4....
3.411.

9%

n .4:
:1.
{83.

an .
3"»er m
:39...
. 1.5
@551.
s

«1
a ‘

g... Jun...

..
.2... .
w.
.23: .-
1. .. 1.5

st 5.
3.!5. J.‘
7‘ ‘ if
‘ (If G:

 

In" LS {s

300;

This is to certify that the
dissertation entitled

Molecular Control of Interfacial Properties:
The Construction, Characterization and Application of
Tailored Interfaces
presented by

Jaycoda Sandor Major

has been accepted towards fulfillment
of the requirements for

Ph . D . degree in Chemistry

flee ¥gf/

/M/ajor prof sor

Date §/// 01/ 0A

MS U is an Affirmative Action/Eq ual Opportunity Institution 0-12771

 

LIBVARY
0i

Michigan mate
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 cJCIFiC/DateDuopSS—ot 5

Molecular Control of Interfacial Properties:
The Construction, Characterization and Application of Tailored
Interfaces

By

J aycoda Sandor Major

A DISSERTATION

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

DOCTOR OF PHILOSOPHY -
Department of Chemistry

2002

ABSTRACT

Molecular Control of Interfacial Properties: The Construction,
Characterization and Application of Tailored Interfaces

By

J aycoda Sandor Major

The work reported in this dissertation has focused primarily on the design of
novel layered interfacial materials where there is discrete control over the resulting
molecular properties of the material achieved as a function of design and assembly. In
the first part of this work, we report on the design of a variety of polymer materials
synthesized with various side groups that can be used in several covalent linking
strategies. As a consequence of design, these polymers are capable of incorporating a
variety of functional constituents within the matrix of the material. The ability to control
the identity/composition of the reactive functionalities selectively within the matrix of the
material affords the direct control over the resulting property of the material. We have
also demonstrated our ability to control the effective loading density of the polymer
adlayers through catalytic and/or dehydration mediation of the various covalent
crosslinking reactions we report. Such control can undoubtedly be used in the design and
improvement of sensors where the ability to increase the number of the sensing elements
within each layer can lead to enhanced sensitivity. This characteristic also allows for the
design of materials where the conductivity of the material can be controlled by the
number elements incorporated within the matrix.

The second part of this dissertation focuses on the assembly and characterization

of these materials. Here, we employ an in-house assembled system for simultaneous

w “Haunt-o9...“

quartz crystal microbalance (QCM) gravimetry and optical ellipsometry — a system that
allows the simultaneous in situ acquisition of isotherm and thickness data. In this work,
we have demonstrated our ability to control adsorption behavior of the interfaces as a
function of the order and chemical identity of the interfacial adlayers. We have also
shown that, by varying the functional groups within each layer, it is possible to control
the adsorption response relative to the single functional group systems. The underlying
implication of this work is that it is indeed possible to design systems that exhibit a
chemical potential gradient where, in this particular case, the gradient is generated by
varying the dipole moment of the individual organic side groups incorporated into the
polymers used in making interfacial layers. The ability to design systems that exhibit a
gradient in some property can find use in separations systems where the desired outcome
is the continuous extraction of some analyte/solute leading to improved separations.
Finally, we have demonstrated that through simple design strategies, we are able
to tailor interfacial material properties with predictable results. We have designed a
copolymer of 4-hydroxyphenymaleimide and a vinyl ether phosponate, where the
phosphonate groups remain protected while the aromatic hydroxyl groups are used in the
covalent crosslinking of interfacial multilayers. Subsequent to adlayer growth, we
deprotect the phosphonate groups and use them for metal ion sequestration. This work

illustrates the range of potential application for these materials.

 

DEDICATION

Dedicated to my parents, Theodore and Christina and
to my other family members.

To my nieces and nephews, this is the first; don’t let it be the last.

-iv-

ACKNOWLEDGEMENTS

I would first like to express my sincere gratitude to my advisor Dr. Gary
Blanchard who has been so instrumental in my scientific and professional development.
You were always so understanding and helpful. I owe you so much, and I hope that at
some point in life, I am able to be of some assistance to you. Thanks to Dr. Simon
Garrett for always listening to me vent about my unpleasant experiences (at least the
major one) at MSU. Thanks to both you and Gary for convincing me to stay at MSU.
You have saved me a few years and a few brain cells. To Dr. Kathy Severin, you were
also so helpful to me, always giving me my last five or ten QCMs. YOU ROCK
KATHY!!! Thanks to all of my committee members for being so understanding.

I would like to express my sincere thanks to Dr. Yvonne Gindt who convinced
me, actually, insisted that I attend graduate school. Thanks for always assuring me that
even the bad times “built character”. You were always willing to listen to me let off
steam and you always had great advice. Thanks for checking in on me over the years,
this level of commitment and concern has truly been appreciated and will never be
forgotten.

To my friends at church that made my stay in Lansing more enjoyable, especially
Bazi, Mitchell, Lulama, Inga, Neliswa, Schvonne, Anne, Brent, Simone, Ruth and all of
the other people whose names escape me at this late hour, thank you. You guys helped
me through some rough times and helped me keep my faith in God intact. To all of my
other friends that have made my stay enjoyable especially Charles and Karen Ngowe,

Dan and Nicole Wampfler, Simona, Dalila, Erik, Fadi, Kam, Matt and Maggie, and

 

everyone else that I don’t remember right now (that old timers disease is kicking in, as
well as the lack of food and sleep) thank you.

Julian and Letitia Newbold, Garnett Burgess, Carla Anderson, Antionette Moxey,
Patrice King, Valderene Gardiner and Deon Stewart, thank you guys for being such great
friends. Even though miles and years have separated us, you have never allowed these
things to affect our relationships. Thanks for racking up the telephone bills or spending
monies to fly all the way from the Bahamas, or wherever you are just to spend some time
with me and to make sure that I was always okay. YOU ARE THE TRUE DEFINITION
OF FRIENDS.

Finally, to all the past and present members of the Blanchard group (most of you
anyway) especially Steve, Michelle, (“yeah baby, yeah ”) Mark, Punit, Lee, Scott, Shawn
and all of you other crazy people that made my transition to, and stay in the Blanchard
lab such an UNFORGETABLE experience - thanks. If I can be of assistance in the
future, you know how to contact me.

Finally, to those individuals who had direct hands in making parts of my stay at
MSU a less than pleasant ordeal — THANK YOU!!! It was those experiences that helped
to strengthen my resolve to work harder and to be successful - it worked. I would also

like to dedicate this work to you. God bless you and I wish you all the best.

-Vl-

Table of Contents

 

Page
List of Tables ......................................................................................... ix
List of Figures ........................................................................................ x
List of Schemes ...................................................................................... xv
Chapter 1. Introduction ........................................................................ 1
1.1 Literature Cited .................................................................... 12
Chapter 2. Strategies for Covalent Multilayer Growth 1. Polymer Design and
Characterization ................................................................... l6
2. 1 Introduction ............................................................... 17
2.2 Experimental .............................................................. 19
2.3 Results and Discussion .................................................. 26
2.4 Conclusions ............................................................... 43
2.5 Literature Cited ........................................................... 44
Chapter 3. Strategies for Covalent Multilayer Growth 2. Interlayer Linking
Chemistry ........................................................................... 47
3. 1 Introduction ............................................................... 48
3.2 Experimental .............................................................. 49
3.3 Results and Discussion .................................................. 52
3.4 Conclusions ............................................................... 78
3.5 Literature Cited .......................................................... 79
Chapter 4. Acid-Enhanced Interfacial Polymer Layer Growth .......................... 82
4. 1 Introduction ............................................................... 83
4.2 Experimental ............................................................. 86
4.3 Results and Discussion ................................................. 89
4.4 Conclusions ............................................................. 104
4.5 Literature Cited ......................................................... 108
Chapter 5. Adsorption Behavior of Polymer Modified Interfaces ..................... 112
5.1 Introduction .............................................................. 1 13
5.2 Experimental ............................................................ 1 15
5.3 Results and Discussion ................................................ 118
5.4 Conclusions ............................................................. 135
5.5 Literature Cited ......................................................... 137

-vii-

Chapter 6.

6.1
6.2
6.3
6.4
6.5

Chapter 7.

7.1
7.2
7.3
7.4
7.5

Chapter 8.
8.1
8.2

Molecular control of Adsorption Properties of Polymer Modified

Interfaces ......................................................................... 139
Introduction ............................................................. 140
Experimental ............................................................ 142
Results and Discussion ................................................. 149
Conclusions ............................................................. 160
Literature Cited ......................................................... 161

Covalently-Bound Polymer Multilayers for Efficient Metal Ion Sorption

............................................................................ 163
Introduction ............................................................. 164
Experimental ............................................................ 166
Results and Discussion ................................................ 171
Conclusions ............................................................. 184
Literature Cited ......................................................... 186
Conclusions and Future Prognosis ............................................ 188
Conclusions ...................................................................... 188
Future Prognosis ................................................................. 190

- viii -

wry-o”:- my

Table 4.1

Table 4.2

Table 4.3

Table 5.1

Table 6.1

Table 6.2

LIST OF TABLES

Slopes of the UV-Visible absorbance data for the addition of HCL and
H230; as a function of concentration for the crosslinking of
poly(NPM-VOB) with adipoyl chloride to form an ester linkage .............. 106

Dependence of ellipsometric thickness on acid addition for amide, ester, urea,
and urethane crosslinked adlayers ................................................. 106

Dependence of absorbance data on acid addition for amide, ester, urea and
urethane polymer adlayers ......................................................... 107

Layer thickness, monolayer adsorbate volume (le) and enthalpy of
desorption (Adcs) extracted from experimental data and fits to BET

isotherm .............................................................................. 136
Calculated dipole moments of selected polymer side groups. Calculations
Performed using Hyperchem V. 6.0 with PM3 parameterization ............ 152
Results of fitting experimental adsorption isotherm data to

equation 6.2 .......................................................................... 160

-ix-

LIST OF FIGURES

Figure 2.1 Polymer structure. (a) Poly(NPM-VOB), (b) Poly(NPM-APVE),
(c) Poly(NPM-4-PC), (d) Poly(NAI) .......................................... 24
(e) Poly(3CPM-hex), (f) Poly(MP-APVE), (g) Poly(4-HPM-VEP) ...... 25

Figure 2.2 FTIR spectrum (KBr pellet) of poly(NPM-VOB) ........................... 28
Figure 2.3 1H-NMR spectrum of poly(NPM-VOB) ....................................... 30
Figure 2.4 FT IR spectrum (KBr pellet) of poly(NPM-APVE) .......................... 31
Figure 2.5 1H-NMR spectrum of poly(NPM-APVE) .................................... 32
Figure 2.6 (a) FTIR spectrum of poly(NPM-4-PC), (b) lH-N MR spectrum of
poly(NPM-4-PC) ................................................................. 34
Figure 2.7 (a) FI'IR spectrum of poly(4NAI), (b) 1H-NMR of pol y(4NAI) .......... 36
Figure 2.8 FTIR spectrum of poly(3CPM-hex) ........................................... 37
Figure 2.9 1H-NMR spectrum of poly(3CPM—hex) ....................................... 38

Figure 2.10 (a) FI‘IR spectrum (KBr pellet) of poly(MP-APVE), (b) 1H-NMR
spectrum of poly(MP-APVE) .................................................. 40

Figure 2.11 (a) FI‘IR spectrum (KBr pellet) of poly(4-HPM-VEP), (b) 1H—NMR
spectrum of poly(4-HPM-VEP) ................................................ 42

Figure 3.1 (a)FTIR spectrum of poly(3CPM-APVE) with amide interlayer linkage
(b) UV-visible spectra as a function of number of layers for
poly(3CPM-APVE) .............................................................. 57

Figure 3.2 Dependence of ellipsometric thickness on number of layers for
poly(3CPM-APVE) with amide interlayer linkage .......................... 58

Figure 3.3 (a) Dependence of ellipsometric thickness on number of layers for poly
(NPM-VOB) with ester interlayer linkage. (b) UV-Visible spectra as a
function of number of layers. Inset: Dependence of absorption maximum
on number of layers .............................................................. 61

Figure 3.4 Poly(HPM-VEP) with ionic and covalent interlayer linkages. (a) Infrared
spectrum of ionically-bound multilayers and (b) IR spectrum of
covalently-bound multilayers. The spectra have been offset for clarity...62

 

Figure 3.5

Figure 3.6

Figure 3.7

Figure 3.8

Figure 3.9

Figure 3.10

Figure 3.11

Figure 3.12

Figure 3.13

Figure 4.1

UV-Visible spectra as a function of numerb of layers. Inset: Dependence
Of absorption maximum on number of layers .............................. 64

(3) Dependence of ellipsometric thickness on number of layers for

poly(NPM-VOB) with ether interlayer linkage. (b) UV-Visible spectrum
for poly(NPM—VOB) with ether interlayer linkage. Inset: Dependence of
absorption maximum on number of layers .................................. 66

IR spectrum for poly(NPM-VOB) with ether interlayer linkage ......... 67

Ether chemistry for modified substrate. (3) FTIR showing C-O-C stretch
of epoxide prior to reaction, (b) FTIR of ring-opened system after reaction
indicating ketone and acid functionalities .................................... 68

(a) IR spectrum of poly(3CPM-APVE) with urea interlayer linkage. (b)
UV-Visible spectra as a function of number of layers for poly(3CPM-
APVE) with urea interlayer linkage. Inset: Dependence of absorption
maximum on number of layers ................................................ 71

Dependence of ellipsometric thickness on number of layers for
poly(3CPM-APVE) with urea interlayer linkage ............................ 72

Dependence of ellipsometric thickness on number of layers for
poly(NPM-VOB) with urethane interlayer linkage ......................... 74

(a) IR spectrum for poly(NPM-VOB) with urethane interlayer linkage. (b)
UV-Visible spectra as a function of number of layers for poly(NPM-VOB)
with urethane interlayer linkae. Inset: Dependence of absorption
maximum on number of layers ................................................ 75

Poly(NPM-4-PC). (a) Structures - with amide and ester linkers. (b) IR
spectrum of a poly(NPM-4-PC) layer bound to the substrate by an ester
linkage and with remaining acid chloride functionalities hydrolyzed by
exposure to air ................................................................... 77

Figure 1. Data for amide-linked poly(3CPM-APVE). (3) Dependence of
optical absorption at 245 nm on number of polymer adlayers. I are data
points for acid-free growth, 0 are data for adlayer growth with 0.145 M
HCl and A are data for adlayer growth with 0.216 M HZSO4. Inset: UV-
Visible absorption spectrum for a four layer assembly of poly(3CPM-
APVE). (b) Ellipsometric thickness of poly(3CPM-APVE) adlayers as a
function number of layers deposited for growth without acid added (I)
and with 0.216 M H2304 added (A ). (c) FT IR spectrum of a two-layer
assembly of poly(3CPM-APVE) .............................................. 92

-xi-

Figure 4.2.

Figure 4.3.

Figure 4.4.

Figure 4.5.

Figure 4.6.

Figure 5.1

Figur 5.2

FI‘IR spectrum of substrate treated with adipoyl chloride showing
essentially complete conversion of the terminal acid chlorides to
carboxylic acids ................................................................... 93

Data for ester-linked poly(NPM-VOB). (a) Dependence of optical
absorption at 231 nm on number of polymer adlayers. I are data points
for acid-free growth, 0 are data for adlayer growth with 0.145 M HCl and
A are data for adlayer growth with 0.216 M H2304. Inset: UV-Visible
absorption spectrum for a four layer assembly of poly(NPM-VOB). (b)
Ellipsometric thickness of poly(NPM-VOB) adlayers as a function number
of layers deposited for growth without acid added (I), with 0.145 M HCl
(0) and with 0.216 M H2804 added (A). (c) FTIR spectrum of a two-
layer assembly of poly(NPM-VOB) ............................................ 95

Data for urea-linked poly(4BPM-APVE). (a) Dependence of optical
growth, 0 are data for adlayer growth with 0.145 M HCl and A are data
for adlayer growth with 0.216 M H2804. Inset: UV-Visible absorption
spectrum for a four layer assembly of poly(4BPM-APVE). (b)
Ellipsometric thickness of poly(4BPM-APVE) adlayers as a function
number of layers deposited for growth without acid added (I), with 0.145
M HCl (0) and with 0.216 M H2804 added (A). (c) FTIR spectrum of a
two-layer assembly of poly(4BPM-APVE) ................................... 98

Data for urethane crosslinked poly(NPM-VOB). (a) Dependence of
optical absorption at 231 nm on number of polymer adlayers. I are data
points for acid-free growth, 0 are data for adlayer growth with 0.145 M
HCl and A are data for adlayer growth with 0.216 M H2304. Inset: UV-
Visible absorption spectrum for a four layer assembly of poly(NPM-
VOB). (b) Ellipsometric thickness of poly(NPM-VOB) adlayers as a
function number of layers deposited for growth without acid added (I),
with 0.145 M HCl (0) and with 0.216 M H2804 added (A). (c) FTIR
spectrum of a two-layer assembly of poly(NPM-VOB) .................... 101

Dependence of ester-bound poly(NPM-VOB) monolayer absorbance at
231 nm on acid concentration for the initial polymer layer. (a) HCl added,
(b) H2804 added. The different concentration dependencies for each acid
indicates the importance of both protonation and dehydration in the ester
formation reaction ............................................................... 103

Structure of molecules bound to QCM surfaces. (a) 6-mercapto-1-
hexanol, (b) poly(4HPM-VEP), (c) poly(4MPA-VEP) .................... 119
(d) poly(4BPM-APVE), (e) poly(3CPM-APVE) ........................... 120

Adsorption isotherm data for uncoated gold QCM. (a) Data and best-fit
line for adsorption of methanol and (b) data and best-fit line for adsorption

-xii-

fl 03”.“;fl‘fl

Figure 5.3

Figure 5.4

Figure 5.5

Figure 5.6

Figure 6.1

Figure 6.2

Figure 6.3

Figure 6.4

Figure 6.5

Figure 6.6

of hexane. The data were fit t the BET model as given in Eqs. 2 and 3.
Results of the fits are presented in Table 5.1 ................................ 124

Adsorption isotherm data for a QCM modified with 6-mercapto-l-
hexanol. (a) Data and best-fit line for adsorption of methanol and (b) data
and best-fit line for adsorption of hexane. The data were fitted to the BET
model as given in Eqs. 2 and 3. Results of the fits are presented in Table
5.1 ................................................................................. 126

Adsorption isotherm data for a QCM modified with poly(4I-IPM-VEP).
(a) Data and best-fit line for adsorption of methanol and (b) data and best-
fit line for adsorption of hexane. The data were fitted to the BET model as

given in Eqs. 2 and 3. Results of the fits are presented in Table
5.1 .................................................................................. 130

Absorption spectrum of poly(4MPA-APVE) on a quartz substrate. The
band centered near 325 nm is the Sz<—So transition for the trans side
groups and the band centered near 240 nm is the Szé—So transition for the
cis side groups. (b) Data and best-fit line for adsorption of methanol (0)
and hexane (o). The data were fited to the BET model as given in Eqs. 2
and 3. Results of the fits are presented n Table 5.1 ........................ 131

Methanol adsorption isotherm data and best-fit lines for a QCM modified
with (a) poly(4BPM-APVE) and (b) poly(3CPM-APVE) with one
polymer layer (0) and two polymer layers (O). The data were fitted to the
BET model as given in Eqs. 2 and 3. Results of the fits are presented in
Table 5.1 .......................................................................... 134

BET isotherm for the adsorption of methanol on S-BPM-NPM and S-
NPM-BPM ........................................................................ 152

BET isotherm for he adsorption of hexane on S-BPM-NPM and S-NPM-
BPM ............................................................................... 153

Isotherms for the adsorption of methanol on S-BPM-CPM and S-CPM-
BPM ............................................................................... 155

Isotherms for the adsorption of hexane on S-BPM-CPM and S-CPM -
BPM ............................................................................... 156

Isotherms for the adsorption of methanol on S-BPM-HPM and S-HPM-
BPM ............................................................................... 157

Forward and backward isotherms for the exposure of S-BPM-CPM to
methanol .......................................................................... 159

- xiii -

Figure 7.1

Figure 7.2

Figure 7.3

Figure 7.4

Figure 7.5

Figure 7.6

Figure 7.7

Ellipsometric data for a four-layered MVE polymer system. Each layer
consists of the MVE polymer and the adipoyl chloride underlayer. The
slope of the best-fit line through the data is 16i1A/layer .................. 173

(a) FT IR absorbance data layers 1-4. The spectra are offset for clarity of
presentation. (b) Carbonyl stretching band absorbance as a function of
number of layers .................................................................. 174

UV-Visible absorbance spectra of MVE polymer layers as a function of
number of layers. Inset: Dependence of the 231 nm absorption intensity
as function of number of layers ................................................ 176

FT IR absorbance data for a four-layered system prior to deprotection of
the phosphonate groups with BTMS (spectrum I), after deprotection with
BTMS followed by hydrolysis (spectrum 11), and after exposure to ZrOClz
(spectrum III). The spectra are offset for clarity of presentation. . . . . 177

Comparison of the four-layered system, after exposure to ZrOClz
(spectrum HI), and after rinsing with absolute ethanol, then water

(spectrum IV). These spectra are offset for clarity of presentation ...... 179

XPS survey scan of a surface with four layers of poly(4HPM-VEP) on a
gold substrate ..................................................................... 181

QCM data showing rapid uptake of Zr4+ in a four-layer film .............. 183

-xiv-

Scheme 2.1

Scheme 2.2

Scheme 3.1

Scheme 3.2

Scheme 3.3

Scheme 3.4

Scheme 3.5

Scheme 4.1

Scheme 6.1

LIST OF SCHEMES

Reaction of maleic anhydride with an aromatic amine to produce the
maleamic acid, followed by ring closure to the maleimide .................. 20

Synthetic route for the polymerization of maleimide and vinyl ether
monomers (top), and for maleimide and vinyl monomers (bottom). Both
reactions are initiated with AIBN ............................................... 21

Reaction schematic and idealized structures of amide-linked MVE layered
assembly. (a) Amine-containing MVE polymer bound to acid chloride
functionalized surface and (b) acid chloride containing MVE polymer
bound to amine-functionalized

surface ............................................................................... 55

Reaction schematic and idealized structures of ester-linked MVE layered
assembly. (a) Alcohol-containing MVE polymer bound to acid chloride
functionalized surface and (b) acid chloride containing MVE polymer
bound to hydroxyl-functionalized surface ..................................... 59

Reaction schematic of a glicidyl ether with an alcohol (top) to form an
ether and an idealized structure of an alcohol-containing MVE polymer
bound to an expoxide functionalized substrate. Subsequent layer growth
requires reaction of the alcohol-containing MVE polymer with a diglicidyl
ether ................................................................................ 65

Reaction schematic of an isocyanate with an amine (top) to form a urea

and an idealized structure of an amine-containing MVE polymer bound to
an isocyanate-functionalized substrate. Subsequent layer growth requires
reaction of the amine-containing MVE polymer with a diisocyanate. . . ..70

Reaction schematic of an isocyanate with an alcohol (top) to form a
urethane and an idealized structure of an alcohol-containing MVE
polymer bound to an isocyanate-functionalized substrate. Subsequent
layer growth requires reaction of the alcohol-containing MVE polymer
with a diisocyanate ............................................................... 73

Top panel, left to right: Structures of pol y(3CPM-APVE), poly(NPM-
VOB) and poly(4-BPM-APVE). Bottom panel: Reaction schemes for the

several crosslinking strategies employed in this work ...................... 88
Interface structures. (a) S-NPM-BPM, (b) S-BPM-NPM .................. 145
(c) S-BPM-CPM, (d) S-CPM-BPM ........................................... 146

-xv-

Scheme 7.1

Scheme 7.2

(e) S-BPM-HPM, (f) S-HPM-BPM ........................................... 147

(a) Monomers used in the synthesis of he MVE alternating copolymer
reported here. (b) Deprotection (activation) chemistry for the MVE
copolymer. In this scheme, the isopropyl groups are hydrolyzed
subsequent to layer assembly ................................................. 167

Idealized schematic of the MVE polymer system showing two layers
connected by a diester linkage ................................................ 169

-xvi-

Chapter 1

Introduction

The past few decades have seen tremendous research effort in the areas of
materials design and assembly where the ultimate goal is molecular scale control over
the macroscopic properties exhibited by the material. Several methods have been

developed for the design and application of thin films chemistry in a variety of areas

3-6

such as chemical separation,“2 biological and chemical sensors, electrode

7-10

passivation, optical information storage, nonlinear optical materials, molecular

ll-l4

recognition and studies of energy transfer. Among the more extensively

employed methods are Langmuir-Blodgett films, self-assembled monolayers,'5‘2' spin

casting/coating, covalent interlayer linkage strategies and metal phosphonate

”‘37 A number of these methods employ a layer-by-layer strategy for

chemistry.
material assembly, where this approach extends the potential applicability of these
materials into a variety of areas and allows for the molecular control of material
property. This level of control is due to the fact that it is possible to introduce
specific chemical functionalities into each individual layer. The ability to control the
identity of the functional groups incorporated into each layer renders accessible the
ability to selectively tune material properties in a relatively well-defined manner — the
major focus of the work we report in this dissertation.

As mentioned above, a variety of methods have been developed for materials

design and assembly. There are however, shortcomings associated with each of these

methods that necessitate the continued development of and improvement upon the
current technologies. One such system that has gained significant attention over the
years has been Langmuir-Blodgett films, based on the work of Irving Langmuir and

8.39 . - -
3 These materials have been used in numerous studies,

Katharine Blodgett.
however, they suffer significantly from lack of robustness, thereby limiting their
application particularly in harsh environments. In this approach, surface monolayers
of amphiphilic molecules preassembled on a liquid surface are adsorbed onto a
substrate as it is withdrawn from or introduced to the liquid subphase. The adsorption
is made possible by the affinity of the head- or tail-groups of the amphiphile to the
substrate or to themselves. The source of the problems associated with this method is
the association of the molecular layers through weak van der Waals interactions that
are subject to disruption, often by rinsing with polar solvents. While this method of
layer formation is typically easy to employ and can be used to prepare a variety of
useful systems where the weak non-selective interaction does not destroy the
functionality of the species being adsorbed, the fact that the intermolecular
interactions are so weak renders L—B films limited in their utility.

A second method that has been studied extensively is alkanethiol-gold self-

4044 In this strategy, surface modification is

assembly pioneered by Allara and Nuzzo.
possible due to the affinity of molecules bearing a terminal thiol for gold and/or silver
substrates. While this method has received much attention, and has been employed
extensively in the design of a variety of materials, it too has proven to be of limited
utility due to the weak association of the thiol group with the surface (energies of

45 .46

interaction AGQ‘cs ~ 20 kJ/mol), consistent with weak chemisorption. A second

limitation of this method is that surface modification is often limited to one molecular
layer due to the selective interaction of the head group with the substrate. Multilayer
assembly is, however, possible with this type of chemistry if the terminal
functionality of the alkanethiol is capable of participating in additional reactions.”5 ‘
An additional limitation of this chemistry is that thiol groups dictate the type of
substrate (typically Au or Ag) that can be used. Despite these shortcomings, this
method has found extensive application in systems where specific surface terminal
groups are desired. Once the surface has been modified, it is possible to perform
further chemistry on the surface that would have not been possible. For instance,
using the alkanethiols to assemble a variety of systems on gold allows
electrochemical characterization of these systems where the gold surface behaves as a
working electrode.

The metal phosphonate cherrristry, based on the pioneering work of the

7,3 . . 3, 9.
and Thompson3 6525 5 60 groups,

M all 0le34.252729.32.52.55 K at223.28.30.35.37,56—58
provides a route to prepare materials that are both thermally and chemically robust.
In this method, a substrate that initially presents surface hydroxyl or silanol groups is
primed by the initial exposure of those groups to a functionalized silane (typically an
amino- terminated silane). The silane-modified interfaces are then exposed to
phosphorus oxychloride and subsequently exposed to aqueous solutions of metal ions
such as Hf“ or Zr“. It should be noted that, even in the absence of silane surface
priming, it is possible to react phosphorus oxychloride directly with the surface
54.55

silanol groups with no apparent compromise in reactivity or surface morphology.

The phosphate/zirconium ion terminated substrate is then exposed to a solution of an

0t,u)-bisphosphonate or an asymmetric (ii-hydroxyphosphonate, where the ionic
interaction between the positively charged metal ion with the phosphate or
phosphonate head group leads to layer deposition. Subsequent to the deposition of
the initial layer, the unreacted phosphonate head groups are exposed to an aqueous
solution of the metal ion followed by exposure to a bisphosphonate. In the case of the
(o-hydroxyphosphonate, the hydroxyl group is reacted with phosphorus oxychloride,
followed by exposure to the aqueous metal ion solution, followed by reaction with
either am-bisphosphonate or (n-hydroxyphosphonate. In this manner, it is possible to
assemble robust multilayer structures.

While metal bisphosphonate/phosphate chemistry has been shown to be robust
(undoubtedly due to the strong, selective interactions) one of the shortcomings of this
chemistry is the lack of stability of the layers parallel to the plane of the substrate. In
work performed by Kohli and Blanchard, a slightly modified version of the zirconium
phosphonate (ZP) chemistry was demonstrated, where an alternating maleimide
functionalized copolymer presenting phosphonate side-groups was used in material
design/assembly.56 In this scheme, the phosphonate functionalities are initially
protected with isopropyl groups, which are then removed by reaction with
bromotrimethylsilane (BTMS), allowing for controlled layer-by-layer assembly. One
of the major advantages of this growth methodology is enhanced structural stability
afforded by the polymer backbone in the plane parallel to the substrate. Additionally,
the maleimide chemistry we employ allows for the introduction of a variety of

functional groups into the polymer layer.

One limitation of metal phosphonate chemistry was revealed in work
performed by Home and Blanchard, where interlayer excitation transport in materials
assembled using zirconium bisphosphonate interlayer linkages was shown to be
inefficient.57 Those authors have offered a possible explanation where the polarizable
ZP interlayer screens dipolar coupling between layers. To this end, the development
of robust chemical routes to multilayer structures that do not involve ionic interlayer
connections might be useful in removing screening effects that serve to minimize
interlayer energy migration. The majority of the work reported in this dissertation
involves the use of a suite of reactions that have been developed for the controlled
layer-by-layer assembly of materials where the interlayer linkages are all covalent in
nature. It should be noted that this work is not a study of energy migration; it is
worth noting however, that the synthetic routes reported herein are very likely
amenable to the design of systems that exhibit this phenomenon.

There are three major thrusts of the work reported in this dissertation. First,
the development of a suite of unique strategies for surface modification is reported.
In these methods, the uniqueness iies not only in the linkage strategies, but also the
demonstrated ability to introduce a variety of functional moieties into the matrix of
the polymer materials that can be used to tune the properties of the material
selectively. Depending on the chromophore incorporated into the layered material, it
is also possible to probe the microenvironment formed within these adlayers.
Secondly, the characterization of these materials is reported. In addition to traditional
methods of characterization such as optical ellipsometry, UV-visible, FTIR and NMR

spectroscopies, an in-house assembled apparatus that combines the techniques of

ellipsometry and quartz crystal microbalance (QCM) gravimetry was employed to
acquire adsorption isotherm data for these systems. By this method, the adsorption of
vapor phase adsorbate molecules is followed in real time and plots of adsorbate
volume as a function of adsorbate vapor pressure at constant temperature generate the
isotherm. From the isotherm data, it is possible to access information about the
energies of interaction between various adsorbates and the interfaces that are prepared
for this study. It is also possible to determine the volume of the monolayer, i.e. the
volume of the adsorbate that is required to cover the interface with a single molecular
layer. The incorporation of an ellipsometer into the apparatus allows the
determination of thickness changes associated with the adsorption of the adsorbate
molecules as a function of vapor pressure. Ellipsometric thickness in concert with
adsorbate volume can be used to acquire information about the specific surface area
of the adsorbent. In the final phase of this project, the application of one of these
interfaces is demonstrated. In that body of work, 4-hydroxyphenylmaleimide was
copolymerized with a vinyl ether phosphonate, where the phosphonate groups were
protected with isopropyl groups. After the polymer has been assembled into the
desired number of layers by crosslinking between the hydroxyphenylmaleimide
functionalities in adjacent layers, the phosphonate groups were deprotected and this
assembly was used in selective metal ion sequestration.

