4 f"‘!

L391:

.

(Id

.$«

1w.
.3) 5..
s. u
1. y: 2.0: 1!;
iv \Jlin, ~!|.J x If“: )\
. $30. .«guu‘ 20...»:

‘1'.-
‘ mamas.
5..

.1 “0.“.
$2.41....

 

 

 

THESIS

25L \

This is to certify that the
dissertation entitled

Design, Synthesis, and Characterization

of Uitra-Thin, Robust Films with
Molecuiar Contggi
y

presente

Punit Kohii

has been accepted towards fulfillment
of the requirements for

Ph.D. degree in CIIQIIII'SLLL.

 

Major professor

Date (9/7//0

MS U i: an Affirmative Action/Equal Opportunity Institution 0-12771

 

Ar

“LIBRARY
Mlchigan State
. University

 

 

 

.__._..

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

 

DATE DUE DATE DUE DATE DUE

 

rq,‘ ‘flg.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

moo mam-peep.“

F..."— .—

Design, Synthesis, and Characterization of Ultra-Thin

Robust Films with Molecular Control
By

Punit Kohli

A DISSERTATION

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

DOCTOR OF PHILOSOPHY
Department of Chemistry

2000

ABSTRACT

Design, Synthesis and Characterization of Ultra-Thin Robust Films
with Molecular Control

By

Punit Kohli

The work reported in this dissertation has demonstrated new ways to design,
synthesize and characterize robust thin films with controlled molecular dimensions. The
central point of this work is to provide the fundamental knowledge and understanding of
layered-material assemblies to advance the field. We are especially interested in
controlled layer-by-layer multilayer assemblies in which the properties of these films
such as film thickness, linear and non-linear optical response, electrical and electronic
behavior, and porosity can be controlled. Ultimately, these ultra-thin films may find
applications in the areas of controlled released delivery systems, nonlinear optical
devices, chemical interfaces and interfacial sciences, separation science, nanoelectronics,
biocatalysis, and biotechnology. There are some requirements for successful films and
coatings. For example, the films must be stable in their surrounding environment in
which are used. Furthermore, the control of the properties of the films is also an
important issue. Keeping these requirements in mind, we used maleimide-vinyl ether
(MVE) copolymers to deposit coatings where we have ability to change the pendant
groups on the maleimide monomers. We used MVE polymers because they form strictly
alternating polymers and their structure and properties are well studied by others. They

also possess excellent thermal and chemical resistant. Since changing the pendant groups

on the succinimide groups gives us a great ability to control various properties makes
MVE copolymers ideal candidates. We have used zirconium phosphonate (ZP) interlayer
linking chemistry which is robust and allows for exquisite control over layered material
assembly. Thus the use of polymer chains and ZP complexation led to quasi-2-
dimensional films where we can control the properties of each layer of the films. These
films are found to be robust to both chemical and thermal treatments. Although these
coatings may find uses in various applications such as permeable films, dielectric
coatings, chemical sensors etc. However due to highly polarizable ZP linkages, the
interlayer excitation-transport is inefficient. The solution to this problem is to use
multilayer assemblies where the interlayer bonding is covalent since the polarizability in
that case would be same as that of rest of the molecules which is covalent organic
molecules. We have demonstrated the formation of multilayers in which the interlayer
moiety is urea. The covalent interlayer bonding and hydrogen-bonding in the
perpendicular direction to covalent bonding makes these layers robust to thermal and
chemical attack. Furthermore, we have also demonstrated that it is possible to use the
combination of these chemistries to produce multilayers where the interlayer linking
chemistry can be changed between ionic and covalent in a well-controlled manner. These
types of assemblies may give properties which may be difficult to obtain using solely

either ionic or covalent chemistry.

“life is good "

iv

Dedicated to My Family, and
In The Memory of My Father, The Late Shri Raghunandan Lal Kohli

Acknowledgments

I firmly believe that personal relations between students and their advisors are as
equally important as the scientific endeavors during graduate studies. I feel very
fortunate to have Dr. Gary Blanchard as my advisor. He has always been there whenever
I needed his help. “Help” in my case included not only his advice about science but also
in the real life. He also gave me “free” hand and encouraged me to carry out new ideas,
sometimes weird ones, that came to my mind.

I would also like to thank my committee members Dr. Alec Scranton, Dr. Gregory
Baker, and Dr. Simon Garrett for their advices during last 3 1/2 years. Particularly, Greg
and his group members help me in synthesis and let me use his lab and equipments
whenever I needed. I also enjoyed his group meetings.

For last three years, I have spent about 12 hours a day or so in the lab and the
people who helped me through these long period of work are the Blanchard group
members. I would therefore like to thank all my group members: Stevo, Joe, Lee,
J aycoda, Scott, Wendy, John, Michelle, Mark, and Shawn along with many past members
of our group. Especially, I had good company of Steve, Joe, and Lee and we had also
great adventures outside of the lab.

Finally, this work would never have been possible without the support of my
family. They have always encouraged me to do whatever I want to do and have
supported both in good and bad times. To my late father who instilled in me the virtues
of hardwork, honesty, and bravery, I thank you from the bottom of my heart and will

forever remember you. I will always remember you all.

vi

Table of Contents

Page
List of Tables ........................................................................................ ix
List of Figures ........................................................................................ x
List of Schemes .................................................................................... xvi
Chapter 1. Introduction .......................................................................... 1
1.] Literature Cited ............................................................................. 14

Chapter 2. Co-Polymerization of Maleimides and Vinyl Ethers: A Kinetic and

Structural Study .................................................................... 1 5

2.1 Introduction ............................................................... 16
2.2 Experimental ............................................................... 1 8
2.3 Results and Discussion .................................................. 22
2.4 Conclusions ................................................................ 42
2.5 Literature Cited ........................................................... 43
Chapter 3. Design and Growth of Robust Layered Polymer Assemblies with
Molecular Thickness Control ..................................................... 46

3. 1 Introduction ................................................................ 47
3.2 Experimental .............................................................. 48
3.3 Results and Discussion .................................................. 50
3.4 Conclusions ............................................................... 64
3.5 Literature Cited ........................................................... 65

Chapter 4. Probing Interfaces and Surface Reactions of Zirconium—
Phosphate/Phosphonate Multilayers Using 31P NMR Spectrometry .. . ....68

4. 1 Introduction ................................................................ 69
4.2 Experimental .............................................................. 70
4.3 Results and Discussion .................................................. 73
4.4 Conclusions ................................................................. 89
4.5 Literature Cited ........................................................... 90

vii

Chapter 5.

5.1
5.2
5.3
5.4
5.5

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
8.3
8.4
8.5

Growth of Maleimide—Vinyl Ether Copolymer Multilayers. Elucidating
the Balance between Metal Ion Complexation and Side Group

Isomerization ....................................................................... 95
Introduction ................................................................ 96
Experimental .............................................................. 98
Results and Discussion ................................................. 100
Conclusions .............................................................. 1 14
Literature Cited ......................................................... 115

Assembly of Covalently-Coupled Disulfide Multilayers on Gold... . .....l 17

Introduction ............................................................... 1 18
Experimental ............................................................. 1 20
Results and Discussion ................................................. 121

Conclusions .............................................................. 136
Literature Cited .......................................................... 138

Applying Polymer Chemistry to Interfaces: Layer-by-Layer and
Spontaneous Growth of Covalently Bound Multilayers. . . . . . . . . . . . ....144

Introduction .............................................................. 145
Experimental ............................................................. 149
Results and Discussion ................................................. 152
Conclusions .............................................................. 166
Literature Cited ......................................................... 167

Design and Demonstration of Hybrid Multilayer Structures. Layer-by-
Layer Mixed Covalent and Ionic Interlayer Linking Chemistry. . . . . 1 71

Introduction .............................................................. 1 72
Experimental ............................................................. l 73
Results and Discussion ................................................. 176
Conclusions .............................................................. 191
Literature Cited ......................................................... 192

viii

Chapter 9. Future Prognosis ........................................................ 197

ix

Table 2.1.

Table 2.2.

Table 2.3.

Table 4.1

Table 4.2

Table 6.1.

LIST OF TABLES

1H and 13’C chemical shift assignments for NPM/DDVE and

NMM/DDVE ................................................................ 34
1H and 13C chemical shifts of various model compounds at the
succinimide/succinic acid/anhydride ..................................... 36

Physical and thermal properties of various maleimide/vinyl ether

copolymers ................................................................... 41
Summary of 31P NMR resonance assignments reported in the
literature. Spectral shifts are measured relative to 85%
H3PO4 ......................................................................... 76
Summary of 3 IP NMR resonance assignments reported in this work.
Spectral shifts are measured relative to 85%
H3PO4 ......................................................................... 88
Elemental Concentrations ................................................ 124

Figure 2.1.

Figure 2.2.

Figure 2.3

Figure 2.4.

Figure 2.5.

Figure 2.6.

Figure 2.7.

Figure 2.8.

Figure 2.9.

LIST OF FIGURES

Chemical structures of the monomers used in this work ................... 23

IR spectra of BMMH and DVBS monomers. (a) 2800 cm" - 3200 cm"I
and (b) 1500 cm" — 1800 cm" .................................................. 25

(a) Variation of band intensities for the monomer alkene CH
stretching resonances during photocopolymerization of BMMH/DVBS.
Times for each spectrum are t = 0 min., t = 30 min., t = 53 min., and t =
12 hrs., in order of decreasing intensity of the ~3100 cm'l bands. (b)
Time dependence of polymerization for BMMH and DVBS with initial
1:1 monomer stoichiometry .................................................... 26

Polymerization time dependencies of monomers for photoinduced and
thermally induced reactions with initial 1:1 BMMH to DVBS monomer
stoichiometry at T = 22°C. 0 = BMMH conversion, thermally induced
reaction, I: DVBS conversion, thermally induced reaction, A = BMMH
conversion, photoinduced reaction, 8 = DVBS conversion, photoinduced
reaction ............................................................................ 27

FT IR measurement of BMMH monomer loss vs. DVBS monomer loss for (a)
[BMMH]0/[DVBS] 0 = 1 at T = 22°C, and (b) [BMMH] o/[DVBS]0 = 1/2 at T =
22°C ........................................................................................ 29

1H NMR determination of BMMH monomer loss vs. DVBS monomer
loss. For all data, total conversion = 5%. (a) [BMMH]0/[DVBS]0 = 0.19
M/0.38 M = 1/2, T = 30°C, slope = ‘1.0 i 0.01, (b) A =
[BMMH]0/[DVBS]0 = 0.19 M/0.063 M = 3/ 1, T = 90°C, slope = 0.93 i
0.09, (c) o = [BMMH]o/[DVBS]0 = 0.53 M/ 1.00 M = 1/ 1.89, T = 90°C,
slope = 0.78 i 0.08 ............................................................... 31

The IH NMR spectrum of NPM/DDVE copolymer in CDC13. Band
assignments are indicated according to the inset structure ................. 33

The 13 C NMR spectrum of NPM/DDVE copolymer system in CDC13.
Band assignments are indicated according to the inset structure ......... 37

The 13C NMR spectrum of thermally induced NPM/DDVE copolymer in
CDC13. Band assignments are indicated according to the inset structure ....... 38

xi

Figure 2.10.

Figure 2.11.

Figure 3.1.

Figure 3.2.

Figure 3.3.

Figure 3.4.

Figure 3.5.

Figure 3.6.

Figure 3.7.

Figure 4.2.

CP/MAS spectra of BMMH/DVBS copolymer. Bottom: Photoinduced
BMMH/DVBS copolymer, Center: Thermally induced copolymer,
Top: 0 - 90 ppm expanded region of thermal copolymer spectrum. . ....39

TGA curves for NMM/DDVE and BMMH/DVBS copolymers ......... 42

Schematic representation of layered growth of poly(NPM-VEP)
indicating direct priming chemistry of silanol groups with POCl3 and
growth of layered assemblies using partial hydrolysis of the -PO3HR
functionalities to control the available concentration of active PO32' sites
layer by layer ..................................................................... 52

(a) Ellipsometric data for ten layers of poly(NPM-VEP). (b) Ellipsometric data
for eight layers of poly(NBM-VEP). The layer thicknesses are 16 A/layer for
poly(NPM-VEP) and 31 A/layer for poly(NBM-VEP) ............................ 54

Absorbance of poly(NBM-VEP) as a function of number of layers. The
bands at 200 nm and at 260 nm both exhibit a linear dependence of
absorbance on number of layers. For the 260 nm band, the absorbance is
~ 0.01 a.u./layer (inset) ......................................................... 56

Schematics of poly(NPM-VEP) and poly(NBM-VEP). The distances indicated
for the oligomers were determined using molecular mechanics calculations.
The structures presented in the Figure are not energetically optimized and are
intended for illustrative purposes only ............................................... 57

X—ray photoelectron spectroscopic (XPS) data for layers of poly(NPM-
VEP). (a) Survey scan indicating the elements present. (b) Ratio of Zr/P
determined from XPS data as a function of number of layers ............. 59

FTIR spectrum of a bilayer of poly(NPM-VEP) on oxidized Si. Band
assignments are as indicated in the text ....................................... 61

”C NMR spectrum of poly(NPM-VEP) grown on high surface area
porous silica. Resonance assignments are as indicated in the inset. The
absence of discernible progressions in the spectrum indicates the
alternating nature of the polymer. The initial ratio of fully to partly
hydrolyzed phosphonate is ~2 monohydrolyzed to 1 dihydrolyzed group
based on NMR integration of iPr methyl group protons relative to phenyl
ring protons ....................................................................... 62

(a) 29Si MAS NMR spectrum of the silica gel prior to reaction with
POC13 (spinning speed 3.5 kHz). (b) 29Si MAS NMR spectrum of the
same sample following reaction with POC13 (spinning speed 3.5 kHz).
The absence of any features near 5 = -200 ppm indicates single point
binding of the phosphorus oxychloride ....................................... 79

xii

Figure 4.3.

Figure 4.4.

Figure 4.5.

Figure 4.6.

Figure 5.1.

Figure 5.2.

Figure 5.3.

Figure 5.4.

Figure 5.5.

31P MAS NMR spectrum of the phosphated and zirconated surface.
There are distinct but poorly resolved resonances at 6 = 0.6 ppm, -11.8
ppm, —14.8 ppm and -19.1 ppm indicating partial complexation of the
surface as well as several different complexation conditions. The bands

marked with asterisks are spinning sidebands (spinning speed 4.0
kHz) ................................................................................ 80

3'P MAS NMR spectrum of the surface with a single layer of partially
hydrolyzed poly(NPM-VEP). The band marked with an asterisk is a
spinning sideband (spinning speed 4.0 kHz) ................................. 84

(a) 31P MAS NMR spectrum of a monolayer of UBPA on primed
SiOx (spinning speed 4.0 kHz). (b) The same sample that has been
zirconated (spinning speed 4.0 kHz) .......................................... 84

(a) 31P NMR MAS spectrum of a single layer of poly(NPM-VEP) after
hydrolysis and zirconation (spinning speed 4.0 kHz). (b) Spectrum of the
same sample with a second layer of poly(NPM-VEP) added (spinning
speed 4.0 kHz) .................................................................... 86

Ellipsometric thickness of poly(MAB-VEP) layers for seven layers. The
zeroth layer is the primer layer on the oxidized Si substrate. The slope of
the best-fit line through these data is 23.5 i 1.1 A/layer .................. 102

Waterfall plot of absorbance for seven poly(MAB-VEP) layers. The
band positions for the first layer (bottom) are shifted from those for
subsequent layers ............................................................... 103

Total chromophore concentration as a function of number of layers. The
data represent the total of the cis and trans concentrations. The best fit
line yields a density of 4.64 x 10'4 chromophores/cmz-layer ............ 104

(a) Absorbance maximum values for cis (0) and trans (o) conformers
as a fimction of number of poly(MAB-VEP) layers. The cis S; (— So
absorption band is centered near 250 nm and the trans S; <— 30 band is at
~ 313 nm. (b) Ratio of [trans] to [cis] as a function of number of
layers ............................................................................. 107

(a) Absorbance spectra of solution phase poly(MAB-VEP) in
CH3CN. For the native form of the polymer (solid line), the dominant
band is at 315 nm. Irradiation of the polymer solution with broadband
UV light produces an enhanced concentration of cis conformers, as

indicated in the dashed spectrum. (b) Absorbance spectrum of a spin-
cast film of poly(MAB-VEP) ................................................. 108

xiii

Figure 5.6.

Figure 5.7.

Figure 6.1.

Figure 6.2.

Figure 6.3.

Figure 6.4.

Figure 6.5.

Figure 7.1.

Figure 7.2.

Figure 7.3.

Calculated dimensions of single repeat unit for poly(MAB-VEP) for the
trans side group (left) and the cis side group (right) ....................... 109

(a) Absorbance spectra for a four layer stack of poly(MAB-VEP)
irradiated with 254 nm light between 1 hour and 50 hours. The spectra
are offset from one another for clarity of presentation. (b) Ellipsometric
thickness of the same four layer stack as a function of UV
irradiation ........................................................................ 1 1 3

Ellipsometric thicknesses of multilayers of 1,6-hexanedithiol (I), 1,8-
octanedithiol (O) and 1,9-nonanedithiol (A). For all cases, the thickness
depends linearly on the exposure time to dithiol solution. Slopes for these
dependencies are 7 [Maya for C6, 8.9 A/layer for C3 and 10.8 A/layer for
C9 ......................................................................................... 122

XPS data in the 82p region for films. All values are referenced to the
C ls line at 285.0 eV. (a) 7-layer C6 film measured immediately afier
formation, (b) 7 layer C6 film exposed to air for 2 weeks, (c) 10 A thick
C3 film following UV irradiation in an N2 atmosphere .................... 126

FTIR spectra of C6. (a) S=O stretch, (b) CH stretching region as a
function of number of layers. The dashed lines are set at 2921 cm'1 and
2852 cm'1 while the dotted lines are set at 2919 cm’1 and 2850 cm'1 . . . 131

Photolysis experiment showing thickness of C3 multilayer film as a
function of UV exposure time. (a) in N2, (b) in air ....................... 133

Cyclic voltammetry data for reduction of Fe(CN6)3' using a gold
electrode covered with mono- and multilayer assemblies. Curves for
monolayers are solid lines and for multilayers are dashed lines. Data
were recorded at (a) pH 3 and (b) pH 10 .................................... 135

(a) Ellipsometric thickness of bilayer assemblies as a function of number
of growth cycles using the chemistry presented in Scheme I. (b)
Ellipsometric thickness of multilayer assemblies as a firnction of number
of reaction cycles using the chemistry presented in Scheme II .......... 153

X-ray powder diffraction data for bulk poly(MDA-MPI) ........................ 155

Calculated structure of two poly(MDA-MPI) oligomers indicating a helical
structure .................................................................................. 15

xiv

Figure 7.4.

Figure 7.5.

Figure 7.6.

Figure 7.7.

Figure 7.8.

Figure 8.1.

Figure 8.2.

(a) FTIR spectrum of 6-mercapto-1-hexanol adsorbed onto gold prior to
subsequent layer growth. (b) F TIR spectra of multilayers of poly(MDA-
MPI) grown on the primer layer shown in (a). Note the grth of the
bands centered at 3030 cm", indicating the addition of phenyl OH

groups ............................................................................ 160

FTIR spectra of multilayers of poly(MDA-MPI) showing the amide I
and II band regions and higher frequency C=O stretches. See text for a
discussion of band assignments. The spectra, in order of increasing
thickness are for multilayers with ellipsometric thicknesses of 85 A, 127
A, 164 A, 209 A, 310 A and 540 A. The multilayers used here were
grown according to Scheme 11 but spectra for multilayers grown
according to Scheme 1 produced identical spectra ......................... 161

FTIR spectra of the 1800 cm'1 to 1250 cm’1 spectral region (a) before and (b)
after reaction with KZCO3 to establish the chemical origin of the 1717 cm'l
resonance indicated by an arrow ..................................................... 162

CP/MAS '3 C NMR spectrum of multilayers of poly(MDA-MPI) grown
of high surface area silica. Band assignments are indicated and
referenced to the inset structure. Bands marked with an asterisk are
spinning side bands, identified by their positional dependence on sample
spinning speed .................................................................. 164

Absorbance of multilayers as a function of number of bilayers added.
The band at 260 nm is assigned to the diphenylurea chromophore. Inset:
Absorbance as a function of number of bilayers for both resonances...166

Ellipsometric thickness as a function of reaction cycle. The first eight
layers are poly(NCPM-VEB), bound together with Zr“. The next 7 data
points are for growth of poly(MPI-MDA), bound together with urea
linkages. The ~100 A step between the ionic and covalent layers is due
to the adsorption of the PEI layer. The four data points for layers 16 — 19
are for poly(NCPM-VEB) grown on the covalent layers .................. 179

(a). FTIR spectrum of poly(NCPM-VEP) layers grown on a gold
substrate and primed with 6-mercapto-1-hexanol. The grth of the
band at 2963 cm'l and the broad feature around 3300 cm" are indicative
of layer growth. The 3300 cm'1 band is characteristic of H-bonded
polymer side groups. (b) F TIR spectrum of the same system in the 2000
cm’1 to 1000 cm'1 region. The band at 1650 cm'1 is characteristic of the
presence of uncomplexed COOH groups. The bands in the 1200 cm'1 to
1100 cm'1 region are associated with the phosphate group ............... 185

XV

Figure 8.3.

Figure 8.4.

Figure 8.5.

Figure 8.6.

Figure 9.1.

Figure 9.2.

Figure 9.3.

Figure 9.4.

FTIR spectra of poly(MPI-MDA) covalent layers grown on poly(NCPM-
VEB) ionic multilayers. The bottom spectrum is of the poly(NCPM-
VEB) layers and the top spectrum is with covalent multilayers added.
The amide I and amide II bands are seen for the covalent multilayers at
1643 cm’l and 1563 cm", respectively. The resonance in the top
spectrum at ~ 3300 cm’] is characteristic of H-bonding of the urea
nitrogens ......................................................................... 186

X-ray powder diffraction pattern for a sample of an 18 layer hybrid
multilayer stack. The absence of sharp features demonstrates the absence
of crystalline structure in these layers ....................................... 187

XPS survey spectrum of an 18 layer hybrid multilayer stack. The data
show the presence of the expected elements (see text) and the absence of
Cl .................................................................................. 188

Absorbance spectra of multilayer stack as a function of layers added. For
the ionic layers, the absorbance maximum is at 243 nm and for the
covalent multilayers, the absorbance maximum is at 260 nm. Inset:
Absorbance at 243 nm (ionic) and 260 nm (covalent) for the multilayer
stack showing linear growth with each type of
chemistry ........................................................................ 190

The structure of MVE copolymers having two different chromophores R.
and R2 to get better understanding of rate of complexation of a metal (e. g.
Zr“, Ca”, Y3+ etc.) with different protonic acids using UV-Vis
spectroscopy ..................................................................... 1 97

The structure of MVE copolymers having two different chromophores R1
and R2 to get better understanding of rate of complexation of a metal (e.g.
Zr“, Ca2+, Y3+ etc.) with different protonic acids using XPS. . . . . . . . ....198
The structure of MVE copolymer where the length of vinyl ether
comomoner is greater than the length of the azobenzene
chromophore .................................................................... 199

The structure of MVE polymer where the trans-to-cis isomerization is
expected to be higher than ZP complexation energy ...................... 200

xvi

Scheme 3.1.

Scheme 4.1.

Scheme 5.1.

Scheme 6.1.

Scheme 7.1.

Scheme 7.2.

Scheme 7.3.

Scheme 7.4.

Scheme 8.].

Scheme 8.2.

Scheme 8.3.

Scheme 8.4.

LIST OF SCHEMES

Synthetic route for poly(NPM-VEP). See text for details .................. 48

Surface reaction sequence studied in this paper. A: Reaction of a silica
surface with POC13. B: Zirconation of A. C: Deposition of a partially
hydrolyzed layer of poly(NPM-VEP). D: Hydrolysis and zirconation of
the poly(NPM-VEP) layer. E: Deposition of a second polymer
layer ................................................................................ 72

Schematic of alternating copolymerization of MAB and VEP. . . . . . . . ....99

Structures of various disulfide derivatives that may be present within the
multilayers. 1 - disulfides, 2 - thiosulfinates, 3 -thiosulfonates, 4 - a-
disulfoxides, 5 - sulfinyl sulfones and 6 - vic-disulfones. All species are
known except 4 ................................................................. 129

Synthetic route for the layer-by-layer formation of urea-linked multilayer
assemblies using the alternating copolymerization of diisocyanates and
diamines ................................................................................. 148

Synthetic route for spontaneous growth of urea-linked multilayer
assemblies using isocyanate/hydrolysis polymerization chemistry. . 1 50

Hydrolysis of an isocyanate to form a carbamic acid, with subsequent
decomposition to produce an amine ................................................. 163

Dehydration of carbamic acids to form an anhydride ..................... 163

Surface functionalization and adsorption scheme for poly(NCPM-VEB
multilayers ...................................................................... l 77

Reaction of poly(NCPM-VEB) layer with SOC12 and PEI to produce an amine
rich surface .............................................................................. 180

Reaction of PEI surface with MP1 and MDA for form covalen
multilayer structures ........................................................... 181

Reaction of MDA-terminated surface with POCl3/ZrOCl2(aq) an
poly(NCPM-VEB) to convert from covalent to ionic interlayer linkin
chemistry ........................................................................ 184

xvii

Chapter 1

Introduction

Modern technology constantly demands new materials exhibiting previously
unrealized properties, and materials-processing methods which yield ever-smaller devices
used as sensors, integrated circuits, and in nonlinear optical and electro-optic
applications. At the present time, there is intensive interdisciplinary research going on in
field of materials chemistry and one indication of its importance is the appearance of new
journals devoted solely to nanoscale materials and architectural construction techniques.

One technique that captured the attention of scientists decades ago was molecular
assembly. In 1917, Irving Langmuir first studied the behavior of amphiphilic molecules
at the air-water interface,l and Katharine Blodgett later introduced the technology of
depositing the monomolecular layers on substrates.2 Since then, organized molecular
assemblies have found use in numerous chemical and physical applications.

Langmuir-Blodgett (LB) films were the first examples of multilayer films with
highly ordered layers and controlled layer-by-layer growth.2 These films have been used
in a wide range of studies, from optical properties of chromophores to ordered

3 However, because of the delicate nature of the van der

polymerizations on a substrate.
Waals forces linking each layer and the inherent difficulty in pre-assembling the
molecules for transfer of each layer, LB films have become limited to mostly studies of
layered assemblies where robustness is not an issue.

