumbers. Eventually, the PLA
carbonyl peak dominated the spectrum and only a single peak at 1767 cm’l was observed
(Figure 5.1, spectrum e). Parallel growth in the methyl stretching peak at 2993 cm’1 and

a decline of the hydroxy peak at 3200—3600 cm"1 also confirm PLA formation.

12]

 

 

Scheme 5.]. Synthetic pathway for PHEMA-g-PLA

0
Au s—(Ctht—O—d—pH—er
CHa

HEMA/CuBr/MeeTREN
CLBr2/2anpr

9 9H3 I
Au S—(CH2)11—O—C—CH CHz —C
0113 COzcl'l'zCHzOHJ (

Dr
”I

0

$15: Sn(Oct)2. Toluene, 95°C

 

 

 

 

 

 

9*‘3 9*
Au 9 CHz—c CHz—(F——Br
COzcl‘bCl-lel-l hm czo m
_ _ _ o _
I
‘7sz
g”? 9 *3

Cs

9 '1
-o-c—c—o

CH; CH;

122

n

4‘ h .In'

 

 

 

 

 

 

 

 

 

 

 

'01

M (9) J
g N (d) J
o
g M
< N (C)

N (b) M

(a) M

4000 3530 3000 2530 2600 1500 1000

Wavenumber (cm")

Figure 5.1. Reflectance FTIR spectra of (a) 2: a grafted PHEMA layer (174 nm) on
gold, (b) 3: a grafted PHEMA(170 nm)-g-PLA(29 nm) layer after 90 min of lactide
polymerization, (c) 3: a grafted PHEMA(173 nm)-g-PLA(127 nm) layer after 180 min of
lactide polymerization, (d) 3: a grafted PHEMA(178 nm)-g-PLA(179 nm) layer after 240
min of lactide polymerization, and (e) 3: a grafted PHEMA(178 nm)-g-PLA(409 nm)
layer after 480 min of lactide polymerization. For clarity, the 2700-3700 cm’1 region in

each spectrum is magnified 5 times relative to the remainder of the spectrum.

123

III. Kinetics of Lactide Polymerization from PHEMA Surface

The growth of the PLA film was followed by monitoring the change in the film
thickness with time. Each sample was characterized by surface reflectance FTIR, and
ellipsometric data were collected at three different spots on the sample. Plotted in Figure
5.2a is the net increase in the thickness of the PLA layer, calculated by subtracting the
thickness of the PHEMA film from the total film thickness. The homogeneous ROP of
lactide follows a “coordination-insertion” mechanism,22 and the data of Figure 5.2a show
the expected linear increase in film thickness with polymerization time, confirming that
the polymerization of lactide from PHEMA is a living process and follows 1st order
kinetics. The data also show a short induction time before the onset of PLA
polymerization. Similar induction times were observed for solution polymerizations of
lactide,23 which may be related to the time needed to form the alkoxy-Sn complex from

Sn(0ct)2 and the hydroxy group of PHEMA.

IV. Control Experiments without surface -OH groups

Two control experiments show that the PLA film only grows from the —OH group
of PHEMA side chain. In the first experiment, the hydroxy groups of PHEMA were
blocked by acetylation with acetyl chloride. Substrates with PHEMA and acetylated
PHEMA films were placed in the same vial with rac-lactide and Sn(Oct)2 at 95 °C for 12
hr. As shown in Figure 5.3, the IR spectra of the PHEMA substrate (spectrum b) shows
the expected changes after ROP of lactide, an increase in the carbonyl peak intensity, its

shift from 1733 cm’1 to 1767 cm’1 and a decrease in hydroxy peak at 3200-3600 cm'l.

124

 

up

500 S

 

 

(6!)

E400 -
5
G)
(D
3 300 ~
8
E
3 200 -
“c’
f, (b A A
.C ..
,_ 100

o A I I T I I I () (J I I

 

0 100 200 300 400 500 600 1000 1500 2000
Reaction Time (min)

Figure 5.2. Plot of PLA thickness increase vs polymerization time: (a) using 174:2 nm
thick PHEMA substrates and (b) using 49:1 nm thick PHEMA substrates. Ellipsometric
thicknesses were measures at three different spots on a sample and the error bars are

smaller than symbols.

However, the IR spectrum of the substrate with acetylated PHEMA shows no sign of
PLA growth (spectrum d). Similarly, the behavior of substrates with PHEMA and
PMMA14 films were compared using the same conditions as before. As seen in Figure
5.4, the polylactide grew from the PHEMA layer, while the PMMA control showed no
PLA growth. Instead of growing PLA, a decrease in IR intensity suggests that some of

PMA chains desorbed from surface under the polymerization conditions.

125

"r

 

 

0.2
(d) After Polymerization

(c) Acetylated PHEMAMAA‘
(b) k After Polymerizatiom

PHEMA W

I

Absorbance

 

I I

4000 3500 3000 2500 2000 1500 1000
Wavenumber (cm'l)

 

I I I

 

Figure 5.3. Reflectance FTIR spectra of (a) a surface grafted PHEMA layer on gold, (b)
after lactide polymerization from PHEMA, (c) a surface grafted acetylated-PHEMA
layer, and (d) after attempted lactide polymerization from acetylated-PHEMA.

126

I A “Una—_

 

0.02

 

 

(d) Mter Polymerizatlofl

 

 

 

 

 

a) a;
o e
5 .
M

o {if n
to
.o
<

(b) After Polymerization A W

(a) , PMMA Control

 

 

 

 

 

I I

4000 3500 3000 2500 2000 1500 1000
Wavenumber (cm")

Figure 5.4. Reflectance FTIR spectra of (a) a surface grafted PMMA layer on gold, (b)
after attempted lactide polymerization from PMMA, (c) a surface grafted PHEMA layer,
and (d) after lactide polymerization initiated from PHEMA.

