I
.. ; .3.
3...:- {sum
$5.5... Whafinn
O. 1. >0.

n1;

. .5
3.... a.
whyfifi. 1:... I
3. $13.55
I 4: ..
55? 35.. :3223.
. I...

6

Sit. 1";

if: 5.11: .1! .

V 51...; aiz.
:9 t5

:1. 5|? 5:

151.... .

iii... z

'9‘

3.5.0251
v ‘3: II):
I 0.

 

 

).
.15....

’13))1...
.2 :17; 3,

 

 

. Q!
(.Iv

‘ .
5,1.9‘. .l i.
infill}; S.)
{5 3-9.! ’\

p3 ..
0 1170.117:
X; .. .

VMS“

cor-

 

‘ ‘ . . , Q‘Kl
. . 2:25,.
‘ mum - . r; .2. .s 2.32
‘ y: . . waéwti 32%....“
. n r». V V , .. , , a
,. . a. .c “a . . V . .
L .13..
[if a: ‘ :E ,
vtl .3 a 3 .

4... (3?. r. .
: \uVI.

 

This is to certify that the

thesis entitled

EFFECTS OF PRECURSOR MOLECULAR HEIGHT ON THE
GRAFTING OF ULTRATHIN HYPERBRANCHED POLY (ACRYLIC
ACID) FILMS

presented by
SANDRA BENC I C

has been accepted towards fulfillment
of the requirements for

M.S. CHEMISTRY

degree in

 

 

(g V
ajor professor /

0-7639 MS U is an Affirmative Action/Equal Opportunity Institution

 

PLACE m 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

 

A% 3 8 2803

 

DEC 0 3 2003

 

ysisfls

 

 

 

 

 

 

 

 

 

6/01 WW

. III: I Ill .\l.| Illl\Il«Ialtl \llIlll iirllrbll.

EFFECTS OF PRECURSOR MOLECULAR WEIGHT ON THE GRAFTING OF
ULTRATHIN HYPERBRANCHED POLY(ACRYLIC ACID) FILMS

By

Sandra Bencic

A THESIS

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

MASTER OF SCIENCE

Department of Chemistry

2001

ABSTRACT

EFFECTS OF PRECURSOR MOLECULAR WEIGHT ON THE GRAFTING OF
ULTRATHIN HYPERBRANCHED POLY(ACRYLIC ACID) FILMS

By

Sandra Bencic

Hyperbranched poly(acrylic acid) (PAA) films are synthesized with a layer—by-
layer grafting procedure that includes activation of a self-assembled monolayer of
mercaptoundecanoic acid (MUA) to an anhydride, grafting of a,a)-diamino-terminated
poly(tert—butyl acrylate) (PTBA) and hydrolysis to a PAA layer. Further grafting on prior
grafts yields hyperbranched films. The limited control of molecular weight and
polydispersities in a,(o-diamino-terminated PTBA synthesized with free-radical
polymerization does not afford precise control over film thickness and coverage of
porous supports. In contrast the use of atom transfer radical polymerization (ATRP)
allows synthesis of polymers with controlled, low polydispersity molecular weight
distributions. ATRP-synthesized PTBA can be derivatized to yield monoamine
terminated PTBA to allow formation of PAA films. An ellipsometn'c investigation of
hyperbranched grth of up to six PAA layers on gold using various molecular weight
PTBA precursors indicates that the film thicknesses (0.6 to 340 nm) increase non—linearly
with the number of layers and linearly with increasing molecular weight. A calculation of
the number of amide bonds formed in each grafting step indicates that grafting yields

(<4%) are higher in the initial layers and substantially lower in the upper layers.

To my family

iii

ACKNOWLEDGMENTS

The warmest thanks go to my advisor Prof. Merlin Bruening, who encouraged me
to apply to this program and made sure my stay in Chemistry at MSU was all about
“getting smarter and excited” while doing research. Thanks for the helpful professional
and personal advice, always given at the right moment. I also thank all my committee
members for their suggestions and support: thanks to Prof. Gary Blanchard who kept
sending me back to my lab to work, thanks to Prof. Warren Beck for all the encouraging
words. I especially thank Prof. Gregory Baker, who kindly allowed me to use his GPC
setup.

I had a great time working with all the people in Bruening’s group: Jeremy,
Milind, Bo, Wenxi, Kangping, Anagi, Dan, Skanth, Jinhua, Keith, Yingda, Jacque, Brian,
Rodney, Tina and Alex. I sincerely thank Jong-Bum and Wenxi for the greatest help,
training and advice on ATRP, GPC and usage of the glove box, and I also thank the
group members for advice and suggestions on my project. Thanks to the wonderful
friends I met in this department, especially Michelle and Scott, Sani and Igor, Maggie and
Matt, and many others. Thanks to the whole faculty and staff in Chemistry.

I owe who I am today to my parents and to the best brother in the world, Tomi.
Thanks for always believing in me. Thank you, Cathy and Aldo, for caring for me from
my first day of stay in the US. I am also grateful to all my friends: Irena, Kannen,
Katarina, Matej a, Lara, Livi, Vesna, Julija, Ursa, Angela and many others,

I owe my happiness to Milind. Thanks for waiting for me to come to meet you at

MSU. Life cannot get prettier than now, with you.

This work was partially supported by a
grant from the National Science iv
Foundation. (CHE-98 1 6 1 08)

TABLE OF CONTENTS

LIST OF TABLES
...vii
LIST OF FIGURES
......................................................................................................... viii
LIST OF SYMBOLS AND ABBREVIATIONS
............................................................................................................ x
Chapter 1
INTRODUCTION
............................................................................................................ 1
Chapter 2
EXPERIMENTAL
........................................................................................................... 12
I. Background on the use of Gel Permeation Chromatography (GPC)
for determining molecular weight distribution
........................................................................................................... 12
11. Background on ellipsometry
.......................................................................................................... 14
III. Materials
.......................................................................................................... 15
IV. Synthesis, derivatization and characterization of PTBA
.......................................................................................................... 16
A. Synthesis of the initiator 2-bromopropionylsuccinimide (BPS)
......................................................................................................... 16
B. Synthesis of poly(tert-butyl acrylate)-propionyl succinimide (PTBA-
PS) -
........................................................................................................... 1 7
C. Synthesis of amino-terminated PTBA (PTBA-R-NHZ)
.......................................................................................................... 22
D. Characterization of the molecular weight distributions of PTBA using
GPC
........................................................................................................... 22
V. Synthesis and characterization of hyperbranched PAA films
.............................................................................................................. 23
A. Synthesis of films '
...23

B. Ellipsometry
................................................................................................................ 24

C. Extemal-Reflection FTIR (ER-FTIR)
Chapter3

RESULTS AND DISCUSSION .
......................................................................................................... 25

I. Molecular weights and polydispersities of PTBA-PS and PTBA-R-NH2
.......................................................................................................... 25
II. Ellipsometric thickness and film growth
.......................................................................................................... 3 1
A. Thickness values obtained using PTBA-R—NH2 varying in molecular
weight
........................................................................................ 3 1
B. Reaction times for grafting of PTBA-R-NH2
........................................................................................ 32
C. Change in thickness with respect to the number of grafted PAA
layers
....................................................................................... 36
D. Relationship between thickness and molecular weights of PTBA-R-
NH2
........................................................................................ 38
III. External Reflection Fourier T ransfonn Infrared (ER-FTIR) spectroscopy
investigation of film growth
................. 41
IV. Yield of amide formation with grafting of new PAA layers
.......................................................................................................... 46
Chapter 4
SUMMARY AND FUTURE WORK
.......................................................................................................... 53
BIBLIOGRAPHY
.......................................................................................................... 56

vi

LIST OF TABLES

Table 1 Reactant amounts, ratios and reaction conditions used for ATRP synthesis of
PTBA-PS with controlled molecular weights. ................................................... 20

Table 2 Molecular weight distribution parameters obtained by GPC for propionyl
succinimide- and amino-terminated PTBA prepared with a variety of M:I ratios. Mthem —
theoretical molecular weight based on M:I ratios, Mn — number average molecular weight,
Mw — weight average molecular weight, PDI — polydispersity index (MW/Mn).
........................................................................................................... 27

Table 3 Ellipsometric thickness of hyperbranched PAA films grafted on gold using
polymers with several different molecular weights. Thicknesses are given for films before
(PTBA) and after (PAA) hydrolysis of each layer. Thicknesses of MUA monolayers (8-
12 A) were subtracted from each film thickness. 32

Table 4 Tentative peak assignments for absorbance bands in activated PAA (A), grafted
PT BA (B) and PAA (C). ........................................................................... 41

vii

LIST OF FIGURES

Figure 1 Formation of a grafted PAA layer through activation of MUA to produce a

mixed anhydride, attachment of a,co-diamino-tenninated PTBA chains by formation of
amide linkages, and hydrolysis of tert-butyl (tBu) groups to carboxylic acids. Adapted
from reference 13. The many underivatized —COOH groups in the MUA layer (after
reaction with HzN-R-PTBA-R-NHZ) are not shown for figure clarity.
............................................................................................................ 3

Figure 2 Scheme of the synthesis of a,a)-diarnino-terminated PTBA using free-radical
polymerization initiated by 4,4’-azobis(4-cyanovaleric acid), followed by end group
derivatization with ethylenediamine (adapted from references 1, 13)....... . . . . . . . . . . . . . . ...6

Figure 3 a) ATRP of TBA using 2-ethyl bromopropionate as an initiator. Et = ethyl; tBu
= tert-butyl; L = pentarnethyldiethylenetriamine (PMDETA). b) Derivatization from
bromide through azide and iminophosphorane to amine. Adapted from references 23, 25.