Building upon the maleimide vinyl ether chemistry used in the Blanchard
group, a new system for multilayer assembly was demonstrated. Here, a suite of
copolymers was prepared using a variety of functionalized maleimide monomers, all

prepared in-house. The maleimides were copolymerized (radical polymerization)

with a variety of allyl—, allyloxy-, vinyl— and vinyloxy— functionalized compounds.
We have chosen to work with maleimides due to the fact that they (1) are easy to
prepare, (2) are very stable, and (3) have been shown to form alternating copolymers,
a feature that allows for greater uniformity/control over material property.
Additionally, the comonomers were chosen due to the variety of functional side-
groups that can be introduced into the polymer, allowing us to realize multilayer
assembly using a variety of interlayer crosslinking strategies, where these crosslinks
are either ionic or covalent in nature. We focus here primarily on covalent
crosslinking of these materials due to the extensive body of literature dealing with
ionic interlayer linkages.

In Chapters two and three of this dissertation, we report the synthesis (Chapter
2) and multilayer assembly (Chapter 3) of the various polymer systems. The major
focus of this project is the ability to prepare materials where we can introduce various
functional moieties into the matrix of the bulk material that can be used to tune or
modify a macroscopic material selectively. We approach this task from two fronts.
First, we design polymer systems that present functionalities capable of participating
in a variety of interlayer linking strategies, where the crosslinking chemistries are all
covalent in nature. Second, by simple substitution of the functional moiety of the
maleimide, it is possible to introduce a wide variety of functional groups into the
matrix of these polymers. A variety of methods have been used in the
characterization of these materials. FTIR (KBr pellet) and lH-NMR spectroscopies
have been employed in the determination of the structures of the polymers, and GPC

was used to determine molecular weights. Subsequent to layer assembly, optical null

ellipsometry was used to determine the thickness of the layers as a function of
deposition cycle for adlayers assembled on either gold-coated or silicon wafers. UV-
visible and FTIR spectroscopies were employed to confirm material deposition and to
verify the identity of the cross-linking chemistry.

In Chapter 4 we report the acid enhanced deposition of polymer materials. In
this work, we have demonstrated the ability to control the loading density of the
chromophore - as evidenced by the absorption spectra and ellipsometry - by simple
acid catalysis and/or dehydration. During layer deposition, concentrated hydrochloric
or concentrated sulfuric acid is added to the reaction vessel and in both cases we
observe significant enhancement in material deposition relative to the corresponding
acid—free reactions. In all cases, we observed an increase in ellipsometric thickness
for the layers as well as an increase in the intensity of the absorption band of the
chromophore in the uv-visible spectra. We have explained these findings in the
context of acid catalysis and/or dehydration. It should be noted that, in all cases
reported, regardless of the chemistry being used for multilayer assembly, the addition
of sulfuric acid led to a higher polymer adlayer loading density. We understand these
findings based on the ability of sulfuric acid to function as both an acid and an
efficient dehydrating agent. In the cases of amide and ester linkage formation, these
reactions proceed with the elimination of water as a by-product. The addition of
hydrochloric and sulfuric acid probably serves to protonate the reactive species, while
sulfuric acid is simultaneously capable of sequestering product water, driving the
reaction toward product formation according to Le Chatelier’s principle. In the cases

of urea and urethane linkage reactions, the concentrated acid again serves to protonate

the reactive species. It should be noted that the reactive intermediate in these
reactions is an unstable carbamic acid that is susceptible to attack by adventitious
water, leading to the degradation of this species. In the case where sulfuric acid is
added, dehydration can thus play a significant role. Finally, we note the dependence
of polymer adlayer density on acid concentration. We found that increasing the acid
concentration eventually led to a decrease in the slopes of the absorbance versus
concentration curves, suggesting that there is an optimal range over which these
reactions are most efficient. We note, however, that increasing the concentration led
to a linear increase in absorbance for the initial layer deposited. The ability to
increase the loading density of the functional species with the material matrix could
prove beneficial in sensor design and the design of conductive or resistive materials,
for example.

The main focus of Chapters 5 and 6 is the characterization of these materials
with respect to their adsorption isotherm. From the adsorption isotherm data on the
various materials, we have been able to demonstrate that, by simply varying the
functional groups within the matrix of the material, we can control the adsorption
energetics of the material. In Chapter 5, we focus on the deposition of single polymer
layers with different functional groups and acquire their vapor phase isotherms. In
Chapter 6, we have prepared the interfaces using multilayers of polymers, where we
have varied the identity of the functional groups within each layer and report the
observe changes in the isotherm response relative to the single layered systems of
Chapter 5. In this work, we recover noticeable changes in the functional form of the

isotherms as well as in the energetics of the systems when different groups are

substituted into the matrix. Another very interesting observation is that when the
order of polymer deposition is reversed, even in these two—layered systems, we
recover substantially different adsorption isotherm data, indicating our ability to
control the kinetic and thermodynamic properties of the system by synthetic means.
We have used semi-empirical calculations to estimate the dipole moment of the
various side-groups we employ in this study and have attempted to rationalize our
findings in the context of adlayer polarity. The ability to control functional group
identity within individual layers lends itself directly to the preparation of interfaces
that exhibit a chemical potential gradient, where these interfaces can be used in
separation schemes capable of continuous extraction. It should also be noted that, in
all cases, we observed no thickness change for our polymer adlayers with adsorption,
suggesting the porous nature of the interfaces. In previous work performed by
Karpovich and Blanchard,58 it was shown that this system was sensitive enough to
determine adlayer thickness changes as a function of partial pressure of the vapor
phase adsorbate molecules. In that work, the adlayers were relatively well-ordered
alkanethiol monolayers. Here, we use alternating copolymers, where these systems,
by their very nature, are less well ordered and more porous.

In Chapter 7 we describe a specific application of one particular polymer
system we have prepared. In this work, we employed a hydroxyphenylmaleimide and
copolymerized this group with a vinyl ether phosphonate. Multilayer assembly was
made possible by generating ester linkages across the hydroxyphenylmaleimide and
using the phosphonate groups for metal ion sequestration. It should be noted that

during layer deposition, the phosphonate groups remain protected, precluding the

-10-

 

formation of phosphoesters. In a second step, these groups are deprotected using

59 We have employed a variety of analytical methods to

bromotrimethylsilane.
characterize these systems. X-ray photoelectron spectroscopy (XPS) data confirm the
sequestration of various metal ions by these systems. Using this polymer and the
various characterization methods, we have been able to demonstrate the selective and
in some case irreversible sequestration of metal ions.

There are three distinct advantages of the materials and assembly strategies
reported in this dissertation. First, as mentioned above, the simplicity of the
maleimide chemistry employed allows the introduction of a variety of chemical
functional side-groups into the matrix of these materials. This ability will be
exploited to develop materials that exhibit a chemical potential gradient. According
to Giddings,60 the creation of a gradient in any particular property of a material can be
used to create a chemical potential gradient. In this work, we generate a gradient in
the dipole moments of the side-groups incorporated into each individual layer.
Second, the side groups of the comonomers allow us to use very simple covalent
interlayer linking strategies such as amide, ester, ether, urea and urethane formation
to assemble these materials into robust multilayer structures. The third advantage is a
direct result of the covalent cross-linking chemistry. We have been able to
demonstrate that, through simple acid catalysis and/or dehydration of the cross—
linking reactions, we are able to control material deposition. This feature can prove

useful in a variety of applications such as sensor design, where increasing the number

density of he sensing elements could lead to enhanced detection.

-11-

1.2 Literature Cited

1. Vrancken, K. C.; VanderVoort, P.; Gillisdhamers, I.; VanSant, E. F.; Grobet, P.; J.
Chem. Soc, Faraday Trans. 1992, 88, 3197.

2. Pfleiderer, 13.; Albert, K.; Bayer, 13.; J. Chromatogr. 1990, 506, 343.

3. Demas, J. N.; Degraff, B. A.; Coleman, P. 8.; Anal. Chem, News &

Features, 1999, 793A.

4. Yoon, H. C.; Kim, H-S. Anal. Chem. 2000, 72, 922.

5. Delamarche, E.; Sundarababu, G.; Biebuyck, H.; Michel, B.; Gerber, C.; Sirgrist,
H.; Wolf, H.; Ringdorf, H.; Xanthopoulos, N.; Mathieu, H. J .; Langmuir, 1996, 12,
1997.

6. Brousseau, HI, L.C.; Aurentz, D. J.; Benesi, A. J .; Mallouk, T. E.; Anal. Chem,
1997, 69, 688.

7. Bakiamoh, S.; Blanchard, G. J.; Langmuir, 2001, 17(11), 3438.

8. Katz, H. E.; Scheller, G.; Putvinski, T. M.; Schilling, M. L.; Wilson, W. L.;
Chidsey, C. E. D. Science, 1991, 254, 1485.

9. Li, D.; Ratner, M. A.; Marks, T. J.; Zhang, C. H.; Yang, J.; Wong, G. K.; J. Am.
Chem. 506., 1990, 112, 7389.

10. Kanis, D. R.; Ratner, M. A.; Marks, T. J. Chem. Rev. 1994, 94, 195.

11. Zak, J .; Yuan, H.; Ho, M.; Woo, L. K.; Porter, M. D.; Langmuir 1993, 9, 2772.

12. Finklea, H. O.; Hanshew, D. D. J. Am. Chem. Soc. 1992, 114, 3173.

13. Sun, L.; Crooks, R. M. Langmuir 1993, 9, 1775.

14. Sun, L.; Kepley, L. J.; Crooks, R. M. Langmuir 1992, 8, 2101.

15. Bain, C. D.; Whitesides, G. M. J. Am. Chem. Soc. 1988, 110, 3665-3666.

-12-

16. Bain, C. D.; Biebuyck, H. A.; Whitesides, G. M. Langmuir 1989, 5, 723.

17. Nuzzo, R. G.; Allara, D. L. J. Am. Chem. Soc. 1983, 105, 4481.

18. Nuzzo, R. G.; Dubois, L. H.; Allara, D. L. J. Am. Chem. Soc. 1990, 112, 558.

19. Nuzzo, R. G.; Fusco, F. A.; Allara, D. L. J. Am. Chem. Soc. 1987, 109, 2358.

20. Xia, Y.; Whitesides, G. M. Angew. Chem, Int. Ed. 1998, 37, 550.

21. Troughton, E. B.; Bain, C. D.; Whitesides, G. M.; Nuzzo, R. G.; Allra, D. L.;
Porter, M. D. Langmuir 1988, 4, 365.

22. Hong, H.-G.; Sackett, D. D.; Mallouk, T. E. Chem. Mater. 1991, 3, 521.

23. Katz, H. E. Chem. Mater. 1994, 6, 2227.

24. Lee, H.; Kepley, L. J .; Hong, H.-G.; Akhter, S.; Mallouk, T. E. J. Phys. Chem.
1988, 92, 2597.

25. Lee, H.; Kepley, L. J .; Hong, H.-G.; Mallouk, T. E. J. Am. Chem. Soc. 1988, 110,
618.

26. Putvinski, T. M.; Schilling, M. L.; Katz, H. E.; Chidsey, C. E. D.; Mujsce, A. M.;
Emerson, A. B. Langmuir 1990, 97, 237.

27. Rong, D.; Hong, H.-G.; Kim, Y.-I.; Krueger, J. S.; Mayer, J. E.; Mallouk, T. E.
Coord. Chem. Rev. 1990, 97, 237.

28. Katz, H. E.; Schilling, M. L.; Chidsey, E. E. D.; Putvinski, T. M.; Hutton, R. S.
Chem. Mater. 1991, 2, 699.

29. C30, G.; Rabenberg, L. K.; Nunn, C. M.; Mallouk, T. M. Chem. Mater. 1991, 3,
149.

30. Ungashe, S. B.; Wilson, W. L.; Katz, H. E.; Scheller, G. R.; Putvinski, T. M. J.

Am. Chem. Soc, 1992, 114, 8717.

-13-

 

_ —-.'
r i
a

 

31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

Vermeulen, L.; Thompson, M. E. Nature 1992, 358, 656.

Yang, H. C.; Aoki, K.; Hong, H.-G.; Sackett, D. D.; Arendt, M. F.; Yau, S.-L.;
Bell, C. M.; Mallouk, T. E. J. Am. Chem. Soc. 1993, 115, 11855.

Frey, B. L.; Hanken, D. G.; Corn, R. M. Langmuir, 1993, 9, 1815.

Yonemoto, E. H.; Saupe, G. B.; Schmehl, R. H.; Hubig, S. M.; Riley, R. L.;
Iverson, B. L.; Mallouk, T. E. J. Am. Chem. Soc. 1994, 116, 4786.

Katz, H. E.; Bent, S. E; Wilson, W. L.; Schilling, M. L.; Ungashe, S. B. J. Am.
Chem. Soc. 1994, 116, 6631.

Thompson, M. E. Chem. Mater. 1994, 6, 1168.

Katz, H. E.; Wilson, W. L.; Scheller, G. J. Am. Chem. Soc. 1994, 116, 6636.
Langmuir, 1.; J. Am. Chem. Soc. 1917, 39, 1848.

Blodgett, K. B.; J. Am. Chem. Soc. 1935, 57, 1007.

Allara, D. L.; Nuzzo, R. G. Langmuir 1985, I, 45.

Allara, D. L.; Nuzzo, R. G. Langmuir 1985, 1, 52.

Nuzzo, R. G.; Zegarski, B. R.; Dubois, L. H. J. Am. Chem. Soc 1987, 109, 733.
Bain, C. D.; Whitesides, G. M. Science 1988, 240, 62.

Strong, L; Whitesides, G. M. Langmuir 1988, 4, 546.

Karpovich, D. S.; Blanchard, G. J. Langmuir 1994, 10, 3315.

Schessler, H. M.; Karpovich, D. S.; Blanchard, G. J. J. Am. Chem. Soc. 1996,
118, 9645.

Sabapathy, R. C.; Crooks, R. M. Langmuir 2000, 16, 1777.

Chechik, V.; Crooks, R. M. Langmuir 1999, 15, 6364.

Lackowsi, W. M.; Franchina, J. G.; Bergbrieiter, D. E.; Crook, R. M. Adv. Mater.

-14-

1999, 11, 1368.

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

51. Zhou, Y.; Bruening, M. L.; Bergbreiter, D. E.; Crook, R. M.; Wells, M J. Am.

Chem. Soc. 1996, 118, 3773.

52. Vermeulen, L. A.; Snover, J. L.; Sapochak, L. S.; Thompson, M. E. J. Am. Chem.
Soc. 1993, 115, 11767.

53. Snover, J. L.; Byrd, H.; Suponeva, E. P.; Vicenzi, E.; thompson, M. E. Chem.
Mater. 1996, 8, 1490.

54. Putvinski, T. M.; Schilling, M. L.; Katz, H. E.; Chidsey, C. E D.; Mujsce, A. M.;
Emerson, A B. Langmuir, 1990, 6, 1567.

55. Bakiamoh, S. B.; Blanchard, G. J. Langmuir 1999, 15, 6379.

56. Kohli, P.; Blanchard, G. J. Langmuir, 1999, 15, 1418.

57. Home, J. C.; Blanchard, G. J. J. Am. Chem. Soc. 1999, 121, 4427.

58. Karpovich, D. S.; Blanchard, G. J. Langmuir, 1997,

59. Major, J. S.; Blanchard, G. J. Langmuir 2001, I7, 1163.

60. Giddings, C. J. Unified Separation Science; Wiley Interscience, 1991.

-15-

Chapter 2

Strategies for Covalent Multilayer Growth 1. Polymer Design and
Characterization

Abstract

We describe the synthesis and characterization of several maleimide-vinyl ether
and maleimide-primary alkene alternating co-polymers capable of being incorporated
into covalently-bound multilayer assemblies. Synthetic control over the monomer
substituents allows substantial versatility in the choice of layer formation chemistry and
in polymer properties. We report on the co—polymerization of substituted allyl-, allyloxy-
vinyl- and vinyloxy- monomers with a variety of N-substituted maleimides. The
substituted maleimides were prepared by reaction of maleic anhydride with selected
aromatic amines. Polymerization yields are typically in the range of 75 to 95%
consumption of the monomers. lH-NMR, FT IR and UV-visible spectroscopies were
used to characterize the polymers. The materials we report in this paper are used to
create a family of covalently-bound multilayer structures, and we report these findings in

chapter 3 of this dissertation.

-l6-

2.1 Introduction

A major effort in the materials community lies in the design, construction and
characterization of novel interfacial thin films, with the goal being the ability to control
the resultant macroscopic interface properties. Such materials have been designed and
demonstrated for specific target applications, including chemical separations,"2 nonlinear
optics,3'5 chemical and biological sensing,6'9 and heavy metal analysis and remediation.”
‘2 These interfacial materials have been applied to surfaces by spin coating, Langmuir-

13-16

Blodgett film formation, monolayer self-assembly (SAMs) and metal phosphonate

”’19 To extend the utility of interfacial chemistry into new areas and to

chemistry.
improve upon current technology, the development of a range of chemical approaches
that are easy to implement and that result in chemically and physically robust materials
will be required. We report here on the synthesis and characterization of several
alternating copolymers that can be used for layered interface growth. The chemical
identity of the side groups of these alternating copolymers renders them capable of
discrete layer growth, where the interlayer linkages are covalent. In addition to having
control over the interlayer linking chemistry, we can also demonstrate control over the
resulting properties of the interface through the judicious choice of maleimide monomer
functionality.

The primary point of the work we present here is the demonstration of two
distinct levels of synthetic and structural control we exercise over the properties of

alternating copolymers. We design alternating copolymers with specific pendant side

groups to gain explicit control over the way in which individual polymer layers can be

-17-

assembled or cross-linked into multilayers. The functionality incorporated into the
polymers allows control over the macroscopic properties of the resulting interface.

A specific aim of this work is to report the synthesis of polymers capable of
covalent interlayer attachment. Earlier work has shown that metal phosphonate interlayer
linkages limited excitation transport because of the polarizable nature of the inorganic

“sheets” that connected the individual layer constituents.20

While this property may be
advantageous for certain applications, it can be a liability for others. To circumvent this
issue, we have devised a means of forming well-controlled multilayer assemblies where
the interlayer linking chemistry is covalent in nature. Because of the number of different
functionalities we use in achieving this goal, we report in this paper on the synthesis and
characterization of the polymers that comprise the individual layers. This paper is not
intended to be a comprehensive review of covalent multilayer assembly, nor is it a study

2' The work we present here is intended to

of energy transfer in confined environments.
demonstrate the structural control and versatility of selected alternating copolymer
systems.

We report the synthesis and characterization of a family of alternating copolymers
that will be used for controlled multilayer assembly.22 The chemistry we use to
synthesize the polymers is based on established maleimide-vinyl ether alternating
copolymerization; we demonstrate here that vinyl and vinyl ether monomers can be used
with equal success. These polymers are robust and afford substantial control over their
properties through the identity of the monomer substituents. We use vinyl and vinyl

ether monomers possessing terminal functionalities capable of participating in amide,

ester, ether, urea and urethane formation chemistry. We report the assembly of these

-13-

polymers into discrete multilayer structures in the following chapter.22 A portion of this
work is dedicated to the synthesis of substituted maleimide monomers because these

moieties influence the properties of the bulk polymer significantly.

2.2 Experimental

We have designed two categories of monomers and polymers that use and build

upon the maleimide vinyl ether polymerization chemistry reported previously by our

-2
group.23 5

Syntheses: Alternating copolymers were prepared by radical polymerization of
the various monomer units as indicated in Scheme 2.1. In the first group of syntheses,
substituted maleimide monomer units are copolymerized with a variety of vinyl and vinyl
ether co-monomers possessing functionalities capable of participating in several different
interlayer linkage schemes.22 In the second group of syntheses, the functional groups of
the maleimide moiety are capable of participating in interlayer linkage chemistry. These
substituted maleimides are copolymerized with the ethylvinylether-Z—
diisopropylphosphonate (VEP) to form a polymer that is also capable of participating in
ionic interlayer linking chemistry. For this type of multilayer growth, control over the
extent of ionic bonding within these materials is determined by stoichiometric

26’” and can potentially be used to mediate (ionic) conductivity in

deprotection chemistry
these materials. We have shown that the presence of the VEP groups can be used to
sequester metal ions, extending the utility of these materials further.

Maleimide monomer preparation: The maleirrride monomers were prepared by

reacting maleic anhydride with a substituted aniline. This reaction results in the

-19-

formation of the corresponding maleamic acid with typical yield of >95% for all
monomers reported here (Scheme 2.1). Cyclization of the maleamic acid is accomplished
by reaction with sodium acetate in acetic anhydride at 70° C for 2-3 hours (Scheme 2.1).
Typical yield for the ring closing reaction is greater than 75%. After ring closure, the
maleimide monomers are collected from ice-cold water with stirring followed by
filtration and rinsed with cold water. Ring closure was confirmed by the absence of a
carboxylic acid proton in the 1H-NMR spectrum. The resulting substituted maleimide is
used in copolymer preparation. Substituted vinyl and vinyl ether monomer were
available commercially (Aldrich), except ethyl vinyl ether diisopropylphosphonate which

was prepared as described elsewhere.24

0 o o
R R R
..+..@_. / @ ——-. me
011
o o 0

Scheme 2.1. Reaction of maleic anhydride with an aromatic amine to produce the

maleamic acid, followed by ring closure to the maleimide.

Copolymerization of the monomer units (Scheme 2.2) was performed according
to the following procedures unless otherwise noted. The constituent monomers were
dissolved in chloroform and 1-25 mol% of the radical initiator 2,2’-azobisisobutyronitrile
(AIBN) was added to the solution. The reaction mixtures were refluxed with stirring at

65°C for 2 - 24 hours, as determined by observing a change in solution viscosity or, in the

-20-

case of the N-phenylmaleimide acrylic acid polymer, a white suspension/precipitate was
obvious after about two hours and yield was ~95%. Reactions were carried out under an
argon atmosphere. Upon completion of the reaction, excess solvent was removed by
rotary evaporation and the resulting solution was diluted with ether or hexanes and stirred
for ~30 minutes to separate residual monomer from the polymer product. It should be
noted that collection of the product from hexanes gives significantly higher yields in all
cases. The resulting polymer was collected by filtration, air-dried and characterized by

UV-visible, FIIR, and 'H-NMR.

R\
o
fifiiif:i>htt +' Qi? -AUU1bL> n
o N o o N o
I \R. I
R R
R!
A + \ AIBN.
o o o o n
i‘ i
R R' R

Scheme 2.2. Synthetic route for the polymerization of maleimide and vinyl ether
monomers (top), and for maleimide and vinyl monomers (bottom). Both reactions are
initiated with AIBN.

-21-

Purification of the polymers was achieved by re-dissolving the product in a small
amount of chloroform and collection from ether or hexane once or twice as necessary.
Poly(N-phenylmaliemide-I -vinyloxy-4-butanol): Poly(N-phenylmaliemide- 1 -
vinyloxy-4-butanol), (poly(NPM-VOB), (Figure 2.1a) was prepared by reacting
equimolar amounts of N-phenylmaleimide with 1-vinyloxy-4-butanol in the presence of
the initiator. Typical yield for this polymerization is ~80%. Purification of the product
was accomplished by dissolution in chloroform and collection from ether or hexane.
Poly(N~phenylmaleimide-3-aminopropyl-1 -vinyl ether): Poly(N-
phenylmaleimide-3-aminopropyl-l-vinyl ether), (poly(NPM-AVPE), (Figure 2.1b) was
prepared by reacting equimolar amounts of N-phenylmaleimide and 3-aminopropyl-l-
vinyl ether in chloroform in the presence of AIBN. Reaction procedures and conditions
are the same as reported above. Typical yield for this polymerization is ~75%.
Poly(N-phenylmaleimide-4-pentenoyl chloride): Poly(N—phenylmaleimide-4-
pentenoyl chloride), (poly(NPM-4PC), (Figure 2.1c) was prepared by reacting equimolar
amounts of N-phenylmaleimide and 4-pentenoyl chloride in chloroform in the presence
of AIBN. Reaction procedures and conditions are the same as reported above. Typical
yield for this polymerization is 50% with 10 mol% AIBN initiator and > 85% with 20
mol% AIBN. Here again, polymer purification was accomplished by re-dissolving the
product in a small amount of chloroform and collecting from ether or hexane. The
polymer possesses acid or acid chloride side groups and the spectroscopic data do not
show evidence for hydrolysis of the acid chloride group even after air-drying the product

for extended periods.

-22-

Poly(4-nitr0phenylma1eimide-allylisocyanate ): Poly(4-nitrophenylmaleimide-
allylisocyanate), (poly(4NAI), (Figure 2.1d) was prepared by reacting equimolar amounts
of 4-nitrophenylmaleimide and allylisocyanate monomers in chloroform in the presence
of AIBN initiator. Reaction conditions and procedures are as reported above, except the
reaction time required for this polymerization was 12-18 hours. Typical yield for this
polymerization is ~80%.

Poly(3-chlorophenylmaleimde and 5-hexe-2-0ne): Poly(3—chlorophenylmaleimde
and 5-hexe-2-one), (poly(3CPM-hex), (Figure 2.1e) was prepared by reacting equimolar
amounts of 3-chlorophenylmaleimde and 5—hexe-2-one in chloroform in the presence of
AIBN (20 mol%). Reaction conditions were the same as those reported above. Typical
yield for this polymerization is ~50%.

Pol y( I -maleimid0pyrene and 3 -aminopr0pyl -1 -vinylether): Poly( 1 -
maleimidopyrene and 3-aminopropyl—1-vinyl ether), (poly(MP-APVE), (Figure 2.11) was
prepared by dissolving equimolar amounts of both monomers in chloroform using AIBN
as initiator (10 mol%). Reaction conditions were the same as reported above. Typical
yield for this polymerization is 90%.

Poly(4-hydroxyphenylmaleimide-vinyl ether diisopropylphosphonate): Poly(4-
hydroxyphenylmaleimide-vinyl ether diisopropylphosphonate), (poly(4—I-IPM-VEP),
(Figure 2.1g) was prepared by reaction of equimolar amounts of 4-
hydroxyphenylmaleimide and vinyl ether diisopropylphosphonate in chloroform, with the

addition of AIBN initiator (10 mol%). Typical yield for this polymerization is 75%.

-23-

 

O 93
O
D
g:—
/\/\O=

c1

N02

Figure 2.1. Polymer structures. (a) Poly(NPM-VOB), (b) Poly(NPM-APVE), (c)
Poly(NPM-4-PC), (d) Poly(NAI)

-24-

NHz

   

OH

- VE ,
Figure 2 1 cont’d. Polymer structures. (e) Poly(3CPM-hex), (f) Poly(MP AP ) (g)
Poly(4-HPM-VEP)

-25-

The reactive functionalities introduced into these alternating copolymers makes the
resulting materials amenable to a variety of deposition strategies where interlayer linking
chemistry can be accomplished by the formation of amide, ester, ether, urea or urethane
functionalities. The functionalized maleimide monomers employed in this work were
prepared in-house as reported previously,28 except N-phenylmaleimide, which is
available commercially, and was used after recrystallization from absolute ethanol. The
reaction schemes for substrate preparation and multilayer assembly are presented and

discussed more extensively in chapter 3.22

2.3 Results and Discussion

The primary goal of this paper is to report the synthesis and characterization of
several alternating copolymers that can be used as layer constituents in the construction
of discretely layered interfacial structures. In this work, the monomer species we use are
capable of forming covalent interlayer linkages and we report on that aspect of the work
in the following chapter.22

We have synthesized and characterized an extensive library of these compounds
and report on a representative selection of seven alternating copolymers here. The
structural versatility of the polymers we report here allows us to achieve control over the
identity of the reactive polymer side group used in covalent interlayer linkage formation.
We also have control over polymer properties by maleimide substitution to incorporate a
variety of functional groups that can be used to determine the properties of the polymer

matrix. The alternating copolymer structural motif of these materials removes the

-26-

substantial structural variability associated with random block copolymers and thus
affords control over the properties of the polymer and resulting interface.24

We discuss the characterization of each polymer next. For all of the polymers,
lH-—NMR and FF IR show no evidence of unreacted C=C-H groups, indicating that
separation of the monomer reactants from the polymer product was essentially complete.
The IR data exhibit no hands for the vinyl ether C=C stretch in the 1660-1610 cm'1
region nor any =C-H wagging resonance in the 900 cm'1 to 1000 cm'l region. Previous
GPC results for polymers prepared under similar conditions indicated molecular weights
of approximately M, = 10,800.

Poly(NPM—VOB): The infrared spectrum of poly(NPM-VOB) contains several
characteristic bands that are useful to the structural characterization of this polymer
(Figure 2.2). First are the characteristic aromatic methylene stretches associated with the
N-phenyl ring between 3000 and 3100 cm'l. The aliphatic CH2 stretches between 2850
cm’1 and 2950 cm'1 are associated with the butyl group and the band centered around
3475 cm"1 is characteristic of hydroxyl stretches. The width of this band is indicative of
intermolecular hydrogen bonding within this polymer.29 There are two carbonyl
stretches, the first centered at 1775 cm‘l and the second at ~1710 cm". This spectral
signature is characteristic of the presence of the maleimide groups and is evident in all of
the polymers we report here. The band at 1775 cm'1 is associated with the symmetric
C=O stretch and the band at 1710 cm"1 is associated with the asymmetric C=O stretch in
the succinimide moiety. We also observe the characteristic C=C stretching modes of the
phenyl ring at 1597 cm", 1500 cm], and1457 cm'1 and the out-of—plane ring bending at

692 cm".29

-27-

absorbance (a.u.)

 

 

 

 

 

 

 

 

 

WW

I A a l

4000 3500 3000 2500 2000 1500 1000 500

 

frequency (cm'l)

Figure 2.2. FTIR spectrum (KBr pellet) of poly(NPM-VOB).

The band centered at 1387 cm‘1 has been assigned to the C-N bond of the tertiary
aromatic amine (aniline).29 This band is centered at 1387 cm'1 rather than the expected
1360 cm'1 due to resonance stabilization of the amine group with the phenyl ring as well
as the presence of the succinimide carbonyl functionalities. The 1189 cm'I band has been
assigned to the maleimide C-N-C stretch.30 Because C—O-C stretches also occur in this

region, it is possible that the intensity of this band is the result of overlap of the C-N-C

-28-

and C-O-C vibrations. This is likely the case based on the noticeably higher intensity of
this band in the vinyloxy- copolymers. The bands at 1073 cm'1 and 1090 cm'1 are also
associated with the C-O-C stretch. The bands at 755 and 734 cm'1 have been assigned to
the C-N symmetric stretching modes.

lH-NMR spectroscopy has also been useful in characterizing the polymers we
report here. An important indicator of the polymerization process is the disappearance of
C=C-H bands of both monomers (5 = 7.1 ppm for the maleimides and 5:62, 3.95 and
4.05 ppm for the protons adjacent to the ether oxygen and the two terminal protons of 1-
vinyloxy-4-butanol, respectively), consistent with essentially complete separation of the
polymer product from the monomers (Figure 2.3). We see no evidence of
homopolymerization of any monomers. Extensive overlap and broadened peaks in some
regions of the spectra makes unequivocal peak assignment impossible, particularly for
aliphatic methylene groups. We observe the protons of the NPM aromatic ring at 8 = 7.2
- 7.6 ppm. The butyl protons adjacent to the hydroxyl oxygen are found at 5 = 4.2 - 4.6
ppm and both the aliphatic protons of the butyl group and the bridging protons between
the maleimide and the vinyloxy- groups are seen as a broad resonance between 5 = 1.0
and 1.6 ppm. The succinimide protons are seen at 5 = 3.3 ppm and the multiplet feature
at 3.6 ppm is assigned to the methine protons adjacent to the ether oxygen. The alcoholic

proton is not seen in the spectrum because of facile proton exchange.