127

 

 

PLA was grown from PHEMA layers of different thicknesses. As shown in
Figure 5.2, the growth rate was extracted from the linear portion of thickness vs. time
curves. The 174 nm PHEMA film grew at 9.8 A/min and while the growth rate of a 49
nm PHEMA film was 2.4 A/min. The ratio of the rates, 4.1, is similar to the ratio of the
film thicknesses, 3.6. This suggests that the hydroxy groups of PHEMA are accessible,
and ROP of lactide can be initiated from most hydroxy groups. If initiation was restricted
to a thin surface layer of hydroxy groups, then the rate of film growth should be
independent of the thickness. Acetylation studies also are consistent with the view that
the —OH group of PHEMA is accessible to reagents. As shown in Figure 5.3c, the
absence of a significant —OH peak in the FTIR spectrum of PHEMA films after reacting
with acetyl chloride indicates nearly complete acylation of PHEMA. In addition, the
decrease in the hydroxy peak at 3200—3600 cm.1 in the FTIR spectrum of PHEMA-g-
PLA (Figure 5.1) is too large to be explained by initiation from a thin surface layer.
Thus, the structure of PHEMA-g-PLA should be viewed as a graft copolymer, with at

least some comb-like character.

Using 49:1 nm thick PHEMA substrates, we extended the polymerization time to
test for linearity in growth at long times. As seen in Figure 5.2b, PLA film growth
eventually ceased at ~10 hr. Examination of the polymerization solution by 1H NMR
revealed a 79:21 mixture of soluble PLA homopolymer and monomer. The formation of
the polymer likely resulted from initiation of ROP by adventitious H20 in the solvent or
on the surface of the glassware, despite efforts to exclude H20 by working a in a drybox,
using dry solvents, and silanizing reaction vials. The termination of growth seen in

Figure 5.2b is consistent with equilibrium polymerization, where the propagation and

128

depropagation rates are identical.”28 The equilibrium concentration of monomer
calculated from the NMR data, [M]eq, is ~ 0.03 M in good agreement with literature
values. For polymerization of L-lactide in 1,4-dioxane, [Mleq ~ 0.15 M at 406 K and 0.06
M at 353 K.26 To prove equilibrium control, we isolated a sample that had grown to a
limiting PLA film thickness (131 nm), and after washing with solvent, we transferred the
sample to a vial containing fresh monomer solution with catalyst. As expected, PLA

growth resumed, and an additional 53 nm of PLA was added in 5 hr.

The ROP of lactide from PHEMA can be stopped the polymerization at a given
thickness, and then reinitiated at a later time to increases the thickness of the PLA layer.
For example, when the samples used to generate the data of Figure 5.2a were re-
immersed in a solution of monomer and catalyst, each sample grew an additional ~ 470
nm of PLA film in 12 hr. In principle, PLA films can be grown to arbitrary thicknesses
as long as accessible initiating sites remain in the PHEMA layer or at the end of PLA

chains.

V. Degradation of PLA film

The hydrolytic degradation behavior of PHEMA-g-PLA on gold was investigated
in phosphate buffer (pH 7.40) at 55 °C. Substrates with surface-grafted PHEMA-g-PLA
were immersed in the buffer solution and were removed at intervals to measure their
ellipsometric thickness and surface reflectance FTIR spectrum. The sample surfaces
quickly became rough, and reliable ellipsometric thicknesses could not be measured after
1 day of degradation. The increase in surface roughness indicates that the degradation

mechanism is not simple surface erosion. Reflectance FTIR, however, could be used to

129

'3 '_ 71? 'ffi Uin'i__'
‘ .

 

characterize the degradation profile of the film. Figure 5.5 shows the evolution of the
FT IR spectra for a single sample during degradation. Shown in Figure 5.5(a) is
spectrum of the 174 nm thick PHEMA film before lactide growth, while 5.5(b) is the
spectrum of the PHEMA-g-PLA film after deposition of 464 nm of PLA. As hydrolytic
degradation proceeded, the methyl stretching peak at 2993 cm"1 and the carbonyl peak at
1767 cm’1 decreased, while the hydroxy peak at 3200—3600 cm"1 increased. After 50
days of degradation, the IR spectrum of the sample 5.50) was almost identical to the
initial PHEMA spectrum 5.5(a), and it showed no further changes for times >50 days.

These results are consistent with selective hydrolysis of the PLA film.

The degradation experiments shown in Figure 5.5 were run using the same
phosphate buffer solution throughout the run. The pH of the buffer solution used in the
first set of experiments decreased from 7.40 to 7.36 in 50 days. We anticipated seeing an
autocatalytic effect29 since degradation of PLA generates lactic acid oligomers that
accelerate the acid-catalyzed cleavage of PLA chains. In a second set of degradation
experiments, we changed the buffer solution every 2 days to ensure a constant pH. The
PLA degradation rates for these two experiments were calculated from the FTIR spectra
of the samples by subtracting the PHEMA carbonyl peak area in spectrum 5.5(a) from
the total area associated with the PLA and PHEMA carbonyls. As noted earlier, PHEMA
shows no evidence of degradation over 50 days. The results, plotted in Figure 5.6, show
a faster degradation rate for samples that were degraded in the same buffer solution,

consistent with a small autocatalytic effect.