The next step in the development of this field was the discovery of self-

assembling materials which, in contrast to LB layers, did not have to be pre-assembled

and they formed more physically stable layers. The first detailed examination of self-

assembled monolayers were published by Allara and Nuzzo in 1985.4 Their studies
showed that certain molecules would chemisorb to a compatible substrate in situ, with
crystalline-like order within the layer. There are many possible combinations of
molecules and substrates that can be used to make self-assembled monolayers (SAMs):
thiols on gold, silver, copper or GaAs,5 carboxylates on aluminum4 or silver,6 and silanes
on glass.7 These systems have been investigated extensively, particularly the thiol/gold
system, and much is known about their formation, structure, and properties. Despite the
extensive discussion in the literature on the potential utility of alkanethiol-gold SAMs,8

9,10

recent work has demonstrated that these assemblies are labile and sensitive to ozone-

mediated oxidative degradation. ' ”2

While there is still much to explore with monolayer
systems, the opportunity to form complex, interfacial structures with this chemistry is _
limited unless specially functionalized molecules are used in the formation of the initial
layer.

In the last fifteen years, some of the focus of interfacial materials research has
shifted to developing materials for covalent multilayer growth. The buildup of multiple,
discrete layers using thiol and disulfide chemistry has not been investigated until
recently,13 and there remains much to be learned about these systems. Silane chemistry
has been used for covalent multilayer growth,7 but this chemistry can be difficult to
control and it is susceptible to unwanted polymerization reactions, precluding the ability
to create uniform, layered interfaces. Metal-phosphonate (MP) organic multilayer
structures have been investigated extensively""30 and, like self-assembling monolayers,

show potential for use in practical applications such as surface modification,3 1‘3 4

16.25.26 38-43

. . 35-37 . . . . . .
electronic devrce, nonlinear optics, and molecular recognition applications.

Phosphonic acids form strong, sparingly soluble complexes with metal ions, giving them
significant advantages over many self-assembled monolayer systems. MP structures are
comparable to SAMs in ease of synthesis, which generally involves immersion of the
functionalized substrate into a solution of the appropriate ((1,0))-organobisphosphonate.
Because two separately deposited components required to build a layer, spontaneous
multilayer growth is precluded. MP multilayers are versatile in a chemical sense because
the identity of individual layers can be controlled selectively as the structure is assembled
and, in this way, chemical potential or optical properties can be control within the system
in three dimensions rather than two. Like other systems that exhibit mesoscopic
organization, metal phosphonate structures have been used successfirlly in many studies,
including optical second harmonic generation, artificial photosynthesis, and light
harvesting.

The work reported in this dissertation has demonstrated new ways to design,
synthesize and characterize robust thin films with controlled molecular dimensions. The
central point of this work is to provide the fundamental knowledge and understanding of
layered-material assemblies to advance the field. We are especially interested in
controlled layer-by-layer multilayer assemblies in which the properties of these films
such as film thickness, linear and non-linear optical response, electrical and electronic
behavior, and porosity can be controlled. Ultimately, these ultra-thin films may find
applications in the areas of controlled released delivery systems, nonlinear optical
devices, chemical interfaces and interfacial sciences, separation science, nanoelectronics,

biocatalysis, and biotechnology.

For thin films and coatings to be successful in practical applications, they have
to be mechanically robust in the environment in which they will be used. Thus these
films and coatings are expected to withstand chemical and thermal attacks. We chose to
use maleimide-vinyl ether (MVE) copolymers for coatings on flat substrates such as gold,
silicon, silica and quartz because these polymers are mechanically and thermally robust
and can also be made with a wide range of physical, optical and electronic properties. In
Chapter 2, we report the stereochemistry and the thermal and chemical stability of MVE
copolymers. We find that the succinimide protons of the polymer backbone have both
cis- and trans-configurations in a statistically distributed manner. Moreover, we
demonstrate the use of bifunctional monomers producing crosslinked MVE copolymers
with enhanced thermal properties.

After having studied their stereochemistry and thermal robustness, we next focus
on the use of these polymeric systems to deposit films where have control over the
properties of each layer. Layer-by-layer deposition of films can, in principle, yield thin
films and coatings where we can control the chemical properties at each deposition cycle.
With this idea in mind, we have constructed multilayer assemblies of MVE polymers
where the vinyl ether monomer has a phosphoester terminal group. This functionality
can be hydrolyzed to phosphonic acid (Chapter 3). The phosphonic acids are used to
form stable coordination bonds with a zirconated substrate to deposit the first layer. The
key to forming multilayers is to use partially hydrolyzed phosphoester MVE copolymers,
where a fraction of vinyl ether phosphoester is unhydrolyzed while rest of the
phosphoester is hydrolyzed to the corresponding phosphonic acid. The unhydrolyzed

phosphoester will not form a complex with zirconium, and can be hydrolyzed once

deposited on the surface using bromotrimethylsilane. Repeated zirconation, deposition
and hydrolysis are used to form multilayer assemblies. Since we deposit the layers one
at a time, we can change identity of each layer. For example, it is possible to deposit
layer one of MV E copolymer having a chromophore with an absorption maximum at 400
nm and a second layer of MVE polymer having chromophore with an absorption
maximum at 600 nm. In this case, we can change the optical response of the interface
just by changing the pendant groups on the succinimide moieties in the copolymer
backbone. Similarly, it is also possible to change the film thickness per layer,
permeability and electronic properties. We thereby have the “tunability” and control over
the properties of multilayer assemblies.

Gaining an understanding of the zirconium-phosphate/phosphonate (ZP)
interlayer bonding is important because it can help in making better and robust
multilayer assemblies. We have used 3|P magic angle nuclear magnetic resonance
spectroscopy (3 lP MAS NMR spectroscopy) to probe the surface reactions and interfaces
of ZP multilayer assemblies (Chapter 4). We chose 31P MAS spectroscopy because: (1)
the chemical shift of 31F is very sensitive to its surrounding environment and is thus a
good indicator of the species present during selected stages in layer-by-layer deposition
reactions; (2) 31P NMR has very high sensitivity as a result of the 100% natural
abundance of 31P. These factors make this spectroscopic method suitable to investigate
the build-up of ZP multilayers. We have found that the interlayer bonding for both
polymeric and alkanebisphosphonate multilayer systems is essentially the same, but
subtle differences in the efficiency of the layer formation arise from the disorder and

permeability that is characteristic of the polymer multilayers.

One application for which three dimensional, layer-by-layer control of multilayer
structure could potentially be useful is in optical information storage. Present-day
storage devices use a single layer medium, and the addition of a useful third dimension
could enhance the storage density of a given volume of the surface. The use of MP
multilayers with chemically and optically distinct layers could be the foundation for such
a three-dimensional structure, but there are aspects of this scheme that could serve to
limit the utility of these materials for information storage. The chemical change
associated with writing or erasing, while still under investigation, will likely be a cis-
trans isomerization. Regardless of the means used to create a molecular “bit”, the writing
process will necessarily be mediated by an excited electronic state to achieve the
necessary spatial selectivity and contrast. During conversion from one logical state to
another, the chromophore will be excited, probably to its SI state, and dissipative
processes from the first excited singlet state can, in principle, serve to limit the
conversion efficiency or definition of the written spot. The conformational change must
be an activated process with a sufficiently high ground state barrier to ensure that the
operation is irreversible under ambient dark conditions. By using simple chromophores
that don’t “switch”, we can observe these dissipative processes without complications
associated with large scale chromophore isomerization. The investigation of MP
materials that show potential for optical switching is underway in our lab. We have
grown multilayers of po1y(4-N-maleimidoazobenzene-c-(2-vinyloxy)-ethylphosphonate),
MAB-VEP, where interlayer connections are made using Zr—bisphosphonate (ZP) ionic
complexation chemistry (Chapter 5). Absorption data on the azobenzene chromophore

side group show constant layer density but a layer-dependent ratio of trans and cis

isomers. Optical null ellipsometry data show a constant average layer thickness despite
the change in conformer ratio. The change in conformer ratio with the growth of multiple
polymer layers results from the steric constraints imposed on the polymer side groups by
the Zr—phosphonate interlayer linkage formation. The ellipsometric measurements point
to the macroscopic disorder inherent in the formation of a polymer layer. Our data
demonstrate that the driving force for metal ion complexation is greater than the
isomerization barrier of ground state azobenzene. Once the layers are formed, the side-
groups do not exhibit any changes in conformer ratio, even when exposed to UV light for
prolonged periods. With this particular copolymeric system, however, it is not possible
to “switch” azobenzene chromophores since chromophores are locked in the cis-
conformation and can not back-isomerize to trans-conformation.

The Blanchard group has incorporated or,c0-bisphosphonated thiophene oligomers
into zirconium-phosphonate (ZP) layered assemblies to understand intralayer and
interlayer optical excitation transport. Interlayer excitation transport measurements on
these same systems showed that the Zr(O3PR)2 interlayer linking moiety screened dipolar
coupling. Horne and Blanchard attributed this effect to the polarizability of the ZP sheets
between layers and have sought ways to design robust layered materials that do not
possess this firnctionality.44’45 This finding has led us to look into new ways to deposit
layers one at a time with covalent interlayer bonding. We have found that or,c0-dithiols
assembled themselves into covalently-coupled disulfide multilayers on gold (Chapter 6).
The linking chemistry between layers is the oxidative formation of a sulfur-sulfur bond
that competes successfully with intralayer S-S bond formation. We have used optical

null ellipsometry, FTIR, X-ray photoelectron spectroscopy (XPS) and cyclic voltammetry

(CV) to characterize the multilayers. Once formed, the multilayers are stable when
washed with 1 M KCl, water, ethanol, CHC13 and n-hexane solutions, before and after
prolonged exposure to ambient laboratory conditions. In addition to the formation of
multilayers, our data point to the efficient oxidation of the interlayer disulfide bond to an
oxidized sulfur moiety where the S-S bond remains intact. Extensive oxidation produces
a sulfonate-terminated surface that reacts with Zr4+ and alkanebisphosphonates to form a
hybrid multilayer assembly.

Because of the direct oxidative thiol-to-disulfide covalent bonding scheme used in
the grth of these systems, layer growth is statistical in terms of overall assembly
thickness, i. e. discrete, single layer growth is not enforceable with this chemistry. These
multilayer assemblies are also susceptible to oxidation of the disulfide linkages to more
highly oxidized species and this oxidation appears to produce lateral S-O-S bonds parallel
to the substrate.

A more general route to the growth of robust, covalent multilayer assemblies is to
take advantage of well-established polymer chemistry. The biggest challenges in
applying polymer chemistry to the controlled growth of interfacial multilayers are to
attach an initial monomer to the substrate and to devise synthetic routes where the
polymerization chemistry terminates after the addition of a single monomer unit.
Comparing the multilayers formed from the two synthetic routes demonstrates the
importance of using step-by-step alternating copolymerization chemistry to control the
growth of multilayers.

The specific chemical systems we used to form covalent interlayer linkages are

based on the reaction of isocyanates with amines to produce urea groups (Chapter 7). A

significant benefit of using the urea moiety as an interlayer linking functionality is its
ability to form hydrogen-bonded networks parallel to the substrate. The formation of
urea linkages to bind polymer layers to silica substrates has also been reported. The
alternating copolymerization of 4,4’-methylenedianiline (MDA) and 4,4’-
methylenediphenylisocyanate (MP1) to form urea linkages between layers affords direct
control over layer thickness. The lateral stability of these multilayer assemblies is
enhanced through intralayer hydrogen bonding between the urea groups. We have also
demonstrated the same urea linking chemistry for multilayers grown using only MP1,
where the formation of reactive amine at each layer is accomplished by deliberate
hydrolysis of the isocyanate to form an amine. The main difference between the
chemistry presented in these two schemes is that the growth of the discrete layers one at a
time is possible with the alternating isocyanate/amine copolymerization chemistry and
spontaneous, uncontrolled growth is achieved with the isocyanate/hydrolysis chemistry.
Our data show that the use of well-established polymer chemistry in the design and
growth of covalently-bonded multilayer structures is feasible, pointing the way to a
generic strategy for covalent layered material design.

After demonstrating both ionic and covalent multilayer chemistry independently,
the next obvious step is to use a combination of these multilayer chemistries to produce
mixed multilayers where the interlayer linking chemistry can be changed between ionic
and covalent in a well-controlled manner. The advantage of synthesizing mixed
multilayer assemblies is that these systems may offer properties which may be difficult to
obtain using solely ionic and covalent chemistries. We have reported on the synthesis

and characterization of layered hybrid interfaces formed using zirconium-

phosphate/carboxylate (ionic) and diisocyanate/diamine (covalent) chemistries (Chapter
8). We have characterized these hybrid multilayer assemblies using optical ellipsometry,
X—ray diffraction and F TER, UV-visible and X-ray photoelectron spectroscopies.

This work taken collectively, points the way toward the use of layered-materials
chemistry in the design of nanoscale structures and, when combined with lithographic
techniques to provide lateral definition and a wide range of device structures can be
envisioned. The central point of this thesis is to provide the fundamental chemical

foundation and understanding for work that will follow in this area.

1.1 Literature Cited

1. Langmuir, I. J. Am. Chem. Soc. 1917, 39, 1848.
2. Blodgett, K. B. J. Am. Chem. Soc. 1935, 57, 1007.

3. Ulman, A. Introduction to Thin Organic Films: From Langmuir-Blodgett to Self-

Assembly Academic Press: Boston, 1991.

4. (a) Allara, D. L.; Nuzzo, R. G. Langmuir 1985, 1, 45. (b) Allara, D. L.; Nuzzo, R. G.

Langmuir 1985, 1, 52.

10

(a) Nuzzo, R. G.; Dubois, L. H.; Allara, D. L. J. Am. Chem. Soc. 1990, 112, 558. (b)

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

(a) Smith, E. L.; Porter, M. D. J. Phys. Chem. 1993, 97, 8032. (b) Canning, N. D. S.;

Madix, R. J. J. Phys. Chem. 1984, 88, 2437.

(a) Gun, J .; Iscovici, R.; Sagiv, J. J. Colloid Interface Sci. 1984, 101, 201. (b) Maoz,
R.; Sagiv, J. J. Colloid Interface Sci. 1984, 100, 465. (c) Maoz, R.; Sagiv, J. Thin

Solid Films 1985, 132, 135.

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

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

10.

ll.

12.

l3.

14.

15.

16.

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

9645.

Zhang, Y.; Terrill, R. H.; Tanzer, T. A.; Bohn, P. W. J. Am. Chem. Soc., 1998, 120,

2654.

Norrod, K. L.; Rowlen, K. L. J. Am. Chem. Soc., 1998, 120, 2656.

Kohli, P.; Harris, J. J.; Blanchard, G. J. J. Am. Chem. Soc. 1998, 120,11962.
Hong, H.-G.; Sackett, D. D.; Mallouk, T. E. Chem. Mater. 1991, 3, 521.
Thompson, M. E. Chem. Mater. 1994, 6, 1168.

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

ll

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

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. F.; Wilson, W. L.; Schilling, M. L.; Ungashe, S. B. J. Am.

Chem. Soc. 1994, 116, 6631.

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

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.

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

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

Chem. Soc. 1992, 114, 8717.

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

Katz, H. E.; Schilling, M. L.; Chidsey, C. E. 1).; Putvinski, T. M.; Hutton, R. S.

Chem. Mater. 1991, 3, 699.

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

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

Putvinski, T. M.; Schilling, M. L.; Katz, H. E.; Chidsey, C. E. D.; Mujsce, A. M.;

Emerson, A. B. Langmuir 1990, 6, 1567.

Rong, D.; Hong, H.-G.; Kim, Y.-I.; Krueger, J. S.; Mayer, J. E.; Mallouk, T. E.

Coord. Chem. Rev. 1990, 97, 237.

12

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

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.
Katz, H. E. Chem. Mater. 1994, 6, 2227.
Dubois, L. H.; Zegarski, B. R.; Nuzzo, R. G. J. Am. Chem. Soc. 1990, 112, 570.

Chidsey, C. E. D.; Bertozzi, C. R.; Putvinski, T. M.; Mujsce, A. M. J. Am. Chem.

Soc. 1990, 112, 4301.
Collard, D. M.; Fox, M. A. Langmuir 1991, 7, 1192.
Chidsey, C. E. D. Science 1991, 251, 919.

Batchelder, D. N.; Evans, S. D.; Freeman, T. L.; Haussling, L.; Ringsdorf, H.; Wolf,

H. J. Am. Chem. Soc. 1994, 116, 1050.
Tarlov, M. J.; Burgess, D. R. F.; Gillen, G. J. Am. Chem. Soc. 1993, 115, 5305.

Wollrnan, E. W.; Kang, D.; Frisbie, C. D.; Lorkovic, I. M.; Wrighton, M. S. J. Am.

Chem. Soc. 1994, 116, 4395 (1994).
Zak, J.; Yuan, H.; Ho, M.; Woo, L. K.; Porter, M. D. Langmuir 1993, 9, 2772.
Obeng, Y. S.; Laing, M. E.; Friedli, A. C.; Yang, H. C.; Wang, D.; Thulstrup, E. W.;

Bard, A. J.; Mich], J. J. Am. Chem. Soc. 1993, 114, 9943.

13

40.

41.

42.

43.

45.

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

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

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

Spinke, J.; Liley, M.; Schmitt, F.-J.; Guder, H.-J.; Angermaier, L.; Knoll, W. J.

Chem. Phys. 1993, 99, 7012.

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

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

14

Chapter 2

Co-Polymerization of Maleimides and Vinyl Ethers: A Structural
Study

Abstract

We report on the thermal- and photoinduced polymerization of the electron donating
divinyl ether di(4—vinyloxybutyl)succinate (DVBS) and the electron accepting 1,5-
bismaleimido-Z-methylpentane (BMMP) and their monofunctional analogs, dodecylvinyl
ether (DDVE), N-methylmaleimide (NMM) and N-phenylmaleimide (NPM). We focus
on the stoichiometry of the monomer constituents consumed in the polymerization and
the structural identity and bonding configuration(s) of the polymer products. For both
BMMP-DVBS and NMM-DDVE, F TIR and 1H NMR data indicate 1:1 stoichiometry for
the resulting polymer. For both systems, the polymers are of the alternating donor and
acceptor form (DA)n. This result holds for both photo-induced and thermally-induced
reactions. Structural NMR studies of polymers synthesized from both mono- and di-
functional monomers indicate the presence of both cis-and trans conformers of the
acceptor at the D-A bond. This finding indicates that thermodynamic factors alone are
not sufficient to explain the ratio of the conformers and either steric factors or features
specific to the transition state play a significant role in determining the conformational

distribution of the polymer.

15

2.1 Introduction

Photopolymerization is one of the most rapidly expanding processes for materials
production. Photopolymerizations have gained prominence in recent years for the
pollution-free curing of polymer films as well as emerging applications in dental
materials, conformal coatings, electronic and optical materials, and high resolution rapid
prototyping of 3D objects. These solvent-free polymerizations proceed very rapidly with
a fraction of the energy requirements of thermally cured systems, and the resulting
polymers possess useful material properties. For example, these reactions can be used to
form highly-crosslinked, thermally stable polymer films exhibiting excellent adhesion,
abrasion resistance and chemical resistance while not emitting or requiring volatile
organic components. In addition to the broad utility of such polymers, the
photopolymerization process itself affords advantages such as very high reaction rates at
room temperature and spatial control of polymerization.

To date, most of the work on UV-initiated photopolymerizations has focused on
free-radical systems based primarily upon acrylate and methacrylate monomers."3 The
kinetics of these free-radical photopolymerizations and the structural properties of the
resulting polymers have been characterized extensively. Acrylate monomers polymerize
rapidly and are modified at the ester functionality, allowing access to materials with a
variety of properties. These systems are, however, not without their limitations.
Monomeric acrylates are relatively volatile, have an unpleasant odor, and present
potential safety hazards. Oligomeric acrylates are much less volatile, but are highly
viscous, making processing a significant challenge. The free radical

photOpolymerizations reactions undergone by acrylates are inhibited by oxygen and, if

16

high polymerization rates are required, the reactions must be carried out under an inert
atmosphere. Finally, these polymerizations require the presence of a photoinitiator,
which places limits on the long-term stability of the monomer formulation and leaves
undesired residues in the final polymer product.

Photopolymerizations mediated by electron donor-acceptor interactions have
emerged relatively recently as a new process“’7 and they are not yet understood fully.
While there is a body of literature concerned with these polymerization reactions,12 there
remain significant ambiguities relating to mechanism and kinetics of formation. The
work we present here focuses on the characterization of the products formed by
photopolymerization of the donor divinyl ether molecule di(4-vinyloxybutyl)succinate
(DVBS) and the acceptor 1,5-bismaleimido-2-methylpentane (BMMP) without use of an
initiator additive. Because cross-linking is extensive in this system (vide infra), we have
used monofunctional monomers dodecylvinylether (DDVE) and N-methylmaleimde
(NMM) for comparison to produce linear (uncrosslinked) chains. In addition, for
purposes of NMR structural resolution, we have also used the monomer N-
phenylrnaleimide (NPM) with DDVE in the formation of non-crosslinked copolymers.
The structures of these monomers are presented in Figure 2.1. The two points we
consider in this paper are the stoichiometry of the monomer constituents in the
photoinduced polymerization and the structural identity of the polymer formed by both
photoinduced and thermally induced reactions. Our data indicate that the polymer is of
the form (DA)... FTIR and 1H NMR data indicate 1:1 stoichiometry for polymer
formation with both the photoinduced and thermally induced reactions. 1H and '3 C NMR

data on the monofunctional monomers DDVE, NPM and NMM, and the corresponding

17

linear polymers, allow the assignment of bands in the spectra of the polymer synthesized
from the difunctional monomers BMMP and DVBS. The data on the BMMP-DVBS
copolymer indicate the bonding of the monomer units exhibits approximately equal
amounts of cis and trans conformers across the unsaturation of the maleimide monomer.
Simple thermodynamic considerations are not sufficient to explain this ratio of
conformers because they differ in energy by ~ 2.3 kcal/mol, and therefore either steric
factors or the identity of the transition state act to mediate the conformational distribution
seen in the polymer product. This is the first report in a series aimed at understanding the
molecular processes at work in photopolymerizations mediated by electron donor-
acceptor interactions. Following the structural characterization presented here, we will
examine the formation kinetics and reaction dynamics of these systems using transient

spectroscopic techniques.8

2.2 Experimental

NMR spectroscopy. 1H NMR spectra were recorded at 300 MHz on Varian
Inova-300 Gemini-2000 and Varian VXR-300 spectrometers and at 500 MHz on a Varian
VXR-500 spectrometer. The proton-decoupled ‘3 C NMR spectra were recorded on
Varian Inova-300, Gemini-2000 and Varian VXR-300 spectrometers at 75.46 MHz and
at 125.75 MHz on a Varian VXR—500 spectrometer. The solid-state CP/MAS ‘3 C spectra
were recorded at 100.58 MHz on a Varian VXR-400 spectrometer with samples spinning
at a speed of either 4.0 kHz or 5.4 kHz. For monomer lH NMR spectra, 64 scans were
usually taken while '3 C NMR spectra required, typically, 2048 scans to achieve a useful
signal-to-noise ratio. For polymers, 1H and l3‘C NMR spectra required 25,000 to 50,000

scans to achieve a useful signal to noise ratio.

18

F T IR spectroscopy. The FTIR spectra (4000 cm'1 - 700 cm!) were obtained using
a Nicolet Magna-IR 550 spectrometer equipped with a N2(1) cooled MCT-A detector.

For kinetic studies, one or two drops of a mixture of BMMP and DVBS monomers were
spread uniformly over a gold coated substrate and the sample was mounted in a grazing
incidence extemal-reflectance accessory. At least 256 scans were taken at 4 cm”l
resolution to achieve an adequate signal to noise ratio.

Molecular Weight Measurements. The molecular weights of copolymers were
determined using a Beckman GPC model 100A with a Waters R401 differential
refractometer detector.

Thermal Analysis. Glass transition temperatures, T3, were determined under N2
atmosphere using a Perkin-Elmer DSC 7 and a heating rate of 20°C/min. The
thennogravimetric analysis was carried out under N2 atmosphere on a Perkin-Elmer TGA
7 with 10°C/min. heating rate.

Chemicals. Maleic anhydride (ARCO Organic), 2-methyl-1,5-diaminopentane,
acetic anhydride, methanol and N-methylmaleimide (Aldrich), benzene (J .T. Baker), and
triethylamine (Spectrum Chemical Mfg. Corp.) were used as received. Chloroform-d
(99.8% with 1% v/v TMS) and db-DMSO (99.9% purity) were purchased from
Cambridge Isotope Laboratories. Di(4-vinyloxybutyl)succinate (DVBS) was donated by
Allied Signal and dodecylvinylether (DDVE) was donated by International Specialty
Products, and both compounds were used as received.

Preparation of 1,5-bismaleimido-2-methy1pentane (BMMP). BMMP was
prepared in two steps,9 with minor modifications as noted below. The first step in the

synthesis of BMMP is the preparation of 2-methylpentane-l ,S-bismaleic acid (MPBMA).

l9

Maleic anhydride (2 mol) in benzene (200 mL) was stirred while being maintained
between 40°C and 45°C. 2-Methyl-1,5-diaminopentane (1 mol) was added dropwise
over a period of one hour. The resulting suspension was stirred for an additional hour. A
white crystalline precipitate was collected by suction filtration and washed with hexanes
and water to yield 2-methylpentane-l,5-bismaleic acid (typ. yield 92 %): m.p.: l3l-132°
C; 1H NMR (300 MHz, dé-DMSO): 8 0.83 (d, 3 H of CH3), 1.10 (m, 1 H), 1.3-1.6 (m, 6
H of CH2), 3.0-3.15 (m, 4 H of CH2 adjacent to two NH), 6.2 - 6.4 (m, 4 H of C=C), 9.1
(q, 2 H of COOH). 1R (cm'l): 3262, 3123, 3084, 2956, 2877, 1706, 1641, 1531, 1410,
1302, 853, 796.

The second step is the preparation of BMMP from MPBMA. 2-Methylpentane-
1,5-bismaleic acid (0.1 mol), triethylamine (0.4 mol) and acetone (100 mL) were added
to a single flask. The solution was refluxed and acetic anhydride (0.3 mol) was added
dropwise. The solution was cooled to room temperature after refluxing for 24 hours.
The resulting dark brown solution was added to 200 mL ice water. After 2 hours of
stirring, a dark brown precipitate was obtained by suction filtration and dried in air. The
crude product was purified using a packed silica gel column eluted with 40% ethyl
acetate/60% hexane. The purified product was recrystallized from 50% aqueous
methanol to yield white crystals of BMMP (typ. yield 28 %): m.p.: 87°-89° C; 1H NMR
(300 MHz, d—CHC13): 5 0.83 (d, 3 H of CH3), 1.10 (m, l H), 1.3-1.6 (m, 6 H of CH2),
3.35-3.5 (m, 4 H of CH2 adjacent to N of two imide rings), 6.66 (d, 4 H of C=C of two
imide rings). l3C NMR (d-CHC13, 75.46 MHz): 5 17.3, 25.8, 31.2, 32.2, 37.8, 43.7,134.9,
134.0, 170.7, 171. IR (cm"): 3452, 3166, 3092, 2969, 2935, 1767, 1699, 1612, 1584,

1410, 1446, 836. Absorption 1km“: 300 nm. Fluorescence Mm: 460 nm.