........................................................................................................... 8

Figure 4 Derivatization of the succinimide-protected initiator end of PTBA with
ethylenediamine to give an amino-terminated polymer. 10

Figure 5 Gel permeation chromatography setup. Adapted from reference 27. 1 3

Figure 6 Reaction scheme for the synthesis of the 2-bromopropionyl succinimide (BPS)
initiator. .............................................................................................. 17

Figure 7 Proton NMR spectra of propionyl succinimide-terminated PTBA (PTBA-PS, a)
and amino-terminated PTBA (PTBA-R-NHZ, b) prepared fi'om a M21 ratio of 25:1. Water
peaks were smaller in spectra of polymers prepared using higher M:
I ratios. ............................................................................................... 21

Figure 8 Variation of the time required for reaction of anhydride groups with amino-
terminated PTBA solutions as a function of molecular weight and the number of PTBA
layers. Values represent an average of reaction times on 3 samples. Error bars represent
the standard deviations. ............................................................................. 33

Figure 9 ER-FTIR spectra of anhydride peaks in an activated 3 layer PAA film before (a)

and after reaction with PTBA-R-NH2 for 17 (b), 21 (c), 25 (d), 29 (e) and 32 (f) hours.
The PTBA-R-NH2 used in this experiment had a molecular weight of 93,300. ............ 35

viii

Figure 10 Thickness of hyperbranched PAA films (a) before (PTBA) and (b) after
hydrolysis of each layer to PAA, as a function of molecular weight. Error bars represent
standard deviations. .................................................................................. 37

Figure 11 Plots of film thickness versus PTBA-R-NI-I2 molecular weight (Mn) for 1-6
layer PAA layers before (top, PTBA was just grafted) and after hydrolysis (bottom).
.......................................................................................................... 40

Figure 12 ER-FTIR spectra of two (a) and five (b) layers of activated PAA, and the same
films after attachment of PTBA (c, d) and subsequent hydrolysis (e, f). The PTBA-R-NH2
used in the synthesis of these films had a M“ value of 3,620. ................................. 43

Figure 13 ER-FTIR spectra of 1-6 PAA (a-f) layers grafted using precursor PTBA-R-
NH2 with a molecular weight of 29,800. ........................................................... 45

Figure 14 (a) Increase of the acid carbonyl absorbance in ER-FTIR spectra plotted versus
the number of grafted PAA layers. (b) Absorbance due to the acid carbonyl peak in PAA
films plotted with respect to the molecular weights of PTBA-R-NH2 used in film
formation. Legends indicate either the molecular weight of PTBA-R—NH2 used for
preparing films or the number of layers in the film. ........................................... 47

Figure 15 A schematic representation of grafting of the first three layers of PAA by
forming xm amide linkages with anhydride-activated carbonyl groups during grafting of
each new m-th layer. The total number of carbonyl groups contained in a film after
performing the m-th grafting step is expressed as cm. ........................................... 49

Figure 16 (a) Representation of the number of amides formed in subsequent steps of

grafting PAA as calculated from the intensity of carbonyl groups. (b) Amide yield plotted
with respect to every grafting step in PAA film formation” 5.2

ix

LIST OF SYMBOLS AND ABBREVIATIONS

ATRP - Atom Transfer Radical Polymerization

BPS — bromopropionyl succinimide

cm — number of acid carbonyls in the film after grafting of m layers
DMEDA — dimethylethylenediamine

DMF - dimethylformamide

DMSO - dimethylsulfoxide

ER-FTIR — External-Reflection Fourier Transform Infrared Spectroscopy
Im — absorbance due to the acid carbonyl group in ER-FTIR spectra after grafting the m-th
layer

MALDI-TOF-MS — Matrix—Assisted Laser Desorption Ionization Time-of-Flight Mass
Spectrometry .

m — number of layer grafted

M:I — monomer to initiator ratio

Min 1" molecular weight of the initiator end

Mn — number average molecular weight

Mtheor — theoretical molecular weight based on M21

MW — weight average molecular weight

MUA — mercaptoundecanoic acid

n —- number of carbonyl groups on one chain

GPC - Gel Permeation Chromatography

PAA — poly(acrylic acid)

PDI - polydispersity index

PMDETA — N, N, N’, N’, N”-pentamethyldiethylenetriamine

PMMA - poly(methyl methacrylate)

PTBA — poly(tert-butyl acrylate)

PTBA-R—NH2 — amino-terminated poly(tert-butyl acrylate)

PTBA-PS — propionylsuccinimide-tenninated poly(tert-butyl acrylate)
PTFE — poly(tetrafluoroethylene)

stOH-HZO — p-toluenesulfonic acid monohydrate

SIMS - Secondary-Ion Mass Spectrometry

TBA - tert-butyl acrylate

tBu -— tert—butyl

THF - tetrahydrofuran

xm — number of amides formed during grafting of the m-th layer

Chapter 1

INTRODUCTION

Ultrathin hyperbranched poly(acrylic acid) "(PAA) films are interesting because
they can be readily derivatized,“2 and thus, tailored for a variety of possible applications.
Such applications may include chemical sensing,l corrosion passivation and inhibition of
electrochemical reactions on metal surfaces?“ entrapment of enzymes,“‘7 control of cell

8'9"°'” separation of gases,12 and purification of protein

growth on patterned surfaces,
solutions. These films are especially attractive because their wide range of thicknesses (6
to 100 nm)3 can be controlled by varying the number of surface-grafted PAA layers. This
work focuses on determining relationships between the molecular weight of amino-
terminated poly(tert-butyl acrylate) (PTBA) precursor chains and the thickness and
structure of hyperbranched PAA films.

Synthesis of hyperbranched PAA on gold occurs by a layer-by-layer grafting
process. Growth of the first layer begins with grafting of amino-terminated poly(tert-
butyl acrylate) onto a mercaptoundecanoic acid (MUA) monolayer that was previously
activated to a mixed anhydride. Hydrolysis of tert-butyl ester groups to carboxylic acids
then yields one grafted PAA layer as shown in Figure 1.'3 Further grafting on prior layers
yields a hyperbranched film. When deposited on gold-coated porous alumina supports,
such PAA films can be used as ultrathin membranes. In that particular case, the low film

thickness allows a high flux of gases or fluids through the membrane as well as

selectivity, provided that the film fully covers the pores of the underlying support.12 The

Figure 1 Formation of a grafted PAA layer through activation of ML’A to produce a

mixed anhydride, attachment of a,o)-diamino-terminated PTBA chains by formation of
amide linkages, and hydrolysis of ten-butyl (tBu) groups to carboxylic acids. Adapted
from reference 13. The many underivatized —COOH groups in the MUA layer (after
reaction with HEN-R-PTBA-R-NHZ) are not shown for figure clarity.

 

 

 

hydrolysis with
p-TsOH-HZO, 55°C

 

ability to simultaneously control film thickness and coverage of pores demands a
thorough understanding of the PAA grafting process.

Koros et a1. ”“5

reported that the most important factor for fabrication of thin fihns
that cover porous substrates is control of the hydrodynamic radius of the polymer chains
that form the film. Coverage without filling of pores occurs when polymers with suitably
high molecular weight are utilized. To cover pores without filling them, the
hydrodynamic radii of polymer chains should be equal to or greater than the radii of
support pores.” In hyperbranched films, grafting of a high number of layers insures
complete coverage of pores, but it also reduces flux. Grafting of only a few layers of PAA
with higher molecular weight constituents may provide both pore coverage and high
fluxes. An understanding of how PAA film thickness relates to the molecular weight of
precursor PTBA chains is thus an important step in the study of hyperbranched
membrane formation.

There are no previous studies on how the molecular weight of precursor PTBA
influences the grth of hyperbranched PAA films. However, studies demonstrating the
impact of molecular weight on film thickness do exist for spin-coated polystyrene fihns
with thicknesses between 2 and 2000 nrn.“"'l7 Work done by Schubertl7 shows that the
film thickness is directly related to the concentration of polymer in solution, spinning
velocity, and molecular weight. Since the spin-coating procedure produces films
physisorbed on surfaces, the same relation might not apply to covalently grafted
hyperbranched PAA films.

Interestingly, ' Bartlett and coworkers'6 showed that film thickness is a linear

function of molecular weight for spin-coated polystyrene films that are covalently

immobilized on silicon surfaces via a photochemical reaction. In that study, residual
polymer was removed after linkage of polystyrene chains to a monolayer on the surface.
Thus this system is somewhat similar to the growth of the first PAA layer on gold.
Growth of subsequent layers of PAA is unique, however, because of the nature of the
hyperbranched process. The increase in the number of active sites upon grafting of every
subsequent layer yields a highly branched structure and allows rapid increases in
thickness.

Precise control of thickness and identification of its relationship with molecular

weight is possible only if polymers are available with controllable molecular weight and

low polydispersity. The synthesis of a,o)-diamino-terminated PTBA reported by
Bruening et a1.”3 is a free-radical process and thus yields a broad, rather uncontrolled
molecular weight distribution. That synthetic scheme is shown in Fig. 2. Although
Bergbreiter and coworkers utilized a purification procedure to reduce polydispersity to
1.9-2.0,7'l8 lower polydispersities are still needed for a careful investigation of film
growth as a fimction of molecular weight. The main reasons for the very broad molecular
weight distribution in free-radical polymerization are fast kinetics and the difficulty of
avoiding side-reactions.

Another source of polydispersity in a,co-diamino-terminated PTBA are different
mechanisms for termination of polymerization (Figure 2). A chain initiated by one end of
an initiator radical (4—cyanovaleric acid radical) can be terminated with a hydrogen

radical (a), another initiator radical (b), or with another propagating chain (0). Formation

rHs $113
CH2=CH + HOOC-CH2CH2—(IZ-i-NzNag-(IZ—CHZCHz-COOH
I

 

 

(|::o CN CN
0 p-Dioxane
120°C, ~20 hrs
TBA
'Nzi
v 5 = H,
CIIHs (I311,
Hooc-CH2CH2-CIZ—EH2—CIIPH-nfl 0, &= ——(l:-—CH2CH2—COOH,
CN C|3=O CN
l
or 1_2_=—[CH2—c1fl—c—CH2CH2—C00H
+ 1,l'-carbonyl diirnidazole |_ n I
. . c_o CN
+ ethylenedlarmne I
in CH2C12 o-|—
HzN-CHZCHz-NH—C—CHZCHz—(lj—EjHZ—clng—nl}:
CN (|3=O
0+
B.’ = H,
rm
0, B.’ = —(|3—-CH2CH2—CO— NH—CH2CH2—NH2,
CN
(fins
0,. flrEcHz—pnjqclz—CHZCHz—Co— NH—CHZCHZ— NH2
(II=O CN
0+

Figure 2 Scheme of the synthesis of or,m-diamino-tenninated PT BA using free-radical
polymerization initiated by 4,4’-azobis(4-cyanovaleric acid), followed by end group

derivatization with ethylenediamine (adapted from references 1, 13).

(a)

(b)

(C)

(a)

(b)

(C)

of molecules from two propagating radicals (c) thus provides an increase in
polydispersity. The molecular weight distribution can also broaden upon reaction of
carboxylic acid-terminated PTBA with ethylenediamine in the second reaction step.
Because ethylenediamine has two amine termini, it can sometimes connect two polymer
chains, thus increasing polydispersity.

To prepare polymers with lower polydispersities, Matyjaszewski et a1.‘9'2°'2"22
developed atom-transfer radical polymerization (ATRP). In this technique, the use of a
controlled catalytic initiation system (Cu‘, Cu"/Ligand) decreases the rates of radical
polymerization, allowing more precise control of molecular weight distribution. Amongst
a variety of polymers prepared by ATRP, synthesis of PTBA was reported with
polydispersities ranging between 1.04 and 1.5.23'24 Moreover, Matyjaszewski and
coworkers prepared amino-terminated polymers from bromine-terminated poly(methyl
acrylate) or poly(butyl acrylate) synthesized by ATRP.”25 The ability to obtain PTBA
with a single amino terminus should be extremely useful for interpretation of amide
formation when grafting PTBA onto surfaces, as it will avoid cross-linking reactions.