-29-

 

 

 

- MLLAC

YTYTTYTTTfiII YfiT‘IfTrTYYYYTY'WYIYTYYIY‘TYYII I "IffiVTY‘ Y‘IV VTWIV I 11Tfi

”1511109 8 7 6 5 4 3 210 - -
chemical shift (Dom)

 

 

 

Figure 2.3. 1H-NMR spectrum of poly(NPM-VOB).

-30..

Poly(NPM-APVE): The FTIR spectrum of this polymer reveals several
characteristic bands in addition to those seen for all of the polymers we report here, as
noted above (Figure 2.4). The band centered at 3408 cm“1 has been assigned to the
associated NH; stretches of the amino group. We also see the NH bending mode

resonance at 1648 cm".

absorbance (a.u.)

 

 

A

 

4000 3500 3000 2500 2000 1500 1000 500
frequency (cm'l)

Figure 2.4. FTIR spectrum (KBr pellet) of poly(NPM-APVE).

-31-

The 1H-NMR spectrum of poly(NPM-AVPE) (Figure 2.5) contains the expected
features such as the OH resonances associated with the phenyl ring between 5 = 7.2-7.6
ppm. The aliphatic methylene bands are contained in the broad resonance between 5 =
1.2-1.8 ppm, succinirrride protons are found at 5 = 3.2 ppm, and the propyl methylene
protons adjacent to the ether oxygen are at 5 = 4.45 ppm. The proton exchange rate for
the amino proton precludes observation of this resonance, in analogy to the hydroxyl

proton resonance for poly(NPM-VOB).

 

 

 

VV '1‘" V v' r—— r

chemical shift (ppm)
Figure 2.5. 1H-NMR spectrum of poly(NPM-APVE).

-32-

Poly(NPM-4-PC): Much of the vibrational and electronic spectral signature of
poly(NPM-4-PC) is the same as that of the other copolymers we report here (Figure
2.6a). We also see the characteristic aliphatic methylene stretches associated with the
pentenyl- portion of the polymer. There are several additional features associated with
the acid chloride moiety in this polymer. For poly(NPM-4-PC), the 1189 cm'1 band is
weaker than that seen for poly(NPM-AVPE) because there are only the C-N-C stretches
for poly(NPM-4-PC). The absence of contributions from the vinyloxy C-O-C band
accounts for the attenuation of this feature. For poly(NPM-4-PC) there is a resonance at
~3450 cm", suggesting intermolecular hydrogen bonding of the hydrolyzed acid chloride
functionality. The carbonyl stretching region is characterized by bands associated with
the succinimide, the acid chloride and some carboxylic acid (hydrolyzed acid chloride).
The 1H-NMR spectrum of poly(NPM-4-PC) (Figure 2.6b) shows no evidence of
hydrolysis of the acid chloride to the acid even after extensive periods of atmospheric
exposure, and this finding is not consistent, at face value with the IR spectra. These two
pieces of information suggest that the FI‘IR (KBr crystals) sample may have been wet.
Aromatic, aliphatic and succinimide protons occur in the same region as seen for the

other polymers.

-33-

absorbance (a.u.)

 

 

W

4000 3500 3000 2500 2000 1500

 

41% 4

1000 500

frequency (cm'l)

 

 

 

wait.

Y ‘Tj f1 T Ti 1' 1' I 1 V V Y Y ‘ V V I I Y ‘ V V Y Y

12"‘1r“'m”"9” 8 7 6 5 4....3....,2...1

chemical shift

Figure 2.6. (a) IR spectrum of poly(NPM-4-PC). (b) 1H-NMR spectrum of poly(NPM-
4-PC).

 

-—4

-34-

Poly(4-NAI): The IR band of the isocyanate group at 2265 cm'1 region is
characteristic of this polymer (Figure 2.7a). We also observe the characteristic aromatic
and aliphatic methylene stretches and the succinimide C=O stretches at 1775 and 1720
cm']. In addition, we observe the asymmetric ArNOz stretch at 1523 cm'1 and the
symmetric ArNOz stretch at 1347 cm'l. We note the lower frequencies of the phenyl ring
as well as the splitting of some of these bands into doublets due to ring substitution. We
also see an intensity decrease in the C-N-C stretching region due to the absence of the C-
O-C functionality in this polymer. The 854 cm'1 band is associated with the aromatic C-
N stretch of the nitro group. We observe a symmetric splitting of the aromatic protons
peak in the 1H-NMR spectrum of poly(4-NPM—AI) which is expected due to the presence
of the N02 group in the 4—position. Para- substitution of the phenylmaleimide leads to
equivalence of the two protons adjacent to the substituent leading to one peak for both
protons (Figure 2.7b). The resonance for the methylene group protons adjacent to the

isocyanate group is in the same region as the bridging methine and succinimide protons,

5 = 3.15 - 3.5 ppm.

-35-

 

absorbance (a.u.)

 

 

 

l n l 1 l 4 l n

4000 3500 3000 2500 2000 1500 1000 500

frequency (cm'l)

 

 

 

10 9 8 7 4 3 2 1
chemical shift (ppm)

Figure 2.7. (a) IR spectrum of poly(4NAI), (b) lH-NMR of poly(4NAl).

—36-

Poly(3CPM-hex): For this polymer, the IR band centered at 3475 cm'1 is likely
an overtone of the carbonyl band (Figure 2.8). We make this assignment because of the
absence of amino or hydroxyl functionality in this polymer. We observe characteristic
aromatic and aliphatic methylene stretches between 2850 and 3100 cm‘I and maleirrride
C=O stretches are seen at 1780 cm'1 and 1715 cm'l. The broadened carbonyl band at
1715 cm'1 is the result of spectral overlap between the maleimide C=O stretch and the
ketone C=O stretch. Aromatic C=C stretches are at slightly lower frequencies than for N-
phenylmaleimide due to the presence of the Cl— substitution at the 3- position. The C-N

stretch of the tertiary amine is seen at 1382 cm'1 for poly(3CPM-hex).

absorbance (a.u.)

 

 

 

 

 

 

 

 

ti

1 l L l

4000 3500 3000 2500 2000 1500 1000 500

frequency (cm'l)

 

Figure 2.8. IR spectrum of poly(3CPM-hex).

-37-

lH-NMR spectroscopy of poly(3CPM-hex) (Figure 2.9) shows the expected
aromatic protons at 5 = 7.2 - 7 .6 ppm. The aliphatic methylene resonances are seen at 5 =
1.5 ppm with the methyl protons adjacent to the C20 functionality occurring at 5 = 2.05
ppm. The succinimide resonances are found at 5 = 3.3 ppm, consistent with the NMR

spectra of the other polymers we report here.

 

 

 

 

 

 

chemical shift (ppm)

Figure 2.9. ‘H-NMR spectrum of poly(3CPM-hex).
Poly(MP-APVE): The IR spectrum of this polymer contains several features

unique to this polymer (Figure 2.10a). We observe the -NH2 stretch at 3277 cm]. The

position of this band suggests that the NH; groups exist primarily in the associated form.

-33-

We note the presence of a high-frequency shoulder on the -NH2 band, indicating some
free -NH2 groups. The aromatic C-H stretches are consistent with the pyrenyl ring
structure. We also observe the characteristic maleimide C=O bands and a broad band at
1632 cm], which we assign to the N-H bending mode(s). This band could also be
present due to some amount of maleirrride ring opening during polymerization. We
observe a relatively strong band at 1194 cm'1 that we attribute to the presence of both the
CDC and C-N-C functionalities in the polymer. The intense band at 844 cm'1 has been
assigned to the out-of—plane C-H bending modes of the polynuclear aromatic moiety.
Examination of the 1H-NMR spectrum (Figure 2.10b) reveals a complex spectral
signature for the aromatic proton resonances at 5 = 8.0 - 8.8 ppm, consistent with the ring
structure of the pyrene moiety. We observe the methylene and succinirrride protons at the
same chemical shifts as seen for the other polymers we report here. There is a small
resonance at 5 = 10.8 ppm, suggesting some succinimide ring opening occurring either as
the result of incomplete ring closure of the monomer or as a result of the polymerization
reaction. Given the integration of this band for 0.3 protons, it appears that ring opening,
if it is proceeding, is not a dominant process. As a point of reference, the 1H NMR
spectrum of the pyrene maleimide monomer shows no evidence of incomplete ring

closure.

-39_

a

Of?-

013-

d

015-

d

011-

 

 

 

Absorbance

OJB— i

 

 

0.2 - fi

OJ —
1

 

 

 

0.0 i 1
4000 3500

 

I I’

I I ' I - I ' I I I
3000 2500 2000 1500 1000 500

Frequency (cm")

 

 

J J I
A.
1 I V 1 V I V Y

,.
I Y Yfi w I T 7 1 T ‘ V T V ‘ ‘ 7 I V I I Y I V I I V I ' 1 ‘ V Y Y I T Y r T fT V V I Y

12 11 10 9 8 7 6 5 4
chemical shift (ppm)

 

“—4
N“.
H

Figure 2.10. (a) IR spectrum (KBr pellet) of poly(MP-APVE), (b) IH-NMR spectrum of
poly(MP-APVE).

-40-

Poly(4-HPM-VEP): The IR spectrum of this polymer shows intermolecular
hydrogen bonding based on the -OH stretching resonance of the phenol hydroxyl group at
~3430 cm'1 (Figure 2.11a). The strength and spectral width of this band obscures the
aromatic methylene stretches. We note the presence of the maleimide C=O stretches.
This polymer exhibits several IR bands that are characteristic of phosphate or
phosphonate. The phosphoester P=O stretch is superimposed on the phenol ring C-O
stretch at ~1200 cm". We also see the P-O-C stretch at 995 cm]. This band is important
in structural assignment because it demonstrates the absence of the oxidation of the
phosphoester groups to the corresponding phosphorus acid. We deliberately keep the
phosphonate groups protected during multilayer assembly so they can be used for metal
ion sequestration once the multilayer structure is formed.31 The 1H-NMR spectrum
(Figure 2.11b) of this polymer reveals an aromatic resonance centered at approximately
7.0 ppm. Again we observe the splitting of this peak due to the presence of the para-
hydroxyl group, where the splitting is a result of ring substitution at the 4- position. We
also observe a peak centered at approximately 9.8 ppm which may be due to ring opening
or, possibly, to hydrolysis of some of the phosphonate groups. However, either of these
processes contribute negligibly due to the integration of the peak for ~ 0.3 relative to the
other peaks. The spectral signature between 1.0 and 4.6 ppm has been assigned to the

various contributions of the vinyl ether phosphonate resonances as reported previously.24

-41-

 

 

 

 

 

 

 

2?
3

Q)

8
.8

‘5 i
8

(<3

4000 3500 3000 2500 2000 1500 1000 500
frequency (cm‘l)
b

 

 

 

chemical shift (ppm)

Figure 2.11. (a) IR spectrum (KBr pellet) of poly(4-HPM-VEP), (b) 1H-NMR spectrum
of poly(4-HPM~VEP).

-42-

2.4 Conclusions

We have reported on the synthesis and characterization of a family of alternating
copolymers possessing a variety of functional side groups capable of forming covalent
linkages for controlled multilayer assembly. Some of the polymers we have prepared are
capable of participating in either ionic or covalent interlayer linking chemistry. We have
demonstrated the ability to introduce a variety of functional groups into the matrix
through substitution of the maleimide N—substituent. This substantial structural
versatility is put to use in the following chapter and we anticipate using these polymers in

the design of chemically selective interfaces.

-43_

25 Literature Cited

1. Vrancken, K. C.; VanderVoort, P.; Gillisdhamers, 1.; VanSant, E. F.; Grobet, P.; J.
Chem. Soc., Faraday Trans. 1992, 88, 3197.
2. Pfleiderer, 8.; Albert, K.; Bayer, E.; J. Chromatogr. 1990, 506, 343.
3. Bakiamoh, S.; Blanchard, G. J.; Langmuir, 2001 17, 3438.
4. Katz, H. E.; Scheller, G.; Putvinski, T. M.: Schilling, M. L.; Wilson, W. L.; Chidsey,
C. E. D.; Science, 1991, 254, 1485.
5. Li, D.; Ratner, M. A.; Marks, T. J.; Zhang, C. H.; Yang, J.; Wong, G. K.; J. Am.
Chem. Soc., 1990, 112, 7389.
6. Demas, J. N.; Degraff, B. A.; Coleman, P. B.; Anal. Chem. News & Features,
1999, 793A.
7. Yoon, H. C.; Kim, H-S Anal. Chem. 2000, 72, 922.
8. Delamarche, E; Sundarababu, G.; Biebuyck, H.; Michel, B.; Gerber, C.; Sirgrist, H.;
Wolf, H.; Ringsdorf, H.; Xanthopoulos, N.; Mathieu, H. J .; Langmuir 1996, 12, 1997.
9. Brousseau, 111, L. C.; Aurentz, D. J.; Benesi, A. J.; Mallouk, T. E.; Anal. Chem. 1997,
69,688.
10. Ng, S. C.; Zhou, X. C.; Chen, Z. K.; Miao, P.; Chan, H. S. 0.; Li, S. F. Y.; Fu, P.;
Langmuir, 1998, I4, 1748.
ll. Beatty, S. T.; Fischer, R. J .; Hagers, D. L.; Rosenberg, E.; Ind. Eng. Chem. Res. 1999,
38,4402.
12. Nooney, R. 1.; Kalyanaraman, M.; Kennedy, G.; Maginn, E. J.; Langmuir, 2001, 17,

528.

-44-

l3.

14.

15.

l6.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

Ulman, A. Chem Rev. 1996, 96, 1533.

Dubois, L. H.; Nuzzo, R. G.; Annu. Rev. Phys. Chem, 1992, 43, 437.

Zhang, Z. J.; Hu, R. S.; Liu, Z. F.; Langmuir, 2000, 16, 1158.

Skulason, H.; Frisbie, C. D.; Langmuir, 1998, 14, 5834.

Lee, H.; Kepley,. L. J .; Hong, H. G.; Akhter, S.; Mallouk, T. E.; J. Phys. Chem. 1988,
92, 2597.

Lee, H.; Kepley, L. J.; Hong, H. G.; Mallouk, T. E.; J. Am. Chem. Soc., 1988, 110,
618.

Putvinski, T. M.; Schilling, M. L.; Katz, H. E.; Chidsey, C. E. D.; Mujsce, A. M.;
Emerson, A. B.; Langmuir, 1990, 6, 1567.

Home, J. C.; Blanchard, G. J.; J. Am. Chem. Soc. 1999, 121, 4427.

Major, J. S.; Blanchard, G. J .; J. Phys. Chem, in preparation.

Major, J. S.; Blanchard, G. J .; chapter 3.

Kohli, P.; Scranton, A. B.; Blanchard, G. J.; Macromolecules, 1998, 31, 5681.

Kohli, P.; Blanchard, G. J.; Langmuir, 1999, 15, 1418.

Kohli, P.; Blanchard, G. J .; Langmuir, 2000, 16, 695.

Jung, M. E.; Lyster, M. A.; J. Org. Chem. 1977, 42, 3761.

Vickery, B. H.; Pahler, L. F.; Eisenbraun, E. J .; J. Org. Chem, 1979, 44, 4444.
Kohli, P.; Blanchard, G. J. Langmuir, 2000, 16, 8518.

Silverstein, Bassler and Morrill, Spectrometric Identification of Organic Compounds,

5th Ed., John Wiley & Sons, New York, 1991.

-45-

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

31. Major, J. S.; Blanchard, G. J. Langmuir 2001, 17, 1163.

-46-

Chapter 3

Strategies for Covalent Multilayer Growth. 2. Interlayer Linking
Chemistry

Abstract

We report a strategy for the covalent assembly of polymer multilayers at
interfaces, where growth is accomplished one layer at a time. The individual layer
constituents are maleimide-vinyl ether alternating copolymers with side groups that
possess reactive functionalities. The identity of the polymer layer side groups determines
the chemistry we employ in interlayer linkage formation. We report on the selective
creation of amide, ester, ether, urea and urethane interlayer linkages. We achieve
controlled multilayer growth using each of these covalent bonding schemes. The
resulting multilayer structures show linear growth in terms of thickness, measured
ellipsometrically, and total mass loading, measured by UV-visible and FIIR

spectroscopies.

-47-

vu—jr‘ '. 7"“T“"tl

 

3.1 Introduction

The design of interfacial materials where there is explicit control over the identity
and ordering of the constituent layers is an area of intense research effort. This chemical
approach to surface modification can find use in sensor design,” chemical separationss'6

and nonlinear optics, 7'8 for example. To date, there have been several means devised to

control layered material growth at interfaces, and among the most widely used are

9.10 11-14
11.

Langmuir-Blodgett film growt thiol/gold monolayer self-assembly and metal
bisphosphonate layer growth.15 '35 Each method has its advantages and shortcomings, so
there is a continued need to identify novel layer growth chemistry where the resulting
materials are relatively well organized and the chemistry binding the layers together is
sufficiently robust to be useful. We describe here the assembly of several polymer
multilayer systems, where the interlayer linking chemistry involves the formation of
covalent bonds. The scheme we report here is, in a sense, the layer-by-layer cross-
linking of a thin polymer film, where the structure of the polymers in a given layer
(parallel to the substrate) can be controlled. The polymers we use in the creation of the
multilayer structures have been published36 and described in detail in chapter 2. Careful
choice of the polymer side groups allows layers to be added to the interface through

3738 allowing for the assembly 0f

either ionic or covalent interlayer linking chemistry,
materials that can function either as ionic conductors or as insulators.

We are interested in covalently-linked polymer multilayers for several reasons.
Earlier work on zirconium bisphosphonate (ZP) ionically-linked multilayer structures

showed that interlayer excitation transport was inefficient by virtue of the polarizable

zirconium bisphosphonate sheets that connect the layers in these materials.35 To control

-43-

interlayer excitation transport in layered interfaces, a necessary first step is the
elimination of these polarizable layers; we replace the ZP functionality with covalent
organic functionalities (vide infra). We have chosen to use polymers as the layer
constituents in this work because of the combined ease of preparation and the degree of
structural versatility available with maleimide-vinyl ether alternating copolymers.39
Polymers provide structural integrity within individual layers, allowing the opportunity to
create relatively porous structures where there is a gradient in some property of the
interface (e.g. dipole moment) along the surface normal axis.40 Such interface properties
are likely to be of use in cherrrical separations and sensing.

The polymer systems we reported in the preceding chapter allow control over the
individual layers through synthetic design.33 The interlayer linking chemistry applied to
these polymers that we report in this chapter allows for a second level of control; we can
choose the means of interconnection between individual layers. In this work we will
focus on an array of covalent interlayer linking chemistries that we have applied to
layered material growth. The resulting materials are all robust and the strategies we
discuss should be applicable to the growth of a variety of other materials in layered
structural motifs. We anticipate that this level of interface structural control will find use

in the design of materials intended for specific purposes, such as the creation of a

chemical potential gradient normal to the substrate plane.

3.2 Experimental

In the preceding chapter, we reported on the design and characterization of

several maleimide-vinyl ether alternating copolymers. We use some of these polymers to

-49-

construct multilayer structures. We describe the procedures for substrate preparation and
then present several strategies for covalent linking of polymer layers. We have used
gold, quartz and silicon substrates for this work, with the choice of substrate being
determined by the measurement to be performed. We used gold substrates for optical
null ellipsometry and FTIR measurements, silica substrates for UV-visible spectroscopy
and oxidized silicon for ellipsometry and Brewster’s angle transmission FTIR
measurements. We have chosen specific chemical systems to demonstrate the chemistry
of interest and have not attempted to report on all of the possible permutations of
polymers and interlayer bonding schemes.

Substrate preparation: Gold, quartz and silicon wafers were prepared by cleaning
in piranha solution (3:1 H2804:H202) for ~15 minutes, followed by rinses with ethanol
and water. The substrates were then dried under a stream of nitrogen gas. The quartz
and silicon wafers were then exposed to a 5M HCl solution approximately 15 minutes
followed by ethanol and water rinse and dried under nitrogen. After cleaning and
preparation, the substrates were exposed to the appropriate solution to give the desired
terminal functionality.

To functionalize the gold surfaces with hydroxyl groups, the cleaned substrates
were exposed to a 60:40 ethanolzwater 10 m1\_/l solution of 6-mercaptohexan-l-ol for ~30
minutes at ~40°C. For the quartz and oxidized silicon surfaces, the surface silanol groups
were used directly and without prior treatment for further functionalization.

To produce amine-terminated surfaces, the gold substrates were exposed to a

60:40 ethanolzwater 10 mM solution of cystamine dihydrochloride for 30 minutes at

-50-

 

~40°C. Quartz and silicon wafers were exposed to ~5% solution of 3-aminopropyl-
trimethoxysilane in toluene at room temperature overnight.

To create a surface terminated with an acid chloride functionality, the substrate
was reacted with an (0t,0))-diacid chloride. Hydroxyl-terminated gold substrates were
exposed to a solution of 10 mL anhydrous acetonitrile, 0.3 mL 4-methylmorpholine and
0.3 mL adipoyl chloride under an inert atmosphere of argon at room temperature for 30
rrrinutes to produce a surface terminated with an acid chloride functionality. Quartz and
silicon wafers were treated in the same way, where the surface silanol groups were
reacted with adipoyl chloride.

To create an epoxide-terminated surface, gold substrates that had been previously
prepared to present a hydroxyl terminal functionality were exposed to a 1% (v/v) solution
of 1,4-butanediglycidyl ether in anhydrous acetonitrile. Concentrated sulfuric acid (0.06
mL) was added drop-wise to the reaction vessel to mediate the epoxide ring opening.
The reaction was allowed to proceed at room temperature for approximately 5 rrrinutes
before rinsing. Silicon and quartz wafers were reacted directly with 1,4-butanediglycidyl
ether in a manner similar to that for the gold substrate. For these substrates, the reaction
is directly with the silanol groups present on the surface.

For isocyanate-terminated substrates, the gold substrates treated with 6-mercapto-
l—hexanol were immersed in a solution 10 mL anhydrous acetonitrile, 0.3 mL 4-
methylmorpholine and 0.3 mL 1,6-diisocyanatohexane and heated to ~40°C for 30
minutes. As with the other reactions, the quartz and silicon substrates were treated in a

similar manner and their native silanol groups were used as the reactive functionalities.

-51-

...n i3

’4“

‘1... a.“

 

Optical Null Ellipsometry: Values of ellipsometric thickness of the deposited
polymer multilayer assemblies were acquired using a Rudolph Auto-EL H optical null
ellipsometer. This instrument is equipped with a HeNe laser operating at 632.8nm. The
Software (Rudolph DAFIBM) used for data collection was acquired from the
manufacturer. The refractive index was taken to be n = 1.54 + 0i for all films. Even
though this value may differ by a small amount depending on the identity of the polymer
film, changes over a physically realistic range affect the recovered thickness negligibly.

U V— Visible Spectrophotometry. The absorption spectra of the multilayer
assemblies were acquired using a Cary 300 UV-Visible spectrophotometer. The layers
were synthesized on quartz and the data were acquired after the deposition of each
individual layer. The scan rate for all measurements was 600 nm/min and the spectral
resolution was 1 nm. The data were processed and plotted using Microcal Origin®
version 6.1 software. In all cases, we recovered R2 values greater than 0.98 for linear fits.

Infrared Spectroscopy. FTIR spectra of the multilayer assemblies on gold
substrates were collected using a Nicolet Magna 750 FT IR spectrometer operating with
Omnic® software. The spectral resolution was set to 4 cm'1 for all spectra collected. An
external reflectance sample mount set at an incidence angle of 80° was used for data
collection. Spectra of samples grown on silicon were acquired using a Nicolet Magna-IR

550 set up to acquire data in the transmission mode using a Brewster’s angle attachment.
3.3 Results and Discussion

In this work, we report the multilayer assembly of polymers that have been
designed with various functional side groups that allow covalent interlayer linkages. The

goal of this work is to demonstrate that careful choice of the functional groups

-52-

incorporated into the copolymers provides accessibility to simple reactions that form
arrride, ester, ether, urea and urethane functionalities, allowing the regular growth of
multilayer systems. In many cases, there is more than one way to generate a particular
covalent interlayer linkage, with the order of the chemical functionalities used being
determined by the native functionality or initial modification of the substrate.

This work is intended to be more than a catalog of covalent interlayer linking
chemistry. It is our intent to demonstrate the facile nature of our approach to the
controlled growth of multilayer interface structures. In all of the examples we present
here, the strategy is essentially the same. We use alternating copolymerization chemistry
where the polymerization reaction is templated from the substrate initially. At each step
of the growth process, the reactive interface cannot grow another layer until it is
presented with the requisite monomer. By presenting the interface with only one type of
monomer at a time, controlled growth of individual, discrete molecular layers is possible.
In cases where the same multilayer is available through multiple routes (save for
orientation of the interlayer linking functionality), we report the reaction(s) that we have
demonstrated experimentally. Depending on the monomer species used in the MVE
polymer synthesis, either the maleimide or vinyl ether side groups can be used for
interlayer attachment.

Amide chemistry. Amide interlayer linkages are created by reaction of an acid
chloride with an amine (scheme 3.1) and we report three viable routes. (3) An amine-
terrninated substrate is exposed to a solution of an MVE alternating copolymer with acid
chloride side group functionalities, resulting in the direct formation of an amide bond

between the polymer side groups and the substrate. Subsequent layers are added by

-53-

". fimzl‘fla’ '1"

exposure of the acid chloride containing polymer (topmost) layer to an (0t,c0)-diamine,
followed by reaction of the aminated surface with the MVE alternating copolymer. (b)
An acid chloride-terminated substrate is exposed to a solution of an MVE alternating
copolymer with amine-terminated side groups. Controlled multilayer growth is achieved
by reacting the polymer amine side group with an (0t,0))-diacid chloride (adipoyl chloride)
followed by exposure to the MVE polymer. This layer growth scheme is identical to the
first save for the orientation of the resulting amide interlayer linkage. (c) An amine-
terrninated substrate is exposed to a MVE alternating copolymer containing acid chloride
side groups, followed by exposure to a MVE alternating copolymer with amine-
terminated side groups. In this scheme, no interlayer linker is necessary. The choice of
initial polymer layer in this scheme is determined by the reactive functionality presented

by the substrate.

-54-

° )L
/U\ + R'—NH2 ——4» /R

 

 

 

Scheme 3.1. Reaction schematic and idealized structures of amide-linked MVE layered
assembly. (a) Amine-containing MVE polymer bound to acid chloride functionalized
surface and (b) acid chloride containing MVE polymer bound to amine-functionalized
surface.

Co-polymerization of 3-chlorophenylmaleimide and 3-aminopropyl-l-vinyl ether
produces poly(3-CPM-APVE),36 a polymer that is capable of participating in amide
interlayer linkage formation. Reacting poly(3-CPM-APVE) with adipoyl chloride leads

to the formation of amide interlayer linkages. The IR spectrum of the amide-linked

-55-

 

multilayer assembly (Figure 3.1a) reveal amide I, H and HI bands, where the amide I
band corresponds to the carbonyl stretching mode centered at approximately 1655 cm".
Amide H and III bands correspond to N-H bending vibrations and C-N stretching

modes occurring at approximately 1595 cm'l and 1421 cm", respectively. In addition,
the band centered at ~3330 cm'1 corresponds to the stretching modes of the free -NH
groups. UV—visible spectroscopy is used to monitor layer deposition, by observing the
aromatic absorbance band centered at 248 nm (Figure 3.1b - inset). Errors were
calculated based on the noise level and found to be less than 2% in all cases reported.
The observed spectral red shift relative to the N-phenylmaleimide copolymer is the result
of the chloro functionality at the 3- position. We observe a loading density for the amide-
linked assembly that is constant with number of MVE polymer layers (Figure 3.1b).
From the ellipsometric data (Figure 3.2) we recover a slope of 19 A/Iayer for the amide-

linked multilayer, consistent with that seen for several other interlayer linking strategies.

-56-

 

 

 

 

 

().()()(i —
31‘
t3
Eg’ ().()()4- —
5 t i
O .
jg ().()Cl2. — E
0 . O O 0 ‘ l ‘ ' ‘ I’IV - l g 1
z1000 3500 3000 2000 1500 1000
hequency (cm‘fi
0.12 —
- (r08- 1
E .
ff 0'10 _ £0.06» ,.
V _ 0 04
Q) 0 .08 g ’ /+
g ' .§ 0.02- i
e 0.06 — 0 00> . . .
8 ' i 0 1 2 3 4
"D — layer number
{1 0.04
0.02 -
0 .00 ‘ ‘ ’f-iAvxjx‘fcm bibm-I

 

 

200

250

300 350

Wavelength (nm)

Figure 3.1. (a) FTIR spectrum of poly(3CPM-APVE) with amide interlayer linkage
(b) UV-Visible spectra as a function of number of layers for poly(3CPM-APVE).

-57-

90
80
70

I 1 T T I—l

40;
30-
201
103-
0’ . . . .
0 r 2 3 4

Layer number

Thickness (A)

 

 

Figure 3.2. Dependence of ellipsometric thickness on number of layers for
poly(3CPM-APVE) with amide interlayer linkage.

Ester chemistry. Ester interlayer linkages are achieved by the reaction of an acid
chloride with an alcohol (scheme 3.2). There are three routes to multilayer formation
using this chemistry. (a) A MVE alternating copolymer with hydroxyl-terminated side
groups is reacted with an acid chloride terminated substrate. Multilayer assembly is
achieved by reacting the polymer hydroxyl side groups with a diacid chloride followed
by exposure to the MVE polymer. (b) The complement of the first route can be used,
where the MVE polymer contains acid chloride terminated side groups and the interlayer

linking moiety is an (0t,0))-diol.

-58-

muar‘rfi-‘fl

0 AL
A + R'—OH —> /R'

 

 

1

Scheme 3.2. Reaction schematic and idealized structures of ester-linked MVE layered
assembly. (a) alcohol-containing MVE polymer bound to acid chloride
functionalized surface and (b) acid chloride containing MVE polymer bound to
hydroxyl-functionalized surface.

To realize this chemistry, we used the alternating copolymer of N-
phenylmaleirrride (NPM) and l-vinyloxy-4-butanol (VOB), poly(NPM-VOB); The
reaction of poly(NPM-VOB) with adipoyl chloride results in an ester—linked multilayer
assembly. Linear growth was seen in the ellipsometric data (Figure 3.3a), with a slope of
21 A/Iayer. As discussed in earlier work, it is not a straightforward matter to evaluate the

thickness of a layer relative to a molecular structure calculation (e.g. molecular

-59-

wmm:m1l

mechanics) because the loading density of the interface is not known quantitatively. The
thickness) we recover experimentally is consistent with other multilayer structures we
report here (vide infra) and have reported elsewhere. The absorbance data indicate a
linear growth as well, (Figure 3.3b - inset), although it is not possible to determine the
absolute loading density from these data because the extinction coefficient at 220 nm for

poly(NPM-VOB) is not known.

-60-

mica . «YEA-til ii

 

 

 

 

 

 

100 I-
90 : ",
80 h ,”’
A 70 " ell’
°§ 60 - I,”
8 I- ’1
50 1" [’1
.4 . -
5 40 — I,’
30 "' ’1”
I- I.’
20 -— ,z
10 —
0 p l l I l
0 l 2 3 4
Layer number
_ 0.08~
g 0.07; '
0.15 - § 0.06L '
A v >
3' g 0.05- .
“i i E 0.04.
G) "“ 0.03-
0 I— .
g 0.10 ,
.8 r 0 I 2 3 4
o layernumber
3
<1 0.05 —
0.00 " 1 A I . Wm .3 . .2” ”44.14
200 250 300 350

Figure 3.3. (a

Wavelength (nm)

) Dependence of ellipsometric thickness on number of layers for

poly(NPM-VOB) with ester interlayer linkage. (b) UV-Visible spectra as a function
of number of layers. Inset: Dependence of absorption maximum on number of layers.