130

 

b (a) PHEMA A W
MI) 50 days

W
W) 36 days M
Ma) 30 days Fm
Na) 25 days Wm
N01) 23 days ‘W
No) 21 days ‘M
No 19 days Wm

 

 

 

 

 

Absorbance

 

 

Mb) PHEMA-g-PLA

 

 

 

 

 

I I I I

4000 3500 3000 2500 2000 1 500 1 000
Wavenumber (cm")

Figure 5.5. Reflectance FTIR spectra of (a) PHEMA (174 run) before PLA growth, (b)
PHEMA (174 nm)-g-PLA(464 run) before degradation, and (CH1) PHEMA-g-PLA films
at various times during hydrolytic degradation. For clarity, the 2700-3700 cm"1 region of

each spectrum was magnified 2 times relative to the remainder of the spectrum.

131

 

During the first 4 days, the PLA carbonyl peak area increased, and then
decreased linearly. Although the cause of the initial increase in peak area is unclear, it
may be related to the macroscopic surface roughness that appears within hours of the start
of the degradation. It is well known that IR peak intensities are dependent on the
orientation of functional groups relative to surface normal,30 and the increased surface

roughness may reflect some change in carbonyl group orientation.

The 9-30 day data are linear and are plotted in the inset in Figure 5.6. The
calculated slopes are —3.06 and —2.52 for the Figure 5.6(a) and 5.6(b) data respectively.
For the first set of data, the initial PLA film thickness was 464 nm, which corresponds to
a degradation rate of ~14.5 nm/day. The second set of data was obtained with a 488 nm
thick PLA film, which degraded at ~12.0 nm/day. Thus the apparent autocatalytic effect

increased the degradation rate by ~20%.

132

 

 

 

 

 

 

140 140

g 120 - gfii "“°

.5 g g IE 1°° (b)
g 100 3.. §§ ii b 80.

5 § ii ( l 60 ‘ (a)

2 80 ~ §§ E i

'5 60 . § 4‘”

x (a) §§ i 20 .

8 o 10 20 30 40
i

E

8

 

 

 

0 10 20 30 40 50 60 70 80
Degradation time (days)

Figure 5.6. Plot of PLA carbonyl peak area vs degradation time: (a) a PHEMA (174
nm)-g-PLA (464 nm) film degraded in the same phosphate buffer solution throughout the
experiment (hollow circles: o) and (b) PHEMA (174 nm)-g-PLA (488 nm) degraded in
buffer that was replaced every 2 days (filled circles: 0). The error bars represent the
standard deviations in determining the carbonyl peak areas in FTIR spectra, which result

from two different choices for the baseline in the FTIR spectra.

133

.1

 

VI. Surface Morphology Study of PLA film during Degradation

An unexpected aspect of the degradation of PHEMA-g-PLA films was the
development of unusual surface domains. These domains form several hours after the
start of the degradation experiment, and then evolve over several days. As shown in
Figure 5.7(a), PHEMA-g-PLA films are smooth when observed using an optical
microscope. However, after 6 hr of degradation, the dried surface was covered with
numerous gear-shaped domains—5.7(h). If the surface is swelled with water and dried
several times, the gear-like structures evolve into star-like domains-5.7(c) and 5.7(d),
with most being five and six-pointed stars. The stars were surprisingly symmetrical with
a circle in the center of each star. With further degradation, the domains evolved into
pentagonal and hexagonal domains-5.7(e) and 5.7(f), and eventually dendrimer-like
structures—5.7(g) and 5.7(h). At long times, the domains developed into flower-like
patterns separated by mottled areas-5.76). Figure 5.8 shows the typical stages seen in

the evolution of the surface features.

The domains depicted in Figure 5.7(b)-5.7(d) are not flat, as can be concluded
from the inability to have the entire image in focus for any one setting of the focus. This
is illustrated in Figure 5.9(a), where the circular features are clearly in the same focal
plane as the substrate, while in Figure 5.9(b), the curved features that give rise to the
geometrical pattern of the domains are above the focal plane of the substrate. The
features and the topography of the sample suggest that the dark lines may correspond to

overlapping PLA layers or folds that reflect light differently than the bulk sample.

134

_ 1?), SR” 0 hi

 

135

1';

()l8hr

 

(h) 24 hr

   

Figure 5.7. Surface images of PHEMA (177 nm)-g-PLA (353 nm) obtained by optical
microscopy during hydrolytic degradation: (a) before degradation and (b)—(i) after
degradation of PLA. Images in this dissertation are presented in color.

136

 

 

 

 

 

 

 

Figure 5.8. Evolution of surface domains during degradation. Images in this dissertation

are presented in color.

What drives the formation of these regularly patterned surface domains? Since
the degradation of PLA was run under hydrolytic conditions and the bottom layer is a
hydrophilic PHEMA layer, direct observation of wet surfaces should provide critical
evidence for explaining the formation and nature of the domains. The wet surface of a
sample that had been degraded for 6 hr was observed by optical microscopy (Figure
5.10). Initially, the surface was decorated with blister-like domains that collapsed as the
surface dried. With further drying, the domains evolved into symmetrical structures.
This process was completely reversible and could be repeated several times before the

domains suffered irreversible damage.

137

 

 

Figure 5.9. Differences in the optical microscopic images of domains as a function of
changes in focal length: (a) image obtained by setting the focus in the plane of the
substrate and (b) image obtained by focusing on the curved features of the domains.

Images in this dissertation are presented in color.

138

(a)

    

«wage; is;
o @Df’g
, 0 G 6 (30

x20

 

Figure 5.10. Evolution in surface morphology during the drying of water-swollen films:
(a)—(d) show successive optical microscopic images taken during the drying process.