In Figure 3a, we show the probable ATRP mechanism for formation of PTBA
when 2-ethylbromopropionate is used as initiator. During ATRP, propagation is
attenuated because the Cu"-pentamethyldiethylenetriamine (PMDETA) complex forms
after initiation and moves the equilibrium for radical formation to the’left. The lower
concentration of free radicals results in a decreased rate of polymerization. The addition
of Cu" to the polymerization mixture is necessary for obtaining low polydisperisties when

synthesizing polymers from low ratios of monomer to initiator, since little Cu" is formed

CH3—CH

a) CH3—CH— Br + Cu(I)L I
<|3=0 *2 (l3=0 + BrCu(II)L
OEt OEt
+ CH2: (EH
C: o
OtBu
CH3—CH—CH2—(IZH— Br CH3—(IZH—CH2—(FH
+
C: O C: O + Cu(I)L C=O C: O BrCu(II)L
I l «=- l |
OEt OtBu OEt OtBu
+ (n-l) CH2: CH
(2: o
OtBu

CH3—(IIH-ECH2—(IZH3—Br CHg—(IZH—ECHz—(IZHE-ECHz—CH
c=o c=o

_ _ + Cu(I)L
C—0 C—0 1 C + BrCu 11 L
l l <——- | | l ( )
OEt OtBu OEt OtBu OtBu

+ N N + PPh
b) H-Br —L> 17.N3 —i> 17-N-PPh3

molysis

17-NH2

17 = CH 3—CH—ECH2—CIJH3;
(|:=o (l:=o
OEt OtBu

Figure 3 a) ATRP of TBA using 2-ethyl bromopropionate as an initiator. Et = ethyl; tBu
= tert-butyl; L = pentamethyldiethylenetriamine (PMDETA). b) Derivatization fi'om
bromide through azide and iminophosphorane to amine. Adapted from references 23, 25.

in the initiation process. Decreases in temperature also decrease ATRP kinetics. Overall,

by control of temperature and the ratios of components in reaction mixtures, synthesis of

low dispersity polymers is possible using ATRP.

The use of ATRP for forming amino-terminated PTBA requires a method for
converting the bromine-terminated PTBA to an amine. Figure 3b shows the bromine end
group derivatization proposed by Coessens et 01.25 The reaction begins with azidation of a
bromine end group with NaN3. Substitution of the azide with PPh,26 yields an
iminophosphorane that is hydrolyzed to amine in an excess of H20. This synthesis was
successful for low molecular weight polymers.23 Because of the ease of hydrolysis of tert-
butyl groups with acids or bases, the conditions for specific hydrolysis of the
iminophosphorane must be extremely mild. However, hydrolysis of iminophosphorane to
amine in water is a slow process. Additionally, in the case of high molecular weight
polymers, precipitation of polymer chains in water impedes a quantitative derivatization
if reaction solutions are not extremely dilute.

Although a correct choice of conditions for hydrolysis might ensure complete
conversion of iminophosphoranes to amines in water, confirmation of amination by end
group characterization is difficult. Characterization of end groups by NMR is suitable for
low molecular weight polymers but becomes problematic in the case of high molecular
weight polymers prepared using high ratios of monomer to initiator. Matrix-Assisted
Laser Desorption Ionization Time-of-F light Mass Spectrometry (MALDI-TOF—MS) is
also reported as an optional technique for end group characterization, but has been
successfully used primarily with polymers having low molecular weight ranges.26
Therefore the existence of amine ends in high molecular weight polymers cannot be
demonstrated with the above mentioned techniques. If a reaction used for end group
derivatization to an amine is fast and provides quantitative derivatization in low

molecular weight polymers, one can assume that the reaction is also valid for high

molecular weight polymers. We used a reaction that fulfilled the requirement of fast and
quantitative derivatization over the entire molecular weight range of ATRP-synthesized
polymers.

Our synthetic scheme for preparing amino-terminated PTBA involves protection
of the carbonyl group on a 2-bromopropionylbromide initiator, use of this initiator for
ATRP of TBA, and subsequent derivatization of the protected group on the initiator to
provide an amine terminus. First, the initiator is derivatized with N-hydroxysuccinimide,
which yields a 2-bromopropionylsuccinimide ester (Figure 4). After ATRP using this
initiator, derivatization with an excess of ethylenediamine (substitution of succinimide
with ethylenediamine on the ester carbonyl) yields amino-terminated PTBA (PTBA-R-

NHz; R = -CH(CH,)—CO-NH-(CH2)2-) as shown in Figure 4. The only drawback of this

CH3—CH—ECH2—(lxfl—nar +H2N-CH2CH2—NH2 —* CH3-CIZH—BiH2—(IZIE—nBr

l
i=0 i=0 0=$ $=°
(I) OtBu NH OtBu
o N 0 (EH?
V are
NH2
PTBA'PS PTBA-R-NHZ

Figure 4 Derivatization of the succinimide-protected initiator end of PTBA with
ethylenediamine to give an amino-terminated polymer.

procedure is the presence of an amide bond in amino-terminated chains. The presence of
this amide bond will complicate monitoring of amidation resulting from the grafting

reaction.

10

Synthesis of amino-terminated PTBA using ATRP yielded narrow molecular
weight polymers with polydispersities between 1.1 and 1.33. Derivatization to an amine
end was successful and was confirmed by NMR for low molecular weight PTBA-R-NHZ.
The presence of amine groups was also confirmed by attachment of chains during grth

of films on MUA monolayers. The thickness of PAA films increased non—linearly with

the number of layers, as in previous reports on a,w-diamino-terminated PTBA prepared
by free-radical polymerization, but film thickness varied approximately linearly with
respect to increasing molecular weight of polymers. We quantified the derivatization of
layers with precursor PTBA-R-NHZ, using information from extemal-reflection Fourier
transform infrared (ER-FTIR) spectroscopy. Using carbonyl intensities we calculated the
number of amide linkages formed upon grafting of every layer. From these results, the
yields of derivatization with amino-terminated PTBA were extremely small (<l%) in the

upper layers and slightly larger (<4%) in the initial layers.

11

Chapter 2

EXPERIMENTAL

I. Background on the use of Gel Permeation Chromatography (GPC) for

determining molecular weight distribution

There are several techniques available for determination of the molecular weights
of polymers. Osmometry, viscometry and light-scattering measurements, for example,
yield average molecular weight values,27 but provide no information about molecular
weight distribution. Mass spectrometry techniques (MALDI-TOF-MS, and Secondary
[on Mass Spectrometry, SIMS) are sometimes used for determining molecular weight
distributions,”28 but these methods are limited in their sensitivity to high molecular
weight polymers. MALDI-TOF-MS was already successful in determining molecular
weights of ATRP-synthesized PTBA,29 but so far, this technique was used only for
polymers with low molecular weights (~ 4,000 Da).

The challenge in determining molecular weight distributions of polymers can be
successfully overcome using Gel Permeation Chromatography (GPC) and subsequent
detection. Provided a large range of polymer standards are available, GPC has little
limitation in the molecular weight ranges it can examine. The principle of GPC is a
separation by size exclusion, and GPC columns are packed with polystyrene or glass
porous beads that entrap small polymer molecules. Molecules larger in size can not enter

these beads and thus have a shorter pathway through the column. Consequently,

12

molecules with low molecular weight travel through a column more slowly than larger
molecules. Elution times can be compared to calibration values for standard polymers
with narrow molecular weight distributions.

Most often, detectors used for characterization of eluted polymer solutions are
based on either a change in refractive index or UV absorption. The refractive index
detector measures the difference in refractive indices between the sample and the solution
used as the mobile phase. A single wavelength or diode array UV detector, detects
transmission intensities of eluted UV-active samples. Figure 5 is a schematic diagram of
a GPC setup with basic components in the system: an inlet, mechanical pump with

pressure gauge, column, feed and collector reservoirs and a detector with data collection.

Injection port

_, __“_.>[ Column h

 

 

 

 

Pump with

y Detector
pressure gauge

 

 

 

 

 

Data system

  

Collector
flask

Solvent
reservoir

Figure 5 Gel permeation chromatography setup. Adapted from reference 27.

Results obtained in GPC analysis are generally expressed with a number-average
molecular weight (Mn), weight-average molecular weight (MW) and polydispersity index

(PDI). They are defined in the following way:

M =Zfl (1)

n ZN:
2(Mi)2Ni

 

M,=——— 2

w ZMiN’. ()

PDI=M"’. (3)
M

II

N) represents the number of chains with mass M,- in the distribution.28 The value of the
peak maximum determined from a chromatogram of an unknown polymer sample

correlates best with the root-mean-square molecular weight, defined as

M m = (M w -M n )“2. Additional corrections are necessary for correlation of Mw or Mn

with the peak maximum in the case of more complicated distributions.” From analysis of
chromatographic peaks and the presented definitions, one can obtain values for MW, Mn

and PD].27

[1. Background on ellipsometry

Ellipsometry is a very sensitive technique suitable for determination of film
thicknesses as low as a few angstroms. The method is based on detection of changes in
the amplitude and phase of incident polarized light after reflection from a sample. The

polarized light is divided into parallel (p) and perpendicular (5) vector components whose

14

properties upon reflection are described by Fresnel equations.30 A definition relates the

ellipsometrically measured angles A (phase) and ‘P (amplitude), and the Fresnel

reflection coefficients (Rp, R) that describe the polarized beamz”
p = —” = tan(‘P)exp(iA) (4).

The Woollam M-44 ellipsometer is composed of a light source (Xe lamp), optical
fiber, beam modulator, polarizer and detector. Samples usually consist of multiple layers
(metal and organic layers) each of which is defined by a refractive index, 71' = n + ik , and
a thickness. Since these sample parameters are not directly measurable, the ellipsometric
experiment requires a model for fitting experimental data. The model contains known
parameters (wavelength of incident light, angle of incidence, thickness of substrate
layers) and unknown parameters such as optical constants (n and k) or film thickness.
After a measurement, measured values of A and ‘1’ are fitted to the proposed values of

thickness and refractive index, and one can obtain accurate values for the film thickness.

111. Materials

T ert—butyl acrylate (TBA, Aldrich, 98%) was washed three times with 0.5 M
NaOH and three times with water, passed through a column of basic alumina and distilled
under vacuum. N,N,N’,N’,N"-pentamethyldiethylenetriamine (PMDETA, Aldrich, 99%)
was distilled under vacuum, and acetone (Aldrich, 99.5%) was distilled at atmospheric
pressure. Reagents for ATRP synthesis (TBA, PMDETA and acetone) were then

degassed with freeze-pump-thaw cycles and introduced into a nitrogen-filled dry box (M.

15

 

Braun; ~3 ppm water, ~20 ppm 02), while CuBr (Aldrich, 99.999%) and CuBr2 (99%,
Aldrich) were introduced in the dry box as received. Dimethylformamide (DMF, Aldrich,
99.8%) was dried over molecular sieves prior to use. 2-Bromopropionyl bromide (97%),
N-hydroxysuccinimide (97%), pyridine (99+%), tetrahydrofuran (THF, 99+%), 2-
propanol (99.5%), ethylenediamine (99.5%), dichloromethane (99.6%), methanol
(99.8%), mercaptoundecanoic acid, 4-methylmorpholine (99%), ethyl chloroformate
(97%), benzene (99+%) and acetonitrile (99.5%) were purchased from Aldrich and used
as received. Ethyl acetate (Spectrum), p-toluenesulfonic acid monohydrate (Spectrum)
and ethanol (Pharmco, 100%) were used as received. Deionized water (18.2 MQ-cm) was
obtained by purification with a Milli-Q system. As substrates for grafting of

hyperbranched PAA films, we used silicon wafers purchased from Siliconquest (dopant

P/Boron, <100>, 2-5 Q~cm, thickness 450-475 pm) that were sputter-coated with 20 nm

of chromium and 200 nm of gold by Lance Goddard Associates.