-61-

-___.,._-m

A slight modification of this reaction scheme has been reported previously.“ In
that work, the reactive polymer side group is attached to the maleirrride monomer and the
vinyl ether monomer is terminated with a diisopropylphosphonate functionality, the
interlayer linking chemistry can proceed through the maleimide monomers. In this
reaction scheme, the monomers are 4-hydroxyphenylmaleimide (HPM) and ethyl vinyl
ether diisopropyl phosphonate (VEP), and the interlayer linking moiety is adipoyl
chloride. We note that phosphoester formation is precluded in this layer growth scheme
due to presence of the diisopropyl protecting groups on the phosphonate functionality.“

The MVE polymer is capable of participating in ionic and/or covalent interlayer linkages.

.O

O

p—I

O
l

a
g 0.008 -
e
g 0.006L
g .
.0
§
.8 b

 

 

 

 

4000 l 3500 ' 3000 2000 ' 1500 . 1000 ‘
frequency (cm-1)

Figure 3.4. Poly(HPM-VEP) with ionic and covalent interlayer linkages. (a) Infrared
spectrum of ionically-bound multilayers and (b) IR spectrum of covalently-bound
multilayers. The spectra have been offset for clarity.

-62-

For ionic multilayer formation, the phosphonate groups are deprotected by
reaction with bromotrimethylsilanef’l resulting in the efficient hydrolysis of these groups
to the corresponding phosphonic acid. Sequential exposure of the interface to M4+ (M =
Zr, Hf) and the MVE polymer, with deprotection chemistry used as required, results in
ionic multilayer formation. Figure 3.4a shows the infrared spectrum of a multilayer of
poly(HPM-VEP), linked by the formation of zirconium bisphosphonate moieties. For
covalent interlayer linkage, the terminal hydroxyl group of the poly(I-IPM-VEP) polymer
is capable of participating in ester, ether or urethane interlayer linkages. Figure 3.4b
shows the IR spectrum of a 4 layer stack of poly(HPM-VEP) assembled using ester
linking chemistry. The ester carbonyl stretching is centered at ~1720 cm”l (superimposed
on the maleimide asymmetric C=O stretch). The broad band centered around 3200 cm']
is an indication of residual water or hydrogen-bonded phosphonic acid functionalities
within the multilayer. The ellipsometric and UV-visible spectroscopic data for this
multilayer indicate linear growth. We recover a slope of 16A/layer for the ellipsometric
data (Figure 3.5). Regression of the UV-visible absorbance data was made possible by

monitoring the aromatic absorbance band centered at 231 nm (Figure 3.5 - inset).

-63-

.r-Kiufi

      

.0
o
[\J
o

absorbance (231 um)

i i 3 4
layer number

 

  

Absorbance (a.u.)

 

g:
8

 

200 A 250 I 300 350
Wavelength (nm)

Figure 3.5. UV-Visible spectra as a function of number of layers. Inset: Dependence
of absorption maximum on number of layers.

Ether chemistry. We have demonstrated ether connections between MVE
polymer layers using a ring-opening reaction of an epoxide with a hydroxyl group
(scheme 3.3). The reaction of an MVE polymer with hydroxyl-terminated side groups
with an epoxide-functionalized substrate in the presence of a strong acid leads to ether-
linked multilayer assemblies. Subsequent exposure of the MVE polymer surface to an
(01,0))-diepoxide yields an epoxide-terminated surface ready for further reaction.

The reaction of an adlayer of poly(NPM-VOB) with 1,4-butane diglycidyl ether
leads to ether linkages. Multilayer assembly was monitored using ellipsometry, and

FI‘IR and UV-visible spectroscopies. A linear response was observed for the regression

-64-

1s citrate-19:1.

of the ellipsometric data (Figure 3.6a), and we recover a slope of 29 A/Iayer. This
thickness is substantially greater than that which we observe for the other multilayer
assemblies suggesting either greater loading density or greater conformational disorder
within the MVE polymer layer. We note that the formation of ether interlayer linkages
appears to be limited to the deposition of two or three layers before terminating. There

are several interesting features contained in the data for this system.

 

j";
R—Q) + R'—OH ——> R O‘R'

H*

WT.—

 

HO

 

 

 

 

Scheme 3.3. Reaction schematic of a glicidyl ether with an alcohol (top) to form an
ether and an idealized structure of an alcohol-containing MVE polymer bound to an
epoxide functionalized substrate. Subsequent layer growth requires reaction of the
alcohol-containing MVE polymer with a diglicidyl ether.

-65-

0.08

0.06

0.04

absorbance (a.u.)

0.02

0.00

100 Z
90
80 :
70 _
60 _
50

40
30
20
10

Thickness (A)

 

 

     

 

 

 

 

r I
0.16 '
_ A. ’ a
g 0.12~
E?
g 0.08» .
" E 0.04»
i 0'000 I 2 3 4
_ layer number
- J A l g l A I
200 250 300 350
wavelength (nm)
b
- .I’
I
I,
... I”
l’
I— I,
1’.
[I
r f”
. L 1 1 1
0 1 2 3 4

Layer number

Figure 3.6. (a) Dependence of ellipsometric thickness on number of layers for

poly(NPM-VOB)

with ether interlayer linkage. (b) UV-Visible spectrum for

polym(NPM-VOB) with ether interlayer linkage. Inset: Dependence of absorption
maximum on number of layers.

First, the UV-visible spectroscopy suggests that there is a much higher loading

density of the chromophore when compared to the deposition of the same polymer

-66-

system using either ester or urethane interlayer linkage (Figure 3.6b), and this finding is
consistent with the ellipsometric data. The anomalously high layer density may be the
result of the way we performed the reaction. We used concentrated sulfuric acid to
facilitate the epoxide ring-opening reaction and to promote ether formation. There is no
analogous reaction step in the formation of ester or urethane interlayer bonds. Repeating
the epoxide ring opening reaction with HCl instead of H2804 yielded lower density

layers.

We observe a similar pattern when the reactions for the ester-, amide- and
urethane-linked system are performed in the presence of H2S04 versus HCl, suggesting
the importance of dehydration by the H2804 in these reactions. We will explore this

issue more fully in a subsequent study. We also note that there is an unexpectedly large

0.02 r-

O

O

p—s
l

absorbance (a.u.)

 

 

 

0.00 Maj! 1 . 1
4000 3500 3000 2000 1500 1000

frequency (cm'l)

 

Figure 3.7. IR spectrum for poly(NPM-VOB) with ether interlayer linkage.

-67-

 

IR resonance near 1700 cm'l for the ether-linked assembly (Figure 3.7). This is not an
expected result because the ring opening reaction should yield a secondary alcohol and an
ether (scheme 3.3). One possible explanation for this observation is the oxidation of the
alcohol to a ketone, and this result is consistent with the absence of a strong OH
stretching resonance (Figure 3.7). To test for the possibility of this reaction, we reacted
an oxidized silicon substrate with 1,4-butane diglycidyl ether. The IR spectrum of the
ether-modified substrate is dominated by resonances associated with the asymmetric C-

O-C stretch centered about 1100 cm'1 (Figure 3.8a).

 

l A l L l 1 IV I l n l n

 

4000 3500 3000 2500 1500 1000

frequency (cm'l)

Figure 3.8. Ether chemistry for modified substrate. (a) FI‘IR showing C-O-C stretch
of epoxide prior to reaction, (b) FIIR of ring—opened system after reaction indicating
ketone and acid functionalities.

-6g-

 

The substrate was then immersed in anhydrous acetonitrile and a few drops of
concentrated sulfuric acid were added to facilitate the ring opening hydrolysis reaction.
The substrate was then heated to ~40° for 30 minutes. The IR spectrum of the reacted
substrate showed a marked decrease in the C-O-C stretch, consistent with the ring-
opening reaction, and the presence of C=O stretching bands of almost equal intensity
(Figure 3.8b). The carbonyl resonance appears as a doublet, with one peak centered at
approximately 1726 cm'1 and the other at about 1745 cm]. These two resonances are
consistent with the formation of an aldehyde and a carboxylic acid, respectively.42 We
observe that this chemistry is difficult to control beyond the deposition of about two to
three molecular layers, at which time, we observe no further growth. If the epoxide
groups are being opened and subsequently oxidized, as the IR data suggest, then they are
rendered capable of participating in interlayer bonding.

It is also possible that the epoxide reaction proceeds in an intralayer fashion,
which would also preclude the formation of interlayer ether linkages. It may be possible
to minimize such side reactions by running these reactions at room temperature, using
strictly anhydrous solvents, and minimizing atmospheric exposure.

Urea chemistry. Urea linkages are made by the reaction of an amine with an
isocyanate (scheme 3.4). An isocyanate-functionalized substrate is exposed to an MVE
polymer with amine-terminated side groups. The surface layer of polymer is reacted with
a ((1,0))-diisocyanate, then with another layer of the same MVE polymer. The
complementary chemistry, where an amine terminated surface is reacted with a MVE

polymer containing isocyanate side groups is also possible.

-69-

R—N=c=o + R'-—NH2 ——> R\ /U\ /R'

Cl
HzN

0%
O O

O

N

‘1‘...

HN

l

 

 

 

 

Scheme 3.4. Reaction schematic of an isocyanate with an amine (top) to form a urea
and an idealized structure of an amine-containing MVE polymer bound to an
isocyanate-functionalized substrate. Subsequent layer growth requires reaction of the
amine-containing MVE polymer with a diisocyanate.

We use poly(3-CPM-APVE) to grow urea-linked multilayer structures. Reacting

poly(3-CPM-APVE) with 1,6-diisocyanatohexane leads to urea formation. The IR band

assignment for the urea-linked multilayer structure is virtually identical to that for the

amide-linked multilayer, save for differences in band intensities, as expected. The IR

spectrum of the amide-linked multilayer assembly (Figure 3.9a) reveal Amide I, H and HI

bands, where the amide I band corresponds to the carbonyl stretching mode centered at

approximately 1655 cm]. Amide II and III bands correspond to N-H bending vibrations

-70-

and C-N stretching modes occurring at approximately 1595 cm'1 and 1421 cm'I

respectively. In addition, the band centered at approximately 3330 cm'I corresponds to

0.005

 

 

 

 

— a

2;: 0.004 -
3
8 0.003 -
5
.0
‘55 0.002 r
.8
cs:

0.001 -

 

4000 3500 3000 2000 1500 ‘ 1000

 

 

       

 

 

-1
frequency (cm ) b
l’ A 0.04. .,.
0.10 "' g r ."
9, 0.03- ,,
f? " {3’ . x ,.
:3 a)
C6 0.08 — Q 0.02“ "’x.
V 5 . ”x
8 E 0.01 - "
5 0.06 e 000.
g ' 0 1 2 3 4
,8 0.04 _ layer number
<
0.02 \
“N.
0.00 I J A ‘t¥.>sAfe;'.r.. .. ,.,...
200 250 300 350

Wavelength (nm)

Figure 3.9 (a) IR spectrum of poly(3CPM-APVE with urea interlayer linkage. (b)
UV-Visible spectra as a function of number of layers for poly(3CPM-APVE) with
urea interlayer linkage. Inset: Dependence of absorption maximum on number of
layers.

-71-

the stretching modes of the free NH groups. This band is significantly larger in the urea-
linked multilayer than in the amide-linked layers, as expected because each urea moiety
possesses two free NH groups. We also observe larger amide II and HI bands for the
same reason. As with the amide-linked multilayers, we monitor the 248 nm absorbance
band - which corresponds to the 3-chlorophenyl maleimide moiety - using UV-visible
spectroscopy (Figure 3%). We observe a higher loading density of the polymer for the
amide-linked assembly than for the urea-linked multilayer. One explanation for this
finding is that urea formation may be less efficient than amide formation under our
reaction conditions (reaction conditions were identical for both deposition schemes). The
ellipsometric data for the urea-linked assembly show a slope of IDA/layer (Figure 3.10).
This result is smaller by a factor of ~2 compared to the amide-linked system and is

consistent with the uv-visible absorbance data.

70 .-
60 l
50 l 3-x
40 ; ,4"

30 L i»
20 - ,”§/’
10 1

O l J l
0 1 2 3 4

Layer number

Thickness (A)

 

:-

Figure 3.10. Dependence of ellipsometric thickness on number of layers for
poly(3CPM-APVE) with urea interlayer linkage.

-72-

Urethane chemistry. The formation of urethane linkages can be accomplished by
the reaction of a hydroxyl group with an isocyanate group (scheme 3.5). We report two

such schemes here. (a) A hydroxyl-terminated substrate is reacted with an ((1,0))-

R—NZCZO + R'—OH ———’ N O
H

 

 

 

 

 

Scheme 3.5. Reaction schematic of an isocyanate with an alcohol (top) to form a
urethane and an idealized structure of an alcohol-containing MVE polymer bound to
an isocyanate-functionalized substrate. Subsequent layer growth requires reaction of
the alcohol-containing MVE polymer with a diisocyanate.

diisocyanate, followed by exposure to an MVE polymer with hydroxyl terminated side
groups. Multilayers are assembled by alternate exposure of the substrate to the
diisocyanate and the MVE polymer. (b) An MVE polymer with isocyanate-terminated

side groups is exposed to a hydroxyl-terminated substrate. This reaction results in the

-73-

formation of a urethane linkage. Multilayer assembly is achieved through alternate
exposure to a diol and the MVE polymer.

We use poly(NPM-VOB) and 1,6-diisocynatohexane to form urethane-linked
multilayer assemblies. The ellipsometric data (Figure 3.11) reveal linear growth with a
slope of 18 Allayer, consistent with that seen for the amide-linked layers and larger by a
factor of ~2 than for the urea-linked multilayers. This finding suggests that the reaction
of isocyanates with alcohols is more efficient than with amines. Growth of the urethane-
linked layers was monitored using FI‘ IR (Figure 3.12a) and UV-visible (Figure 3.12b)

spectroscopies, and regular growth was observed with these methods as well.

100
90;- .
80 7

28:: i
50 ~

40 '-
30 '- {r
20 "- ’

10 -

O p 1 1 I
0 1 2 3 4

Layer number

I' ‘fi

0

Thickness (A)

 

Figure 3.11. Dependence of ellipsometric thickness on number of layers for
poly(NPM-VOB) with urethane interlayer linkage.

In addition to the alternating copolymerization of maleimides and vinyl ethers, we
can also use alkenes as reactive monomers. The interlayer linking chemistry we describe

above is equally applicable to such polymers, and in all cases is determined by the

-74-

m*“”““‘1

I

0.016

I

0.012

I

0.008

absorbance (a.u.)

0.004 ~

0.000 AMA

4000 3500 3000 2000 1500 A 1000
frequency (cm'l)

 

 

 

 

 

 

 

 

E b
o .
2 0.25 r- 3 0.101,“.
“9 ’ o 0.081 .
2: 0-20- s -
g . ,2 0.06 t
g 0.15 — :9: 0.04: x”. ’
2 0.10 _
< . layer number
0.05 L
O . O O " l . I ‘ ~ v‘fi'i’.“t “avg/«,3, .’
200 250 300 350

W avelength (nm)

Figure 3.12. (a) IR spectrum for poly(NPM-VOB) with urethane interlayer
linkage (b) UV-Visible spectra as a function of number of layers for
poly(NPM-VOB) with urethane interlayer linkage. Inset: Dependence of
absorption maximum on number of layers.

-75-

terminal functionalities of the resulting polymer side groups. The co-polymerization of
NPM and 4-pentenoyl chloride, poly(NPM-4PC), is capable of participating in both ester
and amide interlayer linkages as shown in Figure 3.13a. This reaction scheme appears to
be less successful than the same interlayer linking chemistry used for the MVE
alternating copolymers. For poly(NPM-4PC) we see layer growth that is limited to one
or two molecular layers before terminating. One explanation for this behavior is that the
oxidation of the acid chloride groups to the acid followed by intermolecular hydrogen
bonding between the acid functionalities to render these groups less accessible for
subsequent layer deposition. It is also possible for the putative carboxylic acid groups to
hydrogen bond with solvent or adventitious water during rinsing and measurement.
Examination of the IR data (Figure 3.13b) is consistent with these explanations. The
broad —OH stretch centered at ~3100 cm'1 is characteristic of intermolecular hydrogen
bonding. A second piece of evidence in support of this information is the position of the
C=O band in these polymer adlayers. Dimerization of carboxylic acids shifts the
carbonyl stretch to a lower frequency than that of the free carboxylic acid. For a
poly(NPM-4PC) adlayer the C=O stretch is centered at ~1710 cm'l, whereas for an acid

chloride, the C=C stretch is expected to be near 1775 cm".

-76-

 

 

] 1

0.0010
0.0008
0.0006
0.0004 L
0.0002 P
0.0000 _

-00002 - . - . - . - . . .
4000 3500 3000 2500 2000 1500

I I V

 

 

absorbance (a.u.)

 

A

 

1000
frequency (cm'l)

Figure 3.13. Poly(NPM-4-PC). (a) Structures — with amide and ester linkers. (b) IR
spectrum of a poly(NPM-4PC) layer bound to the substrate by an ester linkage and
with remaining acid chloride functionalities hydrolyzed by expose to air.

-77-

3.4 Conclusions

We have demonstrated our ability to design and assemble polymer multilayer
assemblies with layer-by-layer control of the layer identity and interlayer linking
chemistry. We have presented the design and synthetic schemes for the preparation of
these polymers in the preceding chapter. In this paper we have presented our strategies
for assembling these materials into layered structures where the identity of the polymer
side groups allows simple chemical reactions to be used to link the individual layers. The
functionalities formed by this simple chemistry include amide, ester, ether, urea and
urethane. The design and assembly schemes we present here have the distinct advantage
of ease of preparation of the monomers and polymers, ease of assembly of these materials
into multilayers and relatively short deposition times. We have also shown that,
depending on the functional groups of the copolymer, it is sometimes possible to
assemble these materials where prior surface preparation is not necessary, as is the case
for reactions on silicon and quartz wafers. In these cases, the silanol groups serve as the

reactive moiety to which our polymers can be tethered.

-73-

3 .5. Literature Cited

1. Kepley, L. J.; Crooks, R. M.; Ricco, A. Anal. Chem. 1992, 64, 3191.

2 Demas, J. N.; Degraff, B. A.; Coleman, P. B. Analytical Chemistry News & Features,
1999, 723A.

3. Yoon, H. C.; Kim, H-S. Anal. Chem. 2000, 72, 922.

4. Delamarache, E.; Sundarababu, G.; Biebuyck, H.; Michel, B.; Gerber, C.; Sirgrist, H.;
Wolf, H.; Ringsdorf, H.; Xanthopoulos, N.; Mathieu, H. J. Langmuir, 1996, 12,
1997.

5. Vrancken, K. C.; Van Der Voort, P.; Gillis-D’Hammers, I.; Vansant, E. F.; Grobet, P.;
J. Chem. Soc., Faraday Trans. 1992, 88, 3197.

6. Pfleiderer, B.; Albert, K.; Bayer, E. J. Chromatogr. 1990, 506, 343.

7. Katz, H. E.; Scheller, G.; Putvinski, T. M.; Schilling, M. L.; Wilson, W. L.; Chidsey,

C. E. D. Science 1991, 254, 1485.

8. Li, D.; Ratner, M. A.; Marks, T. J.; Zhang, C. H.; Yang, J.; Wong, G. K. J. Am. Chem.

Soc. 1990, 112, 7389.

9. Langmuir, 1.; J. Am. Chem. Soc., 1917, 39, 1848.

10. Blodgett, K. B.; J. Am. Chem. Soc., 1935, 57, 1007.

ll. Ulman, A.; Chem. Rev., 1996, 96, 1533.

12. Nuzzo, R. G.; Allara, D. L. J. Am. Chem. Soc., 1983, 105, 4481.

13. Ulman, A. An Introduction to Ultrthin Films: From Langmuir-Blodgett to Self-
Assembly; Academic Press, Inc.: New York, 1991.

14. Nuzzo, R. G.; Dubois, L. H.; Allara, D. L. J. Am. Chem. Soc., 1990, 112, 558.

-79-

‘l 2.“ IA‘ -.11

‘r-I-‘V ‘

15.

l6.

17.

18.

19.

20.

21.

22.

23.

Lee, H.; Kepley, L. J.; Hong, H.-G.; Mallouk, T. E. J. Am. Chem. Soc. 1988, 110,
618.
Lee, H.; Kepley, L. J .; Hong, H.-G.; Akhter, S.; Mallouk, T. E. J. Phys. Chem. 1988,
92, 2597.
Yonemoto, E. H.; Saupe, G. B.; Schmehl, R. H.; Hubig, S. M.; Riley, R. L.; Iverson,
B. L.; Mallouk, T. E. J. Am. Chem. Soc. 1994, 116, 4786.

Yang, H. C.; Aoki, K.; Hong, H. — G.; Sackett, D. D.; Arendt, M. F.; Yau, S.-L.;
Bell, C. M.; Mallouk, T. E. J. Am. Chem. Soc. 1993, 115, 11855.

Cao, G.; Rabenberg, L. K.; Nunn, C. M.; Mallouk, T. E. Chem. Mater. 1991, 3, 149.
Rong, D.; Hong, H. — G.; Kim, Y. - I.; Krueger, J. S.; Mayer, J. E.; Mallouk, T. E.
Coord. Chem. Rev. 1990, 97, 237.

Thompson, M. E. Chem. Mater. 1994, 6, 1168.

Vermeulen, L.; Thompson, M. E. Nature 1992, 358, 656.

Vermeulen, L. A.; Snover, J. L.; Sapochak, L. S.; Thompson, M. E. J. Am. Chem.

Soc.1993,115,11767.

24.

Snover, J. L.; Byrd, H.; Suponeva, E. P.; Vicenzi, E.; Thompson, M. E. Chem.

Mater. 1996, 8, 1490.

25.

26.

Katz, H. B.; Wilson, W. L.; Scheller, G. J. Am. Chem. Soc. 1994, 116, 6636.

Katz, H. E.; Bent, S. P.; Wilson, W. L.; Schilling, M. L.; Ungashe, S. B. J. Am.

Chem. Soc. 1994, 116, 6631.

27.

Ungashe, S. B.; Wilson, W. L.; Katz, H. E.; Scheller, G. R.; Putvinski, T. M. J. Am.

Chem. Soc. 1992, 114, 8717.

-80-

 

28. Katz, H. E.; Schilling, M. L.; Chidsey, C. E. D.; Putvinski, T. M.; Hutton, R. S.
Chem. Mater. 1991, 3, 699.

29. Putvinski, T. M.; Schilling, M. L.; Katz, H. B.; Chidsey, C. E. D.; Mujsce, A. M.;
Emerson, A. B. Langmuir 1990, 6, 1567.

30. Katz, H. E. Chem. Mater. 1994, 6, 2227.

31. Byrd, H.; Whipps, S.; Pike, J. K.; Ma, J.; Nagler, S. E.; Talham, D. R.; J. Am. Chem.

Soc., 1994, 116, 295.

Alli

 

"v.

32. Byrd, H.; Pike, J. K.; Talham, D. R.; Chem. Mater., 1993, 5, 709.

33. Kohli, P.; Blanchard, G. J. Langmuir 1999, 15, 1418.

m;—

34. Home, J. C.; Huang, Y.; Liu, G.—Y.; Blanchard, G. J. J. Am. Chem. Soc. 1999, 121,
4419.

35. Home, J. C.; Blanchard, G. J. J. Am. Chem. Soc. 1999, 121 , 4427.

36. Major, J. S.; Blanchard, G. J., Chem. Mater. in review.

37. Kohli, P.; Blanchard, G. J .; Langmuir, 2000, 16, 4655.

38. Kohli, P.; Blanchard, G. J .; Langmuir, 2000, 16, 8518.

39. Kohli, P.; Scranton, A. B.; Blanchard, G. J. Macromolecules 1998, 31 , 5681.

40. Major, J. S.; Blanchard, G. J ., Langmuir, in preparation.

41. Major, J. S.; Blanchard, G. J., Langmuir, 2001, 17, 1163.

42. Silverstein, R. M.; Bassler, G. C.; Mon‘ill, T. C. Spectrometric Identification of

Organic Compounds, 5ed., New York, John Wiley & Sons, Inc., 1991.

-81-

Chapter 4

Acid-Enhanced Interfacial Polymer Layer Growth

Abstract

We report on the growth of interfacial multilayer structures formed from
maleimide-vinyl ether alternating copolymers. We have demonstrated that the thickness
and density of these polymer layers can be controlled by adding acid to the interlayer
cross-linking reaction. We have demonstrated this control for several different interlayer
crosslinking strategies, where amide, ester, urea and urethane interlayer covalent bonds
are formed. For all reactions, the addition of concentrated acid during polymer layer
deposition resulted in a two- to four-fold increase in the loading density of the polymer
relative to the acid-free reaction, depending on the acid used and its concentration. These

findings are consistent with acid catalysis (HCl) and/or dehydration (H2804).

-32-

w? l..t_i OYAT‘. j

4.1 Introduction

Designing and constructing polymeric materials where there is explicit control
over the identity of each molecular layer in conjunction with covalent interlayer linking
chemistry is an emerging effort in interface science. The field of layered growth started
with the work of Langmuir and Blodgett,"2 resulting in physisorbed L-B films, advanced

3-13

through the discovery of self-assembling alkanethiol/gold monolayer structures to

14-17 18-34

covalent, ionic and polyelectrolyte multilayers.”46 Such materials have found

47-51 22.52.53

use in chemical sensing, surface second harmonic generation, electronics and
electro-optics.54 A common concern in all of these application areas is the extent of order
and the unifomiity of coverage of the layered assembly. We have developed a means to
deposit maleimide-vinyl ether (MVE) alternating copolymer55 multilayers, one layer at a

5957 and are interested in achieving

time, with covalent interlayer linking chemistry,
control over the density of individual polymer layers. We observe an increase in layer
thickness and, in some cases, density when HCl or H2S04 are added to the crosslinking
reactions used in the polymer layer growth process. This work demonstrates that we
have three distinct types of control over the construction of these materials. First, we can
control the identity of the functional constituents incorporated into the polymer adlayer
by substitution of the polymer side group(s).57 Second, we have control over the covalent
chemistry used to form the interlayer linkage for assembly of multilayer structures.56
Third, the work we report here demonstrates that we can control the loading density of

the polymer at each layer through the addition of HCl or H2SO4 to the crosslinking

reaction.

-83-

1:7“

 

"h" Tu: . 'J; _

The ability to design well-behaved thin film systems where there is some level of
control over the density of specific functionalities within the adlayer matrix could prove
beneficial in areas such as the design of sensor materials, for example, where each
functional group behaves as an individual sensing element. Controlling the density of
functional constituents within a layer could also provide a means to mediate conductivity
or diffusion. Our interest in the design of covalently bound polymer multilayer
assemblies is based on the ability of these materials to mediate interfacial adsorption and
desorption phenomena. We have reported recently on design schemes and deposition
strategies for a variety of alternating copolymers using covalent interlayer attachment
chemistry.”57 We observed that the addition of a small amount of concentrated sulfuric
acid during ether crosslinking between polymer adlayers led to a marked increase in
adlayer absorbance compared to the results for adlayers where no H2804 was added.56
We have undertaken a study to understand this effect more fully and report our findings
here. In addition to the acid-mediated enhancement in the ellipsometric thickness of the
polymer adlayers, we observe a significant increase in absorbance per layer when
interlayer linking chemistry is performed in the presence of H2S04. The thickness
measurements taken by themselves could be indicative of either an increase in amount of
material adsorbed or simply an increase in the extent of structural disorder within the
layers. When these findings are viewed in light of the absorbance data, it is clear that the
change we observe is an increase in adlayer density, and not simply disorder.

The enhancement of adlayer deposition with the addition of acid is not the same
for all crosslinking reactions. To address the comparative importance of acid catalysis

and dehydration in these reactions, we have studied adlayer deposition and crosslinking

-84-

‘."I

 

using H2804 and HCl. HCl enhances acid-catalyzed reactions, but cannot play a role in
dehydration. The presence of H2S04 in these systems could drive these reactions toward
product formation according to LeChatelier’s principle. In the ester and amide formation
reactions, adipoyl chloride is used as the crosslinking agent. In these cases, it is likely
that the reaction proceeds via the elimination of HCl, calling our assertion of adlayer
thickness enhancement by dehydration into question. To better understand this reaction,
we have prepared a separate substrate in a manner consistent with experimental
conditions used in polymer adlayer deposition, except we omitted the polymer layer. We
found that the surface acid chloride groups were converted to the corresponding
carboxylic acid almost exclusively, indicating that the amide and ester formation
reactions were between the amine or alcohol and a carboxylic acid, consistent with the
explanations we offer for the enhanced deposition. The urea and urethane formation
reactions proceed with the formation of a carbamic acid intermediate. The presence of
water in the system would lead to decomposition to an amine and C02. In this case, the
presence of H2804 leads to the removal of adventitious water in the reaction medium. In
all cases, we observed measurably greater deposition of polymer with the addition of
sulfuric acid than with the addition of hydrochloric acid, suggesting that, for these
reactions, dehydration and protonation are both important, but dehydration plays a
dominant role in mediating the reaction efficiency.

We have also examined polymer adlayer deposition as a function of acid
concentration. We assembled ester-linked multilayers of poly(NPM-VOB) using adipoyl
chloride as the crosslinker and varied the concentration of acid added to the reaction. The

adlayer deposition process depends linearly on acid concentration to a point, with the

-35-

..i-FIKIESE A.

win-‘5 1.17.."

growth appearing to saturate at high acid concentrations. We also note that there is a
marked difference between the deposition of the initial polymer layer for the reactions
performed with HCl compared to those done with H2804. We consider the similarities
and differences found for the addition of the two acids during the adlayer deposition
reaction and interpret these effects in terms of the reactions responsible for adlayer

crosslinking.

4.2 Experimental

Synthesis: We use the alternating copolymers poly(N-phenylmaleimide—co-1-
vinyloxy-4-butanol), poly(NPM-VOB), poly(3-chlorophenylmaleimide-co-3-
aminopropyl-l-vinylether), poly(3CPM—APVE) and poly(4-bromophenylmaleimide-co-
3-aminopropyl-l-vinylether), poly(4BPM-APVE), (scheme 4.1 top panel) in this study.
The synthesis and layered interfacial assembly of these polymers has been detailed

elsewhere.”57

These polymers are capable of participating in ester, urethane, ether
(poly(NPM-VOB)) and amide and urea interlayer linkages (poly(3CPM-APVE) and
poly(4BPM-APVE)).

Substrate preparation: Gold coated substrates and silicon and quartz wafers were
cleaned using piranha solution (3:1 HZSO4zH202) for ~15 minutes. These substrates were
then removed and rinsed with ethanol, then water, and dried under a stream of nitrogen.
Silicon and quartz substrates were used in subsequent adlayer deposition reactions
without further modification. For these substrates, the surface silanol groups are used

directly in adlayer deposition reactions. Prior to adlayer deposition, gold substrates were

exposed to 0.01 M 6-mercapto-1-hexanol in 60:40 ethanolezO solution at 40°C for ~30

-86-

rrrinutes. Following activation of the surface with 6-mercapto-l-hexanol, the terminal
hydroxyl groups were used in adlayer deposition reactions. The substrates used for the
formation of ester and amide linkages were first reacted with adipoyl chloride in the
presence of 4-methylmorpholine for 30 minutes at 35°C. The substrates used for the
formation of urea and urethane linkages were reacted with 1,6-diisocyanatohexane in the
presence of 4—methylmorpholine. Following surface modification, the substrates were
exposed to approximately 5 mL of the appropriate polymer solution at a concentration of
10 mM in reagent grade dimethyl sulfoxide. For reactions where acids were added, the
concentration range examined was between 0.073 M and 0.290 M for HCl and between
0.108 M and 0.432 M for H2S04 (Table 4.1). It should be noted that these are the final
concentrations after addition of the acid to the polymer solutions.

Optical Ellipsometry: Ellipsometric thickness data on the deposited polymer
multilayer assemblies were acquired using a Rudolph Auto-EL II optical ellipsometer
operating at 632.8 nm. The software used for data collection and processing was
acquired from the manufacturer (Rudolph DAFIBM). The refractive index of the
polymer adlayers was taken to be n = 1.54 + 0i in all cases. The thickness of the polymer
multilayer assembly was measured after each deposition cycle. Regression of the
ellipsometric data was performed using Microcal Origin® V.6.0 software.