Images in this dissertation are presented in color.

Based on these observations, we propose that surface domain formation is driven
by rapid diffusion of water through pinholes in the PLA layer, leading to swelling of the
PHEMA layer. Pinholes were not observed by optical microscopy in films before
degradation, but may develop as water dissolves catalyst residues that contaminate the

film. Swelling of the PHEMA layer causes localized blistering of the polymer film,

139

 

irreversibly stretching the PLA layer. With removal of the water, the blisters collapse
from the center, forming a circular PLA layer in contact with the underlying PHEMA
film, and a ridge of PLA at the outer edge of the blister. These show up in Figure 5.9 as
the circular features that are in the same focal plane as the substrate, and as the
undulating line above the focal plane. As the film continues to dry, the ridge is under
compression and eventually buckles to give the characteristic geometric patterns of the
dried films. While this scenario is plausible, the details of this process are not known and

require further study.

VII. Conclusions

In conclusion, biodegradable nanometer thick films of polylactide were
synthesized from PHEMA surfaces by surface-initiated ROP. The growth rate of the PLA
layer was linear, and the film thickness was controlled by the polymerization time. PLA
grew from the pendent hydroxy group of PHEMA films, but acetylated PHEMA and
PMMA surfaces failed to initiate polymerization. There was an induction time for lactide
polymerization presumably due to the slow formation of the alkoxy—Sn initiator from
Sn(Oct)2 and the hydroxy group of PHEMA. Once growth began, the ROP followed lst
order kinetics, which is consistent with a living polymerization. The loss of PLA from the
film surface during hydrolytic degradation was linear with time and showed a small
autocatalytic effect. These results support a surface erosion mechanism. Blistering in the
early stages of the degradation process led to the formation of highly symmetrical

structures and fracturing of the PLA layer.

140

 

VIII.

(1)

(2)

(3)

(4)

(5)

(6)

(7)

(8)

(9)

(10)

(11)

(12)

(13)

(14)

References
Otsuka, H.; Nagasaki, Y.; Kataoka, K. Biomacromolecules 2000, I, 39-48.
Metters, A. T.; Anseth, K. S.; Bowman, C. N. Polymer 2000, 41, 3993-4004.

Ouchi, T.; Nozaki, T.; Ishikawa, A.; Fujimoto, 1.; Ohya, Y. J. Polym. Sci, Part
A: Polym. Chem. 1997, 35, 377-383.

Kricheldorf, H. R.; Kreiser, I. Makromol. Chem. 1987, 188, 1861-73.
Song, C.; Chen, H.; Feng, X. Gaofenzi Tongxun 1983, 177-83.
Wallach, J. A.; Huang, S. J. Biomacromolecules 2000, I, 174-179.
Jain, R. A. Biomaterials 2000, 21, 2475-2490.

Schmidmaier, G.; Wildemann, B.; Stemberger, A.; Haas, N. P.; Raschke, M.
Journal of Biomedical Materials Research 2001, 58, 449-455.

Middleton, J. C.; Tipton, A. J. Biomaterials 2000, 21 , 2335-2346.

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.

Moller, M.; Nederberg, F.; Lim, L. S.; Kange, R.; Hawker, C. J.; Hedrick, J. L.;
Gu, Y.; Shah, R.; Abbott, N. L. Journal of Polymer Science, Part A: Polymer
Chemistry 2001, 39, 3529-3538.

Santini, J. T., Jr.; Cima, M. J .; Langer, R. Nature (London) 1999, 397, 335-338.

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

141

(15) Bain, C. D.; Troughton, E. B.; Tao, Y. T.; Evall, J .; Whitesides, G. M.; Nuzzo, R.
G. J. Am. Chem. Soc. 1989, I 1 I , 321-35.

(16) Schlenoff, J. B.; Li, M.; Ly, H. J. Am. Chem. Soc. 1995, 117, 12528-36.

(l7) Lim, D. W.; Choi, S. H.; Park, T. G. Macromol. Rapid Commun. 2000, 21 , 464-
471.

(18) Barakat, 1.; Dubois, P.; Grandfils, C.; Jerome, R. J. Polym. Sci., Part A: Polym.
Chem. 1999, 3 7, 2401-2411.

(19) Iijima, M.; Nagasaki, Y.; Okada, T.; Kato, M.; Kataoka, K. Macromolecules
1999, 32, 1140-1146.

(20) Riley, T.; Stolnik, S.; Heald, C. R.; Xiong, C. D.; Garnett, M. C.; Illum, L.;
Davis, S. S.; Purkiss, S. C.; Barlow, R. J .; Gellert, P. R. Langmuir 2001, 17,
3168-3174.

(21) Barakat, 1.; Dubois, P.; Jerome, R.; Teyssie, P.; Goethals, E. J. Polym. Sci., Part
A: Polym. Chem. 1994, 32, 2099-110.

(22) Stridsberg, K.; Ryner, M.; Albertsson, A.-C. Macromolecules 2000, 33, 2862-
2869.

(23) Yin, M.; Baker, G. L. Macromolecules 1999, 32, 7711-7718.

(24) Lim, H.; Lee, Y.; Han, S.; Cho, J .; Kim, K.-J. J. Vac. Sci. Technol, A 2001, 19,
1490-1496.

(25) Frutos, A. G.; Brockman, J. M.; Corn, R. M. Langmuir 2000, 16, 2192-2197.
(26) Duda, A.; Penczek, S. Macromolecules 1990, 23, 1636-9.

(27) Majerska, K.; Duda, A.; Penczek, S. Macromol. Rapid Commun. 2000, 21, 1327-
1332.