20

BMMP-DVBS Copolymer Synthesis. All copolymers were synthesized without
the addition of a photoinitiator. BMMP and DVBS were heated to 40°C in the absence of
solvent to make a homogeneous solution. The flask containing this solution was
maintained at 40°C and flushed with dry nitrogen for 10 minutes, then sealed.
Photoexcitation was done with either a N2 laser (337 nm) or by a mercury lamp source
(UV output 4.5 mW/cmz). When the N2 laser was used, only BMMP molecules were
excited because 337 nm lies in the tail of the BMMP absorption spectrum.
Photoexcitation of both constituents is possible when the broadband mercury light source
is used. For either excitation scheme, a nominally white product with a sight yellow tint
was obtained. The product was washed sequentially with hot chloroform, hexanes,
methanol and water and dried overnight in vacuo at room temperature. The product was
insoluble in methanol, chloroform, acetone, hexane and cyclohexane.

For the thermally-induced reactions, monomers were heated to 60°C in CHC13 for
24 hours under N2. The solvent was evaporated and the product redissolved in a
minimum amount of acetone followed by precipitation from methanol.

C opolymer Synthesis of N-methylmaleimide/dodecylvinylether. In a typical
reaction, N-methylmaleimide (NMM) and dodecylvinylether (DDVE) were added to a
round bottom flask with a small amount of CHC13 solvent. The CHC13 solvent was
required because NMM is not soluble in DDVE. The rest of the polymerization was the
same as reported above. The reaction yielded a slightly yellow-white polymer which was
purified by precipitation from methanol after dissolving in acetone. The molecular

weight of the polymer, determined by GPC was Mn = 35,000.

21

2.3 Results and Discussion

The purpose of this discussion is to provide the foundation for further studies on
donor-acceptor mediated photopolymerization using transient spectroscopic techniques.
The key points in understanding maleimide—vinyl ether systems are determination of the
reactivity of the monomers with themselves and the stoichiometry of the copolymer.. It
is also important, of course, to understand the chemical structure and extent of cross
linking in the final polymer product. We present data relevant to each of these points in
the sections that follow.

Our first step in understanding this system was to determine whether or not the
individual monomers would homopolymerize in the presence of either light or heat.
BMMP, NPM, NMM, DVBS and DDVE each yielded no detectable homopolymer under
the conditions used to polymerize mixtures of BMMP and DVBS, NMM and DDVE or
NPM and DDVE.

Stoichiometry of the monomers. We have used FTIR and 1H NMR spectroscopy to study
the polymerization of BMMP and DVBS. From the chemical structures in Figure 2.1, it
is clear that there can be several possible structures for the photopolymer product, with
the extent of cross linking being the dominant variable. Before evaluating the cross-
linking, however, we must first understand the intermolecular bonding arrangement that
is responsible for the polymerization in the first place. Since there are several possible
bonding schemes which may appear plausible, it is not necessarily clear that a 1:1
stoichiometry will hold or that an alternating (D-A) repeat unit is appropriate. We

consider the FTIR and NMR data separately.

22

 

l ,5-bismaleimido-Z-methylpentane
O
O ‘ O

O
di(4-vinyloxybutyl)succinate (DVBS)

o o; :O o- [,V---:O

(CH2) N N
l l .

Dodecylvinylether (DDVE) N-phenylmaleimide (NPM) N-methylmaleimide (NMM)

Figure 2.1. Chemical structures of the monomers used in this work.

Infrared data. In the BMMP-DBVS mixed monomer system, two different
reactive functionalities are present, i. e. the unsaturations in the imide ring of the acceptor
(BMMP) and vinyl unsaturations at the termini of the donor (DVBS). To characterize the
kinetic behavior of the system completely, both reactive sites need to be monitored
separately. Our FTIR data shown in Figs. 2a and 2b are the spectra of BMIVLP and DVBS
in the 1500 cm’1 - 1800 cm'1 and 2800 cm'] - 3200 cm'1 regions, respectively. The
BMMP C=C stretching resonances at 1612 cm'1 and 1584 cm’l are highly overlapped

with the intense C=C stretching resonances of DVBS at 1618 cm'1 and 1638 cm’1 (Figure

23

2.23). The C=O resonances, while not overlapped for the two monomers, do not provide
any kinetic information because these functionalities are not consumed in the reaction.
We use these bands as internal standards. We note that it is difficult to define a baseline
for these spectra due to the multiple strong absorptions in this region. For these reasons,
even though the C=C bonds are the reactive functionalities, it is not easily possible to
monitor the progress of the polymerization in the 1500 cm'I to 1800 cm'1 spectral region.
The alkene CH stretching bands between 3100 cm'] and 3200 cm" were chosen for
monitoring the kinetic behavior of the copolymerization (Figure 2.2b) because spectral
overlap and baseline assignment problems are less serious for these resonances. The
3092 cm'l band of BMMP and the 3118 cm'1 band of DVBS were selected because of
their relatively high absorptivity and, while they are overlapped, we have been able to
define a baseline for each of these bands.

In a typical experiment, BMMP and DVBS are added to a gold coated quartz
substrate and the spectrum of the reaction mixture is recorded before
photopolymerization and at selected exposure times. These data are curve fit afier
baseline correction.10 To ensure the uniqueness and reproducibility of the results, the
Gaussian peak width, the number and position of the peaks and the baseline are fixed and

the peak heights are varied to optimize the fit.

24

3092

0.46 BMMP -‘ ‘ a

 

£3; 1‘
8 t.
5
":3 0.36
8
'2
3118 DVBS
0.26 r u a A
3200 3100 3000 2900

frequency (cm")

0.60 n BMMP b

 

 

2.7; DVBS
3

° I
O

.§

§ 0.30 V

.D

<

0.10

1800 1750 1700 1650 1600

frequency (cm")

Figure 2.2. IR spectra of BMMP and DVBS monomers. (a) 2800 cm'1 - 3200 cm'1 and

(b) 1500 cm" - 1800 cm".

25

0.40

0.30

absorbance (a.u.)

 

0.20

 

 

3200 3 100 3000 2900

frequency (cm ")

Figure 2.3. Variation of band intensities for the monomer alkene CH stretching
resonances during photocopolymerization of BMMP/DVBS. Times for each spectrum

are t = 0 min., t = 30 min., t = 53 min., and t = 12 hrs., in order of decreasing intensity of

the ~3100 cm'1 bands.

The fractional conversion of the BMMP and DVBS unsaturations is an indication

of degree of copolymerization and is defined as: H

ar=1—i [1]

A0
where A0 and At represent absorbances at times zero and t respectively. From Beer’s
law, the monomer concentrations C0 and C1 obtain directly. The change in the intensity
of the alkene CH stretching vibrations is shown as a function of time in Figure 2.3.
These spectra are normalized for C=O stretching resonance intensity. The alkene CH

stretching resonance decreases monotonically with UV exposure time. This progression

26

shows that the double bonds of both the BMMP acceptor and DVBS donor have reacted

(Figure 2.4), but their relative rates of reaction are of importance in establishing the

 

 

1 I
:5
.g‘ 0.8 ' A .. A ‘ a f
(I) O
H A
a) 0.6 ' '
E
U 0.4 ' °
I
02 ' ’ -
' I
O bibl- . - - fi
0 10 20 30 40
time (min.)

Figure 2.4. Time-dependence of polymerization of monomers for photoinduced and
thermally induced reactions with initial 1:1 BMMP to DVBS monomer stoichiometry at
T = 22°C. 0 = BMMP conversion, thermally induced reaction, I: DVBS conversion,
thermally induced reaction, A = BMMP conversion, photoinduced reaction, 8 = DVBS
conversion, photoinduced reaction.

stoichiometry of the product. The conversion vs. time data for BMMP and DVBS are
shown in Figs. 5 for initial monomer ratios of [BMMP]0/[DVBS]0 = 1 (Figure 2.53) and
'/2 (Figure 2.5b) at 22°C. For all initial monomer ratios, the disappearance of the alkene

CH resonances reflects a 1:1 reaction stoichiometry.

27

The copolymerization proceeds with UV excitation at room temperature without the use
of a photoinitiator. The initial rate of reaction is high and is dependent on the incident
UV photon flux. For our conditions, within minutes of initial irradiation, more than 60%
conversion of the monomers is seen and after this initially high rate, the reaction slows
considerably. 150 Minutes at the same irradiation intensity is required to achieve 80%
conversion of theunsaturations, although the approach to this conversion is slow and
mostly completed within 15 minutes.

Because we have established earlier that homopolymerization of the monomers
does not proceed measurably, we conclude that the copolymer must be alternate in
nature, i.e. -(DVBS-BMMP)n-. This finding is in agreement with other literature reports
on similar chemical systems.”17 While the UV-induced polymerization rate is very fast
in the first two minutes and achieves close to a limiting conversion after five minutes, the
thermally induced reaction is significantly slower. Within 30 minutes the conversion in
both cases has achieved the same value (~80%) This limiting conversion is in good
agreement with that reported by Martuscelli et a1.ll for polymerizations of bismaleimide
with unsaturated polyester resins. This limiting value is reasonable for highly crosslinked
systems and can be understood in the context of trapping of reactive moieties in the rigid
polymer matrix. Polymerization of NMM and DDVE, where extensive cross-linking is

precluded, achieves greater than 90% conversion.

28

y—a
j

y = (0.95 :I: 0.14)x

   
 
   

u.)

0.75'

.0
Ur

0.25‘

 

DVBS Conversion (a

 

O

0 0.2 0.4 0.6 0.8 1
BMMP Conversion (a.u.)

y = (0.95 d: 0.15)x

 

0.016 '
0.012 '
0.008 '

0.004 '

 

DVBS Consumption (a.u.)

 

0 t u u u

0 0.004 0.008 0.012 0.016
BMMP Consumption (a.u.)

Figure 2.5. FTIR measurement of BMMP monomer loss vs. DVBS monomer loss in the
thermally induced reaction for (a) [BMMP]0/[DVBS] 0 = 1 at T = 22°C, and (b) [BMMP]

0/[DVBS]O = 1/2 at T = 22°C.

29

We next turn our attention to the structural possibilities that exist for this system
consistent with the stoichiometric and reaction constraints that we have established
above. The (BMMP—DVBS)In copolymer is insoluble in many solvents, consistent with
the expected extensive cross-linking in the product. The key point to understanding the
fundamentals of the polymerization reaction is to determine the conformation of the
bonding arrangement between the donor and acceptor functionalities. Specifically, the
orientation of the succinimide protons will be determined by the mechanism of the

- - 4
reaction.12 I ’26

If the reaction is concerted, which one could expect based on the
formation of the complex, the cis conformer is expected. Conversely, if the reaction is
sequential, such that the reaction across the imido unsaturation proceeds in a step-wise
fashion, we would expect the presence of both cis and trans conformers. In the simplest
case, the relative abundance of cis and trans conformers will be given by the relative
thermodynamic stability of the cis and trans conformers. We have synthesized polymers
from NPM and DDVE monomers for this study because the resulting polymer is more
soluble than the BMMP-DVBS polymer and the bonding configuration is expected to be
the same as that for BMMP-DVBS. The solubility of (NMM-DDVE)n allows the use of

higher resolution, solution phase NMR measurements to elucidate the bonding

configuration between the monomer units.

30

.9 .0 .0
b O\ 00
l l l

.0
N
I

O

 

BMMP Consumption (a.u.)

 

O

I I I I

0 0.2 0.4 0.6 0.8
DVBS Consumption (a.u.)

u
.09
UIO\

.0
4;

 

 

'0.0 0.1 0.2 0.3 0.4 0.5 0.6
DVBS Consumption (a.u.)

Figure 2.6. 1H NMR determination of BMMP monomer loss vs. DVBS monomer loss.

BMMP Consumption (a. .)
o o o o
O r-d [\J OJ

For all data, total conversion = 5%. (a) [BMMP]0/[DVBS]0 = 0.19 M/0.38 M = 1/2, T =
30°C, slope = 1.0 i 0.01, (b) A = [BMMP]0/[DVBS]0 = 0.19 M/0.063 M = 3/1, T =
90°C, slope = 0.93 i 0.09, (c) o = [BMMP]0/[DVBS]0 = 0.53 M/1.00 M = 1/1.89, T =

90°C, slope = 0.78 i 0.08.

31

The 1H NMR spectrum of the NPM-DDVE polymer is presented in Figure 2.7.
The NMR peak assignments and proton integrations indicate that the (NPM-DDVE)n
stoichiometry is 1:1, in correspondence with the (BMMP-DVBS)n polymer. 1H and 13 C
NMR spectroscopy of (NPM-DDVE)n can give information about the stereochemistry of
the copolymers formed and thus about the propagation mechanism.

Figure 2.7 shows the 1H NMR spectrum of photoinduced copolymer obtained in
high yield (>90%). The lines are very broad because of high molecular weight of the
copolymer (Mn = 14,310). The resonances in the 2.7 ppm — 3.8 ppm region are
associated with the methylene protons adjacent to oxygen of vinyl ether and the
succinimide protons. These protons are of primary interest in determining the dominant
polymerization mechanism. We have also measured and assigned the '3C NMR spectra
of NPM-DDVE (Figure 2.8) and NMM-DDVE (not shown) and summarize our
assignments in Table 2.1.

The NPM-DDVE copolymer possesses three chiral centers, two in the
succinimide ring and one at the methine carbon of the vinyl ether. These centers can
exist as different diastereomers with different chemical shifts both in 1H and 13 C NMR
spectra. We present in Table 2.2 the chemical shifts of selected compounds that have
been used to compare to the chemical shifts of NPM-DDVE and NMM-DDVE, and
thereby gain stereochemical information about the succinimide moiety. The cis
succinimide/succinic anhydride proton resonances appear downfield relative to those for

the trans conformer. Careful examination of the broad lH resonances between 2.6 and

32

 

 

 

 

 

 

 

 

 

 

~ k g
.0 ( d )
0
CH2 0 .l ' O
_ a
(Ctho N CH2 h
CH3 q m (CH2)s f
P " CH2 1 b+l+f
0 CH2 b
CH3 a
Cbc13 d,“
/ d, trans
{\J I I I I I . IrFI r FT I I I I I T I I I I I I I I I I I I I I I I I
’ \\5.0 4.5 4.0 3.5 3.0 2.5 ppm
‘ i
\\ l
manaoapaq \ I l
\ I, I
\ I;
\ , I
\ \\ h,d I( II /\
\ g m ’1 i
\ I /\ Ill

 

 

j I I I I I i I Y
2 1 m

chemical shift (ppm)

Figure 2.7. The 1H NMR spectrum of photoinduced NPM/DDVE copolymer in CDC13.

Band assignments are indicated according to the inset structure.

33

Table 2.1. ‘H and 13c chemical shift assignments for NPM-DDVE and NMM-DDVE

 

 

atom Succinimde IH Succinimde ”C
Chemical Shift (ppm) Chemical Shift (ppm)
a (DDVE methyl) 0.85 14
b, l, f, k (methylene) 0.9-2.4 (very broad) 32-20
c (methyl of NMM) 2.9 (broad) 27
e (methine of succinimide) 3.2 (cis) and 2.7 (trans) (NPM- 38 and 42 (NPM-DDVE)
DDVE) 37 and 41 (NMM-

Unresolved (NMM-DDVE) DDVE)
d(methine of succinimide) 3.2 (cis) and 2.7 (trans) (NPM- 48-54 (broad) (NPM-

DDVE) DDVE)
Unresolved (NMM-DDVE) 45-53 (broad) (NMM-
DDVE)
h (methylene adjacent to 3.3-3.7 (broad) 61 and 63 (both)
oxygen)
g (methine adjacent to 4.0-4.4 (broad) 74-77 (under
oxygen) CDCl3triplet)
m, n, o, p, q 7.0-7.5 (broad) 127,129,130,133
i, j (carbonyl groups) - 175-179 (NPM-DDVE)

175-180 (NMM-DDVE)

3.8 ppm for NPM-DDVE (Figure 2.7) reveals four distinguishable features. We assign
the features at 3.05 ppm and 2.7 ppm to the succinimide ring cis and trans conformers,
respectively. The cis peak area appears to be larger than that for the trans conformer.
This is, of course, only a qualitative assessment because of the extensive band overlap
present in this region. The NMM-DDVE 1H NMR spectrum (not shown) does not reveal
this same information because the succinimide proton resonances are obscured by those
of the N-methyl group. The ‘3 C spectra of the NPM-DDVE copolymer shows distinct
peaks from succinimide carbon from cis and trans configurations. These data are shown
in Figure 2.8. The succinimide carbon-e resonances between 48 and 54 ppm show two
unresolved peaks that are consistent with cis and trans conformers. Similarly, carbon-d

also shows two moderately resolved peaks at 39 ppm and 41 ppm. Unfortunately, the

34

resolution of these resonances is not sufficiently high to allow their quantitation. It is
clear, however, that both carbon-d and carbon-e in the copolymer exist as both cis and
trans conformers.

As noted above, in addition to the copolymerization of maleimides and vinyl
ethers by photoinduction, thermally induced copolymerization is also significant. It is
important to determine whether or not the copolymer formed by the thermal pathway is
the same as that formed by photoinduction. This information could bear directly on the
initial step in the polymerization reaction of this system and explain in part the
stereochemical data we have determined from the NMR measurements. The '3 C NMR
spectrum of thermally induced NPM-DDVE is shown in Figure 2.9. This spectrum is
essentially identical to that shown in Figure 2.8 for photoinitiated NPM-DDVE.

After understanding the stereochemical information on the non-crosslinked
copolymer, we are now in a position to interpret the solid-state CP/MAS spectrum of the
photoinduced BMMP-DVBS copolymer. The low resolution spectra presented in Figure
2.10 show the solid—state CP/MAS spectrum of the photo- and thermal-induced BMMP-
DBVS copolymer. The resonances are broad and featureless but the chemical shifts of
different bands are consistent with those of the NPM-DDVE and NMM-DDVE
copolymers. Specifically, we can assign the broad shoulder between 19 ppm and 22 ppm
to the methyl carbon-a of BMMP and methylene carbons-d of DVBS. The strong
resonance centered at ~ 32 ppm is due to methylene carbons-h and i from the BMMP and
DVBS monomers. The succinimide carbons-f and f’ do not reveal any splitting because

of the limited spectral resolution.

35

Table 2.2 1H and 13C chemical shifts of various model compounds at the

succinimide/succinic-acid/anhydride.

 

Model Compound

Succinimde/Succinic—
acid/anhydride lH
Chemical Shift (ppm)

15C
Succinimde/Succinic-
acid/anhydride
Chemical Shift (ppm)

 

meso-2,3,-dimethylsuccinic acid
(in CDC13)2°
rac-2,3,-dimethylsuccinic acid
(in CDC13)”
cis-3,4—Dimethyl-N-
phenylsuccinimde (in do-
DMSO)12
trans-3,4-Dimethyl-N-
phenylsuccinimde (in do-
DMSO)12

cis-hexahydro-N-
phenylphthalimide (in CDC13)12
trans-hexahydro-N-
phenylphthalimide (in CDCI3)12
cis-1,2-cyclohexanedicarboxylic
anhydride (in CDCl3)

trans-1,2-
cyclohexanedicarboxylic
anhydride (in CDC13)

Maleic anhydride-
tetrahydrofuran copolymer”
Model compound maleic
anhydride-tetrahydrofuran 1:1
adduct3O

cis-2,3-Dimethyl-N-
ethylsuccinimde3 I
trans-2,3-Dimethyl-N-
ethylsuccinimde3 1

Maleic anhydride-methyl vinyl
ether copolymer32
2-Propenylisocynate—maleic
anhydride copolymer26

2.54

2.65

3.36,28 3.2 27

2.95,28 2.85 27

3.3, 3.5

3.3

2.95

2.41

3.82 (cis) and 3.68 (trans)

42.05
41.03

37.99

42.42

40.06
47.52
45,28 39.8327

47,28 45.6027

46.6, 49.1

44.2

51.6-48.4, 41.1-37.8

57, 59 and 62, 64

 

36

0 ( (LA k 3 )

(“20‘ J ‘ .0 O
.. N
(CH2!) (4: CH2 '1 b.l.f.k
CH3 qs " ' m (CH2)8f
p‘ - n (“2 I g, CDC]3
O CHZ b
CH3 3

a

 

56 48 40

 

 

 

 

 

 

 

 

 

220 200 180 60 140 120 100 80 60 40 20 0
chemical shift (ppm)

Figure 2.8. The 13 C NMR spectrum of photoinduced NPM/DDVE copolymer system in

CDC13. Band assignments are indicated according to the inset structure.

37

 

 

 

 

 

 

 

 

 

 

 

 

i
b, l, f,k
Wit/IO .( Led 113,5 y, g,CDCI3
CH2O i Ii. 0 0
(Ctho N C113 h
J.’
CH3 q m (CHusf
pr . [1 CH2 1
0 CH2 b
CH3 a
i,j
M m. n. o, p. q.r e d a
J M
I “fa“
. . . 'I"I I .. m
180 176 50 40
g _
(WWJ h ° d NJ
NM
200 180 160 140 120 100 80 60 40 20 0

Chemical Shift, ppm

Figure 2.9. The 13 C NMR spectrum of thermally induced NPM/DDVE copolymer in

CDC13. Band assignments are indicated according to the inset structure.

38

I
'Qwatm QJ/\g()\£k/ 0%}:th k d RC 3

ro I i a O, a )1
b
f» Fri tfl‘,
5”“, i,
t \ Expanded 20- 80 ppm
ANV/ 1‘, region of therma y induced copolymer
\
k.k‘ sq.

 

80 60 40 20

 

 

 

 

 

M/ h ' l Thermally induced Copolymer
WWMJ \‘e_‘;AAY d-- st—w' “~12...“ .- AA.
300 250 200 150 100 50 0 chemical shift (ppm)
h , 1 l
f 1‘ ' A
b, b" CC=C ‘ ’ I Photoinduced Copolymer
AM It d
w k, k \
300 250 200 150 100 50 0 chemical shift (PPm)

Figure 2.10. CP/MAS spectra of BMMP/DVBS copolymer. Bottom: Photoinduced
BMMP/DVBS copolymer, Center: Thermally induced copolymer, Top: 0 - 90 ppm

expanded region of thermal copolymer spectrum.

The NMR data are useful, if only qualitatively, to evaluate the stereochemistry of the
polymerization and thereby infer the mechanism. Although unambiguous stereochemical
information on the BMMP-DVBS cross-linked copolymer can not be obtained from the
CP/MAS measurements, the presence of significant amounts of both conformers is clear
and points to the fact that the polymerization reaction is not concerted, despite the
implication of a DA complex as the propagating unit (vide supra).

Recently, Hoyle et al.4 reported on a maleimide-acrylate copolymerization, and their
findings are applicable to the interpretation of our data. Maleimide radicals can initiate

the alternating copolymerization in which maleimide acts both as photoinitiator as well as

39

comonomer. Using 1H NMR spectroscopy, Monnann and Schrnalz showed that protons
on substituted succinic anhydride exhibit more cis than trans conformers. Our semi-
empirical calculations indicate that the trans conformer of N-methyl(2,3-
dimethyl)maleimide is more stable than the cis conformer by 2.3 kcal/mol.

If the mechanism of D-A bond formation involves an intermediate state that bridges
the maleimide unsaturation, then the dominant conformer will be cis. If the bond
formation is mediated either by radical formation or H abstraction, then we expect that
there will be a statistical distribution of conformers with their relative populations being
related to the thermodynamic stability of the two conformers. The fact that we observe
what appears to be more cis than trans conformer in the final polymer can be attributed to
one of several possibilities. The first is that both mechanisms are operative and the
relative contribution of a cyclobutane-like intermediate to produce cis conformer (side-on
attack of the maleimide unsaturation) compensates for the thermodynamic advantage of
the trans conformer in the radical-mediated reaction. A second possibility is that the
transition state for the reaction is not related in an intuitive manner to either of the final
conformers but is such that the probability of forming the two possible conformers is
approximately equal. We note that a steric argument would seem to favor trans over cis
and it is thus unlikely that steric hindrance mediates the reaction.

As noted above, we recover substantially the same result for the thermally induced
copolymerization as for the photoinitiated reaction. This finding would argue that radical
initiation by H abstraction from a photoexcited triplet state, is the initial and most

important step in the reaction to determine the conformational distribution of the final

copolymer.

40

Table 2.3. Physical and thermal properties of various maleimide/vinyl ether copolymers.

 

 

Copolymer Mu Tg (°C) Ti' (°C) T2" (°C)
BMMP-DVBS (Photoinduced copolymer) NA NA 320 680
BMMP-DVBS (“Dark” Copolymer) NA NA 250 < 800
NPM-DDVE (Photoinduced Copolymer) 14310 96 345 430
NPM-DDVE (“Dark” Copolymer) 54054 95 370 430
NMM-DDVE 35000 83 350 420
BMMP-CVE* NA NA 380 < 800
BMMP-DDVE NA NA 350 < 505

. Ti is the temperature at which 10% weight is lost
"Tr is the temperature at which 90% weight is lost.

 

O

O

0 NA: Not Applicable.

* 1,5-bismaleimido-2-methylpentane/cyclohexylvinyl ether copolymer.

Thermal and Physical Properties of Copolymers. The thermal properties of

bismaleimide/divinyl ether copolymers has been reported on before,19‘20

and we present
our data in an effort to allow comparison to earlier work in this area. Table 2.3 shows
different physical and thermal properties of different maleimide-vinyl ether copolymer
systems. Both the photoinduced and thermal-induced NPM-DDVE copolymers show
roughly the same Tg. We are interested in understanding the role of cross-linking in
determining the physical properties of the polymers. We expect that the extensive cross-
linking characteristic of BMMP-DVBS will give rise to thermal degradation properties

that are superior to those for the corresponding non-crosslinked polymer. We show in

Figure 2.11 and Table 2.3 the thennogravimetric analysis (TGA)

41

 

 

 

100‘
804
$3
TX 60“
8
i—l
E.
'5 40‘
3
20'
O ' l r ' r ' r f 1%

0 100 200 300 400 500 600 700 800 900
Temperature (°C)

Figure 2.11. TGA curves for NMM/DDVE and BMMP/DVBS copolymers.

data for BMMP-DVBS and NMM-DDVE. The data demonstrate that the BMMP-DVBS
crosslinked polymer exhibits greater thermal stability than the NMM-DDVE system, as

expected.