UV—Visible Spectroscopy: Absorption spectra of the polymer adlayers were
acquired using a Cary 300 UV-visible spectrophotometer. For these measurements, the
adlayers were deposited on quartz substrates and absorption spectra acquired after each
deposition cycle. Data were collected between 190 nm and 500nm at a scan rate of 600

nm/min with 1 nm spectral resolution.

-87-

puma-E]

.7

NHz NH2
0 O O
O N O O N O O N O
o l
O O O
. . . l?
Cl l4
Br ‘
O O
/U\ + R'NHZ ——’ /U\ ,R' + H20 3
R OH R 171
H
11 ° b
+ R'OH -——> /U\ .+ H0
R OH R Q’R 2
O
)L C
R—NZCZO + R'NHz ——> R\N N’R
I l
H H
O
' d
R—N=C=o + R’OH ———> R\N/U\O/R
l
H

Scheme 4.1. Top panel, left to right: Structures of poly(3CPM-APVE), Poly(NPM-
VOB) and poly(4BPM-APVE). Bottom panel: Reaction schemes for the several
crosslinking strategies employed in this work.

-88-

Infrared Spectroscopy: FTIR spectra of the polymer adlayers on gold substrates
were collected using a N icolet Magna 750 FTIR spectrometer. Spectral resolution was 4
cm'1 for all measurements. An external reflectance sample mount set at an incidence
angle of 80° was used for data collection. Spectra of samples assembled on silicon were
acquired using a Nicolet Magna-IR 550 set up to acquire data in transmission mode at

Brewster’s angle.

4.3 Results and Discussion

We report here on the role(s) of HCl and H2804 in mediating the crosslinking
reactions used in the deposition of covalently linked maleimide-vinyl ether alternating

“'57 The addition of a small amount of acid to the reaction vessel

copolymer multilayers.
mediates the deposition of polymer adlayers. We have investigated the role of acid in the
mediation of amide, ester, urea and urethane bond formation, and we observe adlayer
deposition enhancement in each case. Both spectroscopic and ellipsometric data point to
the acid-enhancement of adlayer thickness, by a factor of between two and three,
depending on the functionality being formed. The chemistry we use for adlayer growth is
deliberately self-terminating at each layer, and the addition of subsequent layers requires
an activation step. For the cases where the thickness and absorbance both increase as a
result of adding acid to the interlayer linking reaction, the data are the result of an
increase in the density of the polymer adlayers. We consider the results for specific

systems next. The uv-visible and ellipsometric data for each system are summarized in

Tables 2 and 3.

-89-

v, Y‘-‘-’T"‘"_‘.. . - . 2‘

Amide crosslinked polymer multilayers: Poly(3CPM-APVE) is bound to the
substrates we use here by first reacting the surface functionality with adipoyl chloride,
then reacting the polymer amine functionalities with terminal acid chlorides.
Spectroscopic and ellipsometric data for this system are shown in Figure 1. Multilayers
of these amine-terminated polymers can be assembled using adipoyl chloride as the
crosslinking agent to form amide interlayer bonds (scheme 4.1a). This reaction is
performed under anhydrous conditions, proceeds with or without the addition of acid and,
for both experimental conditions, we obtain reproducible behavior. Ellipsometry shows
regular layer growth with a slope of 19 Allayer, and uv-visible absorbance spectroscopy
shows a linear increase in adlayer absorbance with a slope of 0.019 a.u.llayer.

The addition of concentrated H2804 to the reaction vessel during layer deposition
gives rise to a measurable change in the layer growth characteristics. In the presence of
0.216 M HZSO4, we recover ellipsometric data showing 30 Allayer and uv-visible
absorbance data showing 0.042 a.u.llayer. As discussed above, the adlayer deposition
chemistry we use ensures the deposition of individual polymer layers. Because there is
an increase in both ellipsometric thickness and absorbance per layer with the addition of
HZSO4 to the layer deposition reaction, the data indicate an increase in adlayer density.

In an attempt to resolve whether or not protonation of a reaction intermediate
accounts for our findings, we have repeated these experiments with the addition of HCl
instead of H2504. In the presence of 0.145 M HCl during the adlayer crosslinking
reaction also gives rise to enhanced adlayer growth, with uv-visible absorbance
spectroscopy yielding a slope of 0.0294 a.u.llayer, intermediate between the anhydrous

and HZSO4-mediated reactions. We were unable to acquire ellipsometry data for the HCl-

-90-

—-— vim-mtmal

mediated reaction because HCl damaged the gold substrate surface, likely due to the
destruction of the chromium adhesion layer. The FTIR spectrum of the crosslinked
adlayers is shown in Figure 4.1. We observe the stretching resonances characteristic of
amides and free NH functionalities at approximately 1655 cm'1 (amide I) and 3300 cm],
respectively, and the bands are identical for adlayers grown without acid and with H2804
and HCl. In all cases, the reaction products are the same. It appears to be
counterintuitive that the addition of HCl (or H2804 for that matter) would enhance the
deposition due to the fact that we use adipoyl chloride as crosslinking agent, which
generates HCl as a reaction product. In order to resolve this issue, we prepared a
substrate under similar experimental conditions to those for polymer deposition and
observed that the acid chloride groups were almost exclusively converted to the
corresponding acid (Figure 4.2). Thus, the layer deposition reaction is between an amine
and a carboxylic acid, and the explanations of dehydration and protonation for acid
addition follow logically. We believe the mechanism of the reaction to be simple
nucleophilic substitution by addition-elimination, where the reaction proceeds by attack

of the acid from the polymer nucleophilic terminal amine, with the elimination of H20.

-91-

w 1-"! at!" up :1

.0
w

  
 
  
 

 

 

 

   

 

 

 

 

E m
0.20 -§ 0'“
h z 0.1»
”“015 - l
g _ 0900 250 300 350 400
<1“ wavelength (nm)
(Cl/0.10 -
< .

0.05 -

0.00' . . . ' 4 ' '
m .. 1 2 3 4
8 125 .. layer number —
g 100.

,o 75 - _

g ..

E 50 - *

8 I-

.9 25 '-

E 0 h l n l n I 1 l
_ 1 2 3 4

.4" 3.5 :- layer number

2 3.0 _-

g 2.5 _-

$2:

.0 ' .

g 1.0 :-

< 0.0 ‘ 4 1 . L .mwr‘“ ‘. 1 . 1 . 1

 

 

4000 3500 3000 2500 2000 1500 1000

frequency (cm'l)

Figure 4.1. Data for amide-linked poly(3CPM-APVE). (a) Dependence of optical
absorption at 245 nm on number of polymer adlayers. I are data points for acid-free
growth, 0 are data for adlayer growth with 0.145 M HCl and A are data for adlayer
growth with 0.216 M H2804. Inset: UV-Visible absorption spectrum for a four layer
assembly of poly(3CPM-APVE). (b) Ellipsometric thickness of poly(3CPM-APVE)
adlayers as a function number of layers deposited for growth without acid added (I) and
with 0.216 M H2804 added (A). (c) FTIR spectrum of a two-layer assembly of
poly(3CPM-APVE).

-92-

 

1.4 -

 

1.2 -
,- 1.0 -
MS? r
5 0.8 —
8 .
g 0.6 -
35 r
g 0.4 :-

0.2

 

 

    

0.0

      

L

3500 3000 2500 2000 1500 1000

freqency (cm'l)

Figure 4.2. FT IR spectrum of substrate treated with adipoyl chloride showing essentially
complete conversion of the terminal acid chlorides to carboxylic acids.

Ester crosslinked polymer multilayers: The attachment of poly(NPM-VOB) to
the substrate and subsequent adlayer deposition proceeds in the same manner as
described above for poly(3CPM-AVPE), with poly(NPM-VOB) reacting with adipoyl
chloride to produce an ester moiety (scheme 4.1b). Spectral and ellipsometric data are
shown in Figure 4.3 a and b. For ester formation, we observe enhanced adlayer density
when either H2804 or HCl are added to the reaction. For the deposition of the polymer
layer in the absence of acid, we recover a growth rate of 22 A/layer from ellipsometric
data, 34 Allayer for growth when the HCl concentration is 0.145 M, and 45 A/layer for

growth when the final H2S04 concentration is 0.216 M For adlayers formed either

-93-

without addition of acid or with H2SO4 addition, UV-visible absorbance results are
consonant with the ellipsometric data, yielding 0.0309 a.u.llayer at 220 nm for acid-free
growth, and 0.0603 a.u.llayer for 0.216 M H2SO4-mediated growth. The absorbance data
for adlayers grown in the presence of 0.145 M HCl yield a slope of 0.0303 a.u./layer, the
same to within the experimental uncertainty as that measured for acid-free growth. This
finding, in conjunction with the ellipsometry data, indicates that the adlayer density does
not change for ester formation, but the polymer morphology does change upon adlayer
exposure to HCl. The experimental data for ester formation suggest that it is not acid
catalysis, but rather dehydration that dominates the ester formation reaction. The
function of H2804 here is to sequester H2O as it is formed during the reaction. We note
that this reaction is somewhat more difficult to control than the other reactions presented
here. In some instances, it was difficult to deposit more than three molecular layers
before large deviations from linearity were noticeable. The FT IR spectrum of the ester-
crosslinked polymer layers is shown in Figure 4.30. Our findings for the growth of ester-
crosslinked adlayers with the addition of acid are consistent with dehydration and
protonation as discussed above for the amide-crosslinked adlayers. We also examined
the growth of ester-crosslinked polymer adlayers without adding acid and with the
addition of morpholine. We observed no significant enhancement for this reaction,
consistent with our assertion that the acid chloride groups are predominantly converted to
the corresponding carboxylic acid (Figure 4.2). Had these groups been acid chlorides, the
presence of the base would have led to an enhancement according to Le Chatelier’s

principle.

-94-

13. :95 “In-1
I

    

 

 

   

"’8 0.4
3
0.25 - ° 03’
,., b .502
0.20 " 2% 01
_ <
0.0». - L . .
0.15 - 200 250 300 350 400
l. wavelength (11m)
F4
(0 0.10 -
V I
< 0.05 -
P

 

 

 

 

 

 

0.00 1 2 3 4
0:5: 200 ' layer number
35’ 150 - i
E . .-
3 100 -
E ' /
E 50 -
8 .
g 0 l J I n l n l
in" l 2 3
MA 2'0 P layer number
o F
E 1.5 :-
D
g 1.0 :-
§ 0.5 :-
2? 0.0

1 l 1 l 1 I n l n J n l

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

Figure 4.3. Data for ester-linked poly(NPM-VOB). (a) Dependence of optical
absorption at 231 nm on number of polymer adlayers. I are data points for acid-free
growth, 0 are data for adlayer growth with 0.145 M HCl and A are data for adlayer
growth with 0.216 M H2504. Inset: UV-Visible absorption spectrum for a four layer
assembly of poly(NPM-VOB). (b) Ellipsometric thickness of poly(NPM-VOB) adlayers
as a function number of layers deposited for growth without acid added (I), with 0.145
M HCl (0) and with 0.216 M H2SO4 added (A). (c) FTIR spectrum of a two-layer
assembly of poly(NPM-VOB).

-95-

— ‘3’“ .“'.‘.".'.'-"“1 . . -1

Urea crosslinked polymer multilayers: Poly(4BPM-APVE) adlayers are bound to
substrates crosslinked by formation of a urea functionality (Scheme 4.1c). Spectral and
ellipsometric data for this system are shown in Figure 4.4. The reaction to form urea
linkages is between the polymer amine group and an isocyanate functionality, present in
the adlayer structure using 1,6-diisocyantohexane. The function of 1,6-
diisocyanatohexane in the formation of adlayer assemblies is analogous to adipoyl
chloride in the reactions described above. When adlayers are grown under anhydrous
conditions, we observe modest adlayer growth, with ellipsometry indicating a thickness
of 10 Allayer and uv-visible absorbance at 238 nm yielding a linear dependence on
adlayer deposition cycles with 0.0043 a.u.llayer. We note that the thicknesses and
absorbance values we recover for this reaction scheme are lower that that recovered for
acid-free ester and amide layer growth, indicating a relatively low polymer adlayer
density under these reaction conditions. We observe adlayer density enhancement with
the addition of HCl or H2SO4 to the crosslinking reaction, with ellipsometry indicating 39
Allayer for growth with 0.145 M HCl and 35 Allayer for growth with 0.216 M H2804.
As with the ester formation, the polymer morphology does play some role in the adlayer
growth, because the uv—visible absorption data yield slopes of 0.0052 a.u.llayer for 0.145
M HCl and 0.0102 a.u.llayer for 0.216 M H2SO4. Thus, while the ellipsometric data
point to similar adlayer thicknesses, the absorbance data point to substantially different
loading densities, even after the difference in acid concentration for the two reactions is
taken into account.

We have established that the reaction chemistry used for forming covalent

multilayers using isocyanate/amine chemistry depends sensitively on the presence of

-96-

“—I‘l. xiii—m!

water.‘6 Under anhydrous conditions, single layers can form in a regular manner and
with the addition of water, adlayer growth becomes much less well controlled. The
addition of concentrated H2SO4 or HCl to the reaction vessel during layer deposition
appears to facilitate urea formation. We understand this effect in the context of H2804
removing adventitious water that would lead to the formation and subsequent degradation
of the unstable carbamic acid functionalities, precluding their participation in urea
formation. The isocyanate carbon is highly electrophilic, making it susceptible to attack
by nucleophiles such as water and amines. Attack at this site by water leads to the
formation of the carbamic acid, which is unstable. Reaction of the carbamic acid with an
amine leads to the formation of a urea, and in the absence of acid this reaction appears to
be inefficient. In these cases, the addition of acid to the reaction likely serves to
protonate the isocyanate group, making it more electrophilic and increasing its reactivity
toward nucleophiles such as alcohols and amines. The enhancement observed with
H2SO4 is, again, probably due to its ability to sequester water that would lead to the
degradation of the carbamic acid. As noted above, the enhancement in absorbance data is
almost a factor of two greater for H2SO4 addition than for HCl addition, even though the
concentration of H2804 is only 1.5 times greater than that of HCl. In an anhydrous

environment, we expect H2SO4 to function predominantly as a monoprotic acid.

-97-

.1

) 3’. «(L11 m

 

 

 

 
 
 

 

 
 

 

 

 

 

 

 

 

0'10 - E 0.10
0.08 -§ 0,,
- s
:30'06 '< 00900 250 300 350 400
a " wavelength (nm) _
V0.04 -
<2 . ’./_//_'
0 02 P 4/4
*

000 l' I . I_ I L I I
.2 1 2 3 4
E: 160 - layer number
8 - i
g 120 r-

.o 80 '

g .

E 40 -

o .

E: O I 4 I n I n I
E 1 l 2 b 3 4
m.“ 4.0 _ ayer num er

2

5 3.0

8

g 2.0

.0

§ 1.0

.D

<

 

 

.9
o

I a I l I I I. I n I

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

Figure 4.4. Data for urea-linked poly(4BPM-APVE). (a) Dependence of optical
absorption at 255 nm on number of polymer adlayers. I are data points for acid-free
growth, 0 are data for adlayer growth with 0.145 M HCl and A are data for adlayer
growth with 0.216 M H2804. Inset: UV-Visible absorption spectrum for a four layer
assembly of poly(4BPM-APVE). (b) Ellipsometric thickness of poly(4BPM-APVE)
adlayers as a function number of layers deposited for growth without acid added (I),
with 0.145 M HCl (0) and with 0.216 M H2504 added (A). (0) FTIR spectrum of a two-
layer assembly of poly(4BPM-APVE).

-98-

 

Urethane crosslinked multilayers: The reaction of poly (NPM-VOB) with 1,6-
diisocyanatohexane produces urethane interlayer linkages (Scheme 4.1d). The substrate,
containing terminal -OH functionality, is exposed to 1,6-diisocyanatohexane in the
presence of methylmorpholine in anhydrous acetonitrile to produce an isocyanate-
terrninated surface. When depositing poly(NPM-VOB) on this activated substrate, we
observe an enhancement in layer density when the adlayer deposition reaction is
performed with the addition of HCl or H2804. The ellipsometric data yield a slope of 12
Allayer for acid-free deposition, 20 Allayer in the presence of 0.145 M HCl and 32
Allayer in the presence of 0.216 M H2804. UV-visible absorbance data of the polymer
adlayers shows a similar trend, with a slope of 0.0093 a.u.llayer for acid-free growth,
0.0265 a.u.llayer for HCl-mediated adlayer deposition and 0.0348 a.u.llayer for H2804-
mediated deposition. This reaction is qualitatively similar to that used for urea interlayer
formation, save for the formation of a urethane moiety. In contrast to the urea interlayer
linking chemistry, both the ellipsometry and absorbance data point to an adlayer density
enhancement that is greater for H2SO4 than for HCl. Similar to acid-free deposition for
the urea crosslinked adlayers, we recover modest adlayer growth, indicating that the
surface reactivity of isocyanate functionalities is not as facile as that of acid chlorides.
As discussed above, the reaction of isocyanates with nucleophiles such as amines and
alcohols is susceptible to competitive reaction with water, resulting in the presence of
carbamic acid terminal functionalities, which would likely decompose to amines.
However, in the presence of concentrated acid we observe an increase in the ellipsometric
thickness. In the case of concentrated H2804, the fact that this species is an efficient

dehydrating agent minimizes the competitive effects of water, allowing for the efficient

-99-

.A" 1:“!-

formation of the urethane moiety. In the case of hydrochloric acid, we would again
expect to see an enhancement due to acid catalysis either in protonation of the imine or in
stabilization of an alkoxide reaction intermediate.

The effect of adding H2804 or HCl to the crosslinking reactions used in the
formation of several polymer adlayer structures is, in general, to increase either the
density and/or disorder of the material deposited at the interface. The crosslinking
reactions we use are simple addition-elimination reactions that rely on nucleophilic
attack. These reactions are typically efficient and in many cases can be catalyzed by the
presence of acid, or driven to completion by the abstraction of water. Both H2804 and
HCl can serve as a source of protons while H2SO4 can also function as a dehydrating
agent. Thus, for reactions that are enhanced more by the presence of H2804 than HCl,
such as ester and urea formation, it is reasonable to expect both acid catalysis and
dehydration contribute to our experimental findings. Because both amide and ester
reactions proceed with the elimination of water, the dehydrating capability of H2SO4
plays a major role in these reactions. While the formation of urea and urethane linkages
does not proceed by water elimination, it should be noted that the intermediate (carbamic
acid) formed in these reactions is susceptible to nucleophilic attack by water, leading to
the decomposition of this species. In this case, the ability of H2804 to remove water
would lead to enhancement in the deposition of these polymers. We can view these
findings in the context of Le Chatelier’s principle, where the removal of either reaction
products (H2O for these reactions) would tend to drive the crosslinking reactions to
completion. These findings thus provide a simple means of controlling the density of

covalently linked polymer adlayers at a variety of interfaces.

-100-

‘1

are???

-_._...‘ - _.._..q

- u“:

0.20»

  
  
 
   

 

 

 

 

 

 

 

 

 

g; 0.15»
0-15 “g 0.10»
E E 0.05»
< .
.4010 l- 0'0 00 250 300 350 400
a + wavelength (nm)
<20.05 -

0.00 I I I I I I I
0;; 1 2 3 4
‘0'; 150 .. layer number
8
g I
g 100 -

.0 I. .1.

.5.

E 50 - *

8

.9 '

a O I I I I I J I

1 2 3 4

MA 2'0 F layer number

i 1.5 r-

§ 1.0 -

B .

‘5 0.5 -

3 .

< 0.0

 

I I I L I I I I I I I

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

Figure 4.5. Data for urethane crosslinked poly(NPM-VOB). (a) Dependence of optical
absorption at 231 nm on number of polymer adlayers. I are data points for acid-free
growth, 0 are data for adlayer growth with 0.145 M HCl and A are data for adlayer
growth with 0.216 M H2SO4. Inset: UV-Visible absorption spectrum for a four layer
assembly of poly(NPM-VOB). (b) Ellipsometric thickness of poly(NPM-VOB) adlayers
as a function number of layers deposited for growth without acid added (I), with 0.145
M HCl (0) and with 0.216 M H2804 added (A). (0) FTIR spectrum of a two-layer
assembly of poly(NPM-VOB).

-101-

We have also measured adlayer deposition density as a function of acid
concentration. In this work, we have assembled ester—linked multilayer systems of
poly(NPM-VOB) using adipoyl chloride as the crosslinking agent, where we varied the
concentration of HCl or H2SO4 as a function of deposition cycle. We performed these
reactions with HCl concentrations in the range of 0.073 to 0.290 M and with H2804
concentrations in the range of 0.108 to 0.432 M. We find that, for both HCl and H2804,
increasing acid concentration initially yields an increase in the slopes of the absorbance
data, and for the highest concentrations of acid used here, we observe a modest decrease
in the slope of the absorption data (Table 4.1). These data suggest that there is an optimal
pH range for the deposition reactions, indicative of a balance between multiple
phenomena which contribute to the adlayer deposition process. For both acids, we
observe an enhancement in the thickness of the initial layer that is related linearly to acid
concentration, with H2804 producing thicker initial adlayers than HCl (Figure 4.6). In
each case, the growth of the initial polymer adlayer is enhanced to a greater extent than
the growth of subsequent adlayers. We understand this effect in the context of the greater
availability of the reactive groups on the substrate surfaces and possibly the reactivity of
the surface-bound functionalities being greater than the functionalities contained in the
crosslinking agents. This latter issue remains open to investigation but, from a
phenomenological point of view, the experimental acid concentration-dependence of
adlayer growth demonstrates the ability to control polymer adlayer density by synthetic

means.

-102-

9
H
N
1

“:11

p

p—I

O
I

 

0.08 -

m_\ “.-

p
E

Absorbance maximum of first adlayer
p o
8 8

 

I I I I I I I

0.1 0.2 0.3 0.4 0.5
acid concentration (M)

p
8

 

.9
0

Figure 4.6. Dependence of ester-bound poly(NPM-VOB) monolayer absorbance at 231
nm on acid concentration for the initial polymer layer. (a) HCl added, (b) H2804 added.
The different concentration dependencies for each acid indicates the importance of both
protonation and dehydration in the ester formation reaction.

—103-

4.4 Conclusions

We have demonstrated the ability to control the density of polymer adlayer
deposition at interfaces using a simple modification to the crosslinking reaction(s) used to
connect adlayers. This new capability provides three distinct levels of control over
interfacial adlayer preparation. We have control over the identity of the functional
moieties incorporated into the polymer used in adlayer deposition by synthetic means.
This level of control is important in the preparation of materials where specific chemical
functionalities are required. We have also demonstrated previously that polymer adlayers
can be assembled in discrete layered structures where each layer is of molecular
dimension. We can connect individual polymer adlayers to one another by means of
either ionic complexation or the formation of covalent crosslinks between polymer side
groups. The third level of control, which is the focus of the work we have reported here,
is the ability to control the density of the polymer adlayers. This new capability bears
directly on our ability to control mass transport properties within these ultrathin
interfacial materials. The control that we have established over adlayer density is
determined by the mechanism(s) of the crosslinking reactions used in multilayer
construction. The addition of H2804 or HCl to the crosslinking reactions gives rise to an
enhancement in the extent of reaction because of dehydration and/or acid catalysis,
depending on the reaction used. For all of the systems examined here we recover
significant enhancement in the absorbance as well as ellipsometric data for the polymer
deposition reactions to which sulfuric acid is added, relative to the results for the
corresponding acid-free reactions. The addition of HCl gives rise to smaller

enhancements in thickness and, sometimes absorbance as well, suggesting that the

-104-

W ‘u‘ “—q

presence of HCl influences the extent of disorder in the polymer adlayer as well as
mediating the crosslinking reaction to adjacent layers. Because these adlayer deposition
reactions are self-terminating at each step, the observed enhancements in ellipsometric
thickness and absorbance are necessarily the result of control over the density of the
polymer adlayer. We anticipate that the addition of H280; or HCl to the polymer adlayer
crosslinking reactions will play a role in determining the interface morphology, and this

is an issue that we intend to address.

-105—

mm “2.1 -m '91?qu

Table 4.1. Slopes of the UV-Visible absorbance data for the addition of HCl and H2804
as a function of concentration for the crosslinking of poly(NPM-VOB) with adipoyl
chloride to form an ester linkage.

 

 

 

 

 

Volume of acid [HCl] Adlayer [H2SO4] (M) Adlayer
added (mL) (M) absorbance absorbance
(a.u.llayer) (a.u.Ilayer)
0.03 0.073 0.0135 0.108 0.0154
0.06 0.145 0.0157 0.216 0.0296
0.12 0.290 0.0097 0.432 0.0273

 

 

 

 

 

 

Table 4.2. Dependence of ellipsometric thickness on acid addition for amide, ester, urea
and urethane crosslinked adlayers.

 

 

 

 

 

 

Crosslink Without acid With 0.073 M With 0.108 M
functionality (Allayer) HC] (A/layer) H2804 (A/layer)
Amide 19 30

Ester 22 34 45

Urea 10 35 39

Urethane 12 20 32

 

 

 

 

 

-106-

“ s «‘57 ‘ 4.“ 7.19“?
I

Table 4.3. Dependence of absorbance data on acid addition for amide, ester, urea and

urethane crosslinked polymer adlayers.

 

 

 

 

 

 

 

 

 

 

 

Crosslink chromophore Am, Without With With
acid 0.073M HCl 0.108 M
(a.u.llayer) (a.u.llayer) H2S04
(a.u.llayer)
o (:1
Amide 245nm 0.0188 0.0294 0.0424
0
o
Ester m—Q 231nm 0.0309 0.0303 0.0603
0
o
Urea IE“ O B" 255nm 0.0043 0.0052 0.0102
0
o
Urethane m© 231nm 0.0093 0.0265 0.0348
0

 

 

-107-

1“: 'M‘nl

4.5 Literature Cited

1. Langmuir, I. J. Am. Chem. Soc. 1917,39, 1848.

N

. Blodgett, K. B. J. Am. Chem. Soc. 1935, 57, 1007.

3. Allara, D. L.; Nuzzo, R. G. Langmuir 1985, I, 45-52.

4. Allara, D. L.; Nuzzo, R. G. Langmuir 1985, I, 52-66.

5. Chidsey, C. E. D.; Porter, M. D.; Allara, D. L. J.Electrochem.Soc. 1986, I33, C130-
C130.

6. Dubois, L. H.; Zegarski, B. R.; Nuzzo, R. G. J. Vac. Sci. & Technol. A-Vacuum
Surfaces and Films 1987, 5, 634-635.

7. Dubois, L. H.; Zegarski, B. R.; Nuzzo, R. G. Proc.Nat. AcadSci. 1987, 84, 4739-
4742.

8. Nuzzo, R. G.; Fusco, F. A.; Allara, D. L. J. Am. Chem. Soc. 1987, 109, 2358-2368.

9. Nuzzo, R. G.; Zegarski, B. R.; Dubois, L. H. J. Am. Chem. Soc. 1987, 109, 733-740.

10. Porter, M. D.; Bright, T. B.; Allara, D. L.; Chidsey, C. E. D. J.Am. Chem. Soc. 1987,
109, 3559—3568.

11. Whitesides, G. M.; Troughton, E. B.; Bain, C.; Holmesfarley, S. R.; Wasserman, S.
R.; Strong, L. H. J.Electrochem. Soc. 1987, 134, C1 lO-Cl 10.

12. Bain, C. D.; Whitesides, G. M. Science 1988, 240, 62-63.

13. Strong, L.; Whitesides, G. M. Langmuir 1988, 4, 546-558.

14. Sagiv, J. J. Am. Chem. Soc. 1980, 102, 92—98.

15. Kohli, P.; Blanchard, G. J. Langmuir 2000, 16, 8518-8524.

16. Kohli, P.; Blanchard, G. J. Langmuir 2000, 16, 4655-4661.

17. Zhang, Z. J.; Hu, R. S.; Liu, 2. F. Langmuir 2000, 16, 1158-1162.

-108-

w‘““r—r~q

 

18

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

Home, J. C.; Huang, Y.; Liu, G. Y.; Blanchard, G. J. J. Am. Chem. Soc. 1999, 121,
4419-4426.

Home, J. C.; Blanchard, G. J. J. Am. Chem. Soc. 1999, 121 , 4427-4432.

Kohli, P.; Blanchard, G. J. Langmuir 1999, 15, 1418-1422.

Lee, H.; Kepley, L. J .; Hong, H. G.; Akhter, S.; Mallouk, T. E. J. Phys. Chem. 1988,
92, 2597-2601.

Putvinski, T. M.; Schilling, M. L.; Katz, H. B.; Chidsey, C. E. D.; Mujsce, A. M.;
Emerson, A. B. Langmuir 1990, 6, 1567-1571.

Hong, H. G.; Sackett, D. D.; Mallouk, T. E. Chem. Mat. 1991, 3, 521-527.

'fu-muuq

Katz, H. B.; Schilling, M. L.; Chidsey, C. E. D.; Putvinski, T. M.; Hutton, R. S.
Chem. Mat. 1991, 3, 699-703.

Vermeulen, L. A.; Thompson, M. E. Nature 1992, 358, 656—658.

Katz, H. E.: Schilling, M. L. Chem. Mat. 1993, 5, 1162-1166.

Schilling, M. L.; Katz, H. B.; Stein, S. M.; Shane, S. P.; Wilson, W. L.; Buratto, S.;
Ungashe, S. B.; Taylor, G. N.; Putvinski, T. M.; Chidsey, C. E. D. Langmuir 1993, 9,
2156-2160.

Vermeulen, L. A.; Snover, J. L.; Sapochak, L. S.; Thompson, M. E. J. Am. Chem.
Soc. 1993, 115, 11767-11774.

Katz, H. E. Chem. Mat. 1994, 6, 2227-2232.

Katz, H. B.; Bent, S. P.; Wilson, W. L.; Schilling, M. L.; Ungashe, S. B. J. Am.
Chem. Soc. 1994, 116, 6631-6635.

Thompson, M. E. Chem. Mat. 1994, 6, 1168-1175.

Vermeulen, L. A.; Thompson, M. E. Chem. Mat. 1994, 6, 77-81.

-109-

 

33.

34.

35.

36.

37.

38.

39.

40.

41.

42.

43

44

45

46

47

48

49

50

Hanken, H. G.; Corn, R. M. Anal. Chem. 1995, 67, 3767-3774.

Hoekstra, K. J .; Bein, T. Chem. Mat. 1996, 8, 1865-1870.

Dubas, S. T.; Farhat, T. R.; Schlenoff, J. B. J. Am. Chem. Soc. 2001, 123, 5368-5369.
Dubas, S. T.; Schlenoff, J. B. Macromolecules 2001, 34, 3736-3740.

Farhat, T. R.; Schlenoff, J. B. Langmuir 2001, I 7, 1184-1192.

Schlenoff, J. B.; Dubas, S. T. Macromolecules 2001, 34, 592-598.

Dubas, S. T.; Schlenoff, J. B. Macromolecules 1999, 32, 8153-8160.

Graul, T. W.; Schlenoff, J. B. Anal. Chem. 1999, 71, 4007-4013.

Schlenoff, J. B.; Ly, H.; Li, M. J. Am. Chem. Soc. 1998, 120, 7626-7634.

Schlenoff, J. B.; Laurent, D.; Ly, H.; Stepp, J. Chem. Eng. Technol. 1998, 21, 757-

759.

. Schlenoff, J. B.; Laurent, D.; Ly, H.; Stepp, J. Adv. Mater. 1998, 10, 347-349.

. Laurent, D.; Schlenoff, J. B. Langmuir 1997, 13, 1552-1557.

. Schlenoff, J. B.; Li, M. Ber. Bunsen-Ges. Phys. Chem. Chem. Phys. 1996, 100, 943-
947.

. Kim, H. N .; Keller, S. W.; Mallouk, T. B.; Schmitt, J .; Decher, G. Chem. Mat. 1997,
9, 1414-1421.

. Everhart, D. S. Chemtech 1999, 29, 30-37.

. Dermody, D. L.; Peez, R. F.; Bergbreiter, D. E.; Crooks, R. M. Langmuir 1999, 15,
885-890.

. Dermody, D. L.; Crooks, R. M.; Kim, T. J. Am. Chem. Soc. 1996, 118, 11912-11917.