IV. Synthesis, derivatization and characterization of PTBA

A. Synthesis of the initiator 2-bromopropionyl succinimide (BPS)
N-hydroxysuccinimide (2.66 g, 22.9 mmol) and pyridine (1.85 mL, 22.9 mmol) were

dissolved in 25 mL of THF, stirred, and cooled in an ice bath. A solution containing 2-

bromopropionylbromide (2.88 mL, 27.5 mmol) dissolved in 25 mL of THF was

introduced dropwise and the reaction mixture was stirred for 12 hours. A white

16

precipitate of pyridine hydrobromide formed and was removed by filtration and washed
with THF. The filtrate was concentrated by rotary evaporation to ~5 mL. The product
precipitated upon addition of 30 mL of water. After filtration, this solid was recrystallized
from 2-propanol, filtered and dried in-vacuo. The reaction scheme for this synthesis is
shown in Figure 6. IH NMR (300 MHz, CDCl,) confirmed isolation of 2—bromopropionyl
succinimide. The spectrum contained peaks corresponding to the methyl group (d, 1.95
ppm), methylenes from the succinimide ring (3, 2.85 ppm) and methine protons (q, 4.6
ppm). Also, the KBr F TIR spectrum of this compound contained an ester carbonyl stretch

at 1740 cm", succinimide carbonyl stretches at 1785 and 1817 cm", and an amide stretch

at1644 cm".
CH3‘CH—BI'
CH B (EH I“)
CH3“ — 1' .d. '—
l=o .. 0 N o 213:, O
. V m .

BPS

Figure 6 Reaction scheme for the synthesis of the 2-Bromopropionyl succinimide (BPS)
initiator.

B. Synthesis of poly(tert-butyl acrylate)-propionyl succinimide (PTBA-PS)
Due to the sensitivity of AT RP to the presence of oxygen, this synthesis was

performed in a dry box filled with nitrogen. Polymers with several different molecular

l7

weights were synthesized following a slight modification of a procedure described by
Matyj aszewski and coworkers.”24 For polymers formed using a monomer to initiator

ratio of 25:1, CuBr catalyst (254 mg, 1.08 mmol) and CuBr2 deactivator (12.6 mg, 0.054

mmol) were dissolved with PMDETA (236 uL, 1.13 mmol) in 0.7 mL of acetone. The
catalyst/deactivator solution was added to tert-butyl acrylate (TBA, 7.85. mL, 54 mmol)
in a round-bottom flask and stirred until most of the solid dissolved. According to
previously developed procedures,” the catalyst and deactivator may not dissolve
completely in the reaction solution. Nevertheless, the actual complexed catalyst and
deactivator in solution represent sufficiently high concentrations to control the kinetics of
living radical formation. Most likely some of the CuBr and CuBr2 are insoluble in
acetone. For monomer to initiator ratios >125, only the catalyst (CuBr) was added.
Because the ratio between monomer and initiator was high in these instances and
therefore conversion of monomer was slower,24 no addition of deactivating CuBr2 was
necessary. The initiator, BPS (540 mg, 2.16 mmol), was dissolved in 0.7 mL of acetone
and then added to the solution containing monomer and the catalyst/deactivator complex
with PMDETA. The reaction mixture was immediately placed in an oil bath at 65°C. Five
additional polymers with higher molecular weights were prepared by following a similar
reaction scheme as for the polymer with a 25:1 monomer to initiator (M:I) ratio. We used
molar ratios between monomer and initiator of 100:1, 200:1, 250:1, 500:1 and 1000:l.
The amounts of reactants used for the ATRP syntheses of PTBA of several molecular
weights, as well as reaction times and temperatures, are listed in Table 1. Upon

completion of polymerization, the ATRP-synthesized PTBA was removed from heat and

18

purified using the following procedure. After addition of 30 mL of THF, the polymer
solution was stirred until homogeneous. To remove the catalyst, the polymer solution was
then passed through a column of basic alumina and eluted with 100-150 mL of THF.
After removal of solvent by rotary evaporation, the round-bottom flask was immersed in
a water bath (45-50°C) and the polymer was dried in-vacuo. Peaks attributed to
resonances of methylene protons from the succinimide ring (5, 2.8 ppm) were observed in
the lH NMR spectra of all polymers except for the one prepared with a monomer to
initiator ratio of 1000:1. Other peaks present in all spectra were assigned to backbone
protons (broad peaks at 2.2 ppm and 1.79 ppm) and tert-butyl groups (5, 1.41 ppm). A
small peak at 1.21 ppm might be attributed to methyl groups in the propionate initiator,
but the intensity was inconsistent in spectra of polymers with higher molecular weights.
The NMR spectrum of PTBA-PS is shown in Figure 7a. The vibrations in the KBr F TIR
spectrum of a low molecular weight polymer (~6,000) contained barely visible shoulders
at 1817 and 1792 cm‘1 assigned as succinimide carbonyl stretches. The polymer
synthesized from a M:I ratio of 125:1 had an Mn of 6,000, lower than expected, because
the polymerization time was too short to react all of the monomer. We did not use this
material for film formation but it served for FTIR characterization of the propionyl

succinimide- and amino-terminated compounds.

19

Table l Reactant amounts, ratios and reaction conditions used for ATRP synthesis of
PTBA-PS with controlled molecular weights.

 

Ratio TBA BPS CuBr CuBr2 PMDETA Acetone T Time

 

 

 

 

 

 

 

M:I rnrnol mmol mmol mmol mmol mol. % °C b
(mL) (mg) (mg) (mg) (uL) (mL)
25:1 54 2.2 1.1 0.054 1.1 25 65 18
(7.9) (540) (150) (13) (240) (1.4)
100:1 54 0.54 0.27 0.014 0.28 25 ‘ 65 26
(7.9) (140) (39) (3) (59) (1.4)
125:] 54 0.43 0.22 0.011 0.23 25 65 12
(7.9) (108) (31) (2) (47) (1.3)
200:1 54 0.27 0.27 - 0.27 25 65 20
(7.9) (68) (39) (56) (1.3)
250:1 54 0.22 0.22 - 0.22 25 60 24
(7.9) (54) (31) (45) (1.3)
500:1 54 0.11 0.11 - 0.11 25 60 48
(7.9) (27) (16) (23) (1.3)
1000: 1 54 0.054 0.054 - 0.054 25 65 123
(7.9) (14) (8) (11) (1.3)

 

20

/ /tert-butyl

HO

succinimide

 

 

 

 

 

 

 

 

 

 

 

 

backbone
methylene protons
3 2 1 ppm
b acetone
3 2 1 ppm

Figure 7 Proton NMR spectra of propionyl succinimide-terminated PTBA (PTBA-PS, a)
and amino-terminated PTBA (PTBA-R-NHZ, b) prepared from a M:I ratio of 25:1. Water
peaks were smaller in spectra of polymers prepared using higher M:
I ratios.

21

C. Synthesis of amino-terminated PTBA (PTBA-R-NHz)

A round-bottom flask filled with a 0.5 M solution of PTBA-PS in CHzCl2 was
degassed and filled with nitrogen three times. Subsequently a 12-hour reaction with a
500-fold excess of ethylenediamine allowed quantitative substitution of the succinimide.
The high excess of ethylenediamine was added to ensure a virtually quantitative
derivatization and reaction of an ethylenediamine with only one succinimide-protected
polymer chain. The excess reactant was removed by extracting three times with saturated
acqueous NaCl. After solvent removal, the polymer was dried in-vacuo, dissolved in 20
mL of THF and separated from any remaining ethylenediamine by precipitation of the
polymer upon addition of an 80 mL mixture of deionized water and methanol (1:1). ‘H
NMR spectra of in-vacuo dried polymer differed from PTBA-PS spectra in the fact that
the peaks due to methylene protons from succinimide disappeared (Figure 7b). We were
not able to assign a peak for methylene protons from ethylenediamine end groups, as
these peaks were likely masked by the peaks assigned to backbone protons. KBr FTIR
spectra of low molecular weight polymer also confirmed this derivatization, because the
vibration previously assigned to the succinimide initiator end group disappeared in

amino-terminated PTBA.

D. Characterization of the molecular weight distributions of PTBA using GPC
Narrow molecular weight poly(methyl methacrylate) (PMMA) samples

(Polysciences) were used as calibration standards in GPC analysis of PTBA. Amongst

commercially available standards, these polymers are the most similar to PTBA in

structure and properties. The calibration curve was generated using THF solutions of

22

standards (lmg/mL) with molecular weights of 6,000, 12,000, 60,000 and 127,000 g mol'
'. Samples of PTBA-PS and PTBA-R-NH2 synthesized at various monomer to initiator
ratios were also dissolved in THF (lmg/mL). For elimination of solid particles and

impurities, solutions of polymers in THF were filtered through Whatrnan 0.2-um

poly(tetrafluoroethylene) (PTFE) syringe filters. Samples and standards (100 11L) were
injected in a rotary sample valve of a GPC chromatograph equipped with a PLgel 20-m
Mixed A column. Tetrahydrofuran (99.9+% inhibitor-free) was used as mobile phase.

The eluent was analyzed by a Waters 410 Differential Refractometer.

V. Synthesis and characterization of hyperbranched PAA films

A. Synthesis of films

Prior to film formation, the surface of a gold-coated silicon wafer was cleaned in a
UV/ozone cleaner for 15 minutes. Immersion of the cleaned substrate in a 2 mM solution
of mercaptoundecanoic acid in ethanol for 30 minutes followed by rinsing with ethanol
and drying with a stream of nitrogen resulted in a monolayer terminated with —COOH
groups. A surface anhydride was obtained after a 10-minute immersion of the wafer in a
solution containing 100 uL of ethylchloroformate and 100 uL of 4-methylmorpholine in
10 mL of DMF. The sample was then rinsed with ethyl acetate, dried under nitrogen and
immersed in a solution of PTBA-R-NH, in DMF (180 mg/ 4 mL). Grafting of PTBA
chains to the surface was allowed to continue until the disappearance of anhydride peaks

was observed in extemal-reflection FTIR spectra. Samples were removed from the PTBA

23

solution, rinsed with ethanol and dried with nitrogen. If their FTIR spectrum showed
reduction of anhydride intensities to <5% of their initial value, we proceeded with the
next step. Otherwise, substrates were again immersed in the PTBA solution. Hydrolysis
of tert-butyl groups was achieved through a one-hour immersion of wafers in benzene
saturated with p-TsOH-HZO (SS-60°C) followed by rinsing with ethanol. Repetition of the
cycle including activation, PTBA grafting and hydrolysis was used to graft up to six PAA

layers.

B. Ellipsometry

Film thicknesses were measured using a J. A. Woollam M-44 ellipsometer that
utilized a Xe light source (LPS-300) and an incident angle of 75°. WVASE32 software
provided fitting of thickness, where the estimated 11 for the polymer film was chosen to be

1.5.