2.4. Conclusion

We have investigated the polymerization of the bifunctional monomers BMMP
and DVBS and their monofunctional counterparts, NMM, NPM and DDVE. Based on
FTIR and NMR data we conclude that the polymerization of both systems proceeds with

1:1 stoichiometry. The structural 1H and 13 C NMR data indicate that neither the

photoinduced or the thermally induced polymerization produce stereospecific products.

42

Rather, the maleimide monomer copolymerizes with vinyl ethers to yield both cis and
trans conformers across its unsaturation. The formation of the same polymer structure by
both means of initiation suggests strongly that a radical mediated initiation step is rate
limiting. With this fundamental structural information in

hand, the next step in characterization of these polymer systems is to understand the
kinetics and mechanism of the actual polymerization event. We are in the process of
developing this understanding using time resolved spectroscopies to monitor the motional
and population dynamics of the maleimide moieties during polymerization. Making the
connection between dynamical data and the extent of cross-linking could provide insight

into controlling the molecular processes that give rise to the bulk properties of the formed

polymer.

2.5 Literature Cited

1. Kloosterboer, J. G.; Adv. Polym. Sci., 1988, 84, l.

2. Roffey, C. G.; “Photopolymerization of Surface Coatings,” Wiley, New York (1981).

3. Reiser, A.; “Photoreactive Polymers,” Wiley, New York, NY (1989).

4. Hoyle, C. E., Clark, S. C.; Jonsson, S., Shimose, M., Polymer, 1997, 38, 5695.

5. Miller, C. W., Hoyle, C. E., Howard C.; Jonsson, S.; Polym. Prepr., 1996, 37, 346.

6. Clark, S. C.; Jonsson, S.; Hoyle, C. E.; Polym. Prepr., 1996, 3 7, 349.

7. Decker, C.; Morel, F.; Jonsson, S.; Clark, S. C.; Hoyle, C. E.; PMSE Prepr., 1996, 75,
198.

8. Kohli, P.; Scranton, A. B.; Blanchard, G. J .; in preparation.

43

9. Wang, Z. Y.; Synth. Comm, 1990, 20, 1607.

10. Marquardt, D. W. J. Soc. Ind. Appl. Math. 1963, 11, 431.

11. Martuscelli, E.; Musto, P.; Ragosta, G.; Scarinizi, G. Polymer 1996, 37, 4025.

12. Olson, K. G.; Bulter, G. B. Macromolecules 1984, 17, 2480.

13. Olson, K. G.; Bulter, G. B. Macromolecules 1984, 17, 2486.

14. Bulter, G. B.; Olson, K. G.; Tu, C. L. Macromolecules 1984, 17, 1884.
15. Jonsson, S.; Ericsson, J. S.; Sundell, P.- E. Shimose, M.; Clark, S. Ch.; Miller, C.;

Owens, J .; Hoyle, C. 13. Conference Proc., RadTech North America, April 28 — May
2, 1 996, pp 377—392.

16. Lee, C.; Hall, H. K. Jr. Macromolecules 1989, 22, 21.
17. Cowie, J. M. G. “Alternating Copolymers”; Plenum: New York, 1985.
18. Barlett, P. D.; Nozalic, K. J. Amer. Chem. Soc. 1946, 68, 1495.
19. Tusuchida, E.; Tomono, T. Makromol. Chem. 1971, 141, 265.
20. Prementine, G. S.; Jones, S. A.; Tirrell, D. A. Macromolecules 1989, 22, 770.

21. Jones, S. A.; Tirrell, D. A. Macromolecules 1986, 19, 2080.

22. Jones, S. A.; Prementine, G. S.; Tirrell, D. A. J. Amer. Chem. Soc. 1985, 67, 5275.
23- (a) Decker, C. Decker, D. Conference Proceedings; RadTech North America, May 1-

5, 1994, pp 602-616; (b) Decker, C. Decker, D. Polymer 1997, 38, 229.

24- Hanna, M. W.; Ashbaugh, A. L. J. Phys. Chem. 1964, 68, 811.

25- Decker, C.; Moussa, K. Macromolecules 1989, 22, 4455.

26- Monnann, W.; Schrnalz, K. Macromolecules 1994, 27, 7115.

27- Hoyle, C. E.; Jonsson, S.; Shimose, M.; Owens, J. Sundell, P. E. ACS Symp. Ser.

1997, 673, 133.

44

28. Smith, M. A. Polym. Prepr. 1988, 29, 337.

29. Pouchert, C. J .; Behnke, J. The Aldrich Library of 13C and 1H FTNMR Spectra, 1st
ed.; Aldrich Chemical Co.: Milwaukee, WI 1993, vol. 1.

30. Regab, Y. A.; Butler, G. B. J. Polym. Sci., Polym. Chem. Ed. 1981, I9, 1175.

31. Renaud, R. N.; Champagne, P. J. Can. J. Chem. 1979, 5 7, 990.

32. Ding, L. Polymer 1997, 38, 4267.

45

Chapter 3

Design and Growth of Robust Layered Polymer Assemblies with
Molecular Thickness Control

Abstract

We report on two alternating copolymers synthesized from maleimide and vinyl
ether monomers, where the vinyl ether possesses a pendant phosphonate functionality.
Partial hydrolysis of the phosphonate groups in conjunction with Zr-phosphonate layer
growth chemistry produces robust, stable polymer multilayer structures where the layer
thickness is 16 A for polymers synthesized with N-phenylmaleimide (NPM) monomer
and 31 A for polymers synthesized with N-biphenylmaleimide (NBM) monomer. We
present the syntheses and characterization of the polymers and layered assemblies
containing up to 10 layers. XPS data on layer-dependent Zr loading suggest interlayer

bonding that is similar to that seen for Zr—alkanebisphosphonate multilayers.

46

3.1 Introduction

The design and synthesis of thin films and chemically modified surfaces has been
an area of intense research activity because of the potential utility of these structures.
Various applications, including optical second harmonic generation,1 chemical sensing,2
electrical or environmental isolation,3 electronic rectification,4 and photoreactivity 5'7
have been either proposed or demonstrated, underscoring the importance of materials
advances in this area. Of particular significance to interfacial materials and thin films is
the ability to grow layered materials where there is good control over the layer thickness
and uniformity. Layer-by-layer deposition of films can provide spatial resolution and
directionality, and both of these structural properties can be critical to the macroscopic
properties of the system. To achieve controlled, layer-by-layer growth requires the
development of efficient and robust means of connecting individual layers. Several
different techniques have been devised for linking individual molecular layers, including

12-18

- . . . . 9 .
covalent,8 H ionic - covalent, coordination,1 charge-transfer,20’21 hydrogen bonding,22

and alternate adsorption of oppositely charged polyelectrolytes.”25

Many of these
methods are well suited to the deposition of multilayers of small molecules and can be
used for the deposition of polymers under certain circumstances. For several of the
linking methods, however, the stability of the resulting structures is limited under
conditions of high temperature or solvent exposure. Only the approaches that use
covalent or strongly ionic interlayer linking chemistry can withstand thermal and solvent
attack. We have combined the advantages of metal phosphonate interlayer linking

chemistry with the physical robustness of vinyl ether-maleimide alternating copolymers26

to produce well controlled polymer multilayer structures. We report here on the synthesis

47

and layer—by—layer deposition of poly(N-phenylmaleimide-(2-
vinyloxy)ethylphosphonate), (NPM-VEP), and poly(N-biphenylmaleimide-(2-
vinyloxy)ethylphosphonate), (NBM-VEP) on oxidized silicon and silica substrates using
Zr-phosphonate interlayer linking chemistry. We have pursued a strategy of layer-by-
layer growth because the resulting materials can, in principle, possess significant
advantages over polymer materials deposited by conventional spin coating methods. For
example, layer-by-layer growth offers better control over the thickness of ultrathin
polymer layers, on the order of tens of A, the formation of macroscopic defects such as
bubbles or pinholes is minimized, and the layers formed can be both chemically and

thermally robust.

agar

fr

0
A Vow /0 n
O/ N \O + P<'OiPr CHC13,AIBN
'Pr 7 / \

70°C, 24 hrs. 0 N O

\
01

Scheme i. Synthetic route for poly(NPM-VEP). See text for details.

 

3.2 Experimental Section

The alternating copolymers of N-phenylmaleimide (NPM) and (2-vinyloxy)-

ethylphosphonate (VEP) and of N-biphenylmaleimide (NBM) and VEP are prepared by

26.27

radical copolymerization using AIBN as the initiator. The reaction between NPM

48

and VEP is shown in scheme 1 and is the same save for the maleimide monomer for
poly(NBM-VEP) synthesis. No monomer homopolymerization was seen for our reaction
conditions. VEP is prepared by the reaction of excess tri(isopropyl)phosphite with 2-
chloroethylvinylether at 170°C for 5 days in an argon atmosphere.28 Distillation of the
product yields ~72% VEP. 1H NMR (300 MHz, d-CHC13): 5 1.3 (d, 12 H of isopropyl),
2.1 (m, 2 H of CH2 adjacent to phosphonate group), 3.9 (m, 2 H adjacent to oxygen of
vinyl ether), 4.0 and 4.2 (dd, 1 H each of vinyl group), 4.6 (m, 2 H of isopropyl group),
6.3 (dd, 1 H from vinyl group adjacent to oxygen).

The monomer N-phenylmaleimide was purchased from Aldrich Chemical Co. and
used after recrystallization from hexanes. N-biphenylmaleimide was synthesized
according to the following procedure. 4-Phenylaniline (1 g) was dissolved in CHC13 (10
mL) and added drop-wise to a solution of maleic anhydride (0.71 g) in CHC13 (5 mL)
over a period of 1 hour. The reaction was allowed to stir for an additional 2 hours. The
resulting amic acid appeared as a bright yellow precipitate and was separated from the
supernatant by filtration. Biphenylamic acid (0.4 g) was cyclized by heating to ~60°C for
2 hours in a solution of 6.4 mL of acetic anhydride and 0.072 g of anhydrous sodium
acetate. The resulting solution was cooled to room temperature and added to 50 mL of
ice water. The N-biphenylmaleimide product was filtered, dried and recrystallized from
hexanes (80% yield). lH NMR (300 MHz, dé-DMSO): 5 7.20 (2H, s), 5 7.46 (5H, m), 5
7.70 (2H, d) 5 7.77 (2H, d).

The copolymer is synthesized by reacting equimolar amounts of the maleimide
and VEP in CHC13 at 80°C under a N2 atmosphere for ~18 hours using AIBN as the

initiator. 1H NMR (300 MHz, (16 — DMSO): 5 1.0 - 1.4 (6 H, VEP isopropyl groups),

49

5 1.8 — 2.2 (2 H, VEP CH2 adjacent to phosphonate group), 5 3.0 — 4.0 (2 H , succinimide
ring + 2 H adjacent to vinyl ether oxygen + 2 H from ethyl group), 5 4.3 — 4.7 (1 tertiary
isopropyl H + l tertiary H adjacent to oxygen), 5 7.0 — 7.4 (5 H, phenyl ring). ‘3 C NMR
(d-CHC13, 75.46 MHz) :5 24, 29.5, 38 - 42, 46 — 54, 71, 77, 125 — 135, 174 —180.

For our experimental conditions, statistically one of the isopropyl groups
terminating each phosphonate oxygen is hydrolyzed to yield a hydroxyl group during the
course of the polymerization. This displacement is likely due to the presence of HCl
formed in CHCl3 solution by AIBN. GPC characterization of the resulting NPM-VEP
polymer shows Mn = 7200, MW = 10800, yielding a polydispersity of 1.5. The reaction of
either poly(NPM-VEP) or poly(NBM-VEP) with bromotrimethylsilane in anhydrous
CH2Cl2 at room temperature for ~2 hours yields a polymer that is partially hydrolyzed,

9 The characterization of

making it useful for the formation of multilayer assemblies.2
these films accomplished using optical null ellipsometry (Rudolph Auto-EL II), 13C
(100.6 MHz) and 31P (161.9 MHz) solid state CPMAS NMR (Varian VXR 400 MHz),
FTIR (Nicolet Magna 550) and UV-Visible (Unicam model UV-2) spectroscopies. We

discuss the results of these measurements in the following section.

3.3 Results and Discussion

The primary focus of this paper is on the synthesis and properties of layered
poly(NPM-VEP) and poly(NBM-VEP) where the layer-by-layer growth of the material is
controlled with molecular thickness resolution. We consider first the formation of the
layers and the chemical basis for layer formation using this polymer. We then consider

the nature of the polymer itself, specifically whether it is an alternating copolymer or a

50

random copolymer, because this issue bears on the extent to which these polymer layers
can be considered homogeneous. We discuss the potentially complex structural issues
associated with the formation of these polymer layers, including the Zr/P ratio in the
films as a function of number of polymer layers. These data in conjunction with FTIR
measurements point to the incomplete reaction of hydrolyzed direct priming chemistry of
silanol groups with POC13 and growth of layered assemblies using partial hydrolysis of
the -P03HR functionalities to control the available concentration of active PO32' sites
layer by layer.

The deposition of polymer layers is shown schematically in Figure 3.1. For
ellipsometry measurements we have used Si substrates. To grow these layered materials,
the silicon substrate is treated to produce 3 ~15 A thick native oxide layer. The resulting
silanol groups are phosphorylated using POC13 in dry acetonitrile, followed by hydrolysis
and zirconation in ZrOCl2 solution. This procedure is a departure from other reports on
the preparation of Si and SiOx substrates for ZP layer growth. Typically the bare
substrates are oxidized (Si) and hydrolyzed (Si and SiOx) in the same manner as we have
done here. Subsequent to the preparation of a hydrolyzed SiOx layer, treatment with
either triethoxyaminopropylsilane or methoxydimethylaminopropylsilane to yield an
aminated surface is the typical procedure. Reaction of the amino-terminated surface with
POC13 and H20 produces an aminophosphonic acid surface. In this work, we treat the
surface silanol groups directly with POC13 then H2O to produce a surface with the same

properties as that achieved with the more widely used Silane-based chemistry.

51

“OH
POC13

 

“”OH

 

 

 

 

 

 

l. ZrOCl2 (2:3 _
WaterzEthanol)
»

2. NPM -VEP
Copolymer in "—0—— P\
acetonitrile, 4V
50°C for 12 hours

 

 

 

 

 

 

fl
——0
BrSiMe3 in CHZCIZ,
2 hrs, 20°C > 81
-—O
3/

 

 

 

 

> Repeat Steps 2 and 3 for additional layers

Figure 3.1. Schematic representation of layered growth of poly(NPM-VEP) indicating
phosphonate groups and thus the possibility of interlayer ZP bonding between non-
adjacent layers. The oxidation states of zirconium and oxygen (Zr4+ and 02') are not

shown for simplicity.

52

Polymer layers are formed on the primed and zirconated substrate by adsorption
of either poly(NPM-VEP) or poly(NBM-VEP) from acetonitrile at 50°C over ~12 hours.
Our XPS data (vide infra) suggest that Zr4+ forms a strong complex with two phosphonic
acid groups of the copolymers while the remaining, partially hydrolyzed phosphonates
are used in the formation of the next layer after displacing the remaining isopropyl groups
with bromotrimethylsilane (Figure 3.1). We have demonstrated this procedure to form
ten layers of poly(NPM-VEP) and eight layers of polymBM-VEP) on oxidized silicon.
We do not intend to imply that these are limiting values for the number of layers that can
be formed with this chemistry. Indeed, we see no evidence that would suggest any
decrease in reactivity with the addition of polymer layers. We present measurements of
the ellipsometric thickness as a function of the number of layers in Figure 3.2. Each data
point represents the average of 27 measurements. Because the true refractive indices of
the system are unknown, we used 11 = 1.54 + 0i as the refractive index for the calculation
of thicknesses. We have used this value of n for other ZP systems. For poly(NPM-
VEP), the lepe of the line is 16.0 i 0.9 A with an intercept of 9.8 i 5.5 A. The intercept
corresponds to the thickness of the oxidized layer on the Si substrate. For poly(NBM-

VEP), we recover a thickness of 31.7 i 1.3 A per layer with an intercept of 12.2 i 6.7 A.

53

00

N

O
I

8888

Ellipsonretric Thickness (A)
§
0

 

 

0‘58

 

0 2 4 6 8 10 12
Number of Layers

Figure 3.2. (a) Ellipsometric data for ten layers of poly(NPM-VEP). (b) Ellipsometric
data for eight layers of poly(NBM-VEP). The layer thicknesses are 16 A/layer for

poly(NPM-VEP) and 31 A/layer for poly(NBM-VEP).

54

The ellipsometric thickness data for these two polymer films can be used in
concert with absorption and FTIR data to provide some insight into the organization of
the layers. The most obvious conclusion that can be drawn from these data is that simple
steric factors play an important role in determining the morphology of the individual
layers. The observed linear increase in thickness with number of layers for poly(NBM-
VEP) is consistent with the linear dependence of the film absorbance on number of layers
for films grown on a quartz substrate (Figure 3.3). The absorptive resonance centered at
260 nm is associated with the N-succinimidobiphenyl chromophore. Finding a linear
relationship between film thickness and number of layers is certainly not a surprising
result for simple ZP systems where the organobisphosphonate layer constituents are
arranged in a sufficiently orderly manner to allow quantitative predictions of layer
thickness to be made and verified experimentally. For these polymers, where the layer
structure and interlayer bonding arrangements are expected to be highly irregular, it may
therefore be somewhat surprising to find that the layer growth appears to be so regular.
The simplest interpretation of these data is that the linear alternating copolymers lie
approximately flat on the interface surface and the density of the layers is significant. We
can estimate the density from the absorbance measurements. For substituted biphenyl
chromophores, the extinction coefficient of the 260 nm resonance is typically on the
order of ~18,000 L/mol cm.30 This value for the extinction coefficient corresponds to an
absorption cross section of 3 x 10'17 cmZ/chromophore. The data shown in Figure 3.3
yield an absorbance of 0.01 per bilayer (single layer on each side of the substrate),

consistent with a surface density of 1.65 x 1014 chromophores/cmz-layer. If half of the

55

polymer phosphonate groups are hydrolyzed at the time of deposition, the interlayer

bonding density would be ~ 8 x 10'3/cm2-layer. Modeling a silica surface as a cubic

   
        

 

 

005. ”w”‘
(109 g 004' wy"’vfl
0.08 < 003. 3.52""
,,g(107 002' ‘wy’wv
3 0.06 0,017“ i 3 it 3
5 layers number of layers

 

 

200 225 250 275 300 325 350
Wavelength (nm)

Figure 3.3. Absorbance of poly(NBM-VEP) as a function of number of layers. The
bands at 200 nm and at 260 run both exhibit a linear dependence of absorbance on

number of layers. For the 260 nm band, the absorbance is ~ 0.01 a.u./layer (inset).

56

O O” A

A k \\’ OH .
Pl © O\\PO}i)l-l ‘0 '
O\\\\/N\70

O
15 A o —— »~/—¥ 21 A
\ O
O/ N \O 2 OXNAO g
P
0.5011 ,K
V I: :1 0011 © 0.0%?“

 

 

Figure 3.4. Schematics of poly(NPM-VEP) and poly(NBM-VEP). The distances
indicated for the oligomers were determined using molecular mechanics calculations.
The structures presented in the Figure are not energetically optimized and are intended

for illustrative purposes only.

close packed array of silanols with a 5 A separation between silanol groups yields a
surface silanol density of 4 x 1014 sites/cmz. This simple calculation certainly
overestimates the silanol density for a flat surface and at the same time fails to account
for surface roughness, and these two oversimplifications act in opposition to one another.
Given that this is intended to be a qualitative estimate, the agreement between measured
chromophore density and estimated bonding site density is remarkable, and it suggests
that something on the order of a third of the available bonding sites on the substrates are

occupied.

57

If the estimate of polymer layer density is close to correct, the measured thickness
of 16 A/layer for poly(NPM-VEP) and 31 A /layer for poly(NBM-VEP) suggests that on
average, each polymer layer is about one molecule thick. Molecular mechanics
calculations on the basic oligomer unit suggest a 15 A per layer thickness for the
poly(NPM-VEP) and 21 A per layer for poly(NBM-VEP), as indicated in Figure 3.4. It
is likely that the thicknesses we measure that are greater than those predicted by
molecular mechanics calculations are reflective of the complex and likely entangled
structure of the polymer molecules within each layer.

An important consideration for these materials is the nature of the ZP chemistry
used to connect polymer layers. One way to determine the nature of interlayer bonding is
through XPS measurements of the layer constituents. The XPS data we report here
(Figures 3.5) suggest that the Zr(PO3R)2 layers are very similar for polymer and
alkanebisphosphonate multilayers. For ZP multilayers formed using bisphosphonated
alkanes, the spacing between the active phosphonate sites is determined either by the
substrate or by the organic gallery constituent but, in either case, phosphonate groups are
sufficiently close to allow the formation of Zr(PO3R)2 sheets that resemble the structure
of solid Zr(PO3R)2.

We show in Figure 3.5a a survey scan of an 8-layer film of poly(NPM-VEP) on
an oxidized Si substrate. We assign the peaks as follows; Si (2p resonances at 98.7 eV
for elemental Si and 102.5 eV for $10,), P (25 at 191.5 eV, 2p at 135 eV), Zr (3d5/2 at
183.5 eV, 3d3/2 at 185.8 eV), C (18 at 285 eV and 289 eV), N (Is at 401.5 eV), 0 (Is at

532.5 eV). In addition , there are Auger resonances for O at ~ 750 eV and C at ~ 1 KeV.

58

 

 

 

10:
9:" a
8.‘ 018 C15
7—
E 6; C Q
A LAuger
E‘l 5' Auger le It 3d
2 4W ‘
3- ”Lev/JAN
2
l
0

:. V 21.)
f- l l l lzr 3MP

1000 800 600 400 200 0
Binding Energy (eV)

 

1—1 1—1 r—A
N b O\
I I I l I I

Zr/P ratio
2
I!)

O O
exec
I
o
o
o

 

0 2 4 6 8 10
number of layers

Figure 3.5. X-ray photoelectron spectroscopic (XPS) data for layers of poly(NPM-VEP).
(a) Survey scan indicating the elements present. (b) Ratio of Zr/P determined from XPS

data as a function of number of layers.

59

There are indications of changes in the interface thickness as a function of layer
addition through the elemental ratios. As the number of poly(NPM-VEP) layers
increases, we observe a change in the relative amount of Zr4+ from 0.46% for a
monolayer to 3.27% for 8 layers. We also observe a change in the Cls/Si2p ratio, which
varies from 3.3 for a monolayer to 22.0 for 8 layers. Both pieces of information point to
layered growth. For the Zr data, the change in measured composition is not due to
changes in the metal ion loading density, but rather is a consequence of the screening of
the substrate by the polymer layers.

We can extract information on the interlayer chemistry from the XPS Zr/P ratio.
The Bein group has reported that, for up to 3 layers of 1,10-decanediylbis(phosphonic
acid) on carbon fibers, the experimental Zr/P ratio is ~1.l3' They have calculated the
Zr/P ratio for an ideal system with stoichiometric amounts of Zr and P and have reported
that the measured ratio should decrease from ~l for a monolayer to about 0.6 for a 3-
layered structure. The Mallouk group calculated a Zr/P ratio of 0.63 i 0.08 for
multilayers formed from a-Zr phosphonate and polycation polymer layered structure,
where they used an inelastic mean free path of 15 A for both Zr and P photoelectrons.32.
We recover experimentally for our polymer multilayers a layer-dependent Zr/P ratio that
is consonant with these predictions. Our data for poly(NPM-VEP) show the same
elemental ratio as that for multilayers of 1,l2-dodecanediylbis(phosphonic acid). We
show these data in Fig. 3.5b. The Zr/P ratio decreases as the number of layers increases
for both systems. For a phosphonated and zirconated substrate, we recover a Zr/P ratio of
1.6 for each substrate. This ratio is the same for direct phosphonate-primed substrates

and for those primed using aminopropylsilane chemistry. These results suggest that the

60

phosphonate group density is substantially the same for both surface treatments. We note
also, as has been considered before, that the elemental ratios recovered depend on
stoichiometry, layer thickness and mean free path of the X-ray photoelectrons. For a
single polymer layer we recover a Zr/P of 1, which decreases to 0.66 for 8 layers. Our
data are fully consistent with other reports on the Zr/P ratio in similar systems, and based
on these data, it appears that the interlayer bonding for the polymer multilayers is the
same as for the presumably more ordered Zr-DDBPA layers. These XPS data are also
significant in that they demonstrate our ability to react the protected isopropyl
phosphonate firnctionalities stoichiometrically. This is a concern due to the statistical

nature of the hydrolysis/deprotection chemistry we use.

 

0.009 ~
1
0.006 —
.1
0.003 _.
0.000 -M g - .

 

2500 2000 1 500 1000
frequency (cm' 1)

3500 3000

Figure 3.6. FTIR spectrum of a bilayer of poly(NPM-VEP) on oxidized Si. Band

assignments are as indicated in the text.

61

To test the stability of the poly(NPM-VEP) and poly(NBM-VEP) multilayers, we
have immersed films of both polymers in n-hexadecane at ~ 100°C and in boiling
ethanol, for 2 hours in each case. These solvents were chosen to test the solubility of the
polymers in both polar and nonpolar environments. In all cases, we find no change in
either ellipsometric thickness or F TIR spectra resulting from exposure to solvent. This
result is consistent with ZP multilayers formed from small molecules. We believe that
the primary reason for the insolubility of the polymer multilayers is the same as that for
the small molecule systems; the Zr(O3PR)2 structure is characterized by its sparing

solubility in most solvents.

1 2 10
11 x (I Y
0 3 N 4 012 ”2 o (clam
3.4 9:635 13 ”2 16 (7):, ,,
8 6 HO II‘ Oct]. CH3 HO !\OH

A 5, 6, V 8,9and* 7 l4 CH3
. .. h 15
/ i [U x ( X/y ~ 2) y
1* l *
// / ii /\'\
/J a“ / um w/ 1 12,14 WW“)
,2 \v u, R J We“ PM" NWMAB
, if \1 0” Vi, ,1
,etfw‘wy'wi «1 iv.”

 

180 160 140 120 100 80 60 4O 20
Chemical Shifi, ppm * denotes spinning side band

Figure 3.7. 13C NMR spectrum of poly(NPM-VEP) grown on high surface area porous
silica. Resonance assignments are as indicated in the inset. The absence of discernible
progressions in the spectrum indicates the alternating nature of the polymer. The initial
ratio of fully to partly hydrolyzed phosphonate is ~2 monohydrolyzed to 1 dihydrolyzed
group based on NMR integration of iPr methyl group protons relative to phenyl ring

protons.