. Yang, H. C.; Dermody, D. L.; Xu, C. J.; Ricco, A. J.; Crooks, R. M. Langmuir 1996,

12, 726-735.

-110-

51. Wells, M.; Dermody, D. L.; Yang, H. C.; Kim, T.; Crooks, R. M.; Ricco, A. J.
Langmuir 1996, 12, 1989-1996.

52. Katz, H. E.; Scheller, G.; Putvinski, T. M.; Schilling, M. L.; Wilson, W. L.; Chidsey,
C. E. D. Science 1991, 254, 1485-1487.

53. Katz, H. E.; Wilson, W. L.; Scheller, G. J. Am. Chem. Soc. 1994, 116, 6636-6640.

54. Batchelder, D. N .; Evans, 8. B.; Freeman, T. L.; Haussling, L.; Ringsdorf, H.; Wolf,
H. J. Am. Chem. Soc. 1994, 116, 1050-1053.

55. Kohli, P.; Scranton, A. B.; Blanchard, G. J. Macromolecules 1998, 31, 5681-5689.

56. Major, J. S.; Blanchard, G. J. Chem. Mat. 2002, in press.

57. Major, J. S.; Blanchard, G. J. Chem. Mat. 2002, in press.

—111-

l- 41. 'u-no ;-.l-I..t1

m

Chapter 5

Adsorption Behavior of Polymer Modified Interfaces

Abstract

We report on the vapor phase adsorption behavior of selected organic
molecules onto/into polymer-modified interfaces. We use maleirrride-vinyl ether
alternating copolymers in conjunction with layer growth chemistry to modify
gold-coated substrates with mono- and multilayer polymer structures. The
identities of the polymer side groups determine the properties of the polymer
interface. We measure the mass uptake of the interface using a quartz crystal
microbalance mounted in a chamber where the partial pressure of the organic
solvent vapor can be controlled. An optical null ellipsometer is used to m0nitor
interface thickness during the deposition process. We interpret the adsorption
data in the context of the BET adsorption isotherm. Our results point to the
adsorption process depending on both polymer identity and thickness, implying
chemical control over the adsorption and desorption kinetics and thus the

thermodynamics of the process.

-112-

i 77' 1___"‘

5.1 Introduction

The development of thin film chemistry over the past 15 years has enabled
advances in several technologically important areas. Work in thin film materials design

. . 4-
3 chermcal sensrng 7,

and synthesis has benefited second harmonic generation,"
separations!“9 and molecular scale electronic device applications. The methods of choice

for the controlled assembly of layered thin films include Langmuir-Blodgett deposition,

10-12 13-15

monolayer self-assembly, ionic and covalent layer growth.”18 The
characterization of these materials has focused on layer thickness, morphology, wetability
and constituent orientation. Among our reasons for interest in thin film materials is their
potential for use in controlling chemical separations. We focus here on the
thermodynamics and kinetics of interaction between vapor phase adsorbates and polymer
interfacial layers. In addition to structural characterization by ellipsometry, and FTIR
and UV-visible spectroscopies, we use an apparatus based on in situ quartz crystal
microbalance nano-gravimetry and optical null ellipsometry that allows us to quantify the
adsorption of organic vapor phase species onto/into polymer modified interfaces. The
combination of in situ mass and thickness data on these layers allows measurement of the
adsorption isotherm for these systems and modeling of the isotherm data provides direct
information on the energetics of adsorption. In some cases, the kinetics of adsorption in
these interfacial materials can be evaluated as well. For example, the net rate of
adsorbate uptake by the polymer monolayers can range from ~1 ng/cmz-s to ~50 ng/cmz-

5, depending on the identity of the polymer. Whether this system-dependent variability

represents a uniform change in both the adsorption and desorption rate constants or a

—113-

 

AM'-A-_m1']
I

W
l

 

 

differential change in one relative to the other can be evaluated by measuring the
adsorption isotherm behavior of these materials.

We use maleirrride-vinyl ether alternating copolymers to construct layers, with
covalent (amide or ester) or ionic (zirconium phosphonate) interlayer linkages to bond
individual layers to one another. To elucidate the behavior of these polymer films in the
presence of organic gas phase adsorbates, we monitor the thickness of the films in situ as
a function of mass uptake using an optical null ellipsometer. The ellipsometric data point
to the absence of measurable swelling of the polymer films, despite the functional form
of the mass uptake data, which reveal substantial interaction between the polymer layer
and the adsorbate. The experimental adsorption isotherm data can be understood in the

context of the Brunauer-Emmett-Teller (BET) model.19

The enthalpy of desorption we
recover from fits of the data to the BET isotherm model depends on both the chemical
identity of the polymer layer(s) and on the adsorbate. The experimental values of AHdes
all lie between 29 and 39 kJ/mol, indicating modest interactions between the adsorbate
and the polymer-modified interface. While such energies fall at the border between what
is considered to be cherrrisorption and physisorption, the reversibility of the adsorption
process demonstrates physisorption to be the dominant mode of interaction. Our data
indicate that, in some cases, capillary condensation occurs for high adsorbate vapor
pressures, and this finding argues for the polymer layers being quite permeable. Of
significance is that, in the case where we examine adsorption for monolayer and bilayer
structures of the same polymer, we find the recovered AHdcs to change with number of

polymer layers. This finding provides direct evidence for differential control over

adsorption and desorption kinetics by means of interface structure.

-114—

v v. - .HIT‘T 01-001.? 1

 

5.2 Experimental

The synthesis and layer-wise assembly of the several polymers we use in this

16"8 These materials were attached to the

work has been described previously.
(evaporated gold) surface of QCMs either covalently or ionically after the device surfaces
had been modified with 6-mercapto-1-hexanol. Covalent interlayer linkages between
polymer layers were formed using adipoyl chloride as a cross-linking agent to form ester
or amide bonds. Ionic interlayer linkages were formed by ionic association between Zr4+
and phosphonate functionalities on polymer side groups. In addition to growth on QCM
substrates, quartz slides were also used for layer assembly to allow characterization using
UV-visible transmission spectroscopy.

Substrates: The QCMs used here are AT-cut crystalline quartz with vapor-
deposited gold electrodes, operating at a resonance frequency of ~6 MHz. The change in
QCM oscillation frequency is related to adsorbate mass uptake through the Sauerbrey
equation,20
= -2Am ( fo )2 n

l
A(flp)/2
where Af = the observed frequency change, Am = the mass change, f0 = the fundamental

Af

 

frequency of the crystal (~6 MHz), n = the harmonic of the fundamental frequency (n = 1
' here), A = exposed area of the QCM (0.33 cmz), u = the shear modulus of quartz (2.947 x
101 I g cm'1 s") and p = the density of quartz (2.648 g cm'3).

Substrate preparation: QCMs and quartz slides were cleaned using piranha
solution (3:1 H2SO42H2O2) for ~15 rrrinutes. The QCMs were then rinsed with absolute
ethanol, then water, and dried under a stream of nitrogen gas. The clean QCMs were

exposed to a 10 mM solution of 6-mercapto-l-hexanol for approximately 30 minutes at

-115-

5.1

mA-ww—gl

 

35°C. This reaction introduces terminal hydroxyl functionality to the QCM surface for
subsequent polymer attachment. Following reaction with 6-mercapto-l-hexanol, the
QCMs were rinsed with ethanol, then water, and dried under a stream of nitrogen gas.
The substrates were then used in the polymer deposition reactions.

Covalent QCM modification: QCMs with gold electrodes coated with o-
mercapto-l-hexanol were exposed to 3% (v/v) 4-methylmorpholine and 3% (v/v) adipoyl
chloride in anhydrous acetonitrile under an inert argon atmosphere for ~30 minutes at
room temperature. The reacted substrates were then rinsed briefly with ethyl acetate,
dried under a stream of nitrogen gas and exposed to a solution containing a maleimide-
vinyl ether polymer synthesized with amine- or hydroxyl-terminated side groups. The
reaction of the substrate acid chloride functionality with the polymer amine or hydroxyl
side groups produces a layer of MVE polymer covalently bound to the QCM electrode,
and we have described this chemistry in detail elsewhere.‘8 We determine the thickness
of the resulting polymer layer using optical null ellipsometry.

Phosphorylation and zirconation of QCMs: Phosphorylation of the QCMs was
achieved by exposing the hydroxyl terrrrinated QCMs to 5% (v/v) collidine and 3.5%
(v/v) phosphorus oxychloride in anhydrous acetonitrile under an argon atmosphere for 30
minutes at room temperature. These substrates were then rinsed with acetonitrile,
acetone and water, and dried under a stream of nitrogen gas. Zirconation was
accomplished by exposing the phosphorylated substrates to a 60:40 ethanolzwater
solution of 10 r_n_M zirconyl chloride octahydrate at room temperature for ~30 minutes.
The zirconated substrates were rinsed with ethanol and water, then dried using nitrogen.

These substrates were then exposed to a solution of an MVE polymer containing

-116-

 

phosphonate-terminated side groups that had been partially deprotected for layer
deposition, as described previously.‘6

Quartz and oxidized silicon substrates: Quartz and silicon substrates were
prepared by first cleaning in piranha solution ~30 minutes followed by ethanol and water
rinses and drying under a stream of nitrogen gas. The resulting surface was characterized
by the presence of silanol groups that are amenable to ionic or covalent layer attachment
chemistry. Preparation of these interfaces was similar to the methods described above
except silanol groups were reacted directly with adipoyl chloride (covalent interlayer
linkages) or POC13 (ionic interlayer linkages). The substrates were then reacted with the
appropriate polymer solution under experimental conditions identical to those described
above for QCM preparation. The quartz substrates were used to acquire UV-Visible
spectra of the polymer films. The silicon and gold-coated (QCM) substrates were used
for FTIR characterization of the polymer adlayers (not presented here).

Adsorption isotherm measurements. The experimental apparatus used to acquire
isotherm data was constructed in-house and has been described in detail elsewhere.2|
This system consists of a QCM housed in an environmentally controlled stainless steel
chamber. Control of the chamber environment is accomplished using a gas blending
system that allows the partial pressure of adsorbate vapor to be varied in a predictable
and reproducible manner. Quartz windows were mounted on the sides of the chamber to
allow optical access to the QCM surface for ellipsometric measurements. ‘ The
ellipsometer was assembled from discrete optical components and operated at 632.8 nm
in the optical nulling mode. The temperature of the chamber, QCM and adsorbate

reservoir was controlled with a circulating water bath (Neslab RTE-111). The frequency

-117-

 

 

 

of the QCM is proportional to the adsorbate mass uptake and the thickness of the polymer
adlayer is monitored in situ and in real time using the ellipsometer. We detemrine the
adsorbate volume from the experimental mass change data, assuming a bulk liquid
density for the adsorbed species.

Optical ellipsometry: The thickness of the polymer-modified interfaces prior to
exposure to the vapor phase adsorbate was measured using a commercial optical
ellipsometer (Rudolph Research AutoEL 11) operating at 632.8 nm. In all cases where
comparisons could be made, the commercial ellipsometer and the home-built instrument
yielded identical results to within the uncertainty of the measurement. Thickness change
calculations for all ellipsometric measurements were performed using Rudolph software
(DAFIBM).

UV- Visible Spectroscopy: UV-visible spectroscopic data were acquired using a
Cary 300 UV-visible spectrophotometer in the 190 — 500 nm range at a scan rate of 600
nm/min with 1 nm resolution. The software used to operate this instrument was provided

with the instrument.

5.3 Results and Discussion

We have assembled several mono- and multilayer structures using maleimide-
vinyl ether alternating copolymers (Figure 5.1) and have characterized their adsorption
behavior on exposure to methanol and hexane. The central issue in this work is achieving
an understanding of how polymer adlayer identity and thickness influences the

thermodynamics and kinetics of interaction with selected adsorbates.

-118-

"‘“ml

 

‘.
t
2.. .
I
t
1,.
s

Figure 5.1 Structures of molecules bound to QCM surfaces. (a) 6-mercapto-l-hexanol,
(b) poly(4HPM-VEP), (c) poly(4MPA-VEP)

-119-

we 1

 

n n
O N O Q N O
0 : o : ‘c1
Br
0 o
9 o
S s

I D.
. , .
--------- ,I......-.,. ._..-..-.

Figure 5.1 cont’d. Structure of molecules bound to QCM surfaces. (d) poly(4BPM-
APVE), (e) poly(3CPM-APVE).

-120-

. “AAA—.1

 

To achieve this understanding we use adsorption isotherm measurements, with
ellipsometry and steady state spectroscopy data being used to provide additional
information on the materials. We use QCM gravimetry to relate the volume of the
adsorbate taken up by thin polymer films and ellipsometry to measure any change in the
thickness of the polymer layer as a result of adsorbate exposure. These data are of a
functional form consistent with the BET adsorption isotherm, and fits of the data to this
model allow us to extract information about the enthalpy of desorption (AHdes) of the
adsorbate, the volume of adsorbate required to cover the interface one molecular layer

thick (le) and, assuming a bulk liquid density of the adsorbate, the effective surface area

'7... .-. ..- 2...?

of the (polymer) adsorbent.
The BET equation relates the adsorbate volume to its partial pressure according to

V ,cz

m

V,,,, = g 5.2
‘ (1-2){1-(1-C)z}
where V422, is the volume of the adsorbate, Vm; is the monolayer volume, c is a term

related to the energetics of adsorption and z is the normalized vapor phase concentration

of the adsorbate.

c = exp((AHd,, — AHW )/RT) 5 3
z = p/p'

The quantity AHdes is the enthalpy of desorption of the adsorbate molecule from the

surface, AHvap is the enthalpy of vaporization at the temperature of the experiment, p is

the vapor pressure of the adsorbate and p* is its saturation vapor pressure. The term c

represents the balance between the enthalpies of desorption and vaporization, and is thus

 

related directly to the energy of interaction. We control the value of z experimentally by

varying the volume ratio of dry nitrogen and “wet” nitrogen using the gas blending

-121-

system. The “wet” nitrogen is saturated with adsorbate by passing dry nitrogen through a
gas-washing bottle filled with the adsorbate liquid. The ratio of these two gas streams is
controlled and introduced to the chamber containing the QCM. We use methanol and
hexane as adsorbates in this report. We consider the behavior of each interface
individually.

Before discussing the experimental data, a word is in order about our choice of
the BET adsorption isotherm model. While there are a variety of adsorption isotherm
models, we have found that the BET isotherm is sufficiently general to allow the range of
experimental data we present here to be considered within the framework of a single
model. Brunauer22 considered that there are five main categories of adsorption isotherms,
ranging from the Langmuir isotherm (Type I) for monolayer adsorption to systems that
exhibit capillary condensation at high adsorbate concentrations (Types IV and V).
Isotherms classified as Types II and III are characterized by multilayer growth that, at
high adsorbate concentrations, are consistent with bulk liquid formation at the surface. In
effect, the difference between bulk liquid condensation and capillary condensation lies in
the porosity of the interface under examination. For the interfaces we examine here,
those characterized by relatively little porosity (e. g. clean gold) exhibit Type II or Type
III behavior while the layered polymer interfaces are characterized by Type IV or Type V
saturation effects. Not surprisingly, the thiol-modified gold interface manifests
intermediate behavior.

Uncoated gold QCM: The dependence of the adsorbate volume on partial
pressure for an uncoated gold QCM substrate is shown in Figures 5.2. By fitting the data

to Eq. 2 and using published values of AHvap for the adsorbate,23 we recover the enthalpy

-122-

 

of desorption, AHdcs = 37.6i0.l kJ/mol for methanol, with the volume of a monolayer le
= 38i2 pL. For hexane, we measure AHdcs = 29.2:03 kJ/mol and le = 179i35 pL.
From these monolayer volume data and with information on the QCM surface area (0.33
cmz),2| and the surface roughness factor (3.7)24 for evaporated gold, we can estimate the
thickness of the adlayers. For methanol, we calculate a monolayer thickness of 3.1 A and
for hexane we calculate a thickness of 14.6 1- 3 A. Molecular mechanics calculations
predict ~3 A for methanol and 8.1 A for all-trans hexane. The recovered estimates for
monolayer thickness are clearly not precise, but they are in reasonable agreement with
expectations, given the nature of the measurements. The functionality of the adsorption
isotherm is most consistent with a BET Type H or Type 111 model for both adsorbates
(Figures 5.2a and 5.2b). One interpretation of such an isotherm is that the surface is
porous, but that is clearly not applicable for this system. We note that the isotherm is
fitted by the BET model for z < 0.6, so determination of Type II (or Type IH) behavior,
which is most strongly characterized for high z, is somewhat speculative. For adsorption
onto gold, no distinct monolayer region is observed, likely because of the fluxional and
associative nature of the adsorbates. The onset of bulk liquid condensation at high 2 is
indicated by the continued increase in Vads. Our data indicate that gold interacts more
strongly with methanol than with hexane, consistent with the well-established hydrophilic

nature of clean gold.

-l23-

 

 

 

300
250'
200*
150
100- r
SOLM 1
0'-
0.0 ‘ 0.2 ‘ 0.4 I 0.6 ' 0.8 A 1.0
b Z=P/P*
300.
250_
200

150
100

I 1
t——o——t

V 1' I

Vads (pL)

 

rm
F—O—d

I

1
I

V8,, (pL)

I

 

L I A l l J A I A I

0.0 0.2 0.4 0.6 0.8 1.0
2 = p/p*

Figure 5.2. Adsorption isotherm data for uncoated gold QCM. (a) Data and best-fit line
for adsorption of methanol and (b) data and best-fit line for adsorption of hexane. The
data were fitted to the BET model as given in Eqs. 2 and 3. Results of the fits are
presented in Table 1.

-124-

 

6-Mercapto-1-hexanol modified QCM: The adsorption isotherms of methanol
and hexane on a QCM modified with a monolayer of 6-mercapto-1-hexanol (Figure 5.13)
reveal substantially different behavior than that seen for uncoated gold. The data shown
in Figures 5.3 (a and b) reveal a significant increase in the volume of a monolayer for
methanol, le = 650 i 55 pL as well as for hexane, th = 896 i 46 pL. In effect, the
addition of the 6-mercapto-l-hexanol layer has increased the “surface area” of the
interface available for interaction with methanol by a factor of 17 and for hexane by a
factor of 5. The implication of this experimental observation is that there is substantial
penetration of the adsorbate into the thiol monolayer and, surprisingly, this effect is more
pronounced for methanol than it is for hexane. We also observe changes in the enthalpies
of desorption for the modified interface. For methanol vapor, we recover AHdeS = 35.4 i
0.3 kJ/mol and for hexane we recover AHdcs = 32.9 i 0.2 kJ/mol. The strength of
interaction between methanol and 6-mercapto-l-hexanol is less than between methanol
and gold, and the opposite is true for interactions with hexane. This is not an unexpected
result based on simple polarity arguments. It is likely that the short reaction time used for
deposition of 6-mercapto-l-hexanol yields a monolayer with significant structural
disorder. A well-ordered surface would present a relatively dense array of hydroxyl
groups at the surface, which would give rise to more favorable interactions with methanol
than with hexane. The increased energy of interaction observed for the hexane indicates
adsorbate interactions with the monolayer methylene groups, a condition that can exist

only in the limit of a disordered monolayer.

-125-

 

 

p—a
O
O
O
I

 

 

 

 

 

 

0.8 10
l - '
100m '
0.6 T 0.8 ' 1.0
2 = p/p*

Figure 5.3. Adsorption isotherm data for a QCM modified with 6-mercapto-1-hexanol.
(a) Data and best-fit line for adsorption of methanol and (b) data and best-fit line for
adsorption of hexane. The data were fitted to the BET model as given in Eqs.2 and 3.
Results of the fits are presented in Table 5.1.

-126-

”mtg-"a

 

Poly(4HPM-VEP) modified QCM: Poly(4-hydroxyphenylmaleirrride-vinyl ether
diisopropyl phosphonate) was attached covalently to QCM surfaces modified with 6-
mercapto-l-hexanol using surface-bound adipoyl chloride to react with the terminal -OH
group of 4-hydroxyphenylmaleimide (Figure 5.1b). The behavior of this polymer toward
methanol and hexane adsorbates is strikingly different (Figure 5.4). For methanol
adsorption we recover le = 894 i 40 pL and AHdes = 36.0 i 0.2 kJ/mol, corresponding
to a factor of 23 increase in effective surface area relative to clean gold. These results,
over the region where the BET isotherm is applicable, point to relatively high
permeability of the polymer and modest strength of interaction with methanol.
Consistent with these findings is the observation that the equilibration required for a
monolayer of poly(4I-IPM-VEP) is fast relative to many of the other polymers we report
here. For high adsorbate partial pressures, we observe a saturation phenomenon, which is
typically taken to indicate the onset of capillary condensation within the polymer layer.
The single poly(4HPM-VEP) layer, where no change of ellipsometric thickness is seen
upon methanol adsorption, appears to be a highly porous entity with a relatively large
internal surface area. The onset of capillary condensation at high adsorbate partial
pressures for such a structure is not surprising owing to the propensity of the adsorbate to
hydrogen-bond. The behavior of this polymer layer upon exposure to hexane is
fundamentally different. There is a small adlayer volume for low adsorbate partial
pressures, with a saturation seen for 2: ~ 0.4. Above z ~ 0.5, the QCM fails to resonate, an
indication of bulk liquid condensation on the surface. This is not surprising since the
phosphonate groups will likely be near the surface, making interactions with nonpolar

adsorbates relatively weak. The picture that emerges for interactions with hexane is that

-127-

Ian-nit

1m?“— *3
I
l

 

 

this adsorbate cannot penetrate the polymer surface substantially, but condenses on the
surface. The average dimension of the pores formed by the polymer layer may be such
that adsorption of methanol into the matrix is favorable while hexane is too large to
penetrate the matrix substantially. For methanol, the value of le for this interface is
large compared to that for poly(3CPM-APVE) and poly(4BPM-APVE), (vide infra),
likely due to the presence of the phosphonate groups with which the methanol can
interact.

Poly(4MPA-VEP) modified QCM: Poly(3,3’-methylphenylazomaleimide-vinyl
ether phosphonate) was bound to the QCM substrate by ionic association chemistry. The
QCM was modified with 6-mercapto-l-hexanol and the terminal hydroxyl groups were
reacted with POC13, then water to produce a phosphate surface. Exposure to Zr4+ then the
polymer produced the modified surface (Figure 5.10). Our intent in using this interface
was to design a material that is highly porous, presenting a large surface area. Given the
rigidity of the azobenzene moiety, we expect the polymer layer to possess significant
void volume and that the strongest interactions should be with a relatively non-polar
adsorbate. The absorption spectrum of this polymer shows that both cis and trans
azobenzene conformers are present (Figure 5.5a), with the ratio being consistent with that
reported previously for multilayers of the azobenzene-substituted MVE alternating
copolymer.25 The UV-visible spectrum thus points to a relatively open polymer structure.

Exposure of the poly(4MPA-VEP) interface to methanol vapor produces an
isotherm that, when fit to the BET model, gives a best-fit le = 744 i 12 pL and AHdes =
35.6 i 0.1 kJ/mol (Figure 5.5b). These results are consistent with several of the other

polymer interfacial structures we present here, indicating significant permeability of the

-128-

 

polymer matrix to methanol despite the presence of the nonpolar dimethylazobenzene
side groups. The phosphonate functionality within the polymer layer is a likely site for
methanol interaction, as suggested above for poly(4HPM-VEP). Exposure of the poly(4-
MPA-VEP) interface to vapor phase hexane yields an adsorption isotherm which gives
best-fit values of le = 4752 i 68 pL and AHdcs = 29.1 i 0.1 kJ/mol (Figure 5b). We
recover a Type V isotherm for exposure of poly(4MPA-VEP) to methanol, with the data
being less amenable to modeling for hexane. For hexane exposure, we were unable to
collect data beyond z ~ 0.4, at which point the QCM ceases to oscillate. This is a
reversible condition; when we reduce 2, we recover the original resonant frequency of the
poly(4MPA-VEP)-modified QCM. Such behavior is taken to be consistent with surface
wetting, i.e., capillary condensation and/or bulk condensation and is consistent with the
high value of le. The confirmation of capillary condensation could be made by
measuring hysteresis in the isotherm. We were unable to acquire hysteresis data due to
the very sharp dependence of Vad, on z. Despite the limits on our ability to apply the
BET model to these data, it is clear that adsorption of hexane is efficient for this polymer
adlayer, in agreement with our assertion that providing a relatively nonpolar and open

structure will favor adsorption of nonpolar adsorbates.

-129-

 

1500

1000

V,,, (pL)

500

 

0.8 1 1.0

 

.5
.8

V... (PL)
8

O
I
o
e
o

 

I I I I I I I I I I I

0.0 0.2 0.4 0.6 0.8 1.0
z=p/p*

Figure 5.4. Adsorption isotherm data for a QCM modified with poly(4HPM-VEP). (a)
Data and best-fit line for adsorption of methanol and (b) data and best-fit line for
adsorption of hexane. The data were fitted to the BET model as given in Eqs. 2 and 3.
Results of the fits are presented in Table 5.1

-130-

I 1'}.- "IA-un'
1.

 

a

0.12

I

0.10
0.08 l
0.06 -
0.04 -
0.02

absorbance (a.u.)

 

 

 

I 'I'

350 ‘ 400 450

A

0.00 . r . .
200 250 300

 

 

h _ wavelength (nm)
800}
600} .O . °
3 ; -
E: 400 -
2.8.
> 200}
0;] I .I I I I I I I I I
0.0 0.2 0.4 0.6 0.8 1.0
z=p/p*

Figure 5.5. Absorption spectrum of poly(4MPA-VEP) on a quartz substrate. The band
centered near 325 nm is the S2t—So transition for the trans side groups and the band
centered near 240nm is the S2t—So transition for the cis side groups. (b) Data and best-fit
line for adsorption of methanol (0) and hexane (O). The data were fitted to the BET
model as given in Eqs. 2 and 3. Results of the fits are presented in Table 5.1.

-131-

IA.“
6

 

I .‘l‘: ‘

 

Poly(4BPM-APVE) modified QCM: Poly(4-bromophenylmaleimide-arrrinopropyl
vinyl ether) was bound covalently to the QCM by formation of an amide bond to an
underlayer of adipoyl chloride that had been bound to the 6-mercapto-1-hexanol (Figure
1d). Exposure of this interface to methanol yields an adsorption isotherm characterized
by th = 425 i 72 pL and AHdcs = 37.1 i 1.1 kJ/mol (Figure 5.6a), similar to that seen for
the uncoated gold substrate and somewhat more than seen for a single layer of the less
polar poly(3CPM-APVE) (vide infra). We observe a decrease in le relative to the 6-
mercapto-l-hexanol modified QCM, presumably due to the blocking of some buried
hydroxyl functionality by reaction with adipoyl chloride. We anticipate some interaction
between methanol and the polymer because of the permanent dipole moment of the 4-
bromophenyl maleimide monomer unit. Semi-empirical calculations using HyperChemqD
show 11. = 3.76 D for 4-bromophenyl maleimide.

Poly(3CPM-APVE) modified QCM: Poly(3-chlorophenylmaleimide-arninopropyl
vinyl ether) was bound covalently to the QCM by formation of an amide bond to an
underlayer of adipoyl chloride that had been bound to the 6-mercapto-1-hexanol (Figure
5.1e). A second layer of polymer is linked to the first using adipoyl chloride as the
covalent cross-linking agent.18 The adsorption isotherm data for poly(3CPM-AVPE)
interfaces provides insight into the structure-dependent adsorption kinetics that
characterizes such systems. Upon addition of a single layer, the adsorption isotherm for
methanol (Figure 5.6b) indicates a monolayer volume of 209 i l pL, which is
substantially less than the 650 pL seen for the 6-mercapto-1-hexanol surface. The
enthalpy of desorption is AHd,s = 35.9 i 0.2 kJ/mol, similar to that measured for the 6-

mercapto-l-methanol surface and less than that seen for the more polar poly(4BPM-

-l32-

 

APVE) layer. These data suggest that the dominant interaction of the methanol adsorbate
is with buried hydroxyl and/or carboxylic acid functionality beneath the polymer layer,
with the reduced monolayer volume being an indication of the fraction of unreacted
hydroxyl and carboxylic acid groups present. The addition of a second polymer layer
yields the same adsorbate monolayer volume, th = 216 i 14 pL, but a larger enthalpy of
desorption, AH“, = 38.5 i 0.7 kJ/mol (Figure 5.6b). We note no ellipsometric thickness
change of the polymer layers on exposure to methanol, consistent with the notion that the
adsorbate interacts most strongly with unreacted portions of the underlayer. The finding
of the same monolayer volume argues for the buried hydroxyl and carboxylic acid groups
acting as the dominant interaction sites for the adsorbate, with the difference in AHdes
being reflective of the structural properties of the polymer. In this picture, the second
polymer layer reduces the permeability of the interface. Because there is a substantial
vapor phase adsorbate concentration, the rate of adsorption is likely quasi-zeroth order
while the rate of desorption is mediated by the ability of the methanol to escape the
polymer layer(s) after adsorption. This model does not argue for a change in the strength
of interaction of the adsorbate with the underlayer, but rather that the apparent change in
AH“,s is a consequence of the change in the rates (not necessarily the rate constants) of
adsorption and desorption into/out of the polymer film. Because we are measuring mass
uptake within the polymer interface, we do not have the experimental resolution to
determine where within the interface the adsorbate is interacting most strongly. Once in
the polymer layer, the methanol has more difficulty escaping from the bilayer than from
the monolayer. In this manner we can use layer structure to mediate the kinetics of

adsorbate uptake and/or desorption.

-l33-

 

 

t—1 r—

O N

O O

O O
r r
o

 

 

0.8 ‘ 1.0
b z=pm*

1200
1000
800

I
O

I
O

I

 

 

(10 (12 (14 (I6 (18 1L0

z = p/p*
Figure 5.6. Methanol adsorption isotherm data and best-fit lines for a QCM modified
with (a) poly(4BPM-APVE) and (b) poly(3CPM-APVE) with one polymer layer (0) and

two polymer layers (O). The data were fitted to the BET model as given in Eqs. 2 and 3.
Results of the fits are presented in Table 5.1.

-134-

 

5.4 Conclusions

The range of substrates and polymer structures we have examined using
adsorption isotherm measurements with methanol and hexane as adsorbates demonstrates
that structural control over selectivity and adsorption/desorption kinetics is feasible. We
interpret our data for single polymer layers in the context of polymer side group and
adsorbate polarity. In many cases we observe adsorption isotherm behavior consistent
with capillary condensation, pointing to the structurally heterogeneous and/or porous
nature of the polymer layers. Modification of our gold QCM electrode interfaces with 6-
mercapto-l-hexanol leads to a significant increase in monolayer coverage relative to the
uncoated gold surface, suggesting disorder in the monolayer. The construction of an
interface using poly(4MPA-VEP) with ionic layer linking chemistry yielded the expected
result for a system possessing substantial free volume. In this case we recover a BET
Type V isotherm, consistent with capillary condensation within the internal matrix of the
polymer. For the case where adsorption of methanol is compared for one and two layers
of poly(3CPM-APVE), we find that the addition of the second layer influences the
effective adsorption and desorption rates, leading to an apparent change in desorption
enthalpy. These experiments demonstrate the feasibility of kinetic control in
diffusion/adsorption over molecular length scales. We anticipate that the use of polymer
multilayers comprised of several different substituted MVE polymers will afford even

greater kinetic control and adsorbate selectivity.

-l35-

 

 

Table 5.1. Layer thickness, monolayer adsorbate volume (le) and enthalpy of
desorption (AI-Ides) extracted from experimental data and fits to BET isotherm.