C. External-Reflection FTIR (ER-FTIR)

Extemal-reflection FTIR spectra were measured with a Nicolet, MAGNA 560
spectrophotometer containing a mercury/cadmium telluride detector and a PIKE grazing
angle (80°) attachment. Polymer films were deposited on gold-coated silicon substrates,
and extemal-reflection FTIR spectra were acquired at an 80° incident angle using 128

scans and 4 cm’I resolution.

24

Chapter 3

RESULTS AND DISCUSSION

1. Molecular weights and polydispersities of PTBA-PS and PTBA-R-NH2

A number of factors such as initiator concentration and the Cul to Cu11 ratio affect
the polydispersity of PTBA-PS. As shown previously in the experimental conditions
presented in Table l, we varied the concentrations of CuBr and Cu"Br2 according to the
amounts of monomer and initiator present in a reaction mixture. In the case of polymers
synthesized from low M:I ratios (25:1 and 100:1), both catalyst (Cu‘) and deactivator
(Cu”) were present. The molar ratio between the initiator, Cu', Cu", and PMDETA was 1:
0.5:0.025:0.525. For polymers prepared from M:I ratios above 125:1, we used only Cu1
and a ratio between initiator, CuI and PMDETA of 1:12]. In these instances we assumed
that the amount of monomer was already high compared to the initiator, therefore
addition of Cu" was unnecessary. Cu" would additionally decrease the rate of formation
of active radicals. The conditions presented above were similar to literature procedures
used for synthesis of low and high molecular weight PTBA. ”22‘2“”

As polymerization should be slower for low ratios of M:I when Cu" is present, we
chose relatively long reaction times for these polymerizations. A correct choice of M:I

ratios and temperatures enabled, in most instances, synthesis of polymers with molecular

weight values close to those predicted by the M:I ratio.

25

Theoretical molecular weights (Mmm) contained in Table 2 were calculated for
both propionyl succinimide- and amino-terminated PTBA using equation 5. Here, n
represents the number of monomers per initiator (monomer molecular weight of 128, 79.9
is the molecular weight of the Br end group). Mini stands for the molecular weight of the
propionyl succinimide (PS) initiator group in PT BA-PS (-CH(CH3)-COO-
N[(CO)2(CH2)2]; Mini = 170) or the ethylenediamine-terminated group (R-NHZ) in PTBA-
R-NH2 («CH(CH3)-CO-NH-(CH2)-NH2; Mini = 115).

M = n-128+M,.m. +799 (5)

them
The calculation assumes termination by bromine, as shown in Figure 4.

Control of the molecular weight of PTBA is not a sufficient condition for
extensive use of these polymers. The synthesis must also be reproducible. We tested the
reproducibility of the synthetic procedure for the lowest molecular weight compound,
which was synthesized fi'om a ratio of M:I of 25:1. As shown in Table 2, similar values of
Mn measured by GPC for the two batches of PTBA-PS (25: l) (a and b) confirmed that the
synthesis was reproducible in this case. The M, values for two different repetitions of the
synthesis differed by less than 8%. However, the error encountered in reproducing higher
molecular weights was larger due to practical synthetic problems. The amounts of
catalyst used in ATRP of high molecular weight polymers were very low (<20 mg) and
there was a possibility of higher weighing errors during preparation and synthesis in a
nitrogen-filled dry box. For example, if the amount of weighed catalyst was lower than

required, a decreased rate of polymerization would have caused an increase in the time

needed for reaction of all monomers and synthesis of polymer with an unexpectedly low

26

molecular weight. A way of overcoming this problem in the future would be preparation
of a concentrated stock solution of Cu1 complexed with PMDETA in acetone that could
be used in smaller amounts for syntheses of polymers with different M:I ratios. In the
case of underweighing the amount of Cu', a polymerization could also last for a time long
enough to ensure reaction of all monomers.

In addition to briefly examining the reproducibility of ATRP, we tested the
reproducibility of GPC experiments for each polymer. GPC analyses for most PTBA-PS
and PTBA-R-NH2 samples were run at least three times. Each value reported in Table 2 is
an average of all GPC measurements taken for a specific polymer. The experimental Mn
values obtained for different molecular weight PTBA-PS samples varied from the .
average values by less than 11%. A difference of up to 18% was observed between the
individual Mn values measured for PTBA-R-NH2 and their average values. The
polydispersities varied with respect to the average values by <6% for propionyl

succinimide-terminated PTBA, and by <1 4% for amino-terminated PTBA.

Table 2 Molecular weight distribution parameters obtained by GPC for propionyl
succinimide- and amino-terminated PTBA prepared with a variety of M:I ratios. Mtheor —
theoretical molecular weight based on M:I ratios, Mn — number average molecular weight,
Mw — weight average molecular weight, PDI — polydispersity index (MW/Mu).

 

 

PTBA-PS PTBA-R-NH,

M:I Mm, M, Mw PDI Mm, Mn M,v PDI
25:1 (a) 3,450 3,620 3,970 1.10 3,395 - - -
(b) 3,450 3,340 3,750 1.12 3,395 — — -
100:1 14,050 14,300 16,200 1.14 12,995 - - -

200:1 25,850 30,800 37,500 1.21 25,795 28,300 32,700 1.15

250:1 32,250 34,000 38,700 1.14 32,195 29,800 35,200 1.20

500:1 64,250 53,800 66,400 1.23 64,195 59,400 78,900 1.33

10001 128,250 98,100 115,100 1.18 128,195 93,300 114,100 1.25

27

Despite the high accuracy of GPC analysis and its suitability for determination of
PTBA molecular weight distributions, we encountered challenges in interpretation of
certain GPC results. In the case of lower molecular weight PTBA-R-NH2 (25:1 and
100:1) we could not observe peaks in chromatograms. One possible explanation for this
might be that there is a high affinity between the amino termini of PTBA-R-NH2 and the
styrene bead packing in the column. The polymer molecules might have been either
retained in the column or eluted with the smallest molecules at the end of the
chromatographic separation. This explanation seems unlikely as we observed small peaks
for amino-terminated PTBA in GPC spectra of polymers synthesized using M:I ratios of
200:1 and higher, and these peaks are similar in position to the corresponding PTBA-PS
polymers. A more likely reason for diminished PTBA-R-NH2 peaks is that the refractive
index detector may be less sensitive to the amino-terminated compounds. We measured
refractive index values for THF and PTBA-PS or PTBA-R-NH2 solutions in THF (1
mg/mL, Mn of 3,620). The refractive indices of these solutions were 1.409, 1.408 and
1.409, respectively. These data certainly suggest that it may be difficult to detect low
molecular weight PTBA-R-NH2 in THF with a refractive index detector. Because we
could not obtain GPC values for low molecular weight amino-terminated PTBA, we
utilized molecular weights of propionyl succinimide-terminated PTBA for studies of
PTBA growth on surfaces in the case of polymers prepared from M:I ratios of 25:1 and
100:1.

Some GPC chromatograms of PTBA-R-NH2 precursors synthesized fi'om M:I
ratios of 200:1 and higher tailed on the side of low molecular weights. This broadening

slightly increases polydispersity in most cases and is likely a consequence of the reaction

28

with ethylenediamine. Although a large excess of reactant satisfactorily ensured only
small increases in polydispersities (<8%), the high reactivity of ethylenediamine most
likely induced linking of a few chains through reaction with both amine groups of
ethylenediamine.

When interpreting Mn, Mw and PDI values, it is relevant to keep in mind that the
hydrodynamic volumes of PMMA standards and PTBA samples are somewhat different.
Consequently, the mobility of samples through the column might differ with respect to
the standards.32 We were able to test the reliability of the calibration obtained from GPC
by comparison with NMR results for the lowest molecular weight polymer. Calibrations
with PMMA standards utilized polymers with molecular weights between 6,000 and
127,000. This was another important reason for comparing Mn of the polymer prepared
from a 25:1 monomer to initiator ratio (Mthem = 3,450) to a molecular weight calculated
from lH NMR intensities as the molecular weight is out of the range of the standards. We
calculated the molecular weight and number of monomer units fiom the integrated peaks
due to the backbone methine (broad peak at 2.2 ppm) and succinimide methylene groups
(5, 2.8 ppm). The calculated M:I ratio of 23:1 corresponds to a molecular weight 3,194
that is 7% lower than the theoretical molecular weight. From GPC, M1, (3,620) was 5%
above the theoretical molecular weight, and the calculated monomer versus initiator ratio
indicated we synthesized a 26-mer. In spite of the previously mentioned uncertainties
concerning the range of standards available, GPC and NMR both show reasonable
agreement between theoretical and experimental Mn values for PTBA-PS with the lowest
molecular weight. Unfortunately, the integration of succinimide and methine peaks in

NMR spectra was possible only for the polymer with the lowest M“.. The intensities

29

attributed to the succinimide initiator in the NMR spectra of higher molecular weight
polymers were on the order of the background noise.

We found reasonably good agreement between Mn values determined using GPC
and theoretical Mn values (Table 2) for polymers as long as Mn values were below 59,400.
In the case of propionyl succinimide-terminated PTBA prepared from higher M:I ratios
(500:1 and 100021), the measured Mn values were significantly lower (16% and 31%,
respectively) than the theoretically predicted molecular weights. Most probably, we
observed such deviation because these reactions were stopped before a 100% conversion
of monomers was achieved.

The polydispersity of PTBA-PS and PTBA-R-NH2 was satisfactorily low for all
polymers synthesized. Polydispersity indices ranged between 1.10 and 1.33. They
increased after derivatization with ethylenediamine in all cases except for the polymers
prepared from a M:I ratio of 250:1. Polydispersities of PTBA-PS prepared fi'om low M:I
ratios (25:1 and 100:1) were comparable to the literature values (1.11-1.34).2“'32'33
Although the molecular weight distributions broadened slightly in higher molecular
weight polymers (Mn>14,300), the PDI values were still quite low. Therefore in all cases,
we were able to synthesize low polydispersity polymers that are suitable for studies of
hyperbranched growth of PTBA-R—NH2 on surfaces.

Reaction yields from ATRP ranged between 67 and 78%, similar to the ones
noted in literature.32 After purification of polymers upon derivatization to amine, loss of 7
to 27% of the material occurred with all but the lowest M,1 polymer. The yield for amino-
terrninated polymer prepared from a M:I ratio of 25:1 was extremely low (~15%) since

the abundant low molecular weight polymer was lost during purification by precipitation.