62

It is important to characterize the polymers within the multilayer structure. We
show in Figure 3.6 the FTIR spectrum of a bilayer of poly(NPM-VEP) adsorbed on a Si
substrate. The bands in the 2700 cm'1 — 3100 cm'1 region are the CH stretching
resonances. While there is a substantial body of literature relating band position to layer
structure for alkanethiols on gold, the analogous information is not available from the
data for this polymer because of the relatively small amount of aliphatic CH2
functionality in this system. Carbonyl group stretching resonances are seen in the ~1700
cm'l - 1800 cm'1 region and the P=O and P-O bands lie in the 1000 cm'l-1200 cm'I range.
While these data confirm the presence of the functional groups and are consistent with
the formation of layers of poly(NPM-VEP), there is little explicit structural information
on the polymer contained within them. In an effort to better understand the nature of the
polymer layers, we have acquired 13 C and 3'P CPMAS NMR spectra of copolymers on
high surface area silica.33 We show in Figure 3.7 the '3 C CPMAS spectrum of
poly(NPM-VEP) adsorbed on silica. The resonances in the region of 10 - 50 ppm belong
to the aliphatic carbons in the polymer backbone and side chains. The carbonyl - and
phenyl - carbon resonances associated with the N-phenylsuccinimide moiety are observed
in the 150 - 180 ppm region and 120 - 140 ppm region, respectively. The form of these
data, especially the absence of discemible progressions, suggests that poly(NPM-VEP) is
an alternating copolymer. We recognize that the characterization of poly(NPM-VEP)
copolymerization by NMR is limited by the spectral linewidths, but these findings are
consistent with the fact that neither monomer homopolymerizes under our experimental

conditions.

63

 

EH1

1'10

[1)

11:

p0:

The characterization of the polymer layers using optical null ellipsometry, FTIR
and '3 C NMR demonstrates the presence of multiple polymer layers. It is also important
to evaluate the extent to which these layers resist chemical and thermal exposure. We
have used a ten-layer sample for these experiments. Exposure of this sample to boiling
ethanol (78°C) for one hour and to boiling hexadecane (120°C) for two hours produced
no significant changes in ellipsometric thickness or FTIR spectra. These data indicate

significant mechanical and chemical stability for these layers.

3 .4 Conclusion

We have synthesized poly(NPM-VEP), a chemically and thermally stable
alternating copolymer. By taking advantage of a strategy of partial, stepwise hydrolysis,
we have demonstrated the ability to form robust layered polymer structures. Because it is
possible to construct these materials with molecular layer resolution, control over the
chemical identity of each layer can be established by simple exposure to different
polymers for each layer. We expect that the N-substitution of the succinimide moiety
will lead to control over polarity of the layers and control over the length of the vinyl
ether phosphonate will allow adjustment in layer thickness. We anticipate these novel
materials to find utility in the design of chemically selective surfaces with controlled

porosity.

64

3.5 Literature Cited

N

Katz, H. E.; Scheller, G.; Putvinski, T. M.; Schilling, M. L.; Wilson, W. L.; Chidsey,
C. E. D.; Science 1991, 254, 1485. (b) Li, D.; Ratner, M. A.; Marks, T. J.; Zhang, C.

H.; Yang, J.; Wong, G. K. J. Am. Chem. Soc. 1990, 112, 7389.

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

Swalen, J. P.; Allara, D. L.; Andrade, J. P.; Chandross, E. A.; Garoff, S.; Israelachvilli,
J .; McCarthy, T. J .; Murray, R.; Pease, R. F.; Rabolt, J. F.; Wynne, K. J. Tu, H.
Langmuir 1987, 3, 932.

Metzger, R. M.; Wiser, D. C.; Laidlaw, R. K.; Takassi, M. A.; Mattern, D. L.; Panetta,
C. A. Langmuir 1990, 6, 350.

Dulcey, C. S.; Georger, J. H., Jr.; Krauthamer, V.; Stenger, D. A.; Fare, T. L.; Calvert,

J. M.; Science 1991, 252, 551.
Calvert, J. M.; Georger, J. H., Jr.; Peckerar, M. C.; Perhsson, P. E.; Schnur, J. M.;

Scheon, P. E. Thin Solid Films 1992, 210/211, 359.

Kumar, A.; Biebuyck, H. A.; Abbott, N. L.; Whitesides, G. M. J. Am. Chem. Soc.

1992, 114, 9188.

. Netzer, L. Sagiv, J. J. Am. Chem. Soc. 1983, 105, 674.

. Tillman, A. Ulman, A.; Penner, T. L. Langmuir 1989, 5, 101.

10. Liu, Y.; Bruening, M. L.; Bergbreiter, D. E.; Crooks, R. M.; Angew. Chem. Int. Ed.

Engl. 1997, 36, 2114.

65

ll. Kohli, P.; Taylor, K. K.; Harris, J. J.; Blanchard, G. J. J. Am. Chem. Soc., 1998, 120,

11962.

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

618.
13. Lee, H.; Kepley, L. J.; Hong, H.-G.; Akhter, S.; Mallouk, T. E. J. Phys. Chem. 1988,

92, 2597.

14. Akhter, S.; Lee, H.; Hong, H.-G.; Mallouk, T. E. White, J. M. Acc. Chem. Res. 1992,

25, 420.

15. Putvinski, T. M.; Schilling, M. L.; Katz, H. E.; Chidsey, C. E. D.; Mujsce, A. M.;

Emerson, A. B. Langmuir 1990, 6, 1567.

16. Katz, H. E.; Schilling, M. L.; Chidsey, C. E. D.; Putvinski, T. M.; Hutton, R. S.

Chem. Mater. 1991, 3, 699.

17. 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.
18. Thompson, M. E. Chem. Mater. 1994, 6, 1168.

19. Bell, C. M.; Keller, S. W.; Lynch, V. M.; Mallouk, T. E. Mater. Chem. Phys. 1993,

35, 225.

20. Shimazaki, Y.; Mitsuishi, M.; Ito, S.; Yamamoto, M. Langmuir 1997, 13, 1385.
21. Shimazaki, Y.; Mitsuishi, M.; Ito, S.; Yamamoto, M. Langmuir 1998, 14, 2768.

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

66

23.

24.

25.

26.

27.

28

29

30

31

32

33

Decher, G.; Hong, J .-D. Makromol. C hem, Macromol. Symp. 1991, 46, 321.

Decher, G.; Hong, J.-D. Ber. Bunsen — Ges. Phys. Chem. 1991, 95, 1430.

Decher, G.; Hong, J.-D.; Schmitt, J. Thin Solid Films 1992, 210/211, 504.

Kohli, P.; Scranton, A. B.; Blanchard, G. J .; Macromolecules, 1998, 31, 5681.
The copolymerization of maleimide and vinyl ether usually results in formation of
alternating copolymers, for example, see: (a) Olson, K. G.; Bulter, G. B.
Macromolecules 1984, 1 7, 2480. (b) Olson, K. G.; Bulter, G. B. Macromolecules
1984, 1 7, 2486.

. Rabinowitz, R.; J. Org. Chem. 1961, 26, 5152.

. McKenna, C. E.; Schmidhauser, J. J. Chem. Soc., Chem. Comm. 1979, 739.

. Berlman,l.B.; .o 1-10-00.0 : f 4. 011a. '

   

Edition, Academic, New York, 1971.
. Hoekstra, K. J.; Bein, T.; Chem. Mater, 1996, 8, 1865.

. Kim, H.-N.; Keller, S. W.; Mallouk, T. E.; Chem. Mater, 1997, 9, 1414.

. The procedure and conditions for adsorption of poly(NPM-VEP) layer on silica
having a surface area of 450 m2/ gm are same as described above for the adsorption of
polymer on Si substrate except that the silica mixture was stirred during

phosphorylation, zirconation and deposition of polymeric layers.

67

Chapter 4

Probing Interfaces and Surface Reactions of Zirconium —
Phosphate/Phosphonate Multilayers Using 33P NMR Spectrometry

Abstract

We report our 3lP magic angle spinning (MAS) NMR characterization of
zirconium-phosphate/phosphonate (ZP) multilayer assemblies grown on SiOx. The
reaction of silica with excess POC13 and treatment with collidine results in both
physisorbed and chemisorbed HXP04("'3) being present at the SiOx surface. The 3 IP NMR
spectrum of zirconium phosphate grown from silica shows no residual Cl-containing
species, indicating essentially complete hydrolysis. The phosphate 31P resonances are
broadened substantially upon complexation with Zr4+ and T. data demonstrate the
broadening to be inhomogeneous in nature. Data on maleimide-vinyl ether copolymer
layers we reported recently [Langmuir, 1999, I5, 1418] confirm that
hydrolysis/deprotection by bromotrimethylsilane is efficient in converting phosphoesters
to the corresponding phosphorus oxyacid in the synthesis of multilayers. Comparison of
polymer multilayers with the relatively more ordered bisphosphonic acid-based
multilayers indicates that essentially the same interlayer bonding chemistry is seen for
both systems, with subtle differences arising as a consequence of the greater available

free volume and more extensive disorder within the polymer.

68

4.1 Introduction

The organization of mono- and multilayer molecular assemblies at solid surfaces
provides a rational approach for fabricating interfaces with a well defined structure,
composition, and thickness. These assemblies may ultimately find application in
nonlinear optical devices,“2 chemical sensing,3 surface passivation,4 photoreactivityf'7
and separationsg‘9 The ability to control interfacial processes has important implications
for both fundamental and technological advances. Of particular significance is the ability
to grow layered materials where there is substantial control over the layer thickness and
uniformity. Layer-by-layer deposition of films provides spatial resolution over
composition and molecular orientation relative to the substrate. Both of these properties
can be critical to the macroscopic properties of the resulting system.

We use metal-bisphosphonate chemistry to produce multilayer interfacial
materials because of its combined simplicity, versatility and robustnessl‘m'27 To grow
well defined layers having low defect densities, it is important to understand surface
reactions at various stages during the formation of multilayers. We report here on the
characterization of layered interfaces formed by adsorption of selected compounds using
zirconium-phosphate/phosphonate chemistry. The substrate upon which the layers are
grown is SiOx in the form of silica gel. We use 31F solid-state magic angle spinning
(MAS) NMR spectrometry to characterize the surfaces after each reaction step. We
chose 3 ]P MAS NMR spectrometry because the chemical shift of 31P is very sensitive to
its surrounding environment and is thus a good indicator of the species present during
selected stages in layer-by-layer deposition reactions. The high sensitivity of the 31P

NMR measurement is the result of the 100% natural abundance of 31P. The data we

69

report here for both polymeric and alkanebisphosphonate multilayers indicate that
essentially the same interlayer linking chemistry is operative for both types of system, but
subtle differences in the efficiency of the layer formation arise from the disorder and

permeability that is characteristic of the polymer multilayers.

4.2 Experimental Section

The alternating copolymer of N-phenylmaleimide (NPM) and (2-vinyloxy)-
ethylphosphonate (VEP), poly(NPM-VEP), was prepared by radical polymerization using
azobisisobutyronitrile (AIBN) as an initiator.33’39’32 VEP was synthesized by the reaction
of tri(isopropyl)phosphite and 2-chloroethylvinyl ether as reported previously. ”’32 NPM
was purchased from Aldrich Chemical Co. and used after recrystallization from hexanes.
POC13, 2,4,6-trimethylpyridine (collidine), anhydrous acetonitrile, bromotrimethylsilane
(BTMS), and AIBN were also purchased from Aldrich Chemical Co. and all were used as
received. Silica gel (230 — 400 mesh) having a surface area of 450 - 550 mz/g was
purchased from Spectrum Quality Products, Inc.

We synthesize multilayers using the metal-phosphonate (MP) chemistry

16-19

pioneered by the Mallouk,30’15 Thompson and Katz groups.”27 Multilayers grown

using MP chemistry are, in many cases, deposited on substrates having surface hydroxyl
groups. The first step in layer formation on such surfaces is typically treatment of the
substrate with ethoxydimethylaminopropylsilane to provide an amine-terminated surface,

24,31

followed by reaction with POC13 and a Lewis base such as collidine. Hydrolysis

yields a phosphonate-tenninated surface. We have reported the formation of robust

33.34

multilayer assemblies containing polymers,32 and small bifunctional molecules using

direct reaction of surface silanols with POC13 as the substrate-priming step. Both

70

ellipsometry and F TIR data reveal no significant differences between the layers grown
with and without the Silane priming chemistry. In both the cases the resulting layers were
not susceptible to removal by polar solvents, nonpolar solvents or aqueous electrolyte
solutions.32’3 3

The direct chemisorption of phosphate on the substrate silanol groups was carried
out by adding 0.4 M POCl3 in dry acetonitrile drop wise to a stirred suspension of silica
gel in 0.4 M collidine in anhydrous acetonitrile under argon. The reacted silica gel was
washed thoroughly with reagent grade acetonitrile and water. The phosphated surface
was zirconated using 5 mM ZrOC12 in 60% ethanol(aq) for l to 2 hrs at room
temperature. The polymer layers were deposited on the zirconated substrate from a 10
mM solution of partially hydrolyzed poly(NPM-VEP) in acetonitrile at 45°C for 12
hours. Partial hydrolysis of the polymer prior to deposition is accomplished by
deprotection of the phosphonate ester groups with 0.5 eq. of BTMS.32 Following initial
layer formation, exhaustive phosphoester hydrolysis to activate the surface-bound
polymer was accomplished using 4 eq. of BTMS in anhydrous CH2Cl2/CH3CN followed
by washing several times with reagent grade CH2Cl2/CH3CN and water. Adsorption of
the second polymer layer is accomplished with the same partial deprotection procedure
described above.32

The 31P MAS NMR measurements were conducted on a Varian VXR 400 NMR
spectrometer with 31P nuclei resonating 161.9 MHz. All spectra are proton decoupled
and nuclei are excited using a 90° pulse between 5 - 7.7 as in duration. Samples were
spun at speeds between 4 and 5.5 KHz. Spinning speeds for individual spectra are given

in the figure captions. The relaxation delays for all spectra were 2 4T1. T1 values for

71

   
  

2*! I I 0H Zl’(OI-I)22
1'2” 70H +3.62: 0 4+ 0.: 9

1’ POCl,/Collidine/H20 2:2 0” Zr 1,. (tummy
'15 ._ . ' " 2: on -———> ._. ,0

3,2,, ~i—OH ___> if; 7201’: 0 .. OK (1

§ - 2,: a 011 o Zr(OH),2+
ras'

 

;- MesSiBr2H20 Poly(NPM-VEP)

   

Zr4+
Zr(OH), ,OiPr
D ,0 rop\'0
Z 0p\"0 OiPr
_ M.,; r 0
if}: Zr(OH)2 loan
5:1 " O .2 i 0
OP,”'0 2 OP\
\ ' .33 OH
0 .,
O Zr(OH)2 19’5"
OP," '0 r=~ ’ OP‘o .Pr
\ 5 r
O
Zr(OH)2

 

Scheme 4.1. Surface reaction sequence studied in this paper. A: Reaction of a silica
surface with POCI}. B: Zirconation of A. C: Deposition of a partially hydrolyzed
layer of poly(NPM-VEP). D: Hydrolysis and zirconation of the poly(NPM-VEP)
layer. E: Deposition of a second polymer layer. The oxidation states of zirconium

(+4) and oxygen (-2) are not shown for simplicity.

72

phosphorylated and phosphorylated/zirconated silica were 0.21 and 0.77 seconds,
respectively. The spinning side bands were identified by acquisition of the spectra at
several spinning speeds; molecular resonances exhibited a constant chemical shift.
Comparisons were made on samples prepared under the same conditions and data were

acquired using the same instrumental parameters. All the 31P chemical shifts reported

here are relative to 85% phosphoric acid (5 = 0 ppm).
4.3 Results and Discussion

The focus of this paper is on understanding the chemical speciation of phosphorus
and its environment at selected points during the synthesis of zirconium-
phosphate/phosphonate layered interfaces on silica gel. We are interested in establishing
the details of multilayer growth from the perspective of the interlayer linking
functionalities. We consider the surface reactions shown in scheme 4.1 sequentially.

The first reaction step, to produce the phosphate-modified surface and shown in
scheme 4.1 A, is the reaction of a silica surface with POCl3 followed by hydrolysis. We
show the 31F MAS NMR spectrum of the reacted silica surface in Figure 4.1. The two
sharp resonances at = 0.60 ppm and = —l 1.8 ppm are associated with two distinct
types of phosphorus (Table 4.1). Morrow et. al.35 have studied the adsorption of PC13
and POCl3 on silica and reported that the reaction of PC13 with silica produces Si-O-P

containing species such as SiOPCl2, SiOP=O(H)(OH), and (SiO)2P=O(H) (Table 4.1).

73

50 40 30 20 10 0 -10 -20 -30 -40 -50
chemical shift (ppm)

 

 

All] AJ‘LHlllll‘lllllllllll

50 40 30 20 10 0 -10 -20 -30 -40 -50
chemical shifi (ppm)

Figure 4.1. (a) 31F MAS NMR spectrum of silica reacted with POCl3. Bands marked
with asterisks are spinning sidebands (spinning speed 4.0 kHz). The bands at 5 = 0.6
ppm and -1 1.8 ppm were fit using Gaussians and the area ratio is 1.5. (b) 31P MAS

NMR spectrum of the same sample after washing with polar solvents. The integrated

hand area ratio has decreased to 0.91 (spinning speed 5.5 kHz).

Because we use an excess of POC13 in the surface reaction, some of POC13 can
react with any water present either in solution or on the SiOx substrate particles to form
phosphorus oxyacids. An unresolved peak at ~ —5ppm could, in principle, be due to
phosphite P-H species (Figure 4.1b). There is, however, no confirming IR evidence of P-

H vibrational modes in our POCl3/collidine/water treated sample.32 Phosphoric acid, the

74

most likely hydrolysis product, is capable of hydrogen bonding with surface silanol
groups“3 and chemisorbed SiOPO3H2. The more intense peak at = 0.60 ppm is
indicative of a physisorbed phosphorus oxide moiety as we discuss below. Keeping all
the parameters and conditions same, the ratio of integrated area of peak at = 0. 6 0 ppm
to the integrated area of peak at = —1 1.8 ppm is reduced from a value of 1.5 (Figure
4.1a) to 0.9 (Figure 4.1b) after successive washings with ethanol, acetone and water.
The interaction of H3P04 and silica is strong enough to withstand this washing procedure
to a significant extent. Recently Murashov et al.33 used ab initio calculations to
determine that silanols can, in fact, hydrogen-bond strongly to phosphate functionalities.
Their calculations suggested a modulation in electron density at the phosphate and silanol
groups in such a way as to strengthen the P-O bond and weaken the P=O and C-0
bonding in adsorbed organophosphate moieties. They calculated the hydrogen bonded
silanol-phosphate complexes to be stabilized by ~14 kcal/mol per hydrogen bond.36 In
our case, although the = 0.6 ppm peak does not decrease greatly after washing with
ethanol, acetone, and water, it is diminished substantially upon reaction with Zr“. This
finding is a clear indication that the strength of the Zr-phosphate interaction exceeds the
~14 kcal/mol H-bonding energy, which is a fully expected result. Physisorbed H3P04
can decrease the extent of chemisorption on silica because the physisorbed phosphoric
acid can block surface silanol groups and preclude covalent bonding between surface
hydroxyl groups and POC13.

We now focus on the species responsible for the resonance at 5 = —11.8 ppm.
Morrow et al.35 and Mudrakovskii et al.37 obtained the 3 IP MAS NMR spectra of 810X

exposed to H3P04 at various temperatures and attributed bands in the —8 to —12 ppm

75

Table 4.1. Surmnary of 3 1P NMR resonance assignments reported in the literature.

Spectral shifts are measured relative to 85% H3P04.

 

Chemical Structure Chemical Shift (ppm) Reference

C13H32 PO(OH)2 weak interactions with Cd33 22.5 50
C13H37 PO(OH)2 free acid 28.5 50
(:an37 PO(OH)2 (s) 31.5 38
can37 POO'(OH)Na+ 26.5 54
c.3113,7 Pofnva")2 23.8 54
C13H37 PO33'(Na+)2 weak interactions with Na+ 28.1 54
canl7 PO33'Zn3+ 33.6 and 32.6 54
C14H26 POf'Zn2+ 33.8 and 32.6 54
C13H37 P032'Zn3+ 33.8 and 32.7 54

- Zr(HPO4)2-H2O, Hp042' -18.7 47,48

- Zr(HP04)2°2H2O, H2P04' -9-4 47,48

- Zr(HPOa)2-2H2O, p043' —27.4 47,48
P(OCH3)3 140 39
OPH(OCH3)2 1 1 39
OP(OCH3)2CH3 32 39
(SiO)2P(H)O -5 40
(SiO)2P(CH3)O 9 with a shoulder at 20 40
(CH3O)PC12 181 41
(CH3O)2PC1 169 41
(CH3)3SiOPO(H)(OH) -1 35
((CH3)3SiO)2PO(H) -13.6,33 44.9.42 44.343 35,42,43
(CH3SiO)3PO -28.3 41
SiOPO3H2 -8 to —12 37
SiOPO(H)(OH) —5 35
(SiO)2P(H)O -16 35

76

region to chemically bound SiOPO3H2 (Table 4.1). Morrow et. al.3 3 also reported
phosphorus oxychloride species bound to silica. Their experiments were conducted
under vacuum so there was little or no water present to hydrolyze the P-Cl bonds. They
observed an interaction between POC13 and silica, and concluded that hydrogen bonding
between surface silanol groups and POC13 accounted for their data. In the work we report
here, where we used excess POCl3 in the presence of a Lewis base, it is more likely that
the silanol groups will react with POC13 to form covalent SiOPOCl2 which can be
hydrolyzed efficiently to yield SiOPO3H2. It is the phosphate chemically-bonded to the
silica substrate that is responsible for the sharp resonance at 5 = -11.8 ppm, in agreement
with the literature.37 This resonance does not decrease in intensity after repeated washing
in polar solvents followed by drying under vacuum, consistent with the 5 = -11.8 ppm
band being associated with chemisorbed phosphate.

We assume complete hydrolysis of the phosphorus oxychlorides either upon
reaction with the SiO,( substrate or immediately subsequent to the formation of the Si-O-P
bond. After the initial reaction with POCl3, we do not attempt to maintain anhydrous
conditions. We can test the assumption of essentially complete hydrolysis
experimentally. The 3 IP chemical shifts characteristic for P-Cl bonds are in the 5 = 150-

3544 - - -
and we find no resonances 1n thrs spectral wrndow. After covalent

180 ppm region,
bonding to the SiOx substrate, the remaining P-Cl functionalities are converted to P-OH
as a result of their reaction with water.

Another important issue relating to the surface bonding of the phosphorus oxides

to SiOx surfaces is the number of bonding sites per phosphate. Multiply bound species

77

such as (SiO)2PO2H and (SiO)3PO are characterized by 3 1P NMR resonances at ~ - 20
and -30 ppm respectively, and we do not observe these bands in our data. We account for
this finding on two grounds. The first is that our reaction conditions provide a
stoichiometric excess of POC13 relative to surface silanol groups, a condition that favors
the formation of single attachment of the phosphorus oxychlorides to surface silanols.
The second reason that we see predominantly single Si-O—P bonding is the unfavorable
steric constrains associated with multiple bonding sites. We note that the 29Si NMR
spectrum of these same samples (Figure 4.2) does not reveal any resonances in the 5 ~ -
200 ppm region that would be consistent with the existence of 5- or 6- coordinate
silicon.45

We next consider the phosphated and zirconated surface, as shown in
Scheme 4.11 B. After reaction of the phosphated surface with ZrOC12 (aq), the silica-
bound phosphates formed a complex with Zr“ of the form of [SiOPO3ZrX2], where X
can, in principle, be OH', Cl'or C2H50' based on the species present in the reaction
vessel. The counter anions X are necessary to maintain charge neutrality in zirconium-
phosphate/phosphonate complexes. Although the concentration of Cl' is 10 mM in the
zirconation solution, much higher than either [OH'] or [C2H50'], previous XPS studies of

3 have not revealed detectable C1' in zirconium-

multilayer assemblies3
phosphate/carboxylate or zirconium-phosphate/sulfonate multilayers, suggesting that OH'
and/or C2H50' are the dominant counter ions in layered assembly growth. The pKal
values for water and ethanol are 14.0 and 15.9, respectively‘36 and, assuming the pH of

ZrOCl2 in 3:2 ethanol:water solution is ~ 4, the concentration of OH' is approximately

twice that of C2H50'. The probability of having OH' as a counter ion for the [SiO(PO3)3'

78

Zr4+] complex is higher than that of having C2H50', assuming the association constants

for these anions with Zr4+ are similar.

 

 

 

0 -25 -50 -75 -100 -125 -150 -l75 -200
chemical shift (ppm)

b

     

-75 -100 -125 -150 -175 -200
chemical shift (ppm)

0 -25 -50

Figure 4.2. (a) 29Si MAS NMR spectrum of the silica gel prior to reaction with POC13
(spinning speed 3.5 kHz). (b) 29Si MAS NMR spectrum of the same sample
following reaction with POC13 (spinning speed 3.5 kHz). The absence of any features

near 5 = -200 ppm indicates single point binding of the phosphorus oxychloride.

79

The zirconated surface exhibits several resonances distinct from those for the non—
zirconated surface (Figure 4.3). The narrow resonance associated with chemisorbed
SiOPO3H2 at = —11.8 ppm broadens on complexation with Zr3+ and two new
resonances are seen at =— 14.8 ppm and = —19.1 ppm. The appearance of these new
resonances is coincident with a substantial loss of intensity in the = 0 .6 ppm resonance,
consistent with either the removal or in situ complexation of physisorbed phosphorus

oxyacids.

 

l A A A A l A A A A l A A A A 1 J_A3 A A l A A A A 1 J; L A_

40 20 0 -20 -40 -6O
chemical shift (ppm)

Figure 4.3. 31P MAS NMR spectrum of the phosphated and zirconated surface. There
are distinct but poorly resolved resonances at 5 = 0.6 ppm, -ll.8 ppm, -14.8 ppm and -
19.1 ppm indicating partial complexation of the surface as well as several different
complexation conditions. The bands marked with asterisks are spinning sidebands

(spinning speed 4.0 kHz).