 

 

 

 

 

 

Methanol Hexane

Interface Thickness AHads AHads
(A) Vm’ (9L) (kJ/mol) Vm‘ 9’” (kJ/mol)

Gold -- 38 1 2 37.6 1 0.1 179 1 35 29.2 1 0.3
6-mercapto-1—hexanol 911 650 1 55 35.4 1 0.3 896 1 46 32.9 1 0.2
Poly(4HPM-VEP) 3512 894 1 40 36.0 1 0.2 -- --
Poly(4MPA-VEP) 4913 4752 1

744112 35.6101 29.1101

68

Poly(4BPM-APVE) 3011 425 1 72 37.1 1 1.1 -- --

 

Poly(3CPM-APVE) 1 34:2
209 1 1 35.9 1 0.2 -- --
layer

 

 

Poly(3CPM-APVE) 2 5613
216114 38.5107 -- --
layers

 

 

 

 

 

 

-136-

 

 

5.5

10.
ll.
12.

13.

Literature Cited

Bakiamoh, S. B.; Blanchard, G. J. Langmuir, 2001, 17, 3438.

Katz, H. E.; Scheller, G.; Putvinski, T. M.; Schilling, M. L.; Wilson, W. L.; Chidsey,
C. E. D. Science, 1991, 254, 1485.

Li, D.; Ratner, M. A.; Marks, T. J.; Zhang, C. H.; Yang, J.; Wong, G. K. J. Am.
Chem. Soc.,: 1990, 112, 7389.

Demas, J. N.; Degraff, B. A.; Coleman, P. B. Anal. Chem. New & Features, 1999,
793A.

Yoon, H. C.; Kim, H-S. Anal. Chem. 2000, 72, 922.

Delamarche, E.; Sundarababu, G.; Biebuyck, H.; Michel, B.; Gerber, C.; Sirgrist, H.;
Wolf, H.; Ringsdtirf, H.; Xanthopoulos, N.; Mathieu, H. J. Langmuir, 1996, 12,
1997.

Brousseau, L. C. III; Aurentz, D. J.; Benesi, A. J .; Mallouk, T. E. Anal. Chem. 1997,
69, 688.

Vrancken, K. C.; VanderVoort, P.; Gillisdhamers, 1.; VanSant, E. F.; Grobet, P.; J.
Am. Chem. Soc. Faraday Trans. 1992, 88, 3197.

Pfleiderer, B.; Albert, K.; Bayer, E. J. Chromatogr. 1990, 506, 343.

Ulman, A. Chem. Rev. 1996, 96, 1533.

Dubois, L. H.; Nuzzo, R. G. Annu. Rev. Phys. Chem. 1992, 43, 437.

Skulason, H.; Frisbie, C. D. Langmuir, 1998, 14, 5834.

Lee, H.; Kepley, L. J.; Hong, H. G.; Akhter, S.; Mallouk, T. E. J. Phys. Chem. 1988,

92, 2597.

—137-

 

14.

15.

l6.

l7.

18.

19.

20.

21.

22.

23.

24.

25.

Lee, H.; Kepley, L. J.; Hong, H. G.; Mallouk, T. E. J. Am. Chem. Soc. 1988, 110,
618.

Putvinski, T. M.; Schilling, M. L.; Katz, H. E.; Chidsey, C. E. D.; Mujsce, A. M.;
Emerson, A. B. Langmuir, 1990, 6, 1567.

Major, J. S.; Blanchard, G. J. Langmuir, 2001, 17, 1163.

Major, J. S.; Blanchard, G. J. Chem. Mater., in review.

Major, J. S.; Blanchard, G. J.Chem. Mater., in review.

Brunauer, S.; Emett, P. H.; Teller, E.; J. Am. Chem. Soc., 1938, 60, 309.

Sauerbrey, G. Z. Z. Phys. 1959, 155, 206.

Karpovich, D. S.; Blanchard, G. J. Langmuir 1997, 13, 4031.

(a) Brunauer, S.; The Adsorption of Gases and Vapors, Vol. 1., Princeton University
Press, Princeton, NJ, 1945. (b) Adamson, A. W.; Physical Chemistry of Surfaces,
Fifth Edition, John Wiley & Sons, Inc., New York, 1990, p. 6097f.

Lide, D. R., Ed.; CRC Handbook of Chemistry and Physics, 71” Edition, CRC Press,
Inc., Boca Raton, FL, 1990.

Abbott, N. L.; Kumar, A.; Whitesides, G. M. Chem. Mater. 1994, 6, 596-602.
Kohli, P.; Rini, M. C.; Major, J. S.; Blanchard, G. J.; J. Mater. Chem, 2001, 11,

2996-3001.

-138-

 

 

Chapter 6
Molecular Control of Adsorption Properties of Polymer Modified

Interfaces

Abstract

In this work we report the assembly of several polymer-modified interfaces where
we attempt to generate a chemical potential gradient in the interface along the surface
normal. To achieve this goal, we have prepared a suite of copolymers capable of
multilayer assembly, where these polymers have been designed to be self-terminating at
the deposition of one molecular layer. The polymers are prepared by radical
copolymerization of various vinyl ether monomers with maleimide monomers that have
been prepared in-house and present a variety of functional side-groups. The
functionalities present on the vinyl ether side groups are used in layer assembly, while the
maleimide side groups are used to generate the chemical potential gradient. In order
realize the chemical potential gradient, we have generated a gradient in the dipole
moment of the maleimide side groups. Measurement of the adsorption isotherms of
several polymer bilayers upon exposure to methanol and hexane reveals behavior that
depends on both the identity and layer order of the interfaces, and we interpret these data

in the context of polymer layer polarity.

139

 

6.1 Introduction

Separation science is one area of chemistry where the chemical nature of the
interface is intimately related to its bulk behavior. In most separation schemes, there are
two different phases (typically one stationary and the other mobile) where the change in
the chemical potential that a solute experiences as it is transferred between the two phases
serves as the driving force for separation. In most cases when the resolution of solute
molecules is poor, attempts at achieving improved separation/selectivity have focused on
applying some external gradient such as temperature”, pressureg’12 or voltage13
programming. Attempts to vary the chemical potential of system have focused on solvent

2'4‘9’14'16 where creating a gradient in solvent “strength” leads to resolution of

gradients,
poorly resolved solutes. In all of these cases, the application of a gradient presents
several challenges to the experimenter, such as the need for additional equipment,
equipment stability (in the case for solvent gradients) costly solvents, and the time
required to optimize the separation. Additionally, the use of high pressures and voltages
could also prove unsafe.

According to Giddings, the generation of a gradient in some property of a system
would lead to a chemical potential gradient within the material.17 In this work, we have
designed and assembled a variety of interfaces where we attempt to create a chemical
potential gradient within the interfaces along the surface normal. To realize this
structural goal, we have employed layer-by-layer assembly of maleirrride-vinyl ether
alternating copolymers onto our interfaces (quartz crystal microbalance, hereafter

referred to as QCM) where we have varied the identity (dipole moment) of the functional

constituent incorporated into each layer. In this way, we generate a gradient in the dipole

140

 

 

moment of the interface. Interfaces possessing an internal gradient in some property
should lead, in principle, to enhanced separations and would eliminate many of the
problems associated with the generation of gradients according to the schemes mentioned
above.

In order to characterize these interfaces, we perform adsorption isotherm
measurements and fit our experimental data to the BET model. We have employed an in-
house assembled apparatus composed of QCM electronics and discrete optical
components that allow simultaneous mass uptake and in situ ellipsometric thickness

18"9 of this system has been described

measurements. The design]8 and operation
elsewhere. Using this apparatus we were able to determine mass uptake as a function of
adsorbate vapor pressure.

We have reported previously (chapter 5) on the adsorption behavior of various
polymer-modified interfaces"). In that work, we established that the predominant
interaction between the adsorbate and the interface occurred at the underlayer as a result
of the adsorbates ability to partition into the interface. We also demonstrated that the
polymer overlayer influences this process, apparently by influencing the desorption
kinetics. In the work we report here, we have set out to demonstrate our ability to
selectively control the partitioning process through design. As mentioned above, these
interfaces have been designed where we vary the identity of the functional groups
incorporated into each layer thereby manipulating the overall property of the interface.

While this body of work represents a substantial departure from the mainstream of

separation science, the results of our studies are clearly indicative of the ability to design

interfacial materials where we are able to control selectivity of the interface by synthetic

141

IN.“

 

means. We demonstrate that, in addition to affecting the properties of the interface based
on the functional groups incorporated during polymer synthesis, varying the order of
deposition of the polymers leads to dramatic changes in the form of the adsorption
isotherm. At this point, the relationship between polymer side group dipole moment and
adsorption behavior remains to be explored fully, but these preliminary data suggest that
we have control over interface adsorption behavior. Further studies are underway to
determine more qualitatively the role of dipole moment on the adsorption properties of

interfaces.

6.3 Experimental

In this work we employ several alternating maleimide vinylether copolymers

20’2“. We assemble

whose synthesis and multilayer assembly was reported previously
multilayers of these polymers onto quartz crystal rrricrobalance (QCM) electrodes (Au)
and have used these substrates to measure the vapor phase adsorption isotherm for these
systems.

Substrate preparation. QCM acquired from McCoy Electronics (P/N 78-18-4).
Their surface is characterized as microcrystalline gold domains that are predominantly
(111). These substrates were cleaned in piranha solution (3:1 H2SO4:H2O2) followed by
rinsing in ethanol, then water. These substrates were then dried under a stream of
nitrogen gas. In order to introduce reactive functionalities onto the surface of the QCM,
we use well-established alkanethiol/gold chemistry. In this scheme, we expose the clean

QCM to a 10mM solution of 6-mercaptohexanol in prepared in a 60:40 ethanol:water

solution. The QCM and solution were warmed to ~40°C for 30-60 minutes following by

142

 

ethanol and water rinses, and drying by nitrogen gas. Once the QCM electrode surfaces
have been hydroxylated they are then exposed to a solution of 3% (v/v) adipoyl chloride
and 3% (v/v) 4-methylmorpholine in anhydrous acetonitrile under an inert argon
atmosphere for approximately 30 minutes at 40°C. After reaction with adipoyl chloride,
the QCMs are removed from the solution and rinsed with ethyl acetate followed by
exposure to the polymer solution.

Multilayer assembly. Subsequent to surface preparation, the QCM is then exposed
to the appropriate polymer solution and warmed to 40°C for 30 minutes. Polymer
solutions are prepared at a concentration of 10 mM in either dimethyl sulfoxide (DMSO)
or a 75:25 DMSO:acetonitrile poly(NPM-VOB). These solutions were reused throughout
the entire experiment with no observed loss of reactivity. Multilayer assembly is
accomplished by reacting the initially deposited polymer layer with adipoyl chloride as
described above followed by exposure to the second polymer solution. This scheme can
be employed to deposit the desired number of polymer layers. In the following
discussion we use the following nomenclature; S-xxx-yyy, where S indicates the
substrate, and the terms xxx and yyy are abbreviations for the polymers used. BPM =
poly(4BPM-APVE), CPM = poly(3CPM-APVE), NPM = poly(NPM-VOB) and HPM =
poly(HPM-VEP). In all cases, adipoyl chloride is used as the interlayer linking moiety.

S-BPM-NPM: This interface (scheme 6.1a) was prepared by first linking one layer
of poly(4BPM-APVE) to the (u-hydroxythiol-functionalized substrate using adipoyl
chloride under the conditions reported above. The initial layer of poly(4BPM-APVE)
was then reacted with adipoyl chloride followed by exposure to a solution of poly(NPM-

VOB).

143

 

 

S-NPM-BPM: This interface (scheme 6.1b - the structural complement to the
above interface) was prepared in an identical manner to S-BPM-NPM except poly(NPM-
VOB) is the initial polymer layer, followed by the addition of a layer of poly(4BPM-
APVE).

S-BPM-CPM: This interface (scheme 6.1 c) was prepared by first depositing one
layer of poly(4BPM-APVE) then capping with poly(3CPM-APVE), using adipoyl
chloride as the crosslinking agent.

S-CPM-BPM: This interface (scheme 6.1d) is the structural complement to the
above system prepared under identical experimental conditions. Here, the layer
deposited initially is poly(3CPM-APVE), which is reacted with adipoyl chloride, then
with poly(4BPM-APVE) to generate the bilayer.

S-BPM-HPM: This substrate (scheme 6.1e) is prepared by initially depositing a
layer of poly(4BPM-APVE) followed by reaction with adipoyl chloride and deposition of
poly(4I-IPM-VEP).

S-HPM-BPM: This interface (scheme 6.1f — the structural complement to the
above interface) was prepared by first depositing one layer of poly(4HPM-VEP) followed
by one layer of poly(4BPM-APVE). It should be noted that in these two interfaces, the
deposition of poly(4HPM-VEP) was accomplished by reaction of the hydroxyl group of
the phenyl maleimide substituent with adipoyl chloride, leading to the formation of an

aromatic ester.

144

 

 

0N0
Eoi
O
0

Br

0

0

ii:
ii
‘i

 

 

l 1

Scheme 6.1. Interface structures. (a) S-NPM-BPM, (b) S-BPM-NPM.

145

 

”—1
-l

Edit—jam Mal? "' "l
‘ .4 —

 

 

 

 

 

 

 

 

Scheme 6.1 cont’d. Interface structures. (c) S-BPM-CPM, (d) S-CPM-BPM.

 

146

 

 

 

 

Scheme 6.1 cont’d. Interface structure. (e) S-BPM-HPM, (f) S-HPM-BPM.

147

We L‘I'I.v: Inn-91m”
. -
I

 

Optical ellipsometry. Measurements of polymer adlayer thickness on the QCM
prior to exposure to the vapor phase adsorbates, were accomplished using a Rudolph
Research AutoEL II commercial optical nulling ellipsometer. This ellipsometer was
equipped with a HeNe laser operating at 632.8 nm. The software (DAFIBM) used to
operate the ellipsometer was provided by Rudolph.

Adsorption isothenn measurements. The apparatus we use for the measurement
of adsorption isotherms has been described in Chapter 5. We note that the results we
obtain with this system are highly reproducible. In almost all cases, reproducibility
between samples is better than 5% and only at very high values of z do deviations
sometimes become significant. In most such cases, these deviations are within 10% from
run to run, even if separated in time by months. This reproducibility is seen for
measurements taken on the same substrates but on different days. For substrates prepared
at different times (often weeks apart) we obtain isotherm data that are reproducible to

within ~5% or less.

148

'l 1.
P

 

 

6.4 Results and Discussion

The focus of the work we report here is the construction and evaluation of
interfaces where we demonstrate the ability to manipulate the adsorption/desorption
properties of the interface in some controlled manner. We have chosen to prepare these
interfaces using various maleimide-vinyl ether copolymers (Chapter 2) and have
assembled these polymers into bilayers using covalent crosslinking strategies that we
have reported on in Chapters 2 and 3. We have chosen to work with these polymers due
to the fact that we are able to introduce various functional groups into the matrix of the
polymer through simple substitution of the maleirrride side-groups, where the preparation
of the maleirrrides have been reported on elsewherezz’zz. Additionally, the ability to vary
the crosslinking chemistry used to assemble the multilayers adds another means of tuning
the properties of the resulting adlayer. A layer-by—layer assembly strategy allows us to
vary the functional constituents incorporated into each individual layer — a feature that we
use to control interfacial behavior.

As mentioned earlier, the ultimate aim of this body of work is the design of
interfaces where we attempt to generate a chemical potential gradient, with the goal of
controlling the selectivity of the interactions between specific analytes and these
interfaces. To construct the chemical potential gradient, we have varied the identity of
the polymer side groups within each layer, where the gradient to be generated is
controlled by the dipole moments of the unreacted side groups. We have used semi-
empirical calculations (PM3 parameterization) to calculate the dipole moments of the
maleirrride side groups and we present these results in Table 6.1. We present the

adsorption isotherm data for the bilayer structures reported here in Table 6.2.

149

 

 

 

Table 6.1. Calculated dipole moments of selected
polymer side groups. Calculations performed using
Hyperchem v. 6.0 with PM3 parameterization.

 

Maleimide side acronym Calculated SO
drpole moment
gm) (D)
0
O
0 Cl
0
O

| -—.~Bt BPM 2.6

O

The adsorption of vapor

phase adsorbates onto the

interfaces has been characterized
in the context of the BET model.
Using this model, we were able to
extract useful information about
our interfaces such as energies of
interaction (enthalpies of
desorption), and monolayer
volume over limited ranges of
adsorbate partial pressure, and
have recovered substantially
different values that depend on (1)

the functional groups incorporated

into the layers, and (2) the order of layer deposition. Based on these data, we have

demonstrated our ability to manipulate adsorption in a qualitatively predictable manner.

We have chosen methanol (polar protic) and n-hexane (non-polar) as adsorbates.

We also attempted these studies using acetone (polar aprotic), but were unable to acquire

useful data for most of our systems, so these studies were discontinued. The adsorption

isotherm data and fits to Eq. 6.2 are presented in Figures 6.1 — 6.6.

S-BPM-NPM/S-NPM-BPM: The adsorption isotherms for these systems reveal

some interesting features. At low methanol vapor pressure, we observe that the isotherms

for both interfaces are similar. As p approaches p*, the isotherm for S-BPM-NPM can be

150

 

 

approximated as a BET Type H isotherm, due to the apparent onset of saturation, while
S-NPM-BPM yields behavior that is more Type H1 in character (Figure 6.1). A Type II
isotherm indicates a strong(er) enthalpy of adsorption, i.e., stronger interaction of the
adsorbate with the adsorbent, while a Type HI isotherm is typically taken to reflect a
weaker interaction between the adsorbate and adsorbent. Fits of these data to the BET
equation produce values of le = 91 1 5 pL and 371 1 24 pL for S-BPM-NPM and S-
NPM-BPM, respectively. Values of c = 6.0 i 1.1 for S-BPM-NPM and 0.45 i 0.05 for
S-NPM-BPM for exposure of these interfaces to methanol, reveal substantially different
enthalpies of desorption; A112,es = 41.9 1 0.5 kJ/mol for S-BPM-NPM and AHdeS = 35.5 1
0.3 kJ/mol for S-NPM-BPM.

When these interfaces are exposed to n-hexane, we measure adsorption isotherms
that are qualitatively different than those seen for methanol (Figure 6.2). The isotherm
for S-BPM-NPM exhibits a distinct region of saturation followed by one of efficient
adsorption. Beyond 2 ~ 0.45, data collection is not possible due to the amount of
adsorbate present at the interface. For S-NPM-BPM, we observe the onset of saturation
near 2 ~ 0.5, consistent with capillary condensation. Fits of these data for low values of z
to the BET equation yield values of Vm. = 113 1 15 pL and c = 9.68 1 3.57 for S-BPM-
NPM and Vm. = 955 1 90 pL and c = 0.78 1 0.10 for S-NPM-BPM. The corresponding
values of AHdes are 37.2 1 0.8 kJ/mol for S-BPM-NPM and AHdes = 30.9 1 0.3 kJ/mol for
S-NPM-BPM. This behavior is the converse of that seen for methanol adsorption when
the data are compared to AHvap for each solvent. For methanol, AHyap = 37.43 kJ/mol at
298 K and for n-hexane, AHvap = 31.56 kJ/mol at 298. K23 This finding indicates that the

top-most polymer layer mediates, but does not

151

control completely, the adsorption behavior of these adlayers. This control over the
effective thermodynamics of adsorption implies differential control over adsorption and

desorption kinetics for these systems that is based on adlayer chemical structure.

 

 

 

16(1): 0 SUB-NPMBPM
. I SUB-BPM—NPM
1400 - 1 .
L nethano adsorption
1200 r O
A 1000-
7i 1
Va, 8(111-
'3 . o
>
600.” f .
400 - 9
~ 8
20) - 0 e 9 G
0
. . . 0 09'
O 900 I I I I I I I I 1
0.0 0.2 0.4 0.6 0.8 1.0

2 (1113*)

Figure 6.1. BET isotherm for the adsorption of methanol on S-BPM-NPM and S-NPM-
BPM.

152

- " I" r 4]. I)
W. —Wr
_Ia

 

 

 

K00-
. ll SUBIWNHWEA
14m_ . O SUB-WW
' hameaqufiar
EDO-
HDO-
jo} wL o O 0
g I 00
> 6D- 09
' o e
403- 00
0
' OO .0
1D- 000 .0
- 00090000.
0 00 1 1 1 1 1

 

 

00 02 04 06 08 10

Figure 6.2. BET isotherm for the adsorption of hexane on S-BPM-NPM and S-NPM-
BPM.

153

S-BPM-CPM/S-CPM-BPM: The adsorption isotherm data we recover for these
polymer adlayers reveal complex, adsorbate-dependent behavior (Figures 6.3 and 6.4).
When S-BPM-CPM is exposed to methanol we recover an isotherm that is characterized
by a plateau region near 2 ~ 0.5, followed be rapid adsorption and saturation. This
behavior appears to be consistent with the two polymer layers functioning as independent
entities, with saturation and possibly capillary condensation occurring for each layer
separately (Figure 6.3). Exposure of S-CPM-BPM to methanol yields adsorption
behavior similar to a BET Type III isotherm, with only a suggestion of saturation at
higher z-values. While saturation at high 2 values occurs at almost the same point for
both systems, the adsorption behavior up to that point is substantially different for the
two systems. Fits of the experimental data to the BET equation yield c values of
0110.08 (AHdes = 31.7 1 4.0 kJ/mol) and 1.081036 (AHdes = 37.6 1 0.7 kJ/mol) for S-
BPM-CPM and S-CPM-BPM, respectively. The monolayer volumes of le = 2800 1
1965 pL for S-BPM-CPM and le = 128 1 28 pL for S-CPM-BPM suggest a porous
polymer structure. S-BPM-CPM appears to manifest BET Type HI behavior in regions
of low z, where we fit these data to Eq. 6.2. We note that fitting these data proved
difficult, as reflected in the significant uncertainty in the resulting values of th and c
(Table 6.2).

When these interfaces are exposed to hexane, we recover behavior that is
characterized by saturation at low 2 values and BET Type H characteristics up to
saturation. Fits of data for both polymer assemblies reveal large le values (Table 6.2)
and small c values; AHdcs = 31.8 1 0.4 kJ/mol for S-BPM-CPM and AHdes = 25.3 1 0.3

kJ/mol for S-CPM-BPM. These interactions are weaker in absolute strength, but similar

154

in magnitude relative to AHvap as those seen for methanol. These findings reveal that
3CPM is experimentally less polar than 4BPM, in agreement with the semi-empirical
calculations of side group dipole moment (Table 6.1). A point worth noting in this
particular case is the fact that the energies of interaction relative to AHvap for each solvent

are reverse for methanol and hexane, as expected.

 

 

 

14(1)—
_ O SUB-BPM—CPM
1201- O SUB-CPM—BPM o
rretlnnoladsorptim o
1000- ~ 0
o
A 80)— o
....l
3 * o
>§ 600- '
' 0
4m- 0000.
o
' 0o 0
21D- 00 ...
r- 900.. .0
00.0
OHj-f-i.l I I I l
00 02 0.4 06 08 10

Z (1113*)

Figure 6.3. Isotherms for the adsorption of methanol on S-BPM-CPM and S-CPM-BPM.

155

W4 lays“ 'mfl

 

- O SUB-Bl’M-(PM
O SUB-(PM-BPM

Vad,(PL)
s 8 es . s

 

 

Figure 6.4. Isotherms for the adsorption of hexane on S-BPM-CPM and S-CPM-BPM.

156

 

S—BPM-HPM/S-HPM-BPM: Exposure of these interfaces to methanol yields BET
Type II behavior for S-HPM-BPM with saturation occurring at z > 0.6 (Figure 6.5). The
isotherm of S-BPM-HPM however, is slightly more complex. There is a significantly
stronger interaction of S-HPM-BPM with methanol, as seen from the low-z behavior for
both adlayers. Fitting the data yields c values of 0.66 1 0.1 (AHdcs = 36.4 1 0.4 kJ/mol)
and 73.5 1 9.0 (AHdcs = 48.1 1 0.3 kJ/mol) for S-BPM-HPM and S-HPM-BPM,
respectively. The fact that the apparently stronger interaction occurs for the case where
the polar adlayer is confined between the substrate and a relatively less polar layer again

indicates differential control over adsorption and desorption kinetics at these interfaces.

 

 

 

1600'- o SUB-BPM-HPM o
e SUB-HPM—BPM
1400 - .
_ methanol adsorptron
1200 - 0
1000 - 0
(a .
v“ 8(1) r O
8 - o o
> 600 - o ' .
.. O
400 1- 00 ° .
O 0
- o
200 - coo , o"
-- 88833” '
O W I I I I 4
0.0 0 2 0 4 0.6 0 8 1 O
z (p/p*)

Figure 6.5. Isotherms for the adsorption of methanol on S-BPM-HPM and S-HPM-BPM.

157

Attempts to acquire data for the exposure of both S-HPM-BPM and S-BPM-HPM
to hexane proved unsuccessful. In both cases, QCM oscillation was not consistent with
adsorption. In some instances, we observed that the frequency of the QCM increased,
behavior consistent with mass decrease, suggesting the possible destruction of the
polymer layers. At high values of z, the QCM oscillation frequency was unstable, and
upon removal of the adsorbate stream, we recovered the original (z = 0) oscillation
frequency. In contrast to the other adsorbate/polymer adlayer systems, S-HPM-BPM and
S-BPM-HPM do not appear to achieve a reproducible equilibrium condition with hexane
adsorbate. Clearly, further work will be required to understand the reason for this
anomalous behavior.

One issue that requires consideration is the assumption that we are working in the
thermodynamic limit with these adsorption measurements. The experimental data point
to differential control over the adsorption and desorption kinetics of these polymer
adlayers based on chemical structure and ordering. The issue of differential control over
kinetic processes calls into question our assumption of operating in the thermodynamic
lirrrit. At the very least, it is clear that we are operating under a steady-state condition,
but our experimental data indicate that we are, in fact, in the thermodynarrric limit. We
show in Figure 6.6 the forward and reverse adsorption isotherms for methanol adsorption
on S-BPM-CPM. These data do not demonstrate any significant hysteresis. While it is
true that hysteresis in adsorption isotherms can be seen, even in the thermodynamic limit,
depending on the nature of the adsorbate-adlayer interaction, the absence of hysteresis

demonstrates that we are operating in the limit where the adsorption and desorption

158

processes are occurring under equilibrium conditions.

1400-
1200- 8
1000-

8004

Vads (DU

600- O
400- 0°93

1
200- O.

 

 

, . .
0.0 0.2 0.4 0.6 0.8 1.0

Figure 6.6. Forward (P ) and backward (D) isotherms for the exposure of S-
BPM-CPM to methanol.

159

Table 6.2. Results of fitting experimental adsorption isotherm data to Equation 6.2.

 

methanol hexane

 

V1111 C Vm] C

 

S-BPM-NPM 9115 pL 6.01l.1 113115 pL 96813.57

 

S-NPM-BPM 371124 pL 0.451005 955190 pL 07810.10

 

S-BPM-CPM 280011965pL 0101008 9681113 pL 1.111018

 

 

 

 

 

 

 

 

 

S-CPM-BPM 128128 pL 1.081036 169218 0.081001
S-BPM-l-IPM 911193 pL 0.661010
S-I-IPM-BPM 11712 pL 73519.0

 

6.5 Conclusions

In this work, we have demonstrated our ability to control the adsorption and
desorption kinetics of polymer interfacial materials differentially by means of the
chemical identity of the polymer adlayers and the order in which they are deposited. We
have observed significant differences in the effective interaction energies of these
systems and the shapes of the isotherms recovered in most cases these shapes are
consistent with the expected trends. These data are a first demonstration that we can
control the adsorption properties of these interfaces and, ultimately, their selectivity

through careful design and ordering of polymeric adlayers.

160

 

6.6 Literature Cited

10.

ll.

12.

l3.

14.

15.

16.

17.

18.

19.

20.

Kanazawa, H.; Kashiwase, Y.; Yamamoto, K.; Matsushima, Y.; Kikuchi, A.; Sakurai,

Y.; Okano, T. Anal. Chem. 1997, 69, 823.

. Lukulay, P. H.; McGuffin, V. L. Anal. Chem, 1997, 69, 2963.

Li, J.; Carr, P. W. Anal. Chem.1997, 69, 837.

Go, H.; Sudo, Y.; Hosoya, K.; Ikegami, T.; Tanaka, N. Anal. Chem. 1998, 70, 4086.
Li, J .; Carr, P. W. Anal. Chem. 1997, 69, 3884.

Li, J .; Carr, P. W. Anal. Chem. 1997, 69, 2202.

Thompson, J. D.; Carr, P. W. Anal. Chem. 2002, 74, 1017.

Lan, K.; Jorgenson, J. W. Anal. Chem. 1998, 70, 2773.

MacNair, J. E.; Patel, K. D.; Jorgenson, J. W. Anal. Chem. 1999, 71, 700.

Tolley, L.; Jorgenson, J. W.; Moseley, M. A. Anal. Chem. 2001, 73, 2985.
MacNair, J. E.; Lewis, K. C.; Jorgenson, J. W. Anal. Chem. 1997,69, 983.
McGuffin, V. L.; Chen, S-H. Anal. Chem. 1997, 69, 930.

Hutterer, K. M.; Jorgenson, J. W. Anal. Chem. 1999, 71, 1293.

Taylor, M. R.; Teale, P.; Westwood, S. A.;.Perrett, D. Anal. Chem. 1997, 69, 2554.
Yan, C.; Dadoo, R.; Zare, R. N.; Rakestraw, D. J.; Anex, D. 8. Anal. Chem. 1996,
68, 2726.

Huber, C. G.; Choudhary, G.; Horvath, C. Anal. Chem. 1997, 69, 4429.

Gidding, C. G. Unified Separation Science; Wiley Interscience, 1991.

Karpovich, D. S.; Blanchard, G. J. Langmuir, 1997, I3, 4031.

Major, J . S.; Blanchard, G. J. Langmuir, in review.

Major, J. S.; Blanchard, G. J. Chem. Mater. in press.

161

21. Major, J. S.; Blanchard, G. J. Chem. Mater. in press.
22. Kohli, P.; Blanchard, G. J. Langmuir 1999, 15, 1418.

23. Lide, D. R. CRC Handbook of Chemistry and Physics, 82nd Ed. ,' CRC Press, 2001-
2002.

162

Chapter 7

Covalently-Bound Polymer Multilayers for Efficient Metal Ion Sorption

Abstract

We report on the design, synthesis and characterization of a polymeric multilayer
assembly that can act as an efficient sorbent for selected metal ions. We use a covalent
layer-by-layer deposition scheme for the construction of multilayer assemblies using the
alternating co-polymer of 4-hydroxyphenylmaleimide and ethylvinylether-Z-
diisopropylphosphonate. The maleimide-vinyl ether (MVE) polymer layers are reacted at
the maleimide hydroxyl functionality with adipoyl chloride to form an ester linkage, and
the terminal acid chloride functionality is reactive toward subsequent deposition of
another MVE polymer layer. The vinyl ether isopropylphosphonates remain protected
during layer growth. Once layer growth is complete, the isopropylphosphonates are
deprotected using bromotrimethylsilane (BTMS) to activate the multilayer for metal ion
sorption. The uptake of Zr4+ by the multilayer is confirmed by ellipsometry, FI‘IR, X-ray
photoelectron spectroscopy (XPS) and quartz crystal microbalance (QCM) gravimetry.
The XPS data show ~ 13 atom % of the multilayer is Zr4+ after exposure to a 5 mM
ethanolic solution of zirconyl chloride, and QCM data show that metal ion uptake is fast.
The films also show changes in FTIR spectra and ellipsometric thickness, indicating

substantial structural changes associated with metal ion uptake.

-l63-

 

7.1 Introduction

Gaining the ability to design and construct interfaces with significant control over
their macroscopic properties has attracted substantial research interest. This effort has
been aimed largely at optimizing molecular-level organization, with Langmuir-Blodgett
films, alkanethiol-gold self-assembled monolayers (SAMs)“4 and metal phosphonate
multilayers}7 being the most widely studied examples. The array of chemical
functionalities that can be introduced into the layers, as well as the range of assembly and
interlayer chemistry that can be accessed affords great versatility in the properties of the
resulting materials.