142

'!» . '.< ml ’35 it”

(28)

(29)

(30)

Nishida, H.; Yamashita, M.; Endo, T.; Tokiwa, Y. Macromolecules 2000, 33,
6982-6986.

Siparsky, G. L.; Voorhees, K. J .; Miao, F. J. Environ. Polym. Degrad. 1998, 6,
3 1-41.

Porter, M. D.; Bright, T. B.; Allara, D. L.; Chidsey, C. E. D. J. Am. Chem. Soc.
1987, 109, 3559-68.

143

‘A. 1r- -‘h-c

Chapter 6. Experimental

I. General Details
l-l. Materials

THF (Aldrich, 99%) was distilled from calcium hydride followed by a second
distillation from sodium/benzophenone ketyl. Acetonitrile (Merck, HPLC grade) was
passed through an activated basic alumina column, and fractionally distilled over calcium
hydride. Triethylamine (Aldrich, 99.5%) was distilled from calcium hydride under an
argon atmosphere at reduced pressure. Methyl acrylate (MA) (Aldrich, 99%) and methyl
methacrylate (MMA) (Aldrich, 99%) were distilled from powdered KOH under ambient
pressure. 2-Hydroxyethyl methacrylate (HEMA) (Aldrich, 97%) was passed through
basic alumina columns twice to remove inhibitor. After purification, all solvents and
monomers were transferred to Schlenk flasks, de-gassed using three freeze-pump-thaw
cycles and then transferred into a drybox. Tris[2-(dimethylamino)ethyl]aminel
(Me6TREN) and 4,4'-di-n-nonyl-2,2'-bipyridine2 (anbpy) were prepared by literature
procedures with some modification. 11-Mercapto-1-undecanol (MUD) (Aldrich, 97%), 2-

bromopropionyl bromide (Aldrich, 97%), Cu(I)Br (Aldrich, 99.999%), Cu(I)Cl (Aldrich,

99.999%), and Cu(II)Br2 (Aldrich, 99.999%) were used as received.

[-2. Synthesis of MegTREN

A solution of formaldehyde (43.37 g of a 37% solution in water, 0.534 mol of

formaldehyde) was placed in a 250 mL round-bottomed flask with an attached water

144

 

 

condenser. The solution was cooled in an ice-water bath, and tris(2-aminoethyl)amine
(Aldrich, 96%, 9.77 g, 0.067 mol) was added drop-wise to the formaldehyde solution.
Some solid material formed during the addition. Following the drop-wise addition of
formic acid (88 %) (69.88 g, 1.37 mol), the solution was heated with a heating mantle and
refluxed for 12 hrs. The evolution of C02 gas was good evidence for the formation of the
imine intermediate. The reaction mixture was cooled to room temperature and HCl (26.7
mL) was added to form the HCl salt of Me6TREN. Removal of the solvent under
vacuum gave 47.4 g of a yellow solid. The solids were dissolved in de-ionized water and
the pH of the solution was adjusted to ~12 with aqueous KOH. The top layer was
extracted with diethyl ether (2x100 mL) and the combined organic layers were dried over
Na2SO4 and KOH pellets. Removal of the ether gave an oily product (14.56 g), which
was purified by vacuum distillation (68—69 °C; 60 mmHg) to give 13.2 g (86 %) of
MegTREN. 1H NMR (300 MHz, CDCl3) 5 2.53-2.60 (m, 6H), 2.30-2.37 (m, 6H), 1.80 (s,

18H). 1“‘C NMR 6 57.4, 53.1, 46.0. GC-MS m/z = 172, 72, 58.

L3. Synthesis of anbpy

Dry THF (85 mL) and dry diisopropylamine (9.20 mL, 65.1 mmol) were placed in
a 500 mL round-bottomed flask under a flow of Ar. The mixture was cooled to -78 °C in
a dry ice-acetone bath, and n-BuLi (24.8 mL of a 2.5 M solution in hexanes, 62.0 mmol)
was added dr0p-wise. After 15 min, the bath was removed and the solution was allowed
to warm to 0 °C. The reaction mixture again was cooled to -78 °C, and then a solution of
4,4'-dimethyl-2,2'-dipyridyl (5.00 g, 27.1 mmol) in THF (120 mL) was slowly added via

a cannula. After stirring the mixture overnight, l-bromooctane (15.8 g, 81.9 mmol) was

145

added and the reaction mixture was allowed to warm to room temperature. The solution
was then poured into cold brine (250 mL), extracted with EtOAc (2x100 mL), and the
combined organic layers were dried over Na2804. Removal of the solvent gave 15.8 g of
crude product, which was purified by recrystallization from CH3CN/Et0H (1:1) to give
10.2 g of anbpy (92 % yield). 1H NMR (300 MHz, CDC13) 6 8.55 (d, 1H), 8.11 (s, 1H),
7.12 (m, 1H), 2.65 (t, 2H), 1.64 (quintet, 2H), 1.24 (m, 12H), 0.92 (t, 3H). 13C NMR 8

148.9, 124.1, 121.6, 35.6, 32.2, 30.8, 29.5, 29.4, 29.30, 29.27, 22.8, 14.2.

1.4. Analytical Methods

Reflectance FTIR spectroscopy was performed using a Nicolet Magna-IR 560
spectrometer containing a PIKE grazing angle (80°) attachment. Ellipsometric
measurements on polymer films were obtained with a rotating analyzer spectroscopic
ellipsometer (model M-44; J.A. Woollam) at a 75° angle of incidence. Polymer molecular
weights were measured using two PLgel 1011 mixed-B columns and a PLgel 3p mixed-E
column in series, and a Waters R410 differential refractometer detector. THF was used as
the eluent at a flow rate of 1 mIJmin at 35 °C, and monodisperse poly(methyl

methacrylate) standards (Polysciences, Inc.) were used to calibrate the molecular weights.
The molecular weights of PMA standards were 350,000, 127,000, 30,000, and 6,000

g/mol.