30

After the entire evaluation of GPC data, we chose average Mn values as the
nominal molecular weights for each polymer sample. Average Mn values of PTBA-PS
(for the two lowest molecular weight polymers) and PTBA-R-NH2 (for all other
polymers) were carried on for further use and evaluation for reactions on surfaces. Thus
the prepared polymers with narrow molecular weight distribution used in grafting of PAA
films had the following molecular weights: 3,620, 14,300, 28,300, 29,800, 59,400 and

93,300.
11. Ellipsometric thickness and film growth

A. Thickness values obtained using PTBA-R-NH2 varying in molecular weight
PTBA films were grafted onto self-assembled monolayers of MUA activated as
mixed anhydrides, and then hydrolyzed to PAA. Repeating the process of surface
activation, reaction with amino-terminated PTBA and hydrolysis yielded up to 6 layers.
Table 3 contains average thicknesses for films composed of up to six PAA layers. We had
difficulty in fitting ellipsometric parameters for films with thicknesses over ~5000 A,
probably because of increases in surface roughness. Therefore we deposited and inspected
only up to 5 and 4 layers when grafting PTBA with M“ values of 59,400 and 93,300,
respectively.
After grafting PTBA films from solutions used for 7-10 days, we observed an
increase in solution viscosity. As a consequence, films grafted from such solutions were
thicker than those from fresh solutions by up to 25%. For this reason, solutions of PTBA-

R-NH2 with molecular weights above 14,300 were prepared fresh for every deposition of

31

a multiple-layer film. Solutions used for grafting amino-terminated PTBA with molecular

weights of 3,620 and 14,300 were used for up to three weeks. Average thicknesses were

obtained from ellipsometric measurements of films grafted on three separate substrates

from three freshly prepared solutions. We averaged thicknesses on each substrate from

ellipsometric measurements at three different spots on the surface. A discussion of

thickness values is given in sections C and D.

Table 3 Ellipsometric thickness of hyperbranched PAA films grafted on gold using
polymers with several different molecular weights. Thicknesses are given for films before
(PTBA) and after (PAA) hydrolysis of each layer. Thicknesses of MUA monolayers (8-
12 A) were subtracted from each film thickness.

 

 

 

 

PTBA 3,620 14,300 28,300 29,800 59,400 93,300

layer
1“ 18.4 1 2.4 38.3 1 9.9 53.8 1 1.9 53.2 1 0.9 59.9 1 4.4 65.9 1 3.0
2"d 47.9 1 3.1 147 1 24 284 1 20 255 1 21 409 1 82 488 1 54
3'd 86.5 1 7.4 318 1 90.2 806 1 35 735 1 13 1460 1 279 2000 1 168
4‘h 136 1 17 542 1115 1390 1192 1240 1 63 2940 1 538 4610 1 498
5‘h 188 1 19 869 1 65 2250 1 352 1970 1 67 4810 1 872 -
6‘h 244 1 32 1160 1 209 2970 1 461 2800 1 162 - -

PAA

layer
1“ 6.17 1 1.1 16.6 1 5.3 21.3 11.9 20.3 1 2.4 22.4 1 3.0 26.5 1 3.8
2“d 38.8 1 1.3 78.4 1 22 152 1 9.2 136 1 12 203 1 43 251 1 23
3'd 58915.8 197151 500121 460142 8171140 10801105
4‘h 102 1 13 403 1 70 954 1 99 905 1 61 1910 1 366 2830 1 435
5‘h 149117 6101115 16401257 15601122 34201601 -
6‘h 210130 9041169 23301325 22101118 - -

B. Reaction times for grafting of PTBA-R-NHz

The time necessary for derivatization of anhydride layers with PTBA-R-NH2 varied

amongst polymers with different molecular weights. We monitored this time using FTIR

spectra by evaluating when anhydride peaks had decreased to <5% of their initial values.

32

Derivatization of anhydride groups using the lowest molecular weight PTBA (3,620)
occurred within one hour, as did derivatization of every first layer irrespective of the
molecular weights of polymer. For subsequent layers the time required for reaction of
anhydride groups increased when grafting PTBA from solutions of polymers with M“
values of 14,300 and above. As an example, the time necessary for reacting >95% of the
anhydride groups when grafting a sixth PTBA layer using PTBA-R-NH2 chains with a Mn
of 29,800 was 49 hours. The time required for reaction as a function of the number of

layers is shown in Figure 8.

 

 

 

 

 

 

 

 

60
o Mn=3,620
50 « o Mn= 14,300 §
7 Mn=28,300
40 7 v Mn=29,800
3 . Mn=59,400
V 30 7 D Mn=93,300
<1)
.§ 20 .
i” l
10 ~ Q 9
o
0 1 D Q 8 9 O O
0 l 2 3 4 5 6 7

Number of layers

Figure 8 Variation of the time required for reaction of anhydride groups with amino-
terminated PTBA solutions as a function of molecular weight and the number of PTBA
layers. Values represent an average of reaction times on 3 samples. Error bars represent
the standard deviations.

33

The longer times needed for reaction of anhydride groups with higher molecular
weight polymer precursors and after deposition of several layers are consistent with film
structure. As noted in the literature,” hyperbranched grth begins with a high degree of
branching in the first layers, and thickness rapidly increases. The polymer strands likely
become more tightly packed during grafting of subsequent layers and anhydride groups
may be less accessible to active amine ends. Films also are thicker after the deposition of
more layers or when using high molecular weight polymers, so diffusion of reagents to
anhydride groups will take more time. Additionally, when polymer chains are longer, the
solution contains a lower molar amount of amine groups that are capable of permeation
into the fihn and reaction with anhydride groups. The rate of grafting to the surface is
therefore expectedly slower in higher molecular weight PTBA-R—NH2 solutions and after
deposition of more layers.

Another important fact to consider is the hydrolysis of anhydride groups to PAA,
which must also contribute to the disappearance of anhydride peaks in the FTIR
spectrum. In fact, hydrolysis is probably the primary reaction contributing to the
disappearance of anhydride groups. Even though carefully dried prior to use, the solvent
(DMF) may contain small amounts of water that can hydrolyze anhydrides over long
reaction times. DMF also contains some impurities (amines) that might be responsible for
catalysis of hydrolysis. Hydrolysis would certainly be slower in thicker films. Figure 9
contains an example of F TIR spectra that show the reaction of anhydrides as a function of
time. Here it is visible that the‘ anhydride peaks of an activated third PAA layer
disappeared quantitatively only after a 32-hour reaction with PTBA-R-NHZ. Most of the

anhydride (>70%) reacted in the first 17 hours, while the reaction slowed down in the

34

latter hours. The slowing of the reaction may indicate the need for reagents to penetrate to

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

groups deeper in the fihn.
0.2 -1
Energy (cm )ester C=O .
anhydride

f A
O.)
0
g e
..D
H
8
g d A
<1:

c

b

a M A y.

1900 1800 1700 1600

Energy (cm’l)

Figure 9 ER-FTIR spectra of anhydride peaks in an activated 3 layer PAA film before (a)
and after reaction with PTBA-R-NH2 for 17 (b), 21 (c), 25 (d), 29 (e) and 32 (f) hours.
The PTBA-R-NH2 used in this experiment had a molecular weight of 93,300.

35

C. Change in thickness with respect to the number of grafted PAA layers
In agreement with previous studies,” we observed that thicknesses of PTBA and
PAA fihns increased non-linearly with the number of layers. This applied to all of the
molecular weights used in our study. Figures 10a and 10b show how the thickness of
these films varies with the number of layers for several different polymer molecular
weights. In all the instances, thicknesses increased non-linearly in the first 2-3 layers and
linearly thereafter. Thicknesses of PAA fihns decreased by 32-66% upon hydrolysis in
the first 1-3 layers and by 14-39% in the upper (4-6) layers due to loss of tert-butyl
groups.
We compared PAA thicknesses to previously reported values for hyperbranched
PAA films prepared from or,c0-diamino-terminated PTBA with higher polydispersity and
a Mn value of 14,600.13 Thicknesses of one to seven layers of PAA reported in the above
mentioned work were 33, 120, 213, 453, 600, 1000 and 1330 A, respectively. These
values fall very close to thicknesses for PAA fihns prepared from a polymer with a
molecular weight of 14,300 in our experiments (Table 3). Therefore, grafting of a
polymer with a lOwer polydispersity does not have a large impact on the thickness, but

solutions containing lower polydispersity chains react more slowly with anhydride

groups. In high polydispersity a,(0-diarnino-terminated PTBA the broader molecular
weight distribution provides both long and short chains, the latter being able to react or

catalyze reactions with the anhydrides in shorter times.

36

 

1400

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

PTBA M =3,620
A 1200 - ' ( " ) a
m o PAA (Mn=3,620)
g 1000 d v PTBA(Mn=14,300)
35‘; v PAA (Mn=l4,300) i
E” 800 ~
<
v
a: 600 r
U)
<1)
g 400 ~
.2 i
4:" -
E—« 200 1 6 a 8 3
0 - 5 8
O l 2 3 4 5 6 7
Number of layers
6000

. PTBA (Mn=29,800) b
4;; 5000 P =

0 AA (Mu 29,800)
g y PTBA (Mn=58,400)
a 4000 i v PAA (Mn=58,400)
OD PTBA M =93,300
fl 3000 ' ( " )
<: 1:1 PAA (Mn=93,300) §
v 0
a 2000 i I o
0 0
g o
o 1000 " i O
J: g o
1— 0 _ D

0 1 2 3 4 5 6 7

Number of layers

Figure 10 Thickness of hyperbranched PAA films (a) before (PTBA) and (b) after
hydrolysis of each layer to PAA, as a function of molecular weight. Error bars represent
standard deviations.

37

D. Relationship between thickness and molecular weights of PTBA-R-NH2

After evaluation of the layer-by-layer growth of hyperbranched PT BA, we studied
how the increase in thickness is related to the molecular weight of the preCursor PTBA-R-
NHZ. As mentioned in the introduction, although there are no reports about the
dependence of hyperbranched polymer film growth on the molecular weight of precursor
polymers, there is a study about spin-coated and photoreacted covalently bound thin films
of polystyrene on silicon.16 In that study, polystyrene with molecular weights ranging
from 115,700 to 1,815,000 was covalently bound by photoreaction to a monolayer of
silane-fiinctionalized perfluorophenyl azide. The thickness of one deposited layer ranged
from 2-60 rim and increased linearly with molecular weight.

Although in our case we used multiple grafting steps, and we grafted amino-
terrninated PTBA to the surface via an amide linkage, we also observed a fairly linear
relationship between thickness and molecular weight. Figure 11 represents the linear
increase in thickness with molecular weight for PTBA and PAA films with molecular
weights ranging from 3,620 to 93,300. The data points show a satisfying fit to linearity
from the first to sixth layers.

The reproducibility of layer thickness was especially good for low molecular
weight films. As expected, we encountered higher standard deviations in thicknesses of
the fifth and sixth layers. We mentioned previously that the viscosity of polymer
solutions increases after several days of usage. In the case of high molecular weight
PTBA-R-NHZ, after the deposition of the first 3-4 layers, an increased solution viscosity

caused less reproducible film thicknesses. Nonetheless, the average thicknesses of the

38

films with 3-6 layers did not deviate significantly from the linear curve fit. One can easily

conclude that thickness increases linearly with the molecular weight of precursor PTBA.

39

 

8000

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0 PTBA 1
1;? v PTBA2
E V PTBA3
o 6000 7 - PTBA4
a u PTBAS
on o PTBA6 _
s:
33 4000 —
m
m
E ,
.2 2000 q '-
.c:
P-4 /
0 _...IA'; I 4 f ‘T .
0 2011103 401103 60x103 80x103 1001103
Molecular weight
5000
o PAA l
o PAAZ
”v? 4000 - ' PAA3
E v PAA4
e I PAA 5
H 0 PAA6
g}, 3000 —
g .-
3 2000 -
28’
0
'F" 1000 7 . v,-
1—"5 / ‘
Z... . 41.4 44——
0 5.3, I n ' 3‘ I -
0 201103 401103 60x103 80x103 100x103
Molecular weight

Figure 11 Plots of film thickness versus PTBA-R-NH2 molecular weight (Mn) for 1-6
layer PAA fihns before (top, PTBA was just grafted) and after hydrolysis (bottom).