80

Clayden,47 Nakayama,48 and Morgan and workers49 studied 3 IP isotropic chemical
shift values of phosphate groups in a—Zr(HP04)2 as a function of phosphate group
deprotonation. They found that as the phosphate group is successively deprotonated, the
3'P resonance shifts to lower frequencies. For example, the (H2POa)', (HP04)3’, and
(P04)3' groups in or-Zr(HP04)2 appeared around = - 10, -20, and -30 ppm respectively
(Table 4.1). Another factor that can also affect resonance position is the extent of
hydrogen bonding between phosphate groups and other adsorbed molecules.
MacLachlan et al.49 observed that the difference between anhydrous and hydrated -
Zr(I-IPO4)2 was ~2 - 4 ppm, with the hydrated and presumably hydrogen-bonded forms
being shifted downfield of the anhydrous form. Comparing the data in Figure 4.3 to data
in the literature, we assign the resonance at = - 19.1 ppm to (SiOPO3)3’ complexed with
Zr‘3337339 and the =—14.8 ppm resonance to hydrated (SiOPO3va2O)3'Zr.49 Due to
complex surface topology and roughness of silica after reaction with POCl3, there may be
some residual, uncomplexed phosphate (SiOPO3H2) that exists following zirconation.
The presence of residual SiOP03H2 would account for the resonance at = —11.8 ppm.
It is important to note that there also remains a small amount of physisorbed H3P04 (5 =
0.6 ppm) following zirconation, and its existence is likely due to surface morphology
effects as well.

With the identities of the individual 31F resonances determined for phosphated
and zirconated SiOx, we are in a position to consider the factors that contribute to the
spectral width of the resonances in the -5 ppm to -35 ppm spectral region. The width of
NMR resonances following the formation of zirconium-phosphate complex provides

insight into the structural freedom of the phosphate functionality in the ZP complex.

81

Comparing the 31P NMR spectra of -Z r(HP04)2 obtained by Nakayama48 and
MacLachalan49 with the spectrum of our ZP complex (Figure 4.3) shows the full width at
half maximum (F WHM) of the dominant resonance to be ~25 times narrower than our
spectrum. The FWHM of the 31P spectrum of -Zr (HPQ)2 is ~ 0.7 - 0.8 ppm33‘49 while
the F WHM of our spectrum is broadened from ~ 3.0 ppm (Figure 4.1b) to ~ 18 ppm after
complexation with Zr4+ (Figure 4.3). This result is not without precedent. The spectra of
solid octadecylphosphonic acid (ODPA)compared to that of a 125-bilayer ODPA
Langmuir — Blodgett (LB) film containing Cd3+ ions also shows the broadening of
phosphonic acid 31P resonances in the LB films as a result of complexation with Cd”.50
The breadth of the resonances following complexation with zirconium could be
due to any of several factors. One likely contribution is the presence of overlapping
resonances characterized by a distribution of chemical shifts, the result of different types
of surface bonds and/or binding sites of phosphates and zirconium on the surface.
Hydrogen bonding can alter the isotropic chemical shift of phosphate by 3 - 5 ppm.3 3 ’3 1
Rothwell et al.51 reported 31P spectrum of monocalcium phosphate, Ca(H2POa)2-H2O in
which they found two 3 IP resonances separated by ~4.5 ppm. They attributed these
resonances to the hydrogen bonding between two inequivalent H2P04' groups. In their
model, one H2P04' accepts two hydrogen-bonding protons, one from a water molecule
and the other from the second H2P04' , which can be considered a hydrogen-bond
donor.5 1 If our data are understood in the context of inequivalent H-bonds, the exchange
time for the different types of environment formed by the H-bonding phosphate groups

must fast at room temperature relative to the speed of the measurement. Our results are

not inconsistent with H-bonding playing a role in the line broadening we observe upon

82

complexation. In our case, there also exists the possibility of hydrogen bonding between
the terminal hydroxide/ethoxide of the ZP complex with other neighboring
hydroxide/ethoxide groups and/or water/ethanol molecules. The evidence of hydrogen
bonding comes from FTIR data on a bilayer of poly(NPM—VEP) on silicon showing a
broad resonance in the 3200 to 3700 cm'1 region.32

To help resolve the issue of line broadening upon zirconation, we have measured
the 3'P spin-lattice relaxation times (T1) for both phosphated (Figure 4.1b) and
phosphated/zirconated silica (Figure 4.3). We found T1 = 0.21 sec for the phosphated
substrate and T1 = 0.77 sec for the phosphated/zirconated surface. This finding suggests
that the additional linewidth seen for the complexed phosphate must be due to
inhomogeneous broadening, presumably dominated by an unresolved and rapidly
exchanging distribution of environments. If the complexation step simply slowed the T1
relaxation time and did not add to the distribution of sites, we would expect a
significantly narrower resonance for the complexed system.

The change in T1 as a function of complexation suggests that the motional
freedom of the surface-bound OPO; groups is restricted upon zirconation. It is not
immediately apparent that the restriction of motion is limited to a specific rotation, for
example, and other effects could contribute to the change in T1 we observe. Restrictions
on proton exchange within or between adjacent phosphate groups could give rise to the

51.52

change in T1. Such motions have been observed in KD2P04 in the paraelectric

33 Unfortunately, given the signal-to-noise ratio of our data, it is not possible to

phase.
separate the various contributions to the additional 31P linewidth in the complexed

system.

83

 

IAAAAIJLAAJAAAAlAAAAlAAAAJAAAAIAALLAIAJJAJ‘ILAlll‘l‘

100 80 60 40 20 0 -20 -40 -60 -80 -100
chemical shift (ppm)

Figure 4.4. 31P MAS NMR spectrum of the surface with a single layer of partially

hydrolyzed poly(NPM-VEP). The band marked with an asterisk is a spinning sideband

 

(spinning speed 4.0 kHz).
a
b
40 20 0 ~20 -40 -60

chemical shift (ppm)
Figure 4.5. (a) 31P MAS NMR spectrum of a monolayer of UBPA on primed SiOx

(spinning speed 4.0 kHz). (b) The same sample that has been zirconated (spinning speed

4.0 kHz).

84

The next step in layer growth is addition of the first layer of poly(NPM-VEP), as
shown in scheme 4.1 C. Figure 4.4 shows the 3 IP MAS NMR spectrum after adsorption
of the first layer of partially hydrolyzed poly(NPM-VEP). The dominant feature
associated with the addition of the polymer is the appearance of a new, broad resonance
at ~ 28 ppm. This feature is attributed to a phosphonate ester and/or uncomplexed
phosphonic acid pendant groups on the poly(NPM-VEP) polymer. The solution phase
31P NMR spectrum of poly(NPM-VEP) contains only one broad band in the = 2 4 - 28
ppm region. The width of this peak in the solution phase spectrum suggests that, in
addition to slow motion of the polymeric chains, there are many conformations of the
phosphonate ester and/or phosphonic acid and that inter- and intrarnolecular hydrogen
bonding plays a significant role in the dynamics of the polymer. With the addition of the
polymer layer, the relative intensity of the 31P resonance at = 0.6 ppm diminishes
fiirther, consistent with its assignment as physisorbed phosphoric acid on silica. The
bands in the = -10 to -20 ppm region also become less well-resolved, suggesting an
increase in the distribution of available environments upon complexation.

At this point it is useful to compare the data for the polymer layers with the
corresponding information on ZP layers formed using alkanebisphosphonates. We
compare the 31P NMR spectrum of a relatively well-ordered monolayer of 1,11-
undecylbisphosphonic acid (UBPA) (Figures 5) with a poly(NPM-VEP) monolayer. The
upfield region of monolayer UBPA spectrum shows features similar to those of
poly(NPM-VEP) monolayer (Figure 4.6a). For the UBPA layer, the band at 5 = —11.8
ppm is due to SiOPO3H2, and/or mono deprotonated ZP complex, i. e. Si(OPO3H')Zr(OH'

)3. The 5 = —18.4 ppm and 5 = —14.5 ppm bands arise from SiOPO33' and hydrogen —

85

bonded (SiOP03)3' in the ZP complex respectively (Figure 4.6b). A broad, relatively
weak band in the 5 = 10 — 30 ppm region is due to uncomplexed or weakly complexed
phosphonic acids (Table 4.1).

We also find a band at = 22.9 ppm in the UBPA monolayer (Figure 4.5b). The
UBPA mono- and multilayers are relatively ordered compared to the polymer layers, but
IR data on the aliphatic portions of the layers indicate the extent of organization is less

than that of alkanethiol/ gold monolayers.5 3 The presence of free phosphonic acids is

A.IAAAILAAALJAAAIIIAAIAAIlllll

40 20 0 -20 -40 -60
chemical shift (ppm)

Figure 4.6. (a) 3'P NMR MAS spectrum of a single layer of poly(NPM-VEP) after
hydrolysis and zirconation (spinning speed 4.0 kHz). (b) Spectrum of the same sample

with a second layer of poly(NPM-VEP) added (spinning speed 4.0 kHz).

86

most likely due to the terminal groups at the top of the interface that remain to be
zirconated. Recently, Fanucci et al.30 reported 3 'P NlVIR spectrum of Langmuir —
Blodgett films of bisphosphonic acid acids and they found that the weak interactions
between free phosphonic acids and nearby Cd2+ ions shifts the isotropic 3‘P resonances
slightly upfield. Similarly, mono- and disodium salts of octadecylphosphonates are
shifted upfield from octadecylphosphonic acid.54 We consider that, from the point of
view of the phosphonate groups, essentially the same chemistry is operative, save for the
protection/deprotection steps used in the polymer (vide infra), for both types of layered
material.

We consider next the protection/deprotection chemistry of the polymers that
allows us to grow multiple, uniform polymer layers. This step is depicted as the
transition between C and D in scheme 4.1. To hydrolyze the remaining phosphonate
ester present in the monolayer of poly(NPM-VEP), we react the layer with 4 eq. of
bromotrimethylsilane (BTMS) in anhydrous CH3CN for 12 hours. We use a
stoichiometric excess of BTMS at this point in the layer growth because we want to
achieve complete deprotection. To within our detection limit, the broad resonance at ~
28 ppm associated with the phosphonate ester is removed by BTMS/hydrolysis and
subsequent zirconation, indicating essentially complete deprotection. A new shoulder at

= -23.8 ppm appears as a result of the reaction (Figure 4.6a). We tentatively assign
this band to complexation of the new phosphonic acid functionalities in the polymer that
are able to interact with any available (i.e. incompletely complexed) zirconium. Further
study is needed to understand the origin of this band more fully. The Page group55

reported recently that the hydrolysis of phosphoesters in ZP multilayers formed on silica

87

Table 4.2. Summary of 31P NMR resonance assignments reported in this work. Spectral

shifts are measured relative to 85% H3P04.

Chemical Structure Chemical Shift (ppm)

 

(1) Treatment of silica with POCl3/collidine/water 0.60, -11.8
(2) (1) + ZrOCl2-8H2O 0.60, -11.8, -l4.5, -l9.1

28 (broad), 0.60, -l 1.8, -14.8, -
(3) (2) + 1 Layer of poly(NPM-VEP)
19.1

(4) (3) + BTMS hydrolysis + ZrOCl2.8H2O 0.60, -11.8, -14.8, -19.1,~ -23.8
(5) (4) + 2nd Layer ofpolyWPM-VEP) 21.2, -11.8, -14.8, -18.0, -25.4
CH2=CHOCH2CH2P(O)(OCH(CH3) 2) ~ 27, — 28 (singlet, sharp)
Poly(NPM-VEP) — solution phase ~ 24, — 28 (broad)
H203P(CH2)11PO3H2 (UBPA) ~ 28.7 (sharp)

with BTMS is less efficient than our data indicate. They used relatively more ordered
materials, however, that are less porous than our polymer multilayers. The availability of
the phosphoester functionalities in their layers is very likely lower than it is in our
polymers. The spectrum of the bilayer shown in scheme 4.1 E is shown in Figure 4.6b.
There are few new features present in this spectrum. The peak at ~ —20 ppm
characteristic of the complex formed between zirconium and phosphonic acids on
polymer chains, corresponding to the analogous peak in Figure 4.4. A new peak at =
21.2 ppm is likely associated with free phosphoesters in the second layer. The position of

the band suggests that weak interactions of these phosphoesters with zirconium ions may

88

be present. The addition of additional polymer layers is expected to produce results that

are consistent with the addition of the second layer.
4.4 Conclusion

We have reported the step-by-step layer growth of poly(NPM-VEP) and UBPA
using 31P NMR spectrometry and summarize our results in Table 4.2. We resolve the
several forms of phosphorus oxyacids that are present during the reaction steps shown in
scheme 4.1. Our data indicate that, for the initial reaction of PDQ; with SiOx,
physisorption of H3POa (from hydrolyzed POCl3) competes efficiently with
chemisorption processes. Subsequent washing with polar solvents does not remove the
physisorbed species completely, and we understand this effect in terms of the relatively
strong hydrogen-bonds formed between silanols and phosphorus oxyacids.36 Reaction of
the surface with Zr“ is required to remove the non-chemically bound phosphate. For
layer growth, despite some differences in the details, there is broad similarity between the
31P NMR data for the poly(NPM-VEP) and UBPA layers, indicating that essentially the
same chemistry is operative in both types of layered material. The differences between

the 3 IP NMR response of these two materials arise primarily for steric reasons.

89

4 .5 Literature Cited

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

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

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

4. Swalen, J. P.; Allara, D. L.; Andrade, J. P.; Chandross, E. A.; Garoff, S.; Israelachvilli,
J.; McCarthy, T. J.; Murray, R.; Pease, R. F.; Rabolt, J. F.; Wynne, K. J.; Tu, H.
Langnudrl987,3,932.

S. Calvert, J. M.; Georger, J. H., Jr.; Peckerer, M. C.; Perhsson, P. E.; Schnur, J. M.;
Scheon, P.; E. Thin Films 1992, 114, 9188.

6. Dulcey, C. S.; Georger, J. H.; Krauthamer, V.; Stenger, D. A.; Fare, T. L.; Calvert, J.
M. Science 1991, 252, 551.

7- Kumar, A.; Biebuyck, H. A.; Abbott, N. L.; Whitesides, G. M. J. Am. Chem. Soc.
1992,114,9l88.

8- Vrancken, K. C.; Van Der Voort, P.; Gillis-D’Hammers, 1.; Vansant, E. F.; Grobet, P.;
J. Chem. Soc., Faraday Trans. 1992, 88, 3197.

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

10. Lee, H.; Kepley, L. J.; Hong, H.-G.; Mallouk, T. E. J. Am. Chem. Soc. 1988, 110,
618.
11. Lee, H.; Kepley, L. J .; Hong, H.-G.; Akhter, S.; Mallouk, T. E. J. Phys. Chem. 1988,

92, 2597.

90

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22-

23.

24.

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. — 1.; 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.

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

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

Katz, H. E.; 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.

Ungashe, S. B.; Wilson, W. L.; Katz, H. E.; Scheller, G. R.; Putvinski, T. M. J. Am.
Chem. Soc. 1992, 114, 8717.

Katz, H. E.; Schilling, M. L.; Chidsey, C. E. D.; Putvinski, T. M.; Hutton, R. S.

Chem. Mater. 1991, 3, 699.

91

25.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38

39

40.

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

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

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

Olson, K. G.; Butler, G. B. Macromolecules 1984, 1 7, 2480.

Olson, K. G.; Butler, G. B. Macromolecules 1984, 17, 2486.

Rabinowitz, R. J. J. Org. Chem. 1961, 26, 5152.

Home, J. C.; Blanchard, G. J. J. Am. Chem. Soc. 1996, 118, 12788.

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

Bakiamoh, S. B.; Blanchard, G. J. Langmuir, in press.

Flory, W. C.; Mehrens, S. M.; Blanchard, G. J. J. Am. Chem. Soc., in preparation.

Morrow, B. A.; Lang, S. J.; Gay, 1. D. Langmuir 1994, 10, 756.
Murashov, V. V.; Leszczynski, J. J. Phys. Chem. A, 1999, 103, 1228.

Mudrakovskii, I. L.; Mastikhin, V. M.; Kotsarenko, N. S.; Shmachkova, V. P. Kinet.

Catal. (Engl. T ransl.) 1988, 29, 165.

Gao, W.; Dickinson, L.; Grozinger, C.; Morin, F. G.; Reven, L.; Langmuir, 1996, 12,
6429.
Gay, 1. D.; McFarlan, A. J.; Morrow, B. A.; J. Phys. Chem, 1991, 95, 1360.

Lang, S. J.; Gay, 1. D.; Morrow, B. A.; Langmuir, 1995, 11, 2534.

92

41. Mark, V.; Dungan, C. H.; Crutchfield, M. M.; Van Wazer, J. R.; Top. Phosphorus

Chem, 1967, 5, 227.

42. Livantsov, M. V.; Prishchenko, A. A.; Lutsenko, I. F .; J. Gen. Chem. USSR (Engl.

Trans), 1985, 55, 1976.

43. Brazier, J. F.; Houalla, D.; Wolf, R.; Bull. Soc. Chim. Fr., 1970, 1089.

44. Bogatyrev, V. M.; Brei, V. V.; Chuiko, A. A. T heor. Exp. Chem. (Engl. T ransl. )
1988, 24, 603.

45. Bernstein, T.; Fink, P.; Mastikhin, V. M.; Shubin, A. A.; J. Chem. Soc. Faraday
Trans. 1 1986, 82, 1879.

46. Vollhardt, K. P. C.; Schore, N. E. “Organic Chemistry”, W. H. Freeman and
Company, Ed.2, 1994, page 244.

47. Clayden, N. J ., J. Chem. Soc. Dalton Trans. 1987, 8, 1877.

48. Nakayama, H.; Eguchi, T.; Nakamura, N.; Yamaguchi, S.; Danjyo, M.; Tsuhako, M.
J. Mater. Chem. 1997, 7, 1063.

49. MacLachlan, D. J .; Morgan, K. R. J. Phys. Chem. 1990, 94, 7656.

50. Fanucci, G, E.; Bowers, C. R.; Talham, D. R. J. Am. Chem. Soc. 1999, 121, 1088.

51. Rothwell, W. P.; Waugh, J. S.; Yesinowski, J. P. J. Am. Chem. Soc. 1980, 102, 2637.

52. Blinc, R.; Piré, J. J. Chem. Phys. 1971, 54, 1535.
53. Hong, H.-G.; Sackett, D. D.; Mallouk, T. E.; Chem. Mater., 1991, 3, 521.

54. Gao, W.; Dickinson, L.; Morin, F. G.; Reven, L. Chem. Mater. 1997, 9, 3113.

93

55. Neff, G. A.; Page, C. J.; Meintjes, E.; Tsuda, T.; Pilgrim, W. C.; Roberts, N.;

Warren, W. W., Jr. Langmuir 1996, 12, 238.

94

Chapter 5

Growth of Maleimide-Vinyl Ether Copolymer Multilayers.
Elucidating the Balance between Metal Ion Complexation and Side
Group Isomerization

Abstract

We have grown multilayers of poly(4-N-maleimidoazobenzene-c-(2-vinyloxy)-
ethylphosphonate), MAB-VEP, where interlayer connections are made using Zr-
bisphosphonate (ZP) ionic complexation chemistry. Absorption data on the azobenzene
chromophore side group show constant layer density but a layer-dependent ratio of trans-
to-cis isomers. Optical null ellipsometry data show a constant average layer thickness
despite the change in conformer ratio. The change in conformer ratio with the growth of
multiple polymer layers results from the steric constraints imposed on the polymer side
groups by the Zr-phosphonate interlayer linkage formation. The ellipsometric
measurements point to the macroscopic disorder inherent to the formation of a polymer
layer. Our data demonstrate that the driving force for metal ion complexation is greater
than the isomerization barrier of azobenzene. Once the layers are formed, the side-
groups do not exhibit any changes in conformer ratio, even when exposed to UV for

prolonged periods.

95

5.1 Introduction

There has been significant recent interest in the design and growth of mono- and
multilayer thin films because of then potential utlllty ln chemlcal sensrng’ surface

modification?"3 and device applications.4 From a more fundamental perspective, gaining
control over surface chemical identity, structure and thickness is a challenge to the
ingenuity of the chemistry community. Many chemical routes to the growth of layered
structures at interfaces have been demonstrated, using both crystalline and amorphous
substrates. In most of these cases, the dominant growth axis is perpendicular to the
substrate surface because individual molecules are used as the fundamental building
block for these multilayer interfaces. Intralayer interactions between constituents have,
in many cases, been considered to be coincidental and of secondary importance.
Recently, several groups have recognized the importance of structural organization

within individual layers and they have demonstrated layered growth where the greatest
degree of structural integrity resides within the layers}.13 Structural integrity parallel to
the substrate has been demonstrated for siloxane-based layers,5.lo electrostatically-bound
polymer multilayers”.12 and polymers containing pendant phosphonate groups, where

. . . . . l3 . . . .
metal phosphonate interlayer llnklng chemistry 18 used. Our interest lies 1n this latter
family of interfaces because of the structural integrity both between and within layers and
the chemical versatility of the polymers used in the construction of the layers.

We have reported recently on the photoinduced polymerization of maleimides and

vinyl ethers.'4 This family of polymers is thermally stable and, with the appropriate

choice of monomer, the properties of the resulting polymer can be adjusted to be suitable

96

for many applications. Among the maleimide-vinyl ether (MVE) polymers we have

. . . . . 13 . .
made are those contalnlng pendant protected phosphonate functronalrtres. Usmg Silane-

5. r .
based deprotection chemistry,l H we have been able to grow multilayers of MVE

polymers using Zr-phosphonate (ZP) interlayer linking chemistry. For these materials,
the adjustment of their properties is achieved by controlling the length of the vinyl ether
phosphonate monomer and with the identity of the maleimide N-substituent. For the
maleimide monomer substituents, it is synthetically a simple matter to attach
chromophores or molecules capable of isomerization and probe the optical properties of
these substituents once incorporated into the polymer matrix. In this work we have used
4-N-maleimidoazobenzene (MAB) and (2-vinyloxy)-ethylphosophonate (VEP) as
monomers in the synthesis of poly(4-N-maleimidoazobenzene-c-vinyl ether ethyl
phosphonate), MAB-VEP. Absorption data on the azobenzene chromophore side group
show that the density of each layer is the same but the ratio of trans-to-cis conformers
decreases with the addition of polymer layers. Optical null ellipsometry data show that
the average layer thickness remains constant despite the change in conformer ratio. The
change in conformer ratio and constant layer thickness with the addition of polymer
layers points to the two different length scales that absorption and ellipsometry sense.
The absorption data are the result of local steric limitations imposed on the azobenzene
side groups by the formation of Zr(RPO3)2 linkages. The driving force for metal ion
complexation to produce polymer interlayer linkages exceeds the ground state (So)
isomerization barrier of azobenzene. Once the layers are formed, the side-groups do not

exhibit changes in conformer ratio, even after prolonged periods of UV exposure. On a

97

more macroscopic level, the constant ellipsometric thickness per layer points to the

disorder inherent in thin films made from polymers.
5.2 Experimental

Surface priming chemistry. Fused silica substrates were cleaned by immersion in
piranha solution (Caution! Piranha solution is extremely corrosive and is a potent
oxidizer) for 10 minutes and rinsed with flowing distilled water, followed by hydrolysis
in 2 M HCl for 5 minutes, and a final distilled water rinse. Samples were dried with a N2
stream and were primed immediately. The substrates were primed with a solution of 0.2
M POC13 and 0.2 M collidine in anhydrous CH3CN for 20 minutes at room temperature,
followed by rinses with CH3CN and water. The resulting surface was exposed to Zr“ by
immersion in a 5 mM solution of ZrOCl2 in 60% aqueous ethanol for 30 minutes.

Synthesis. The alternating copolymer of 4-N-maleimidoazobenzene (MAB) and

(2-vinyloxy)-ethylphosophonate (VEP) was prepared by radical copolymerization using

3
azobisisobutyronitrile (AIBN) as the initiator.l The reaction between MAB and VEP is
shown in Scheme 5.1. We observed no monomer homopolymerization for our reaction

conditions. VEP was prepared by the reaction of excess tri(isopropyl)phosphite with 2-

chloroethylvinylether at 170°C for 5 days in an argon atmosphere.17

The monomer MAB was prepared by reacting 4-phenylazoaniline (0.1 mol) with
maleic anhydride (0.1 mol) in CHCl3 at 25°C for 12 hours. The resulting amic acid was
filtered and dried under vacuum. The amic acid was cyclized to the corresponding
maleimide by reaction with acetic anhydride and sodium acetate under N2 at 70°C for 4

hours. The crude precipitate was recrystallized twice from absolute ethanol to yield the

98

pure product (70% yield). NMR: ('H, 300 MHz, in DMSO-d6) 5=7.25 (s, 2H), 7.6-7.7
(m, 5H), 7.95 (d, 2H), 8.05 (d, 2H).

The copolymer was synthesized by reacting equimolar amounts of MAB and VEP
in CHC13 at 60°C under a N2 atmosphere for 18 hours using AIBN (5 mol %) as the
initiator. The resulting polymer was dissolved in a minimum amount of CHC13 and
precipitated from ether twice. NMR: ('H, 300 MHz, in DMSO-d6) 6.9-8.0 (9H) 4.4 (2
H), 3.0-4.0 (4H), 2.0-2.6 (2H), 1.4-1.8 (2H), 1.0-1.4 (6H). (13C, 75.48 MHz, in DMSO-
(16) 5:24, 30-34, 42, 50, 68-74, 78,120-135, 174-180. We estimate Mn = 3700 from end
group analysis using 1H NMR data. For our experimental conditions, statistically one of

the isopropyl groups terminating each phosphonate oxygen was hydrolyzed to yield a

OiPr OiPr
0: 13’0in 0? 13’0in

 

A
N 0 ———. m
0 O O N O O N O O N O
1 ), AIBN
+ P\'""OiPr
OiPr
No
N N, N. N.

Scheme 5.1. Schematic of alternating copolymerization of MAB and VEP.

99

hydroxyl group during the course of the polymerization. This displacement is likely due
to the presence of HCl formed in CHC13 solution by AIBN. The reaction of poly(MAB—
VEP) with bromotrimethylsilane in anhydrous CH2Cl2 at room temperature for ~2 hours
yielded a polymer that was partially hydrolyzed, making it capable of forming multilayer
assemblies. After deposition, the surface-bound polymer is reacted again with
bromotrimethylsilane to deprotect the remaining isopropylphosphonate sites and allow
complexation with Zr“.

Measurements. The characterization of these films accomplished using optical
null ellipsometry (Rudolph Auto-EL II), 13C (75.48 MHz) and 1H (299.9 MHz) liquid
phase NMR (Varian lnnova 300 MHz) and UV-Visible (Unicam model UV-2)
spectroscopies. The semi-empirical calculations used to estimate the repeat unit

dimensions were performed using the PM3 parameterization with HyperchemTM v. 4.0.