The promise of interfacial chemistry lies in its potential application in areas such

8,9 11

as bio-sensors/bio-recognition, optical second harmonic generation)“ chemical

sensors”’13 and separations.'4"5 The ability to assemble thin films with molecular-scale
control over layer thickness and uniformity necessitates the use of well-defined and well-
controlled reaction schemes. We have reported previously on the growth of layered
polymeric interfaces using maleimide-vinyl ether (MVE) chemistry, where Zr-
bisphosphonate (ZP) interlayer linking chemistry is used to achieve layer-by-layer
growth.‘6 In that work, the identity of the maleimide substituent determines the properties
of the resulting multilayer interface.

We report here on a different use of MVE alternating copolymers in the
construction of multilayer assemblies. Our present focus is on the design, synthesis and
characterization of a covalently-bonded multilayer structure where the polymer vinyl

ether side groups are free to interact strongly with any metal ions they come into contact

with. We construct the MVE copolymer using 4-hydroxyphenyl maleimide and

-l64-

 

ethylvinylether-2-diisopropylphophonate. The unique design aspect of this work is that
we connect the polymer layers through the 4-hydroxyphenylmaleirnide side groups
instead of using the well-established metal ion coordination to the phosphonate side
groups. We use adipoyl chloride to form covalent diester interlayer linkages.

In contrast to the conventional ZP interlayer linking chemistry, where the
phosphonates are sequestered by the formation of metal-bisphosphonate complexes, we
use the sidegroups in these polymer multilayers in their deprotected form as active sites
for metal ion uptake, and report here on the interaction between deprotected polymer
multilayers and Zr“. To circumvent the formation of phosphoesters during layer growth,
the phosphonate groups of the copolymer are protected by isopropyl functionalities.
Even if these linkages form to some extent, the deprotecting agent bromotrimethylsilane
(BTMS), which we use to activate the multilayers, has been shown to de-esterify
phosphoesters.l7 After the desired number of deposition cycles, four in this case, the
phosphonate groups are deprotected using BTMS. We have chosen to demonstrate this
chemistry using four layers as a matter of convenience. We have grown thicker layers
and see no evidence to indicate a loss of reactivity beyond four layers. Deprotection
results in the facile hydrolysis of the phosphonate groups, rendering a multilayer film
capable of efficient metal ion uptake.

Exposure of the deprotected multilayer to ZrOCl2 solution results in rapid metal
ion uptake. Ellipsometry, FTIR and X-ray photoelectron spectroscopy (XPS) data each
reveal significant changes in the multilayer structure after sorption of Zr“. XPS
measurements show that ~13 atom % of the surface, is comprised of Zr“, consistent with

the concentration of Zr“ we would expect for saturation of the multilayer assembly.

-165-

These data point to essentially irreversible uptake of the metal ion. XPS data on systems
exposed to the ZrOC12 solution for extended periods of time yielded the same
concentration as those exposed for short times. Quartz crystal rrricrobalance (QCM) data
demonstrate the uptake of the Zr“ is fast, with a steady state loading density being
achieved within seconds of exposure to metal ion solution. We report here on the design,
synthesis and characterization of these multilayer assemblies that act as metal ion

“sponges”.

7.2 Experimental

Syntheses: The synthesis of maleimide-vinyl ether alternating copolymers has

“5'18 In this work, we use 4-hydroxyphenylmaleimide as one

been described before.
monomer. It is synthesized from maleic anhydride and 4-hydroxyaniline and the reaction
chemistry is the same as that described for similar, substituted maleimides.16 The vinyl
ether phosphonate was prepared by reacting tri(isopropyl)phosphite with 2-
chloroethylvinylether according to a method described by Rabinowitz.l9 This reaction
was carried out under an inert argon atmosphere and refluxed at 170°C for five days. The
maleimide-vinyl ether alternating copolymer was prepared by radical copolymerization

using azobisisobutyronitrile (AIBN) as the initiator (scheme 7.1a).'8‘20 We observe no

homopolymerization of the monomers used in this work.

-l66-

 

 

0 o
“ AIBN/CHC13
24 Hrs, 60°C
iPrO/P
OH OiOPr
o } BTMS/CHgCN
2 Hrs, 45°C
/P\‘\‘0 P\=o
IPrO OiPr HO 0H
CH1 (”1

Scheme 7.1.

(a) Monomers used in the synthesis of the MVE alternating copolymer
reported here. (b) Deprotection (activation) chemistry for the MVE copolymer. In this

scheme, the isopropyl groups are hydrolyzed subsequent to layer assembly.

-l67-

 

Polymer Layer Deposition: Quartz and gold substrates, including QCMs, were
cleaned using piranha solution (3:1 H2SO4:H2O2) and rinsed with ethanol, then with
distilled water prior to layer deposition. The gold substrates were first exposed to a 10
mM solution of 6-mercapto-1-hexanol in a 60:40 ethanol:water solution for 1 hour at
45°C. Quartz substrates were exposed to a 5 M HCl solution for 30 minutes. Both
substrates were rinsed with ethanol and water, then dried under a stream of nitrogen and
exposed to adipoyl chloride in anhydrous acetonitrile under an argon atmosphere. The
substrates were then removed and rinsed with ethyl acetate, dried under a stream of
nitrogen and exposed to a 10 mM solution of the MVE polymer in DMSO for 1 hour at
45°C. The resulting polymer layer was reacted with adipoyl chloride, then immersed in
the polymer solution and the process repeated as required to deposit the desired number
of layers (scheme 7.2).

Deprotection/hydrolysis of the diisopropylphosphonate groups: Once the
multilayer synthesis is completed, the polymer layers, which contain
diisopropylphosphonate side groups, were exposed to bromotrimethylsilane (BTMS) in
anhydrous acetonitrile under an argon atmosphere for two hours. The substrates were
then removed from the BTMS solution and immersed in acetonitrile for 30 minutes,
rinsed with ethyl acetate and dried under a stream of nitrogen. After rinsing the
substrates, ellipsometric measurements were taken, followed by the exposure of the

substrates to a 5 mM ethanolic ZrOC12 solution for at least one hour.

-168-

 

o o
.. 6%
fl Q:~r\/\O
o/\/P.. 0.. o
I;
O O O
N
19.8319
0 o
.H mm o
R IV..\/\
o/\/p...6 0..
I, o
0 0 0

\/\/>\3

H0

 

 

Idealized schematic of the MVE polym system showing two layers
-169-

connected by a diester linkage.

Scheme 7.2.

Optical Null Ellipsometry: Ellipsometric thickness measurements of the layers
deposited on gold were made with a Rudolph Auto-EL H optical null ellipsometer
operating at 632.8 nm. Rudolph DAFIBM software was used for data collection and
processing. For all films, the refractive index was taken to be n = 1.54 + 0i.

Infrared Spectroscopy: For the multilayer assemblies on gold, FTIR spectra were
acquired following each individual deposition cycle. IR spectra were also collected prior
to multilayer reaction with BTMS and ZrOCl2. For all measurements, a Nicolet Magna
750 FT IR spectrometer was used, and for all data presented here, the instrumental
resolution was set to 4 cm". The data were acquired using an external reflectance sample
mount, set to an incidence angle of 80°.

UV—Visible Spectrophotometry: UV-visible absorption data were acquired for
samples grown on quartz substrates. Spectra were acquired after each deposition cycle
using a Cary 300 UV-visible spectrophotometer. The wavelength range scanned was 190
nm - 500 nm, at a scan rate of 600 nm/min. The collected data were plotted using
Microcal Origin 6.0® software.

X-Ray Photoelectron Spectroscopy (XPS): XPS data were acquired using a
Perkin-Elmer Physical Electronics PHI 5400 X-ray photoelectron spectrometer. This
system was equipped with a Mg X-ray source operated at 300 W (15kV, 20 mA). The
carbon (Cls) line at 284.6 eV was used as a reference in determining the binding energies
of the various species.

Quartz Crystal Microbalance (QCM): Time-resolved solution phase QCM data
were acquired using a Hewlett-Packard 53131A 225 MHz universal frequency counter

with data acquisition programmed using National Instruments LabVIEW® programming

-170-

 

language. The in-house built QCM holder was connected to a resonant oscillator circuit
(Maxtek, Inc.) powered by a 5V DC supply. A Neslab RTE-111 refrigerated circulating
water bath was used for temperature control at 303 K 1 0.01 K. The mounted QCM is
immersed in a jacketed beaker containing 100 mL of mixed hexanes at 30°C. The
solution is stirred vigorously. Once a baseline oscillation frequency is established, 1.0
mL of a 5 mM solution of zirconyl chloride in 60:40 ethanol:water is introduced. Data
collection is terminated after a new steady-state frequency had been established, typically

within a minute.

7.3 Results and Discussion

The focus of this work is on a novel covalent growth strategy for MVE polymers
that allows their use as efficient metal ion sorbent materials. In this work, the ZP
chemistry usually employed in the connection of layers is used instead as a chemically
reactive site for metal ion complexation. The novel aspect of the work we report here lies
in our ability to assemble these systems in a discrete manner such that the phosphonate
groups are available at a later time for metal ion sequestration. The maleirrride monomer
we use in polymer synthesis possesses a terminal hydroxyl functionality which we react
with adipoyl chloride to form interlayer linkages. Because we use a protected
phosphonate-containing vinyl ether, formation of a phosphoester functionality does not
compete efficiently with the formation of the diester. With this chemistry, the formation
of porous multilayers is facile and subsequent reaction of the protected phosphonates
with BTMS results in essentially complete deprotection (Scheme 7.1b). Several groups

have explored the use of covalent ester and amide interlayer linking chemistry“24 for the

-l71-

 

growth of thin films. The method we report here uses diester formation for layer
adhesion, resulting in rapid growth of robust layers. Using this chemistry, the assembly
of a four-layered system takes about eight hours.

Optical null ellipsometry and FTIR and UV-visible spectroscopies all confirm the
deposition of the polymer layers. The ellipsometry data (Figure 7.1) demonstrate a linear
dependence of interface thickness on number of layers, with a slope of 16 1 1 A per
adipoyl chloride-MVE polymer layer growth cycle. Each ellipsometric data point is the
average of 40 individual determinations at different locations on the sample surface.
Semi-empirical calculations for the layer unit suggest a thickness of 24 A for the fully
extended structure. The fact that we recover 16 Allayer experimentally, which is the
same as seen for layers of phenyl-substituted maleimide-containing polymer,16 is likely
the result of the substantial lack of order in these layers. The ellipsometric measurements
represent the average of the thickness distribution for the sampled area, and there is likely
heterogeneity over length scales smaller than those sampled by the ellipsometric
experiment. We have also used FT IR to monitor the layer-by-layer deposition of the
polymer, and present these data in Figure 7.2a. We monitor the carbonyl stretching
region between 1770 cm”I and 1650 cm'1 because both the maleirrride monomer and
diester interlayer linkages contain these functionalities. Unfortunately, these bands are
not resolved. The C=O absorbance depends linearly on the number of layers deposited

(Figure 7.2b).

-l72-

 

8

   
   

slope = 16 1 l A/layer

\)
O
' 1

interoept=41lA

ellipsometric thickness (A)
8 8 8 ES 8

 

 

10-
0 1 1 1 1
0 l 2 3 4

number of layers

Figure 7.1. Ellipsometric data for a four-layered MVE polymer system. Each layer
consists of the MVE polymer and the adipoyl chloride underlayer. The slope of the best
fit line through the data is 1611 Allayer.

-173-

 

0.010

I

Layer 3

Layw
m
0.000 -

4

0.005

Absorbance (a.u.)

 

 

 

 

3500 3250 3000 270,00 1750 1500 1250 1000

frequency (cm’l)

 

 

0.004 1- b
0.003 -
3?
a
Q)
g 0.002 ~
:6
e
O
a
0.001 -
0.000 1 . 1 1
0 1 2 3 4

number of layers

Figure 7.2 (a) FTIR absorbance data layers 1-4. The spectra are offset for clarity of
presentation. (b) Carbonyl stretching band absorbance as a function of number of layers.

-174-

 

The UV-visible spectroscopic data were collected for each layer deposited and
they also yielded a linear response per layer (Figure 7.3). The dominant absorption band
is associated with the hydroxyphenyl succininride moiety in the polymer backbone and its
maximum is centered at 231 nm. We are not aware of any reports of the extinction
coefficient for this chromophore, precluding our ability to estimate the loading density of
the polymer layers on the surface.

We use both ellipsometry and FT IR to monitor the deprotection/hydrolysis of the
isopropylphosphonate groups. After the deprotection reaction, the ellipsometric
thickness of the four-layer interface decreased from 83 1 6 A to 44 1 1A (the
uncertainties are 110). We believe that this decrease in thickness is the result of
replacing the relatively bulky isopropyl groups with the smaller hydroxyl groups. This
reduction in thickness can be accounted for only in part by the change in physical size of
the phosphonate terminal functionalities. For the deprotected phosphonic acid
functionalities, the hydroxyl groups are capable of participating in hydrogen bonding.
Hydrogen bonding between the hydroxyl groups would naturally result in a more
compact structure, leading to a decrease in the thickness, as we observe. A key issue
relative to the structural freedom within the polymer layers is the ability of the
deprotected phosphonate groups to come into close proximity with phosphonate groups
from adjacent layers. This is because of the ability of deprotected phosphonates to H-
bond. Interlayer H-bonding can affect the thickness of the films substantially, while not
affecting the availability of the phosphonate groups for metal ion complexation. The
dramatic change in thickness upon deprotection argues for significant structural freedom

within the polymer matrix.

-175-

fl“_‘7‘“—j

 

0.020 »

 

 

 

 

 

 

 

0.035 .
E 0015-
C:
0.030 ,7, ’
8 0.010-
< .
0.025
:5
3 0.020
0 .
5 0.015
0 .
3 0.010
<1
0.005
0.000
-0'm5 I I I I I I L I I A I I I I L 4 J
200 225 250 275 300 325 350 375 400
Wavelength (nm)

Figure 7.3. UV—Visible absorbance spectra of MVE polymer layers as a function of
number of layers. Inset: Dependence of the 231 nm absorption intensity as a function of
number of layers.

The FTIR data also confirm the deprotection/hydrolysis reaction based on the

significant changes seen in the organophosphorus spectral region (Figure 4a).

 

Unambiguous assignment of the features in this spectral region is not possible due to
extensive overlap in the ~1415 cm’l - 1085 cm'1 region, but it is known that the position

of these resonances is sensitive to the nature of the substituents attached to these groups.

-l76-

A significant difference is observed in the spectrum after deprotection and exposure to
the ionic solution (Figure 7.4). The -OH stretching region at ~3600 cm'l exhibits a

4
pronounced change upon exposure to Zr +.

 

 

 

0.015 -
1
0.010 -
32‘
:3
Q)
g 0
~ .005 -
8
E
< ”A
0.000 fi/A‘“

 

I I I III I I I
3500 3000 2000 1500 1000

fiequency(cn15

Figure 7.4. FTIR absorbance data for a four-layered system prior to deptrotection of the
phosphonate groups with BTMS (spectrum I), after deprotection with BTMS followed by
hydrolysis (spectrum H), and after exposure to ZrOCl2 (spectrum H1). The spectra are
offset for clarity of presentation.

The interfacial assembly we report here changes substantially upon exposure to
Zr“. After immersion in a 5 mM ethanolic ZrOCl2 solution, the polymer multilayer
increases to 137 1 23 A, an increase of more than three times the original thickness of the
deprotected polymer film suggesting substantial structural disruption upon metal ion
sequestration. This increase in thickness resulting from Zr“ uptake is likely related to

the incorporation of non-stoichiometric waters of hydration in the multilayer matrix. The

FTIR spectrum of the deprotected and zirconated polymer multilayer is shown in Figure

-l77-

 

7.5. After rinsing the zirconated multilayer with 100% ethanol, then with distilled water,
we see a decrease in the thickness of the interface to 110 A, which we attribute to the
removal of non-stoichiometric waters of hydration incorporated into the multilayer
structure during Zr“+ uptake. An FTIR spectrum taken after the ethanol/water rinse
confirms the loss of water, as seen by the decrease in the ~3400 cm'1 and ~1500 — 1600
cm'l regions (Figure 7.5). Once this decrease has occurred, however, no further thickness
change was observed, even after heating the interfacial assembly at 50°C in 100% ethanol
for one hour. FTIR spectra taken before and after heating in 100% ethanol were
idenficaL

One issue of potential concern is whether or not the polymer swells upon
exposure to ethanolic solution. We measured the ellipsometric thickness before and after
immersion of the sample in a 60:40 ethanol:water solution and observed no thickness
change to within the experimental uncertainty, indicating that these polymers do not
swell upon exposure to the solvent system when metal ions are not present. When these
same substrates are exposed to a solution of ZrOCl2 in 60:40 ethanol:water, the 100 A
thickness increase for the four layer interface is reproduced readily.

Examination of the interface FTIR spectrum after Zr“ uptake revealed significant
changes in the hydroxyl stretch region of the spectrum. One significant change is the

shift in the peak maximum from ~3200 cm'1 to ~3400 cm’1 (Figure 7.5) with processing.

-l78-

 

I—

0.008

0.006

0.004

absorbance (a.u.)

 

0.002

 

0.000 " 1 I 1 JII . 1 I 1
3500 3000 2600 1500 1000

frecuency (cm")

Figure 7.5. Comparison of the four-layered system, after exposure to ZrOCl2 (spectrum
III), and after rinsing with absolute ethanol, then water (spectrum Iv). These spectra are
offset for clarity of presentation.

The reason for the prominence of this band in spectrum III of Figure 7.5 is the
presence of liquid water in the polymer multilayer subsequent to exposure to metal ions.
When rinsed with ethanol, we observe a decrease in this peak (spectrum IV, Figure 7.5),
coincident with a decrease in the ellipsometric thickness, consistent with the removal of
non-stoichiometric water from the polymer film. After heating to 50°C in 100% ethanol

for 1 hour, no further change was observed in the FTIR spectrum, again consistent with

the ellipsometric data.

-179-

We have studied the uptake of Zr“ in the polymer multilayers using both XPS
and quartz crystal microbalance gravimetry. We consider the XPS data first and present
a survey scan of a zirconated four layer assembly in Figure 7.6. For the area sampled,
approximately 12.9 atom % of the interface is Zr“.25 This finding suggests that most of
the Zr“ is complexed by one phosphonate group. XPS data also confirm the presence of
C1 in the multilayer assembly in some of the samples, and the meaning of its presence is
not completely clear. We offer two possible explanations for this observation. First, the
presence of Cl may be consistent with the maintenance of macroscopic charge neutrality
within the polymer matrix. We note that the observed Cl concentration is not present in a
stoichiometric amount. Second, the Cl could be due to the presence of unreacted adipoyl
chloride used as an interlayer linker. We note that hydrolysis of the acid chloride
functionality is known to be very efficient, making this explanation unlikely. In some
cases, Cl was observed prior to exposure to ZrOCl2 and in other cases no Cl was detected
even after exposure to the metal ion solution, making a conclusive explanation
impossible. We note the apparent absence of a P resonance in these data. XPS is
relatively insensitive to P, however, 3 1P NMR data of polymer multilayers bound to silica
and suspended in solution26 (not shown) reveal the presence of P in these multilayers.
We note that the XPS indication of 1:1 complexation is consistent with the increase in
layer thickness upon metal ion sorption. If 2:1 complexation dominated, it would cross-

link the polymer matrix, precluding an increase in thickness.

-180-

 

 

 

 

70000 C Auger
O Auger
60000r
50000- 015
' N 1
§ 40000- S
8 \ Zr 3p
0 30000- 1 C Is Zr3d
I c1 y
20000- \ P 2p
1 Au 4f
10000-
0 I I I I I I I I I I

1000 800 600 400 200 0

binding energy (eV)

Figure 7.6. XPS survey scan of a surface with four layers of poly(4HPM-
VEP) on a gold substrate.

We consider the QCM data next. This measurement records the change in quartz
crystal microbalance resonant frequency as a function of time. The injection of an
aliquot of Zr“ causes a change in the resonant frequency of a polymer-coated QCM

suspended in solution and we present these data in Figure 7.7. The response of a QCM to

-181-

 

changes in mass is typically quantitative in the gas phase, but not in solution due to the
dielectric properties of the surrounding medium. Our interest in these data is not in terms
of the quantitative mass uptake (this information is available from the XPS data), but
rather in the time-course of the QCM response. The kinetics of mass uptake for these
systems can be understood in terms of the BET adsorption isotherm and thus we can
obtain valuable information on the interaction of the metal ions with the polymer matrix.
We provide the QCM data in this paper only as a demonstration of rapid metal ion uptake
and will treat these data in greater detail in a subsequent paper. We have found that the
uptake of Zr“ and Ca2+ (not shown) is very fast for this polymer multilayer structure.
We note that, among the complexities associated with extracting quantitative information
from these data is the change in the thickness of the polymer layer as a result of metal ion
complexation. Detailed modeling of this effect will need to be considered before useful

kinetic information can be extracted from the data.

-182-

-1m 1-

Af (Hz)

-150 _

 

 

 

time (sec)

Figure 7.7. QCM data showing rapid uptake of Zr“ in a four-layer film.

7.4 Conclusions

We have reported on two key points in this paper. First, we have demonstrated

facile covalent layer-by-layer growth for MVE alternating copolymers. Second, we have

-183-

demonstrated that these materials can be used as efficient metal ion sorbents based on the
XPS loading data and the QCM kinetic data. We provide ellipsometric, spectroscopic
(FTIR, UV-visible and XPS) and gravimetric (QCM) data to support facile covalent layer
construction and metal ion uptake. The ellipsometric, FTIR and UV-visible
spectroscopic data demonstrate linear layer growth and the XPS and QCM data provide
the mass and kinetic uptake information for Zr“ sorption from solution. The XPS are
consistent with predominantly 1:1 P03: : Zr“ complexation, with the presence of Cl
supporting this assertion.

We observe a large ellipsometric thickness increase upon uptake of Zr“ and
speculate that the basis for this thickness change is (1) structural rearrangement within
the polymer matrix to accommodate the presence of the metal ions and (2) the presence
of nonstoichiometric water that is somehow associated with the incorporation of the
metal ion. These data are not consistent with polymer swelling upon exposure to the
solvent system we use. A substantial fraction of the thickness change can be eliminated
by washing of the metal-containing polymer multilayer with ethanol, supporting the
assertion that nonstoichiometric water is associated with the initial thickness increase on
exposure to metal ions.

Several issues remain to be explored for these polymer multilayers. One
important consideration is that, for four polymer layers, the polymer matrix appears to be
sufficiently open to allow ready access by metal ions. How the polymer matrix
permeability will change with the growth of additional layers will ultimately determine
the use of these materials for large-scale rapid metal ion sorption applications. A second

vital issue that needs to be addressed is a detailed understanding of the metal ion uptake

-184-

 

kinetics. Gaining this understanding will be complicated by the changes in the polymer
matrix associated with the incorporation of the metal ions. Understanding the forward
and reverse kinetic processes will allow us to determine the thermodynamic properties of
these layers and thereby understand the extent to which metal ion uptake is reversible.

These issues are under investigation.

7.5 Literature Cited

1. Ulman, A. Chem. Rev. 1996, 96, 1533.

2. Dubois, L. H.; Nuzzo, R. G. Annu. Rev. Phys. Chem. 1992, 43, 437.

3. Zhang, Z. J.; Hu, R. S.; Liu, Z. P.; Langmuir 2000, I6, 1158.

4. Skulason, H.; Frisbie, C. D. Langmuir 1998, 14, 5834.

5. Lee, H., Kepley, L. J., Hong, H. G., Akhter, S., Mallouk, T. E. J. Phys. Chem. 1988,
92, 2597.

6. Lee, H., Kepley, L. J., Hong, H. G., Mallouk, T. E. J. Am. Chem. Soc., 1988, 110, 618.

-185-

7. Putvinski, T. M.; Schilling, M. L.; Katz, H. E.; Chidsey, C. E. D.; Mujsce, A. M.;
Emerson, A. B.; Langmuir 1990, 6, 1567.

8. Yoon, H. C.; Kim, H-S. Anal. Chem. 2000, 72, 922.

9. Delamarche, E.; Sundarababu, G.; Biebuyck, H.; Michel, B.; Gerber, C.; Sigrist, H.;
Wolf, H,; Ringsdorf, H.; Xanthopoulos, N.; Mathieu, H. J .; Langmuir, 1996, 12,
1997.

10. Katz, H. E.; Scheller, G.; Putvinski, T. M.; Schilling, M. L.; Wilson, W. L.; Chidsey,

 

C. E. D.; Science, 1991, 254, 1485.
11. Li, D.; Ratner, M. A.; Marks, T. J.; Zhang, C. H.; Yang, J.; Wong, G. K.; J. Am. 1
Chem. Soc. 1990, 112 7389.
12. Brousseau, 111, L. C.; Aurentz, D. J.; Benesi, A. J.; Mallouk, T. E. Anal. Chem. 1997,
69, 688.
13. Brousseau, 111, L. C.; Mallouk, T. E. Anal. Chem. 1997, 69, 679.
14. Vrancken, K. C.; VanderVoort, P.; Gillisdhamers, I.; VanSant, E. F.; Grobet, P.; J.
Chem. Soc., Faraday Trans. 1992, 88, 3197.
15. Pfleiderer, B.; Albert, K.; Bayer, E. J. Chromatogr. 1990, 506, 343.
16. Kohli, P.; Blanchard, G. J. Langmuir, 1999, 15, 1418.
17. Chouinard, P. M.; Bartlett, P. A. J. Org. Chem. 1986,51, 75.
18. Kohli, P.; Scranton, A. B.; Blanchard, G. J.; Macromolecules 1998, 31, 5681.
19. Rabinowitz, R. J. Org. Chem. 1961, 26,5152.
20. Olson, K. G.; Butler, G. B.; Macromolecules 1984, I 7, 2480.
21. Kohli, P.; Blanchard, G. J .; Langmuir 2000, 16,4655.

22. Beyer, C.; Bohanon, T. M.; Knoll, W.; Ringsdorf, H.; Langmuir 1996, 12, 2514.

-l86-

23. Duevel, R. V.; Corn, R. M. Anal. Chem. 1992, 64, 337.

24. Van Ryswyk, H. B.; Turtle, E. D.; Watson-Clark, R.; Tanzer, T. A.; Herman, T. K.;
Chong, P. Y.; Waller, P. J.; Taurog, A. L.; Wagner, C. E.; Langmuir 1996, 12, 6143.

25. The concentration of Zr“ was calculated using the raw experimental data and the
sensitivity factors for the elements found. The corrected data indicate that 12.9% of
the atoms detected by XPS are Zr“ in the four-layer films.

26. Kohli, P.; Blanchard, G. J .; Langmuir, 2000, 16, 695.

-l87-

Chapter 8

Conclusions and Future Prognosis

The work presented in this dissertation has focused mainly on designing
materials and new assembly strategies, where we are able to manipulate material
property in a controlled manner. We have demonstrated three distinct levels of
control over these materials. First, we have demonstrated our ability to control the
identity of the functional groups incorporated into the matrix of the material by
simply varying the substituted aniline used to prepare the maleimides. Second, by
choosing the appropriate vinyl or vinyl ether co-monomer, we have been able to
demonstrate several simple covalent routes to multilayer assembly. The third level of
control demonstrated in this work is the ability to control the loading density of these
materials at the interface. This level of control is realized through simple acid
catalysis and/or dehydration of the various covalent interlayer linking reactions. The
ability to control the loading density of the functional moieties at the interface can
prove useful in the design of materials that can be used as sensors and in conductive
interfaces.

With these levels of control demonstrated, we next focused on the design of
specific interfaces where we attempt to control the adsorption/desorption behavior of
the interface. In this body of work, we attempt to design interfaces where there is a
“built-in” chemical potential gradient. Here, we create a gradient in the dipole
moment of the side-groups incorporated into the layered polymer matrix, where the

layer-by-layer chemistry employed to assemble these materials allows the

188

 

introduction of different moieties in each layer. Subsequent to polymer multilayer
assembly, we measure the vapor phase adsorption isotherm for the interfaces in the
presence of methanol, acetone and hexane. To determine the adsorption isotherms,
we employ an apparatus designed and assembled in the Blanchard lab that allows the
simultaneous acquisition of adlayer mass and thickness change as a function of
adsorbate vapor pressure. These data in concert allows us to determine the energetics
of adsorbate-adsorbent interaction, monolayer coverage and specific surface area. It
should be noted that our inability to determine the specific surface area for our
systems speaks directly to the porous/disordered nature of our interfaces.

We have chosen to fit our adsorption data to the BET equation and recovered
values consistent with physisorption of the adsorbate molecules on the interfaces. In
all cases, the adsorption was reversible by simply shutting off the adsorbate vapor
supply to the interface. We have shown from fits of our data to the BET isotherm
model that we do have the ability to control the adsorption/desorption properties of
the interfaces that we prepare (Chapter 6). This control was realized when we (1)
varied the side group functionalities in the layers and (2) when we varied the order of
layer deposition for our mixed polymer interfaces. Additionally, we also
demonstrated our ability to design interfaces for specific purpose. One example is the
assembly of 4-hydroxyphenylmaleimide-co-ethyl diisopropylphosphonate vinyl ether
into multilayers, where the layer assembly is accomplished though covalent linkages.
Once we have the layers assembled, we remove the isopropyl groups and use the

interface to sequester metal ions from solution.

189

8.2 Future Prognosis

The work we present in this dissertation serves as an opening to a variety of
new areas that will be under further exploration within the Blanchard group. With
respect to the preparation of interfaces characterized by a chemical potential gradient,
the next logical step is the assembly of more complex structures and the measurement
of their adsorption isotherms. The aim of this next step is the assembly of three and
four layered stacks and varying the position of the different polymer layers. One key
issue with respect to design, is how best to assemble the layers to create an interface
that behaves in a predetermined manner. We want to determine, in a more qualitative
way the effects of the dipole moment gradient on the adsorption properties of the
interfaces. It should be noted that, while we represent the gradients as being one
continuous profile, this is not the case due to the fact that the dipole moments of the
side-groups can be significantly different from layer to layer and intralayer dipole-
dipole interactions remain to be accounted for. The difference in the dipole moments
of the side-groups will likely lead to a stepwise variation, and the real issue now
becomes how large or small do these steps need to be before noticeable change or
control is realized. While it may seem rather intuitive that the larger the difference
between the dipole moment, the greater the effect should be, this may not necessarily
follow. If the steps are significantly large, then the isotherms may actually become
very complex as we may begin to observe mixed mechanisms (e.g. absorption,
partitioning and/or capillary condensation) due to the fact that the layers begin to
behave as individual entities within the matrix. To this end, a more smooth and

regular decrease may lead to a more uniform gradient profile. Before moving on to

190

ram:

further studies, the first step should be to estimate the dipole moments of the side
groups using semi-empirical calculations. Based on the results of these calculations,
interfaces can then be designed to vary the slopes of the gradient profiles and to
determine the best profile to optimize interfacial behavior in a predetermined manner.

A second area that needs to be explored is the use of a more extensive library
of adsorbates to determine the isotherms. With a body of data that allows us to
determine how best to assemble the interface to achieve a desired behavior, and
having information that tells us how these interfaces respond to specific adsorbate
molecules, we can begin to design specific separation systems. If we begin to think
about designing chromatographic stationary phases, for instance, we can optimize the
system in a quick and easy manner by performing these types of measurements rather
than having to pack columns and optimize the separations. While these isotherm
determinations are gas-solid measurements, it is also possible to perform liquid-solid
measurements as well. In this scheme, a technique such as surface plasmon
resonance (SPR) can be used to acquire the isotherm of liquid phase adsorption at the
interface. Here, it is possible to determine mass transport of bulk liquid adsorbates in
a manner similar to QCM measurement.

Another issue worth exploring is surface organization as a function of adlayer
density. In chapter 4 we noted the acid enhanced deposition of polymer materials
through acid catalysis and/or dehydration. In order to get a better understanding of
these interfaces it would be useful to determine how these materials
assemble/organize on the surface. To perform these measurements, a technique such

as AFM or SEM can be used. From these measurements, we would be able to

191

 

determine the formation (or not) of islands on the surface. Additionally,
electrochemical measurements, as well as isotherm measurements could prove useful,
where we can determine the role of the adlayer in mass transport of an electroactive
of vapor phase species to the interface. These types of measurements would allow

for the design and optimization of the interfaces with specific functions.

 

192

   

l11111111111I