40

111. External Reflection Fourier Transform Infrared (ER-FTIR) spectroscopy

investigation of film growth

External reflection F TIR spectroscopy in principle allows monitoring of
derivatization of functional groups in polymer films. To interpret spectra, we assigned
peaks in grafted PTBA, PAA and activated PAA layers to specific functional groups
using analogous absorbance bands in the literature."3"3 Below, we present these

assignments (Table 4).

Table 4 Tentative peak assignments for absorbance bands in activated PAA (A), grafted
PT BA (B) and PAA (C).

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Wavenumber Assignment Compound Ref.

(6111")

3319 NH stretch A, B, C 1,35
3192-3083 OH stretch C l, 35
2983, 2975 CH3 antisymmetric, symmetric stretch A, B 35, 36, 37

2933 CH2 antisymmetric stretch A, B, C 35, 36, 37

2866 CH2 symmetric stretching A, B, C 35, 36, 37
1817, 1767 carbonyl (anhydride) A l, 13, 35

1735 carbonyl (ester) B 1, 3, 13, 35

1740 carbonyl (acid) C 1, 3, 13, 35
1674, 1643 amide I A, B, C 1, 3, 13, 35

1537 ON stretch and C-N-H in—plane bend A, B, C l, 3, 13,

(amide II) 35

1448 CH2(C=O); CH2 deformation, amide III A, B, C 6, 35
1394, 1370 tert-butyl; symmetric bending B 13, 35
1256, 1157 00 stretch A, B, C 13, 35

 

 

 

 

41

 

Figure 12 shows ER-FTIR spectra for the cycles involved in the deposition of the
third (a, c, e) and sixth (b, d, 1) layer of PAA. The spectra demonstrate the activation of a
previous PAA layer with formation of anhydrides (a, b), complete disappearance of
anhydrides upon reaction with PTBA-R-NH2 (c, d), and hydrolysis to PAA (e, f). The
vibrations noted in Figure 12 were observed in the spectra of PAA fihns made using all
the different molecular weights of PTBA-R—NHZ. The anhydride bands at 1817 and 1767
cm" (a, b) confirm the activation of the PAA fihn, and replacement of the bands with the
ester carbonyl peak at 1735 cm" (0, d) demonstrates PTBA grafting. After hydrolysis the
appearance of the broader carbonyl (acid) peak at 1740 cm'1 (e, f) suggests formation of
carboxylic acids. Quantitative hydrolysis of PTBA to PAA was also confirmed by the
disappearance of two sharp bands at 1394 and 1370 cm", attributed to symmetric bending
of tert—butyl groups as well as bands at 1256 and 1157 cm'1 due to the C-0 stretch of the
ester. The small remaining peaks at 1258 cm'l and 1180 cm’1 are likely due to the newly
formed -COOH groups.°

There are several bands that could appear due to amide linkages formed upon
grafting new layers onto underlying layers. Similar to the spectra of previously reported
PTBA and PAA films,"13 we observed absorbance peaks at 1674 and 1643 cm“.
According to the literature,” one of these two bands might correspond to the amide I
peak. For comparison to these peaks, we immersed an anhydride-activated 6-PAA film in
a 10 mM solution of N, N’-dimethylethylenediamine (DMEDA) for one hour. The result
was an increase in thickness (from 276 A to 362 A) and a nearly two-fold increase of both
possible amide I bands (1674 and 1643 cm"). We are as yet unsure of the assignment of

these two peaks. Other absorbance bands at 1537 cm" and at 1448 cm’1

42

 

amideI amide II

   
   
   
 
 
 

0.02 acid C=O
amide III

 

tert-butyl

ester C=O

Absorbance

 

 

anhydride

I l I l I l

3500 3000 2500 2000 1 500 1000

 

 

 

 

 

Energy (cm-1)

Figure 12 ER-FTIR spectra of two (a) and five (b) layers of activated PAA, and the same
films after attachment of PT BA (c, d) and subsequent hydrolysis (e, f). The PT BA-R-NH2
used in the synthesis of these films had a M11 value of 3,620.

43

might be amide II and amide III (C-N stretch and C-N-H bend) peaks, respectively. One
goal of synthesizing polymers with only one amino terminus was to use them to examine
amide formation and the disappearance of anhydride peaks upon grafting PTBA-R-NHZ.
Since we are unsure of these peak assignments, however, we could not use these spectra
for studies of amide formation.

Figure 13 shows spectra of 1 to 6 PAA layers prepared using PTBA-R-NH2 with a
molecular weight of 29,800. As noted above, the intensities of possible amide I bands
(1674 and 1643 cm"), as well as acid carbonyl peaks (1740 cm") increased with the
number of layers. The increase of the possible amide I peaks relative to the carbonyl peak
suggests that the number of amide linkages formed was higher in the upper grafted layers.
The high amide intensity after depositing the sixth layer of PAA, for example, would
indicate a very high yield of amide formation upon grafting of PTBA-R-NHZ. However,
the carbonyl intensity should increase simultaneously with increases in amide peaks due
to grafting of new chains and the addition of new carbonyl groups, but this was not
observed in our spectra.

One could explain the anomalously large peaks in the amide I'region in three
ways. First, specific orientations of amide groups might increase their intensity. As films
become more tightly packed, amide groups may be parallel to the surface. We doubt this
possibility as these films are not highly ordered. A second explanation for large amide

bands is that the presence of hydrogen bonding between C=O and NH groups might

44

 

0.1

amide I
region

M.

A

 

 

 

Absorbance

 

 

 

if

 

 

 

 

3 500 3000 2500 2000 l 500 1000

Energy (cm'l)

Figure 13 ER-FTIR spectra of 1-6 PAA (a-f) layers grafted using precursor PTBA-R-
NH2 with a molecular weight of 29,800.

45

increase intensities.35 The third and most reasonable explanation for high amide peaks is
that traces of ethylenediamine may be present, and these could react with anhydride
groups. This would be especially prevalent for low molecular weight polymers that are
difficult to purify. Indeed, we see relatively larger amide peaks for low molecular weight
polymers. The presence of ethylenediamine might also explain the large amide peaks in
later layers because this small molecule may selectively penetrate the film to react with

anhydride groups.

V. Yield of amide formation with grafting of new PAA layers

As stated previously, the non-linear increase in film thickness due to the grafting
of PAA shows that the highest branching occurs in the initial 2-3 layers, followed by a
somewhat linear increase in thickness of the upper layers. Similarly, we observed an
analogous change in acid carbonyl absorbances with respect to the number of PAA
layers. Figure 14a shows that carbonyl absorbances increase non-linearly upon depositing
first few layers and then somewhat linearly, with an exception of films with higher
thicknesses (~3400 A). In thicker films, absorbance is no longer linear with thickness so
the linear relationship between thickness and acid carbonyl intensities does not occur. A
plot of carbonyl intensities versus molecular weight also shows a linear relationship for

thinner films (Figure 14b). This is consistent with ellipsometric data.

46

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

r7 25
:3 _
3 . Mn:3’620 . a
m 20 _ o Mn—l4,300
erg . Mn=29,800 i
D v Mn=59,400 %
o 15 ‘ . Mn=93,300 %
5
"D i i
‘2 10 ~
0 i
U)
a i e
v— 5 n
a 9 i
O 9 Q Q .
U
0 l 2 3 4 5 6 7
Number of layers
A 25
=3 0 PAAl b
3 o PAA 2
c8 20 7 v PAA3
g v PAA4 9
o n PAA 5
g 15 - 1:1 PAA 6
+3 10 - i i
O i
U)
a 1 i
a» 5‘ 1 .
o g 0 O Q
"g 0 - O O O
U
0 20x103 403th3 60x103 80x103 100x103
Molecular weight

Figure 14 (a) Increase of the acid carbonyl absorbance in ER-FTIR spectra plotted versus
the number of grafted PAA layers. (b) Absorbance due to the acid carbonyl peak in PAA
films plotted with respect to the molecular weights of PTBA-R-NH2 used in film
formation. Legends indicate either the molecular weight of PTBA-R-NH2 used for
preparing fihns or the number of layers in the film.

47

Intensities of carbonyl vibrations in thin films can be used to estimate the yields
of amide formation upon deposition of each layer of PAA. Because the number of
carbonyl groups increases when linking new chains of amino-terminated PTBA onto the
surface, one can calculate the number of amide linkages formed in each layer of PAA
from the carbonyl absorbance. We use the known number of carbonyl groups (71) present
in a PTBA-R—NH2 n-mer to develop an expression of the number of amide linkages (xm)
formed by grafting an m-th layer. The values for n can be calculated from molecular
weights of every polymer, resulting in values of 28, 111, 221, 232, 464 and 729 for
polymers with Mn values of 3,620, 14,300, 28,300, 29,800, 59,400 and 93,300,
respectively.

The number of carbonyls contained in one strand in the first grafted layer (c,) is
equal to n and the number of amides formed by attachment of a strand (x1) corresponds to
a value of 1 for all molecular weight polymers. Since we are interested in amides formed
upon grafting of precursor PTBA-R-NH2 to an already existing underlying PAA layer, we
began evaluating the number of amide linkages formed on a single strand already
attached to MUA by one amide. Figure 15 is a scheme representing how amide linkages
form on a single strand of PAA throughout the first three grafting steps. The values for cm
include all the carbonyls present after every m-th layer is added to the underlying layers
of PAA, while xm expresses the number of amides that form during a single step of
grafting of strands of PTBA-R-NH2 to the underlying anhydride groups.

The number of amides formed upon attachment of the second layer (x2) relative to
one strand in the first layer can be related to the total number of carbonyls (c2) by

equation 6.

48

Carbonyls:

 

 

 

 

 

o o
o. .
C
I. . g .
. o
.0 . . . ..
o o o
. . C. .C
a . g . .
O D . c
. o . o. . .- 3
o o u
.- ,'
[I
C
J c
L! U 2 A
u C,
K Au substrate K
=x1
Anrides:

Figure 15 A schematic representation of grafting of the first three layers of PAA by
forming xm amide linkages with anhydride-activated carbonyl groups during grafting of
each new m-th layer. The total number of carbonyl groups contained in a fihn after
performing the m—th grafting step is expressed as cm.

49

c2 =(n—x,)+(n-x2) (6)
The first term shows that the number of carbonyl groups in the initial chain was reduced
by the number of new amide attachments while the second term represents the number of
newly deposited carbonyl groups. A ratio of the IR carbonyl intensities in the first two
layers, I/I,, is expressed by equation 7:

I 32. n—x2+n-x2

 

I—jzcl = n-xl (7)
Subsequent expressions of c3, c4, cm are:
c,=(n-x2+n-x2)—x3+nox3, (8)
c4=(n—x2+n-x2—x3+n-x3)-x4+n-x4,... (9)
cm = cm_l «x»l +n'xm (10)

Also, the number of carbonyl groups and the intensities relate as shown in equation 11:

 

 

m = m (11)

By combining equations 10 and 11 one can solve for the xm value in every grafting cycle
if the number of carbonyl groups in the previous polymer layer is known, remembering

that the relative number of carbonyl groups in the first layer is 1.