146

 

 

II. Kinetic Studies of ATRP
II-l. General Procedures for Surface-initiated Polymerization

The preparation of monomer solutions, polymerization reactions, and the initial
washing steps for surface polymerizations were done in a drybox filled with helium. To
increase the solubility of the Cu(II)Br2/ligand complex, monomer (methyl acrylate (MA)
and methyl methacrylate) solutions were prepared in 1:1 (v/v) CH3CN/THF.
Polymerizations designed to investigate the dependence of polymer growth on the
concentration of Cu(II)Br2 were run with different catalytic systems over a wide range of
catalyst concentrations. For polymerizations with no added Cu(II)Br2, Cu(I)Br/MegTREN
(0.1 mM) was used as the catalyst. The catalytic system for polymerizations with added
Cu(H)Br2 was a mixture of Cu(I)C1/Me6TREN (0.1 mM) and Cu(II)Br2/Me6TREN (0.03
mM). The surface-initiated polymerization of MA was run at Cu(I)Cl/Me6TREN = 40
mM, 2 mM, 0.1 mM, 5x10"3 mM, 2.5x10"4 mM, and so on. The monomer concentration
(2M) and the ratio of Cu(II)Br2 to Cu(I)Cl (0.3) were the same for all polymerizations,
and all polymerizations were run at ambient temperature (~23 °C) inside a drybox. After
polymerization, the surfaces of the samples were washed with a fresh mixture of solvent
and THF, and then the ellipsometric thickness of the polymer film was measured at two

different spots on the sample and averaged.

II-2. Preparation of Monomer/Catalyst Solution

A 40 mM catalyst solution was prepared by dissolving CuCl (0.0396 g, 0.4 mmol)

and CuBr2 (0.0268 g, 0.12 mmol) in 5 mL of a 1:1 mixture of CH3CN/I'HF. Other

147

 

 

catalyst solutions were prepared by dilution. MA (2.0 g, 0.02 mol) was added to each
catalyst solution (5 mL) and then mixed solvent was added to make the total volume 10
mL. 2M MA solutions were prepared by adding MA (1.72 g, 0.02 mol) to the catalyst
solution and diluting to 10 mL. After the addition of each reactant, the reaction vials were

swirled vigorously to give homogeneous solutions.

II-3. Polymerization for Kinetic Study

MA and MMA were polymerized from initiator-immobilized surfaces using five
different concentrations of Cu(DCl/MesTREN. In each reaction, the ratio of Cu(II) to
Cu(I) was 0.3. For polymerizations run at a catalyst concentration of 40 mM, 5 samples
were immersed into the same reaction vial and samples were taken out at 30 min, 1 hr,
2hr, 4 hr, and 8 hr. To avoid scratching the surface of the films, samples were stacked in
the vial as shown in Figure 6.1 to prevent the gold face of the samples from touching.
After removing the sample from the reaction vial, the surface was washed sequentially in
6 vials of fresh solvent. The first 3 vials contained 1:1 mixtures of CH3CN/THF and the
last 3 vials contained THF. The samples were transferred from vial to vial quickly to
avoid drying of the surface. After the last washing step, the surface was allowed to dry.
The samples were removed from the dry box and the sample surfaces were rinsed with
c0pious amounts of EtOAc, EtOH, and de-ionized water, and then were dried under a
flow of N2. Using the same procedure, polymerizations were run at Cu(DCVMeGTREN

concentrations of 2 mM, 0.1 mM, 5x10‘3 mM, and 2.5x10'4 mM.

148

 

 

   

‘1‘“ Qin‘ ._ ,1:

Figure 6.1. Digital images of five stacked initiator-immobilized samples in a reaction

vial. Images in this dissertation are presented in color.

II-4. Solution Polymerization Procedures

Three solutions were prepared for solution phase ATRP of MMA: a 4.6 M
solution of MMA, a 4.8 mM solution of Cu(I)Br/MegTREN, and a 1.6 mM solution of
Cu(II)Br2/2 equivalents of anbpy in 1:1 (v/v) CHgCNzTHF. The starting time for
polymerizations was defined as the time when ethyl 2-bromoisobutyrate, the initiator,
was transferred via cannula into a Schlenk flask containing the monomer and catalyst.
After mixing, the concentration of the initiator was 23 mM. The polymerization reaction
was stirred vigorously at 25 °C, and small samples (~2 mL) were removed using a
syringe to measure the conversion from monomer to polymer. The samples were weighed
and filtered through a basic alumina column to removing the copper catalyst and ligand,

thus quenching the polymerization reaction. The column was rinsed 5 times with THF,

149

and then the combined solutions were evaporated to dryness under reduced pressure.
After drying under vacuum, the mass of the recovered polymer was used to calculate the

conversion, and the molecular weight and poldispersity were measured by GPC.