5.3 Results and Discussion

We are interested in understanding the nature of the matrix formed by MVE
polymers in layers held together using ZP ionic coordination chemistry. With a sufficient
understanding of the structural factors that determine the polymer matrix properties, we
will be able to control interface characteristics such as porosity, polarity, and extent of
crystallization. MVE polymers with pendant azobenzene sidegroups are useful for

probing local organization in layered polymer interfaces because of the well understood
. . . . . 18-26 ,
spectroscopic and isomerization properties of the azobenzene chromophore. Usmg

. . . 3
the chemistry we have reported earller to construct polymer multllayers,l we have grown
up to 7 layers of MAB-VEP. The spectroscopic and ellipsometric response of the

resulting multilayer provides insight into the strength of interlayer interactions and the

100

role of long-range disorder in these materials. MAB-VEP exhibits a constant average
layer thickness (23.5 A/layer) with increasing number of layers (Figure 5.1). This
measurement of interface thickness, which provides a spatial average over a ~l mm
diameter spot size, senses macroscopic disorder in the polymer layers. The absorption
data point to the local steric limitations placed on the azobenzene side groups by the
formation of ZP linkages. These data together provide a self-consistent picture of
polymer layer morphology. We consider each piece of information individually.

The ellipsometric data for MAB-VEP growth (Figure 5.1) do not fit perfectly to a
straight line, with a slight deviation from linearity for the first layer or two. There are
several possible explanations for this finding. The first layer may experience a different
dielectric environment than subsequent polymer layers and, as a result, the complex
refractive index of this layer is different from the others. This explanation is equivalent
to proposing that there is an absorptive component to the optical response at 632.8 nm
only for the first layer. Despite the band shifts we see for the first layer, there is no
evidence to support this explanation. It is also possible that the loading density first layer
of polymer is lower than that for subsequent layers. The layer density data argue against
this explanation (vide infra). The third explanation is that the macroscopic morphology
of the first layer(s) is slightly different than it is for subsequent layers, although the
uncertainty in the ellipsometric data make substantiation of this point problematic.

In the absence of other information, the ellipsometry measurements do not offer
much chemical insight because it is not clear whether the data in Figure 5.1 result from a
Change in the structure of the polymer layer or simply from variations in the loading

density of the polymer within the layer. This ambiguity can be resolved by examining

101

the linear optical response of the multilayers. We show in Figure 5.2 the absorption

spectra of the poly(MAB-VEP)—modified interface as a function of number of layers

180

160

p—d
J}.
O

p—A
N
O

100

00
o
I—O—I

ellipsometric thickness (A)

60
I

40

20

0 _._.,. . .

 

0 1 2 3 4 5 6
number of polymer layers

Figure 5.1. Ellipsometric thickness of poly(MAB-VEP) layers for seven layers. The
zeroth layer is the primer layer on the oxidized Si substrate. The slope of the best-fit line

through these data is 23.5 i 1.1 A/layer.

102

absorbance

0.20

0.16

0.12

0.08

0.04

0.00

200 I 300 I 400

l

 

‘ 500 I 600
wavelength (nm)

Figure 5.2. Waterfall plot of absorbance for seven poly(MAB-VEP) layers. The band

positions for the first layer (bottom) are shifted from those for subsequent layers.

103

added. There are several noteworthy features contained in these data. First, there is a
general trend toward increasing absorbance with number of layers. We can use these
spectra to estimate the layer loading density (Figure 5.3), indicating a constant amount of
polymer is being adsorbed per layer.

A second significant feature in the data shown in Figure 5.2 is that the bands for

the first layer appear to be shifted relative to those for subsequent layers. We understand

 

6x1015 -
5 10,5 slope =(4.642t0.l6)x1014/cm3/layer
x _.
4.3
E .
4:
E 4x1015 -
'5
f3 .
E
d.)
23 3x1015 -
o
O
E .-
8
2x1015 -
. intercept = (1.89i0.07)x1013/cm3
IXIOIS l l l l l r l

 

 

0 1 2 3 4 5 6 7

number of layers

Figure 5.3. Total chromophore concentration as a function of number of layers. The
data represent the total of the cis and trans concentrations. The best fit line yields a

density of 4.64 x 10'4 chromophores/cmZ-layer.

104

this phenomenon in terms of the local dielectric response sensed by the azobenzene
chromophore. The dielectric response of the SiO,( substrate is substantially different than
that of a polymer layer and, because there is only one layer of polymer present, the extent
of side group confinement will necessarily be different than for the multilayers.

The third point of interest is that the ratio of the absorbance maximum near 313
nm to that near 250 nm depends sensitively on number of layers present at the interface.
To understand the significance of this observation, assignments must be made for these
two bands. For azobenzene chromophores it is well known that the S2 So transition for
the trans form is centered at 313 nm ( = 25,000 M'lcm'l, = 4.15x10'l7 cm2) and the

82 S) transition for the cis conformer occurs at 250 nm ( = 13,000 M'lcm'l, =

2.16x10'l7 cmz)“.25 Thus the ratio of the two bands we observe is related directly to the
relative amounts of cis and trans conformers in the multilayer assemblies. We show in
Figure 5.4a the maximum absorbances for the cis and trans isomers and in Figure 5.4b
we show the ratio of the isomer concentrations as a function of number of layers. The
fact that this ratio changes with the number of polymer layers means that the azobenzene
is confined significantly by its local environment and the driving force for the formation
of the multilayer assembly must be greater than the energetic cost associated with
isomerization of the chromophore in its ground state. We note that this is true even for
the first layer, where [trans]/[cis] = 0.59. For solution phase azobenzene, essentially no
cis conformer is present.

Careful inspection of the data shown in Figure 5.2 reveals that for multilayers
there is an apparent band tail to the red that is not obvious for the first several layers. We

understand these data as follows. The Si <— 80 transition for both conformers of

105

azobenzenes is centered near 440 nm and the total extinction coefficient for the SI - S0
band in ~ = 2,500 M"cm'l ( = 4.15x10'18 cmz). Because this is a weak, broad
resonance and we are measuring the response of only 7 molecular layers at most, we
expect the S) (— S0 band to be weak and poorly resolved. The second possible reason for
the observation of a red-edge band tail is the substantial disorder and thus conformational
distribution that is necessarily present in these multilayers. The resolution of the extent
to which these two factors give rise to the observed spectral profiles requires further
investigation.

Because of the complexity of the absorption data presented above, it is important
to provide some frame of reference to understand the optical response of this polymer
more fully. We show the absorption spectra of the polymer in solution (Figure 5.5a),
where the azobenzene side groups are expected to be predominantly all-trans and in a
spin-cast film of the polymer (Figure 5.5b), where We expect a significant amount of
trans conformer. For both of these spectra, the trans band is centered at ~ 315 nm, with
the cis resonance near 250 nm. The resonance near 200 nm seen in all spectra is
apparently intrinsic to the polymer. Irradiation of the polymer solution at 315 nm
produces azobenzene cis conformers (Figure 5.5a dashed spectrum), resulting in a change
in the absorption intensity of both 250 nm and 315 nm bands. Based on this experiment,
it is clear that the bands we observe for the multilayer are associated with the two
polymer side group conformers. For the spin-cast film there is a larger contribution from
the trans conformer than is seen for the ZP layers. This is an expected result based on the
local structural restrictions imposed in the layered polymers by the formation of ZP

linkages.

106

0.09 '
0.08 ' ,
0.07 ' ClS /
0.06
0.05 '- /
0.04 '
0.03 '
0.02
0.01
0.00 4 2 . 1 . . .

0.7

\m
o

absorbance
1 fr
0

/j\
0).
i.
\

 

fIfi I I

 

I O
._i
N
b»)
.3;
Lil
O\
\1

number of layers

0

C\
1

I
0"

.O
{I}
1

.O
h
1

[tranj/ [cis]

L 1 l I l

3 4 5 6 7
number of layers

 

_o
w
T

 

.9

N
O
NI-

Figure 5.4. (a) Absorbance maximum values for cis (0) and trans (o) conformers as a
function of number of poly(MAB-VEP) layers. The cis S2 +— 80 absorption band is
centered near 250 nm and the trans S2 <— 80 band is at ~ 313 nm. (b) Ratio of [trans] to

ICJS] as a function of number of layers.

107

 

 

 

 

. a
S
:5
Q)
U
c:
N
.0
§
.0
<
O 1 fi‘ """"" + ‘ _ I
200 300 400 500 600
Wavelength (nm)
b

 

Absorbance (AU)

 

L.
0.00 — . . 1 1 VVVV

200 300 400 500 600
Wavelength (nm)

 

Figure 5.5. (a) Absorbance spectra of solution phase poly(MAB-VEP) in CH3CN. For
the native form of the polymer (solid line), the dominant band is at 315 nm. Irradiation
of the polymer solution with broadband UV light produces an enhanced concentration of

cis conformers, as indicated in the dashed spectrum. (b) Absorbance spectrum of a spin-

cast film of poly(MAB-VEP).

108

The linear absorption data provide the most obvious signature that the ratio of the
two azobenzene side group isomers in changing with the addition of polymer layers.
These data are consistent with semi-empirical calculations of the polymer repeat unit
(Figure 5.6). These calculations indicate a 21 A length for the trans form and a 17 A
length for the cis form. A dimer of each, with side groups oriented in opposite directions

would yield a trans layer thickness of 42 A and for cis, 34 A. We recover

0” A OH
on i

OZP’ OzP/OH
O O

 

/ \
O N O
N\\/©
N

i’ 17A

\N v

 

21A

Figure 5.6. Calculated dimensions of single repeat unit for poly(MAB-VEP) for the

trans side group (left) and the cis side group (right).

109

experimentally a slope of 23.5 A/layer, suggesting substantial compression of the
polymer side groups. The comparison of experimental data to semi-empirical
calculations provides a useful consistency check with the steady-state absorbance data.

We turn now to a discussion of why we observe the change in azobenzene side
group conformation upon polymer layer deposition. In solution, the trans form of the
azobenzene side group is the thermodynamically stable conformer (Figure 5.53). It is
thus unexpected that only 60% of the sidegroups are trans in the first polymer layer. The
reason for this finding must be that the driving force for the formation of the ZP ionic
linkage is greater than the ground state isomerization barrier of the size group, and some
of the side groups are converted to the cis isomer to accommodate ZP association. We
note that for the spin-cast film, we recover a predominantly trans conformation since
there is no opportunity for sterically restrictive complexation to proceed in the spin-cast
matrix. Our data on polymer layer formation imply that we are forming a kinetic product
with the ZP-bound polymer layers because there is sufficient structural freedom in the
polymer backbone to allow rotation of the maleimide unit about its bond to the pendant
ether moiety. We ascribe the formation of this kinetic product to the fast ionic bonding
kinetics of the ZP chemistry.

The ground state barrier for the isomerization of azobenzene is thought to be on
the order of 50 kcal/mol, although an accurate determination of this quantity remains to
be reported. We can only estimate the strength of RPO3-Zr-O3PR group formation

because the literature on the solubility of zirconium phosphates and phosphonates is so
sparse. We are aware of a reported value of KSp = 10'132 for Zr3(PO4)4,27

3 Zr+4 + 4 PO4'3 ~——*‘ Zr3(PO4)

110

but we consider this value to be suspect because of the difficulties associated with
determining such a small quantity. If we assume that this value is correct and apply it to
the formation of RPO3-Zr-O3PR, where the metal to ligand ratio is different, we can
estimate the equilibrium constant for RPO3-Zr-O3PR formation based on the
concentration of free Zr“. We infer that Keq = 1044 for the reaction
Zr+4 + 2 RPO3‘2 : Zr(RP03)

and based on this estimate, we calculate AG = -60 kcal/mol. This is admittedly a
qualitative estimate, and the ratio of the complexation energy to the isomerization energy
is consistent with a [trans]/[cis]= 0.43. From the experimental ratio of [trans]/[cis] =
0.59, assuming 50 kcal/mol is correct for the azobenzene So isomerization barrier, AG ~ -
98 kcal/mol for ZP linkage formation. There is thus considerable uncertainty in the free
energy of formation for a ZP linkage but, regardless of the exact value, it exceeds the
isomerization barrier for So azobenzene.

One potential complication to the above explanation is that the ZP complexation
process is itself mediated by the photoinduced isomerization behavior of the side groups.
In other words, the ZP complex formation proceeds only when ambient light “toggles” a
side group to the cis conformation and the complex cannot form if the side group remains
trans. In order to evaluate this possibility, we have synthesized poly(MAB-VEP)
multilayers in the dark, where photoisomerization of the azobenzene side groups by
exposure to room light cannot proceed. We find from ellipsometric measurements (632.8
nm) that the polymer layers form equally well under either dark or light lab conditions.
The formation of the ZP interlayer complex is thus not dependent on side group

photoisomerization.

lll

It is useful to consider the values of the [trans]/[cis] ratio and how they
correspond to fractional concentrations of each isomer. For the first layer, where
[trans]/[cis] = 0.59, the surface is comprised of 63% cis isomers and 37% trans isomers.
As discussed above, even for the first layer there is considerable steric restriction placed
on the azobenzene side group. For the second layer, [trans]/[cis] = 0.43, corresponding
to a surface that is, overall 70% cis and 30% trans. Since the loading density is constant
for each layer (Figure 5.3), it is tempting to speculate that the second layer is 77% cis and
23% trans. This ratio holds relatively constant for subsequent layers. Unfortunately, it is
not possible to extract the layer-by-layer fractional concentrations of cis and trans
isomers in this manner because of inter-layer cooperative effects. The fact that the trans
absorbance decreases with the addition of a second layer demonstrates that the treatment
of these polymers as discrete and non-interacting is not valid. In other words, the
addition of a polymer layer reduces the amount of trans conformer in the layer to which it
is adsorbed.

One prediction of these findings is that, once the polymer matrix is formed,
interconversion between isomers is not facile. The very forces that cause the
isomerization to proceed in the first place act to prevent back-isomerization. To test this
hypothesis, we have irradiated a four-layer assembly with 254 nm light for an extended

period of time. Excitation of the 82 <— 80 transition of either cis or trans azobenzene

gives rise to efficient isomerization in the solution phase.24a25 Our irradiation of the S;
(— So transition for the cis side group conformer at 254 nm was intended to produce the

more stable trans conformer. We observed no change in the spectral profile of the four

112

(114
(112

(104

Absorbance (AU)

0.02 b

150

)—‘ D—‘ —l
8 8

fl

ellipsometric thickness (A)
5

100

0.10 b
0.08
0.06

200 250

N
O

 

 

b
I I I I I I I I l

300 350 400 450 500 550 600
wavelength (nm)

 

l—la

 

I. I 1 I 11 I [VLJ

I 4

0 100 200 300 400 500 600 30003500

time (min)

Figure 5.7. (a) Absorbance spectra for a four layer stack of poly(MAB-VEP) irradiated

with 254 nm light between 1 hour and 50 hours. The spectra are offset from one another

for clarity of presentation. (b) Ellipsometric thickness of the same four layer stack as a

function of UV irradiation.

113

layer assembly as a function of UV irradiation (Figure 5.7a). As a check on the
spectroscopic data, we measured the thickness of the 4-layer stack ellipsometrically
(Figure 5.7b). We found no change in layer thickness over the irradiation time. These
data demonstrate that the azobenzene conformers are locked in place, once deposition has
occurred, and the rigidity and steric restriction imposed on the side groups is sufficient to

prevent isomerization in the formed polymer matrix.

5.4 Conclusion

We have found that the conformation of the pendant side group of a polymer multilayer
assembly depends sensitively on the details of polymer interlayer linking chemistry and
on the steric restrictions imposed by the resulting polymer matrix. Using maleimide-
vinyl ether polymers with pendant azobenzene sidegroups, we determined that the ratio
of cis to trans isomers varies in a regular manner with the number of polymer layers. Our
linear absorption data demonstrate the change in conformer ratio with increasing layers,
and the polymer adsorption density is constant for each layer. The ellipsometric data
point to the macroscopic disorder inherent to the deposition of an amorphous polymer
film. The driving force for ZP interlayer linking chemistry is greater than the energetic
penalty associated with conformational change in the (ground state) side group. From
our data we estimate the free energy of formation for the ZP interlayer linkage to be -60
kcal/mol .<_ AG 5 -100 kcal/mol. We anticipate that the use of longer chain vinyl ether
monomers will provide a less restrictive environment for the maleimide substituents and
this control over polymer structure provides a direct means of controlling the optical

response of this family of materials.

114

5.5 Literature Cited

H

. Kepley, L. J .; Crooks, R. M.; Ricco, A. J. Analytical Chemistry 1992, 64, 3191-3193.
2. Swalen, J. D.; Allara, D. L.; Andrade, J. D.; Chandross, E. A.; Garoff, S.;
Israelachvili, J .; McCarthy, T. J .; Murray, R.; Pease, R. F.; Rabolt, J. F.; Wynne, K.

J .; Yu, H. Langmuir 1987, 3, 932-950.

D)

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

4. Li, D. Q.; Ratner, M. A.; Marks, T. J .; Zhang, C. H.; Yang, J .; Wong, G. K. Journal of
the American Chemical Society 1990, 112, 7389-7390.

5. Wirth, M. J .; Fatunmbi, H. 0. Analytical Chemistry 1992, 64, 2783-2786.

6. Fatumnbi, H. 0.; Bruch, M. D.; Wirth, M. J. Analytical Chemistry 1993, 65, 2048-
2054.

7. Wirth, M. J .; F atunmbi, H. 0. Analytical Chemistryl993, 65, 822-826.

8. Fairbank, R. W. P.; Xiang, Y.; Wirth, M. J. Analytical Chemistry 1995, 67, 3879-
3885.

9. Wirth, M. J.; Fairbank, R. W. P. Journal of Liquid Chromatography & Related
Technologies 1996, 19, 2799-2810.

10. F airbank, R. W. P.; Wirth, M. J. Journal of Chromatography a 1999, 830, 285-291.

11. Keller, S. W.; Kim, H. N.; Mallouk, T. B. Journal of the American Chemical Society
1994, 116, 8817-8818.

12. Keller, S. W.; Johnson, S. A.; Brigham, E. S.; Yonemoto, E. H.; Mallouk, T. B.

Journal of the American Chemical Society 1995, 117, 12879-12880.

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

115

14.

15.

l6.

l7.

l8.

19.

20.

21.

22.

23.

24.

25.

26.

27.

Kohli, P.; Scranton, A. B.; Blanchard, G. J. Macromolecules 1998, 31, 5681-5689.
Jung, M. E.; Lyster, M. A. Journal of Organic Chemistry 1977, 42, 3761-3764.
Vickery, B. H.; Pahler, L. F.; Eisenbraun, E. J. Journal of Organic Chemistry 1979,
44, 4444-4449.

Rabinowitz, R. Journal of Organic Chemistry 1961, 26, 5152.

Asano, T.; Okada, T.; Shinkai, S.; Shigematsu, K.; Kusano, Y.; Manabe, 0. Journal
ofthe American Chemical Society 1981, 103, 5161-5165.

Asano, T.; Yano, T.; Okada, T. Journal of the American Chemical Society 1982, 104,
4900-4904.

Asano, T.; Okada, T. Journal of Organic Chemistry 1984, 49, 4387-4391.

Biswas, N.; Umpathy, S. Journal of Chemical Physics 1997, 107, 7849-7858.

Han, S. W.; Kim, C. H.; Hong, S. H.; Chung, Y. K.; Kim, K. Langmuir 1999, 15,
1579-1583.

Lednev, 1.; Ye, T. Q.; Abbott, L. C.; Hester, R. E.; Moore, J. N. Journal of Physical
ChemistryA 1998, 102, 9161-9166.

Lednev, I. K.; Ye, T. Q.; Matousek, P.; Towrie, M.; Foggi, P.; Neuwahl, F. V. R.;
Umpathy, S.; Hester, R. E.; Moore, J. N. Chemical Physics Letters 1998, 290, 68-74.
Nagele, T.; Hoche, R.; Zinth, W.; Wachtveitl, J. Chemical Physics Letters 1997, 272,
429-495.

Rau, H. Journal of Photochemistry and Photobiology A: Chemistry 1988, 42, 321-
327.

Dean, J. A. Lange's Handbook of Chemistry; 14 ed.; McGraw-Hill, Inc.: New York,

1992.

116

Chapter 6
Assembly of Covalently-Coupled Disulfide Multilayers on Gold

Abstract

We report on the spontaneous organization of up to eight covalently attached
layers formed on gold from solution phase (or,(0)-dithiols 1,6-hexanedithiol (C6), 1,8-
octanedithiol (C3) and 1,9-nonanedithiol (C9). The linking chemistry between layers is
the oxidative formation of a sulfur-sulfur bond that competes successfully with intralayer
S-S bond formation. We have used optical null ellipsometry, FTIR, X-ray photoelectron
spectroscopy (XPS) and cyclic voltammetry (CV) to characterize the multilayers. Once
formed, the multilayers are stable when washed with 1 M KCl, water, ethanol, CHCl3 and
n-hexane solutions, before and after prolonged exposure to ambient laboratory
conditions. In addition to the formation of multilayers, our data point to the efficient
oxidation of the interlayer disulfide bond to an oxidized sulfur moiety where the S-S
bond remains intact. Extensive oxidation produces a sulfonate-terrninated surface that

reacts with Zr4+ and alkanebisphosphonates to form a hybrid multilayer assembly.

117

6.1 Introduction

Alkanethiol/gold self-assembled monolayers (SAMs) are a highly reproducible
and well characterized model system for understanding organic monolayer interfaces."2
Much of the pioneering work on the synthesis and macroscopic characterization of these
SAMs was done in the 1980’s by Whitesides, Nuzzo and Allara.3'28 External reflection
FTIR spectroscopy has been used extensively to study SAMs and from these
measurements we know that, for long alkanethiols (>10 carbons), the predominantly all-
trans chains are oriented at ~30° with respect to the surface normal.””25 This
organization was shown to decrease only for short chain alkanethiols, where interchain
interactions were insufficient to induce macroscopic ordering. Surface wetting studies
showed that the polar character of the interfaces depends on the chemical identity of the
organic terminus of the surface modifier. Polar-capped modifiers produce hydrophilic
surfaces while alkanethiol surfaces are hydrophobic. Helium diffraction studies of these
SAMs indicate that the organization intrinsic to the substrate is carried through to the

26-28

ends of the aliphatic chains. More recently, a wealth of atomic microscopy

measurements (STM, AFM) have shown that the thiol sulfur head groups adsorb in a

pattern related directly to the structure of the gold substrate, and that the monolayer

2940

exhibits large scale structural change over several minutes. Our recent

thermodynamic data indicate that the labile nature of the SAM is a direct consequence of

4L42

the small free energy of adsorption for these systems, and this finding, based on

3 In more

gravimetry, is in excellent agreement with the relevant electrochemical data.4
recent years, there have developed many potential applications for SAMs including

lithography, contact printing, corrosion resistance and biotechnology.44 We are able to

118

use this considerable body of knowledge to understand the formation of the multilayers
we report here.

The wealth of information on SAMs and their intrinsic high degree of ordering
make them an appealing starting point for more complicated chemical structures. Despite
their desirable features, perhaps the most significant limitation of SAMs is their inability
to form robust, covalent multilayer structures without the use of reaction schemes

5

requiring several steps to form each layer.4 While there have been several reports of

multilayer formation of thiols on gold, the layers are thought to be physisorbed upon one

“'47 There is passing reference in the literature to the

another and not bonded covalently.
formation of dithiols into layered assembliesfg'49 but we are not aware of any detailed
reports on the formation of covalent multilayers directly from dithiols on a gold substrate.
We report here on the synthesis and characterization of covalent multilayers formed on
gold from the (a,0))-dithiols 1,6-hexanedithiol (C6), 1,8-octanedithiol (C3) and 1,9-
nonanedithiol (C9), from ethanol and hexane solutions. The linking chemistry between
layers is the oxidative formation of a S-S bond, and this process competes successfully
with intralayer S-S bond formation. We do not find evidence for “looping” where both

thiol termini bond to the gold substrate. The simple synthetic route we report here is

reminiscent of the metal-phosphonate (MP) chemistry pioneered by the Mallouk,”55

56.57 58-63

Thompson and Katz groups, and the cobalt-isocyanide chemistry developed by
the Page groupf’4 except our approach is to use direct covalent chemistry without a metal
coordination center. In addition to the formation of multilayers, our data indicate
efficient oxidation of the S-S and SH groups within the layers to a family of oxidized

sulfur moieties. We consider the various forms of oxidized sulfur that can be expected

119

for this structural motif. Our findings are consistent with recent reports on the facile

“5'67 and offer a potential new route to the formation of

ozone-mediated oxidation of thiols
sulfonate-terminated surfaces. We have shown that for oxidized multilayer surfaces, it is
possible to grow additional layers using metal ion coordination chemistry. This growth
can occur only when the terminal functionality is a sulfonate. The S-S bonds within the

multilayers appear to “protect” the Au-S bond from oxidative attack based on photolysis

experiments we report here.

6.2 Experimental Section

Substrates and Reagents. The gold slides used as substrates were made by
evaporation of 200 nm of Au on 20 nm of Ti on Si(100) wafers.68 1,6-Hexanedithiol,
1,8-octanedithiol and 1,9-nonanedithiol were purchased fi'om Aldrich Chemical Co. and
used as received. The multilayers were formed on gold by immersing the substrate in a
10 mM ethanolic solution of the dithiol at 20 :l: 1°C for 12 hours or in a 1 mM solution of
the dithiol in n-hexane for 1 hour at the same temperature. Immersion for longer periods
produced more than one statistical layer for synthesis from ethanol solution, as measured
ellipsometrically. Solutions were purged with dry N2 and the multilayer interfaces were
stored either in air or under dry N2 and covered with aluminum foil to minimize their
exposure to UV 1i ght."5

Optical null ellipsometry. Layer thicknesses were measured using optical null
ellipsometry. The instrument is a Rudolph Auto-EL II equipped with a He-Ne laser light
source operating at 632.8 nm. The software used for data acquisition and reduction is

from Rudolph. For all measurements, the layer refractive index was assumed to be u =

120

1.45 + 01'. For thin films such as these, the dependence of the recovered thickness on the
real part of the refractive index is modest for physically realistic values of n.

Infrared Spectroscopy. The vibrational spectroscopic response of the multilayers
was characterized using a Nicolet Magna 550 FTIR spectrometer. Resonances in the C-H
and S=O stretching regions were measured using a Harrick grazing incidence attachment
with a beam incidence angle of ~80° with respect to the sample surface normal. Spectral
resolution for all measurements was 2 cm].

X-ray photoelectron spectroscopy. XPS measurements were made on multilayer
samples using a PHI Model 5400 X-ray spectrometer. The X-ray source is the AlKor line
and all values reported are referenced to the Cls line at 285.0 eV.