I c
x: "'—1-"“1 12
.[, 1.3 ()

m—l

 

 

In calculating the amount of amide formation, we used areas of the experimentally
obtained peak at 1740 cm‘1 from spectra of films with different numbers of layers. The

number of amides formed in each grafting step (xm) is shown in Figure 16a. The

50

calculated number of amides was low in the second grafting step, then it increased in the
subsequent two steps and leveled off. More amides formed in these initial steps when
films were formed using precursors with higher molecular weight. The number of amides
was smallest in films formed with precursor PTBA with an M“ value of 3,620, as would
be expected from the smaller thickness of these films.

The yield of amide formation was simply calculated as a ratio between the
number of newly formed amides (xm) and the number of the carbonyl groups present in

xm

 

the previous film (cm_,): % yield = 100. The yield of amide formation decreased

m—l
upon grafting of every new PAA layer (Figure 16b). A >3% yield was observed only in
films synthesized using the smallest molecular weight polymer (Mn = 3,620) and was the
highest in the initial grafting step. Trends showing amide yields in fihns grafted with
higher molecular weight polymers were similar. Less than 1% of anhydride groups were
derivatized with PTBA-R-NH2 during the third and subsequent steps when using PTBA-

R-NH2 with molecular weights above 3,620.

51

N
£11

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

”E
35 a o Mn=3,620
m o Mn=14,300
'8 20 d D M -29 800
,,_‘ D V n_ 9
g . Mn=59,400
‘4—4 15 — I :1 Mn=93,300
O I
H
<1)
E 10 -
S a ‘7 v 9
> 5 - u v
’5
cc 8 O
E) 0 . o 9 U o
l 2 3 4 5 6 7
4
M = 2
f: b . o n 3,6 0
é . 0 Mn = 14,300
D 3 ~ ' Mn=28,300
E 0 v Mn=59,400
9 . Mn=93,300
e\° 2 -
v
'2 o
.2 ' -
f; 1 . .
f2 ' 9 O 0
3
O
5 0_ 2 . .
1 2 3 4 5 6 7

Number of grafting steps (m)

Figure 16 (a) Representation of the number of amides formed in subsequent steps of
grafting PAA as calculated fiom the intensity of carbonyl groups. (b) Amide yield plotted
with respect to every grafting step in PAA film formation.

52

Chapter 4

SUMMARY AND FUTURE WORK

The use of ATRP for synthesis of propionyl succinimide-terminated PTBA
allowed preparation of polymers with tailored molecular weights. NMR and FTIR
confirmed the polymerization of TBA as well as successfiil reaction of these polymers
with ethylenediamine to yield amino-terminated PTBA. The polydispersities of PTBA-PS
and PTBA-R-NH2 measured by GPC ranged from 1.10 to 1.33. We were able to prepare
polymers with Mn values of 3,620, 14,300, 28,300, 29,800, 59,400 and 93,300.

These polymers were used for studying the dependence of the ellipsometric
thickness of hyperbranched PAA fihns on the number of PAA layers and on the
molecular weight of polymers. Thicknesses increased with the number of layers non-
linearly in the beginning (2-3 layers) and approximately linearly thereafter. These results
were analogous to previously reported trends for grth of much higher dispersity 01,00-
diamino-terminated PTBA. Film thickness increased approximately linearly with the
molecular weights of PTBA-R-NH2 used in these experiments. The times needed for a
complete derivatization of anhydrides increased with molecular weight and the number of
layers. This reflects a decrease in reaction rate with film thickness as well as a lower
concentration of amine groups in solutions of high molecular weight polymers.

The formation of amides upon grafting of every new layer of PAA could not be
quantified using the FTIR intensities of amide bands. Instead, we calculated the relative

number of amides formed from the experimental absorbances due to acid carbonyl

53

groups. Acid carbonyl intensities increased with the nrunber of layers and with increasing
molecular weights, as would be expected from the increases in ellipsometric thicknesses
of the same films. The calculated yield of derivatization to amides was very low (<4%)
and decreased with the number of layers. This is consistent with the notion that the high
branching in films occurs primarily in the first three layers.

A future inspection of these polymers will include a study of the coverage of
porous supports using films prepared from polymers with different molecular weights.
Eventually, porous supports coated with films with different numbers of PAA layers will
be inspected with scanning electron microscopy (SEM). Top views will provide
information on pore coverage, while side views of these ultrathin membranes will
confirm thickness values. Although successful pore coverage of 0.02-um alumina
supports previously involved a deposition of a 40-nm thick film,12 grafting of only few
layers of PAA using polymers with higher molecular weight might provide complete
coverage of pores. The film thickness required for coverage of pores might be lower than
in the case of fihns grafted from or,w-diamino-terminated PTBA with high polydispersity.
This might enable higher fluxes through membranes.

With a greater understanding of the growth of amino-terminated PTBA on gold
and alumina supports, several applications may be possible in the future. If derivatized
with perfluorooctyl amine, these films might serve as protecting layers for preventing
corrosion of metals, or as selective layers for separating gases, liquids or ions. Another
possible application of these membranes might be the use of specifically derivatized

ultrathin layers for entrapment of biomolecules by affinity filtration. Often the drawback

54

of ultrathin membranes in affinity filtration is clogging of pores in the early stage of
filtration of biomolecules. With control over thickness and the coverage of pores, the
present synthesis of ultrathin PAA membranes might allow better control of partial pore

coverage and thus provide affinity filters with high selectivity and high permeability.

55

BIBLIOGRAPHY

56

Bibliography

(1) Bruening M. L., Zhou Y., Aguilar G., Agee R., Bergbreiter D. E., Crooks R. M.
Langmuir 1997, 13, 770.

(2) Bergbreiter D. E., Tao G. Journal of Polymer Science: Part A: Polymer Chemistry
2000, 38, 3944.

(3) Zhou Y., Bruening M. L., Liu Y., Crooks R. M., Bergbreiter D. E. Langmuir
1996, 12, 5519.

(4) Zhao M., Zhou Y., Bruening M. L., Bergbreiter D. E., Crooks R. M. Langmuir
1997, 13, 1388.

(5) Zhao M., Bruening M. L., Zhou Y. , Bergbreiter D. E., Crooks R. M. Israel
Journal of Chemistry 1997, 37, 277.

(6) Franchina J. G., Lackowski W. M., Derrnody D. L., Crooks R. M., Bergbreiter D.
E. Analytical Chemistry 1999, 71, 3133.

(7) Derrnody D. L., Peez R. F., Bergbreiter D. E., Crooks R. M. Langmuir 1999, 15,
885.

(8) Aoki A., Ghosh P., Crooks R. M. Langmuir 1999, 15, 7418.

(9) Lackowski W. M., Ghosh P., Crooks R. M. Journal of the American Chemical
Society 1999, 151, 1419.

(10) Ghosh P., Crooks R. M. Journal of the American Chemical Society 1999, 121,
8359. '

(l l) Ghosh P., Amipour M. L., Lackowski W. M., Pishko M. V., Crooks R. M.
Angewandte Chemie International Edition 1999, 38, 1592.

(12) Nagale M. P., Kim B. Y., Bruening M. L. Journal of the American Chemical
Society 2000, 122, 11670.

(13) Zhou Y., Bruening M. L., Bergbreiter D. E., Crooks R. M., Wells M. Journal of
the American Chemical Society 1996, 118, 3773.

(14) Koros W. J., Fleming G. K. Journal of Membrane Science 1993, 83, 1.

57

(15) Rezac M. E., Koros W. J. Journal of Applied Science 1992, 46, 1927.
(16) Bartlett M. A., Yan M. In Proceedings of the ACS Division of Polymeric
Materials: Science and Engineering; ACS, 2000; Vol. 83, p 451.

(17) Schubert, D. W. Polymer Bulletin 1997, 38, 177.

(18) Peez R. F., Derrnody D. L., Franchina J. G., Jones S. J ., Bruening M. L.,
Bergbeiter D. E., Crooks R. M. Langmuir 1998, 14, 4232.

(19) Wang J .-S., Matyjaszewski K. Journal of the American Chemical Society 1995,
11 7, 5614.

(20) Matyjaszewski K., Patten T. E., Xia J. Journal of the American Chemical Society
1997, 119, 674.

(21) Kajiwara A., Matyjaszewski K. Macromolecules 1998, 31 , 5695.

(22) Matyjaszewski K., Teodorescu M., Acar M. H., Beers K. L., Coca S., Gaynor S.
G., Miller P. J ., Paik H. Macromolecular Symposia 2000, 15 7, 183.

(23) Coessens V., Nakagawa Y., Matyjaszewski K. Polymer Bulletin 1998, 40, 135.
(24) Davis K. A., Matyjaszewski K. Macromolecules 2000, 33, 4039.

(25) Coessens V., Matyjaszewski K. Journal of Molecular Science - Pure and Applied
Chemistry 1999, A36, 667.

(26) Kim H.-J., Lee J .-K., Park S.-J., Ro H. W., Yoon D.Y. Analytical Chemistry 2000,
72, 5673.

(27) Allcock H.A., Lampe F. W. In Contemporary Polymer Chemistry; Hall, P., Ed.:
Englewood Cliffs, NJ, 1990; p 394.

(28) Cheng H., Cornelio Clark P. A. , Hanton S. D. Journal of Physical Chemistry A
2000, 104, 2641.

(29) Cheng G., Simon P. F. W., Hartenstein M., Muller A. H. E. Macromolecular
Rapid Communications 2000, 21, 846.

(30) Pedrotti F. L., Pedrotti L. S. In Introduction to Optics; Hall, P., Ed.; Englewood
Cliffs, N.J., 1987.

(31) Collins R. W., Kim Y.-T. Analytical Chemistry 1990, 62, 887A .

58

(32) Davis K. A., Charleux B., Matyjaszewski K. Journal of Polymer Science: Part A:
Polymer Chemistry 2000, 38, 2274.

(33) Matyjaszewski K., Gaynor S. G., Qiu J., Beers K., Coca S., Davis K., Muhlebach
A., Xia J ., Zhang X. ACS Symposium Series, 765 (Associative Polymers in Aqueous
Media) 2000, 52.

(34) Lackowski W. M., Franchina J. G., Bergbreiter D. E., Crooks R. M. Advanced
Materials 1999, 11, 1368.

(35) Lin-Vien D., Colthup N. B, Fateley W. G., Grasselli J. G. In The Handbook of
Infrared and Raman Characteristic Frequencies of Organic Molecules; Academic Press,
1991. -

(36) Zamborini F. P., Crooks R. M. Langmuir 1998, 14, 3279.

(37) Porter M. D., Bright T. B., Allara D. L., Chidsey C. E. D. Journal of the American
Chemical Society 1987, 109, 3559.

59

IIIIIIIIIIIIIIIIIIIII

111111111121111111112111911111111i11