II-5. Polymerization With or Without Stirring

To carry out ATRP from surface initiators with stirring, a glass reactor with 8
slots was designed to hold gold-coated silicon wafers. Figure 3.9 shows the reaction
vessel loaded with 8 samples and a magnetic stirring bar at the center. To avoid contact
between the magnetic stirring bar and the samples, a glass 0-ring was inserted between
the stirring bar and the sample holder. The 0-1ing ensures smooth stirring and prevents
damage to the gold surfaces. The reaction solvent was 1:1 (v/v) CH3CN/THF containing
MMA (2M), Cu(I)Br/MeeTREN (2 mM), and Cu(II)Br2/2anbpy (0.6 mM).
Polymerizations were run at 23 °C in a drybox. For kinetic studies, samples were
regularly taken out, washed with the co-solvent mixture and THF, and then the
ellipsometric thicknesses of polymer films were measured. A duplicate set of surface-
initiated ATRP was run under the same reaction conditions as described above, but this

time with no stirring. The samples were treated and measured in the same way.

II-6. Simulation of Polymer Brush Growth at Different Catalyst Concentration

Microsoft Excel was used to simulate the growth of polymers from surfaces.
Each term in equation 3 (see Chapter 3) was simulated from time = 0 seconds to

approximately eight hours. Since some simulations required over three billion rows in the

150

spreadsheet, a Visual Basic macro was used to collect every 65532"d data point for
graphing, and then restart the calculation. The simulation required the input of 5
variables: k3, kd, kt, initiator concentration, and dt, the time between iterations. k8, kd, and
kt were taken from values reported in the literature.3 The initiator concentration was
chosen to be 2 mol/L. The simulations included the range of Cu(I)/Cu(II) concentrations
used for polymer film growth. The value of dt used in each simulation depended upon the
Cu(D/Cu(II) system that was being simulated, and was selected to allow the smallest
number of iterations without causing the simulation to oscillate out of control. The radical
concentration at the conclusion of each step was multiplied by dt, and then summed over
the whole polymerization time. The radical concentration versus time was plotted at five

different catalyst concentrations.

111. Synthesis of Triblock Copolymer Brushes
III-l. Preparation of Substrate Immobilized with Initiator

Gold-coated Si wafers (200 nm of gold sputtered on 20 nm of Cr on Si (100)
wafers) were cleaned in a UV/03 chamber for 15 min, immersed in de-ionized water for
15 min, and dried under a flow of N2 just before use. A SAM of MUD was obtained by
immersing the gold-coated Si substrate in a 1 mM MUD solution for 1 day, and the MUD
SAM was washed with copious amounts of ethanol and de-ionized water, and dried under
N2. The ellipsometric thickness of the MUD layer was 12 i 1 A. To produce the initiator-
anchored surface, 1 (Scheme 3.1), the MUD SAM was immersed for ~10 sec in 5 mL of

0.12 M triethylamine in dry THF at ~0 °C after which 5 mL of 0.1 M 2-bromopropionyl

151

 

bromide (2-BPB) in dry THF was added drop-wise while gently agitating the
triethylamine solution. During the addition of 2-BPB, the reaction vial was tilted at
different angles to ensure that the reaction surface was well exposed to the incoming 2-
BPB solution. This initiator-immobilization reaction was carried out in a helium-filled
drybox. Upon addition of 2-BPB, a white precipitate formed immediately, which
presumably was triethylamine hydrobromide. After the reaction (~3 min), the surface was
washed with fresh THF several times and allowed to dry. Once the substrate was taken
out of the drybox, it was rinsed sequentially with EtOAc, EtOH, and de-ionized water,
and dried under N2. Initiator immobilization was apparent from the appearance of a

carbonyl peak at 1743 cm'1 in the reflectance FTIR spectrum of this film.

III-2. Procedures for Block Formation

The preparation of monomer solutions, polymerizations, and Q & R procedures
were done in a drybox filled with helium. 2M solutions of MA, MMA, and HEMA were
prepared in 1:1 (v/v) CH3CN2THF. Each solution contained monomer,
Cu(I)Br/Me6TREN (2 mM), and Cu(II)Br2/2(anbpy) (0.6mM). Each block of the
polymer brush was synthesized by dipping substrates 1, 2, or 3 (Scheme 4.2) in the
appropriate monomer solution, and then quenching the polymerization as described
below. Polymerizations were run at 25 °C for the first and second blocks (PMA and
PMA) and at 40 °C for the third block (PHEMA), and the reaction times for formation
of each block were 20 min, 40 min, and 5 hrs, respectively. After polymerization, the
substrate was removed from the solution and sequentially immersed in three vials of

CH3CN-THF (1:1) and three vials of THF. This procedure avoids drying of the film and

152

 

 

consequent deposition of solids that are difficult to remove from the substrate surface.
For the synthesis of heptablock copolymers, the polymerizations were carried out in a 2M
solution of monomer in 1:1 CH3CN-TI-IF, 1 mM CuBr/MegTREN, and 0.3 mM of
CuBr2/2anbpy at 24 °C. To compare Cu(II)Br2 quenching with solvent washing,
polymerizations were either quenched or solvent rinsed after 15, 25, 40, 60, 90, and 150

min of reaction time and then re-initiated.

III-3. Quenching Procedure

To quench polymerizations, 3 vials of 0.02 M Cu(II)Br2/2(anbpy) in 1:1
CH3CNle-IF were prepared, and substrates were transferred sequentially into the vials.
The substrates were then washed in 3 vials containing the cosolvent to remove Cu(II)Br2
from the surface. After examination by reflectance FTIR spectroscopy, substrates were
immersed in a solution containing another monomer/catalyst mixture to form the next
block. When simple solvent rinsing was used rather than Cu(II)Br2 quenching, the

substrates were sequentially immersed in six separate vials containing CH3CN2THF.

III-4. Detachment of Polymer Brushes from Gold Surfaces

Polymer brushes were detached from gold surfaces by immersing the polymer-
coated substrates in a 4 mM solution of 12 in CH2C12 for 15h at room temperature. After
rinsing the surface with fresh CH2Cl2 and drying under a flow of N2, the reflectance FTIR
spectrum was measured to confirm that the polymer chains were detached from the

surface. The initial CH2C12 solution was collected and transferred to a pear-shaped flask

153

that was connected to a closed vacuum line. The solution was frozen in liquid N2, and
after opening the vacuum line, the bath was withdrawn, and the solvent was slowly
removed under reduced pressure. After evaporation of CH2C12, a purplish thin coating
remained on the inside of the flask. The flask was kept at 60 i 10 °C in an oil bath and as
12 was removed by sublimation under vacuum, the thin film coating on the flask walls
became pale yellow. For GPC measurements, all of the remaining material in the flask
was dissolved in ~150 11L of THF and injected into the instrument. Assuming full
recovery of the polymer from a 380 A-thick PMMA film on ~8 cm2 of gold surface, the
concentration of detached polymer brushes in 150 11L of THF should have been ~0.2

mg/mL.

IV. Synthesis of Biodegradable Nanometer Thick Fihns of PLA
IV-l. Preparation of Monomer Solution and Substrates

A saturated solution of lactide was prepared in a drybox by stirring an excess of
lactide in toluene (120 mL) at room temperature for one hour. The homogeneous solution
was decanted from undissolved lactide, and the concentration of the lactide solution was
determined by gravimetry to be ~0.13 M. Sn(0ct)2 catalyst (0.0844 g) was added to the
lactide solution (100 mL) to give a catalyst concentration of ~2.1 mM. PHEMA substrate
2 (Scheme 5.1), was synthesized by surface-initiated atom transfer radical
polymerization. To prepare monomer solutions, Cu(I)Br (4.5 mg), Cu(II)Br2 (1.6 mg),
and MesTREN (8.9 mg) were dissolved in 5 mL of CH3CN2THF (5:1, v/v) in a drybox.

HEMA (15 mL, 0.123 mol) was added and the solution was diluted to 35 mL with the

154

gnu-lain;

 

cosolvent. The concentrations of each component were: Cu(I)Br/Me6TREN (0.9 mM),
Cu(II)Br2/Me6TREN (0.2 mM), and HEMA (3.5 M). The initiator-immobilized substrates
were immersed in the monomer solution and polymerized for one day at 40 °C. After
polymerization, the samples were taken out of the monomer solution and immersed in
anhydrous DMF to wash out catalyst and monomers. The sample surface was rinsed with
EtOAc, EtOH, and de-ionized water, and then dried under a flow of N2. The PHEMA

films were characterized by ellipsometry and surface reflectance FT -IR.

IV-2. Ring-opening Polymerization from PHEMA Surfaces

In a helium-filled drybox, PHEMA substrates, 2, were immersed in the lactide
(0.13 M) and Sn(0ct)2 catalyst (2.1 mM) solution. Silanized glass vials (15 mL) with
silicon rubber caps were used as reaction vessels to avoid polymerization from surface
silanol groups. The vials were suspended in an oil bath from copper wires, and a hot plate
was used to heat the bath to 95 °C. Vials were removed from the oil bath at various
intervals, and the samples were removed from the vial and were sequentially immersed in
a series of vials containing toluene, ethyl acetate, ethanol, and de-ionized water to clean
the surface. The surface was dried under a flow of N2 and characterized by reflectance

FTIR and ellipsometry.

Acetylated PHEMA samples were prepared by reacting PHEMA with acetyl
chloride in dry THF. A PHEMA substrate was placed in a 20 mL glass vial and 8 mL of

dry THF were added. While the vial was gently shaken, 2 mL of acetyl chloride was

added drop-wise. After completion of the addition of acetyl chloride, the surface was

155

 

washed with fresh THF and the process was repeated twice to ensure maximum blocking
of the hydroxy groups on PHEMA side chains. The acetylated PHEMA sample was used

as a substrate for ROP of lactide as describe above.

IV-3. Hydrolytic Degradation of PLA film

To study the hydrolytic degradation of PI-IEMA-g-PLA on gold, substrates were
immersed in phosphate buffer solution (pH 7.40) at 55:01 °C. At predetermined times,
the sample was removed from the buffer, washed thoroughly with de-ionized water, and
dried under a flow of N2. The sample was characterized by surface reflectance FTIR and
optical microscopy, and then the sample was returned to the proper buffer solution. We
ran two sets of hydrolytic degradation experiments to test for an autocatalytic effect in the
degradation. In the first, a PHEMA (174 nm)-g-PLA (464 nm) substrate was degraded in
the same buffer solution throughout the degradation process. The second used a PHEMA
(174 nm)-g-PLA (488 nm) substrate, and the buffer solution was replaced every 2 days.
The degradation rate was measured by surface reflectance FT-IR to monitor the decrease

of the PLA carbonyl peak with time.

IV-4. Observation of PI-IEMA-g-PLA Films During Hydrolytic Degradation

Changes in surface roughness of the polymer films during the hydrolytic
degradation of PHEMA-g-PLA were observed using a Nikon 0ptiphot2-POL polarizing

optical microscope equipped with a video camera. Digital images of the dried polymer

156

 

surface and changes in the surface morphology during the drying process were captured

as graphic files.

V. References
( 1) Ciarnpolini, M.; Nardi, N. Inorg. Chem. 1966, 5, 41-4.
(2) Matyjaszewski, K.; Patten, T. E.; Xia, J. J. Am. Chem. Soc. 1997, 119, 674-680.

(3) Matyjaszewski, K.; Paik, H.-j.; Zhou, R; Diamanti, S. J. Macromolecules 2001,

34, 5125-5131.

157