Electrochemistry. Cyclic voltammograms were recorded using a standard three-
electrode cell containing an Ag/AgCl (3 M KCl) reference electrode and a Pt-wire
counter-electrode. The working electrode was the gold coated substrate used in the
synthesis of the multilayer, mounted in a sealed plastic holder. The exposed area of the

gold substrate/multilayer was 0.1 cm2

. A CH-Instruments electrochemical analyzer was

used to perform the cyclic voltammetry (CV) measurements at a scan rate of 0.1 V/s.68
Photolysis. Multilayer samples were photolyzed with a Hg lamp in the presence

of air or N2 in an optically isolated mount. The Hg light source produces UV output with

an intensity of 4.5 mW/cm2 (all lines). The samples were mounted 45 mm from the lamp

for all photolysis experiments.

6.3 Results and Discussion

The primary issues we focus on in this paper are the formation of multilayer

structures using (or,c0)-dithiols, determining the nature and chemical speciation of the

121

interlayer bonding, and the response of these layers to exposure to selected solvents and
UV light. It is important to note that, when we discuss multilayer assemblies in the
context of these materials, we are not referring to the discrete deposition of individual
uniform layers at a given point in time. Rather, because of the interlayer linking
chemistry used in the synthesis of these assemblies, the growth is statistical, and the
thickness measurements represent an average of a given number of layers in the area
illuminated by the ellipsometer beam.

Optical null ellipsometry is a useful method for the characterization of these
multilayers. The slope of the best-fit line through the ellipsometric thickness data yields

an average thickness of 7 A per layer for C6, 8.9 A per layer for C3 and 10.8 A per layer

60
50
40
3O

20

ellipsometric Thickness (A)

10

 

 

 

 

oh
\I
00

0 l 2 3 4 5

Number of Layers
Figure 6.1. Ellipsometric thicknesses of multilayers of 1,6-hexanedithiol (A), 1,8-

octanedithiol (O) and 1,9-nonanedithiol (I). For all cases, the thickness depends linearly
on the exposure time to dithiol solution. Slopes for these dependencies are 7 A/layer for

C6, 8.9 A/layer for C8 and 10.8 A/layer for Co.

122

for C9 (Figure 6.1). Molecular mechanics calculations predict lengths of 9.3 A, 11.8 A
and 13.0 A for all-trans Cg, Cg and C9, respectively. With the assumption of a 30° tilt
angle, we predict an 8 A layer thickness for C6, 10.2 A for Cg and 11.3 A for C9. It may
be tempting to relate any small differences between calculation and experimental results
to slight variations in the average tilt angle. Because we have not determined the
fractional coverage of the surface and we know that multilayer growth does occur for
these systems, it is not possible to distinguish between tilt angle and surface coverage
contributions to the ellipsometric data. This limitation, in addition to the good agreement
between experiment and calculation, precludes our ability to discern any even-odd effect,
which was the initial reason for our use of C6, Cg and C9. The regular growth of these
layers argues strongly against significant “looping” of dithiols on the gold surface and
against efficient intralayer S-S bond formation between neighbors. The uncertainties in
the ellipsometric readings are significant relative to the absolute values (Figure 6.1),
consistent with the formation of regions containing different numbers of adlayers. Such
heterogeneity is an unavoidable consequence of the statistical nature of this
polymerization chemistry and the extended immersion times required to form these
structures. Our ellipsometric data demonstrate the formation of multiple layers of dithiol.
We are interested in understanding the nature and chemical speciation of the interlayer
bonds, and, indeed, whether or not they exist. To address these points, we offer two
bodies of data. The first is wet chemical data in concert with ellipsometry to determine
the susceptibility of the multilayers to solvent attack. Significant solubility would imply
the absence of covalent interlayer bonding. The second body of data is focused on

Understanding the chemical identities of the interlayer linking functionalities. This

123

information is provided by XPS measurements. We discuss these bodies of information
separately.

To establish whether layer constituents are bound to adjacent layers covalently or
are physisorbed, we have performed a series of washing experiments. The experimental
procedure was to measure the thickness of the multilayer assembly ellipsometrically,
immerse it in a solution for a specific period of time (typically one hour), removal,

Table 6.1. Elemental Composition

 

¢

Element Au4f Cls Ols 82p Cls/ Cls/ Cls/ Ols/ Sr/S0
Composition-3 (%) (%) (%) (%) Au4f SZp Ols 82p

Sample ~11

 

2 layers C6 (16 A, 27.1 56.0 12.0 4.8 2.1 11.7 4.7 2.5 1.5
2 weeks old)

 

6 layers C6 (42A, 10.7 60.6 20.0 8.7 5.7 7.0 3.0 2.4 5.3
2 weeks old)

 

7layers Cg (48A, 8.5 66.5 18.6 6.4 7.8 10.4 3.6 2.9 8.3
Relatively fresh)

 

7 layers Cg (45A, 9.6 60.0 22.8 7.6 6.2 7.9 2.6 3.0 3.3
2 weeks old)

 

4 layers C9 (42 A: 13.3 68.0 9.4 8.5 5.1 8.0 7.2 1.1 6.4
Air)

 

3 layers C9 (34 A, 14.6 66.3 11.4 7.7 4.5 8.6 5.8 1.5 6.2
N2)

 

 

3 layers C3 28.4 54.1 13.7 3.9 1.9 13.9 4.0 3.5 1.6
(Photolysis, 11 A,
Air)

 

 

3 layers Cg 22.9 59.2 13.0 5.3 2.9 11.2 4.6 2.5 1.5
(Photolysis, 10 A,
D12)

 

 

 

 

 

 

 

 

 

 

 

0 Sr/So represents the ratio of sum of areas of reduced form of sulfur having the
peaks at 161.7 and 163.6 eV to the oxidized form of sulfur with a peak at 168.3

eV. All the peak positions are referenced to C l 5 binding energy of 285.0 eV.

124

rinsing and drying of the sample, and ellipsometric measurement of the thickness. The
solutions used for these experiments were neat n-hexane, absolute ethanol and 1 M
KCl(aq). These solutions were chosen because they represent non-polar, polar and ionic
solvating systems. For all experiments, we observed no change in the ellipsometric
thickness of the multilayer, indicating that the interlayer bonding is covalent and not
physical. This is a physically reasonable conclusion based on the strength of hydrogen-
bonding interactions that characterize thiols. The reported value of dimer formation in
H2S is 1.7 - 1.8 kcal/mol.""‘70 Van der Waals interactions are unlikely to be sufficiently
strong to account for our findings either, and it would be difficult to envision a layered
structure dominated by these interactions given the orientation of the initial layer,
assuming it to be the same as that of alkanethiols on gold.

With the covalent nature of the interlayer bonding supported by the washing
studies, we consider the chemical speciation of that bonding, to the extent that such
information can be obtained. X-ray photoelectron spectroscopy (XPS) is sufficiently
sensitive for this purpose and this technique provides significant information on the
oxidation state(s) of the adsorbed species. We show in Table 6.1 the relative
concentrations of the elements present in the substrate and interface. Before describing
these results, we offer a few cautions on the interpretation of XPS data. The atomic
composition determined experimentally is sensitive to the energy of the incident X-ray
beam, variations in photoionization cross-section with chemical structure, the take-off
angle, the elemental distribution perpendicular to the surface, and the actual composition
and density of the monolayer. Thus elemental composition derived from XPS are, at

best, semi-quantitative.7l’72 Experimentally, because we use the same conditions for all

125

experiments, our results vary little but the band ratios offer the best comparisons between

samples. By comparing selected ratios for multilayers exposed to different experimental

O

 

 

 

 

ON-hCNWOONhOOOOON-hc‘xoo

_

 

 

A

 

132 A 1i0 ‘ 168 166 A 1&4 I 182 ‘ 1&0
binding energy (eV)

Figure 6.2. XPS data in the S2p region for films. All values are referenced to the Cls

line at 285.0 eV. (a) 7-layer Cg film measured immediately afier formation, (b) 7 layer

Cr) film exposed to air for 2 weeks, (c) 10 A thick Cg film following UV irradiation in an

N2 atmosphere.

126

conditions such as UV light or air, it is possible to examine the effect of synthesis and
sample treatment on the extent of S-S group oxidation.

To establish the susceptibility of these multilayers to oxidation, we have investigated
changes in formed layers as a function of exposure to both air and light (Table 6.1). The
most notable feature in these spectra is the dependence of the 82p resonances on sample
history (Figs. 2). These data show the progressive oxidation of the S-S fimctionalities.
For fresh multilayer samples, we observe the dominant 82p resonance to be at 163.6 eV
with a weak sideband at 161.7 eV (Figure 6.2a). The 163.6 eV resonance is associated
with gold-bound thiolate and interlayer disulfide bonds. There is not sufficient spectral
resolution with the system we used to distinguish the contributions from these two forms
of sulfur. The 161.7 eV band is known to arise from gold-bound thiolate at a site that is
inequivalent to the 163.6 eV thiolate.”75 The reported S2p chemical shifi of a thiolate in
alkanethiol SAM on gold,46 docosanethiol,7° didocosyl disulfide,76 and sulfonate77 are
161.7, 163.6, 163.3 and 168.3, respectively. Figure 6.2b shows the 82p region of a seven
layer C6 film on a gold substrate. In addition to the thiolate peaks at 163.6 eV and 161.7
eV, we find an additional peak at ~168.3 eV that is seen only after exposure to air.
Exposure to air causes slow oxidation with time, and even afier two weeks exposure
(Figure 6.2b), there remains a significant amount of reduced S at 163.6 eV. The data in

Figure 6.2c, showing the effect of exposure to UV light, demonstrate that the oxidation of
the S-S moiety is, at the very least, assisted by UV light. These findings are consistent
with the work of the Pemberton,"7 Bohn"5 and Rowlen"6 groups on the oxidative attack of
the Au-S bond by O3. It is likely that photoproduced 03 is the dominant oxidizer in our

Work as well, although the details of the photodegradation of the multilayers point to the

127

initial photolysis of the S-S bond (vide infra). The photolysis data indicate a reduction in
multilayer thickness for photolysis in N2 and in air. What remains unresolved from the
ellipsometric data is whether the dominant process is oxidative Au-S bond cleavage and
subsequent desorption or S-S bond cleavage within the assembly. We discuss these data
in more detail below.

The oxygen concentration in the multilayers, as measured by XPS, increases after
exposure to air for extended periods. For example, maintaining a seven layered gold
substrate in laboratory conditions for about two weeks, the Cls/Ols ratio decreased from
3.6 to 2.6 (Table 6.1). Thiols on gold are oxidizable to some types of SOx species.65'(’7'78
The Sr/SO ratio reduces significantly from a value of 8.3 to 3.3 for a 7 layer C6 film after
keeping the sample in ambient laboratory conditions for 2 weeks. Despite this oxidation,
these films are stable with respect to washing in polar, nonpolar and ionic solutions.

It is important to consider the R-S-S-R moiety in more detail. Organosulfur
compounds are known to exist in a variety of different oxidation states.”82 Most
experiments on organosulfur systems focus on either the fully reduced thiol or the hilly
oxidized sulfonate forms. The initial step in the interlayer bonding chemistry we report
here is the oxidation of two thiols to form a disulfide, and our time-resolved experiments
on the grth of these layers point to the need of an oxidizer to allow the reaction

proceed. Specifically, for growth from n-hexane, a single layer forms within minutes,
and is stable in solution despite the presence of excess dithiol. Once the substrate is
removed from solution, exposed to air and re-introduced to the solution, another layer

forms and remains stable. This same behavior is not observed for growth from ethanol

128

solution, presumably because of the greater solubility of oxygen in that solvent. Once the

oxidation process has been initiated, the product distribution will necessarily be

i? i.) i? ‘1? i? i? ‘1? ICI)
R—s—s—R R—S—S—R R—s—s—R R—s—s—R R—lSl—S—R R—fi—fi—R
o o o o
1 2 3 4 5 6

Scheme i. Structures of various disulfide derivatives that may be present within the
multilayers. l - disulfides, 2 - thiosulfinates, 3 -thiosulfonates, 4 - a-disulfoxides, 5 -

sulfinyl sulfones and 6 - vic-disulfones. All species are known except 4.

statistical. Under conditions of excess oxidizer, the form of the thiol will be a sulfonate,
and this functionality terminates the multilayer growth process, unless certain metal ions
are introduced to the reacting system. There are, however, a variety of sulfur-oxide
compounds that exist where the S-S bond remains intact or where an S-O-S functionality

”'82 The S-S moiety can exist in the form of disulfides, thiosulfinates,

is formed.
thiosulfonates, or-disulfoxides, sulfinyl sulfones and vic-disulfones (scheme i). In
general, the oxidation of the disulfide moiety proceeds through the series 1 to 6, with
only a-disulfoxides (4) being unstable.

The XPS data are not the only indication of oxidized sulfur in these multilayer
assemblies. A prominent IR band at 1056 cm'1 is consistent with the SO stretching mode
(Figure 6.3a). We believe that the thiols at the multilayer/air interface have been

oxidized to 80x, consistent with recent reports on the oxidative sensitivity of

alkanethiol/ gold monolayers."5

129

In an effort to verify the chemical identity of at least the most extensively
oxidized form, we have used ionic coordination chemistry. The grth of metal

50"” It is also possible to form hybrid metal phosphonate-

phosphonates is well known.
sulfonate structures using Zr“, and we have used this chemistry as a test for the presence
of R803" on the surface of a multilayer exposed to air for an extended period of time and
for which immersion in dithiol solution does not produce additional layer formation.
When this film is immersed in a 5 mM ZrOCl2 solution for 0.5 hr. followed by immersion
in a 1 mM 1,12-dodecanebisphosphonic acid (DDBPA) ethanolic solution for 12 hrs., a
10 A thick layer is formed on top of the original multilayer. For a full monolayer of
DDBPA, we expect a thickness of ~17 A. Our data indicate not only the presence of
RSOg' at the multilayer surface, but also sub-monolayer coverage. To verify that the 10
A thickness is not a consequence of the hybrid interlayer bonding scheme, we formed a
second DDBPA layer on top of the first and recovered the same 10 A/layer thickness.
This finding reinforces our assertion that the RS03' is present at less than a full layer.
After having discussed the chemical nature between interlayer bonding and the
partial oxidation of the multilayers, we now consider the XPS and ellipsometry data on
photolyzed multilayers formed from Cg in air and N2 atmosphere. Table 6.1 reports the
elemental composition of photolyzed Cg multilayers. After 5.6 hours of photolysis, the
thickness of the SAM is decreased to 10 A from 28 A (photolysis in air) and 22 A

(photolysis in N2) (Figure 6.4). The photolysis data exhibit two distinct time

dependencies. The first is characterized by a rapid decrease in thickness with initial

130

0.006
0.005 _
0.004 _
0.003
0.002 r
0.001 _
0.000 _

-log10(R)

 

 

. l 1 4 J n I A l L l A l L L L l n J

1300 1200 l 100 1000 900 800

frequency (cm'l)

  
         

 

 

 

0.015 i b
9% 0.010.. E '
:55“ : :
2 5 5

0.005.. E 5

0.000%? 2 \g—té-

 

3230 3000 2930 2900 2850 2s00
frequency (cm‘)

Figure 6.3. FTIR spectra of Co. (a) S=O stretch, (b) CH stretching region as a function
of number of layers. The dashed lines are set at 2921 cm"1 and 2852 cm'I while the

dotted lines are set at 2919 cm'] and 2850 cm'].

131

exposure to UV light, followed by the second regime, where there is either no change for
photolysis under N2 (Figure 6.4a) or a slow decrease in thickness for photolysis in air
(Figure 6.4b). We believe that the initial loss of thickness is due to direct photolytic
cleavage of the S-S bond and the slower decrease in slope for the air-exposed sample is
associated with the oxidative degradation of the Au-S bondbs‘“ Figure 6.2c shows the
82p region after photolysis which is fitted to three bands each having a doublet
corresponding S2p3/2 and S2p1/2 of an area ratio of 2 :1. Three distinct peaks at 161.7,
163.6 and 168.3 eV are seen and their area ratio is 10:20:19. Even after photolysis of
multilayers for 5.6 hours, the layers are stable to ethanol washing, indicating that the
remainder of the SAM is covalently bonded.

Ideally, the characterization of these multilayers would include a measurement of
the distribution of sulfur oxide species within the layers, but this task is, at present,
beyond the capabilities of XPS or vibrational spectroscopies. For XPS measurements,
the spectral resolution required to achieve this level of speciation is not available. For
Raman and IR spectroscopies, the limiting factor is primarily sensitivity. The most
relevant resonance to study would be the S-S stretch in the region of 500 cm".23 For IR
spectroscopy, the S- S stretch would be nominally forbidden for the disulfide and, in any
event, the MCT-A detector required for monolayer IR measurements cuts off near 700
cm'l, rendering this spectral region inaccessible. Raman spectroscopy could also, in
principle, be used, but it is not clear that the S-S stretch is a sufficiently strong Raman
scatterer to provide a useful signal from a few molecular layers. Surface enhanced

Raman scattering may prove useful in future investigations, but we do not have sufficient

132

28;
24-
.0:
.6; 1
.2: l

Ellipsometric Thickness (A)

 

141 I I A l 1 J

. I 1
0 60 120 180” 4000 6000 3000
Time (sec)

 

O
h
i-
b

32

28

24

TfiIrTfii

20
16 1'" § i

0 i l 1 I J J 1 1 II I 1 I L I n I
Ir

0 60 120 180 3000 6000 9000 12000
Time (sec)

 

Ellipsometric Thickness (A)

 

Figure 6.4. Photolysis experiment showing thickness of Cg multilayer film as a function

of UV exposure time. (a) in N2, (b) in air.

133

control over the gold surface morphology at the present time to make these experiments
feasible.

Measurement issues aside, any experimental information on the distribution of
sulfiir oxide species in the multilayer assembly is not likely to tell the complete story on
the structure of these systems. There will likely be a variation in the extent of oxidation
of the S-S bonds as a function of depth within the multilayer. The primary reason for this
variation will be the permeability of the multilayer to oxidizers. Understanding the depth
profile of each sulfur oxide form could provide additional insight into the chemical limits
on the maximum film thickness achievable with this chemistry.

In addition to the interlayer linking chemistry, we can also gain some insight into
the extent of organization of the aliphatic portions of the dithiol multilayers. The
organization of the aliphatic chains can be probed using FTIR spectroscopy. The position
of the methylene C-H stretching resonances is a sensitive measure of the order within the
SAM.2 For C6, the first two layers yield the peak position of asymmetric and symmetric
CH2 stretching modes at 2921 cm'1 and 2852 cm], respectively while for more than two
layers, these resonances are shifted to lower frequencies, 2919 cm'1 and 2850 cm"
(Figure 6.3b). For the first two layers, the chains are partially disordered, consistent with
data on alkanethiols indicating that for chain lengths less than 9 carbons, interchain
ordering is not well established.2 For the thicker multilayers, the positions of the bands
are the same as those seen for long chain SAMs, indicating that the dominant
conformation of the aliphatic chains is fully extended, all-trans and in a quasi-crystalline
environment. We also note that, as the number of layers increases, the width of the bands

at 2919 and 2850 cm'1 decreases. The peak position and width of the CH stretching

134

resonances indicate that order within the films increases with the number of layers. The
narrowing of the resonances with increasing thickness suggests either cooperative,
multilayer structural organization or a variation in layer density with If the latter

explanation is operative, the formation of islands would be required to achieve the

 

 

 

 

 

0.5 I 0.4 A 0.3 I 0.2 I 0.1 ‘ 0.02 0.1
Potential(V)

 

 

 

 

 

. I l I 1 I j l l I 1 I
0.5 0.4 0.3 0.2 0.1 0.0 -0.1
Potential (V)
Figure 6.5. Cyclic voltammetry data for reduction of F e(CNg)3 ' using a gold electrode
covered with mono- and multilayer assemblies. Curves for monolayers are solid lines

and for multilayers are dashed lines. Data were recorded at (a) pH 3 and (b) pH 10.

135

density enhancement. For Cg and C9 multilayers we observe resonances only at 2919 cm'
I and 2850 cm", even for single layers.

Electrochemical methods are well established in the characterization of
alkanethiol/Au SAMs.43 The electrochemical data point to a potentially complex surface
morphology for these layers. Cyclic voltammograms of 5 mM Fe(CNg)3' in 1 M
Na2SOg(aq) were recorded using gold electrodes coated with monolayers or multilayers
of C9 dithiol. These CVs show a reduction in the current allowed through the bilayer
relative to the monolayer, at both pH 3 and pH 10 (Figure 6.5) This is an expected result,
indicating that the layers do indeed provide significant coverage of the Au surface. To
gauge the extent of coverage requires the examination of absolute currents, and there is
the potential for misinterpretation associated with unaccounted-for leakage paths and an
incomplete understanding of the surface roughness, thereby allowing only an estimate of
the expected tunneling current. These difficulties notwithstanding, currents greater than
10 uA are seen near -0.lV. These currents exceed tunneling currents by orders of
magnitude and indicate that our layered assemblies are characterized by substantial
defects. We believe that the dominant form of defect is likely a vacancy, given the
currents we measure. The form of the single monolayer scans, showing the presence of
plateau peaks, is characteristic of an array of microelectrodes, consistent with a pinhole

structure.

6.4 Conclusion

The results presented here demonstrate the facile formation of robust multilayer
films where interlayer bonding proceeds through a covalent S-S linkage. The chemistry

is spontaneous and, at the high dithiol concentrations we use, formation of non-looped

136

layers is favored and interlayer oxidative thiol —> disulfide chemistry competes
efficiently with intralayer S-S bond formation. XPS measurements reveal that the S-S
moiety is readily oxidized without loss of overall multilayer robustness, consistent with
the presence of multiple sulfur oxide species with the S-S bond intact. UV photolysis
measurements indicate the susceptibility of the S-S bond to photolytic cleavage. Based
on our ability to form a hybrid (RSO3-Zr-03PR) layered structure, this approach to
multilayer growth will likely provide a useful means of constructing more functionally
complex, ordered interfaces than has been attainable with alkanethiol/gold monolayer

chemistry.

137

6.5 Literature Cited

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

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

L»)

. Dubois, L. H.; Zegarski, B. R.; Nuzzo, R. G. ;J. Am. Chem. Soc., 1990, 112, 570.

P

Chidsey, C. E. D.; Bertozzi, C. R.; Putvinski, T. M.; Mujsce, A. M.; J. Am. Chem.

Soc.,1990, 112, 4301.

5. Chidsey, C. E. D.; Science, 1991, 251, 919.

6. Dubois, L. H.; Zegarski, B. R.; Nuzzo, R. G.; J. Chem. Phys, 1993, 98, 678.

7. Biebuyck, H. A.; Whitesides, G. M.; Langmuir, 1993, 9, 1766.

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

9. Camillone, N. 111; Chidsey, C. E. D.; Eisenberger, P.; Fenter, P.; Li, J.; Liang, K. S.;
Liu, G.-Y.; Scoles, G.; J. Chem. Phys, 1993, 99, 744.

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

11. Folkers, J. P.; Laibinis, P. E.; Whitesides, G. M.; Langmuir, 1992, 8, 1330.

12. Chidsey, C. E. D.; Loiacono, D. N.; Langmuir, 1990, 6, 682.

13. Nuzzo, R. G.; Korenic, E. M.; Dubois, L. H.; J. Chem. Phys, 1990, 93, 767.

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

15. Whitesides, G. M.; Laibinis, P. E.; Langmuir, 1990, 6, 87.

16. Bain, C. D.; Troughton, E. B.; Tao, Y.-T.; Evall, J.; Whitesides, G. M.; Nuzzo, R. G.;

J. Am. Chem. Soc., 1989, 111, 321.

17. Bain, C. D.; Evall, J.; Whitesides, G. M.; J. Am. Chem. Soc., 1989, 111, 7155.

138

l8.

19.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

Bain, C. D.; Whitesides, G. M.; J. Am. Chem. Soc., 1989, 111, 7164.
Strong, L.; Whitesides, G. M.; Langmuir, 1988, 4, 546.
Troughton, E. B.; Bain, C. D.; Whitesides, G. M.; Nuzzo, R. G.; Allara, D. L.; Porter,
M. D.; Langmuir, 1988, 4, 365.
Bain, C. D.; Whitesides, G. M.; Science, 1988, 240, 62.

Porter, M. D.; Bright, T. B.; Allara, D. L.; Chidsey, C. E. D.; J. Am. Chem. Soc.,
1987, 109, 3559.
Nuzzo, R. G.; Zegarski, B. R.; Dubois, L. H.; J. Am. Chem. Soc., 1987, 109, 733.
Allara, D. L.; Nuzzo, R. G.; Langmuir, 1985, 1, 45.

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

Chidsey, C. E. D.; Liu, G.-Y.; Rowntree, P.; Scoles, G.; J. Chem. Phys, 1989, 91,
4421.

Chidsey, C. E. D.; Liu, G.-Y.; Scoles, G.; Wang, J .; Langmuir, 1990, 6, 1804.
Camillone, N. HI; Chidsey, C. E. D.; Liu, G.-Y.; Putvinski, T. M.; Scoles, G.; J.
Chem. Phys, 1991, 94, 8493.

Bucher, J. P.; Santesson, L.; Kern, K.; Langmuir, 1994, 10, 979.

Liu, G.-Y.; Salmeron, M. B.; Langmuir, 1994, 10, 367.

Sondag-Huethorst, J. A. M.; Schonenberger, C.; Fokkink, L. G. J .; J. Phys. Chem,
1994, 98, 6826.

Kim, Y.-T.; McCarley, R. L.; Bard, A. J .; Langmuir, 1993, 9, 1941.

McCarley, R. L.; Dunaway, D. J .; Willicut, R. J .; Langmuir, 1993, 9, 2775.

McCarley, R. L.; Kim, Y.-T.; Bard, A. J.; J. Phys. Chem, 1993, 97, 211.

139

35.

36.

37.

38.

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

Edinger, K.; Golzhauser, A.; Demota, K.; Woll, Ch.; Grunze, M.; Langmuir, 1993, 9,
4.

Kim, Y.-T.; McCarley, R. L.; Bard, A. J .; J. Phys. Chem, 1992, 96, 7416.

Yeo, Y. H.; McGonigal, G. C.; Yackoboski, K.; Guo, C. X.; Thomson, D. J .; J. Phys.
Chem, 1992, 96, 6110.

Alves, C. A.; Smith, E. L.; Porter, M. D.; J. Am. Chem. Soc., 1992, 114, 1222.

Gao, X.; Zhang, Y.; Weaver, M. J .; J. Phys. Chem, 1992, 96, 4156.

Widrig, C. A.; Alves, C. A.; Porter, M. D.; J. Am. Chem. Soc., 1991, 113, 2805.
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.

Widrig, C. A.; Chung, C.; Porter, M. D.; J. Electroanal. Chem, 1991, 310, 335.

Xia, Y.; Whitesides, G. M.; Angew. Chem. Int. Ed., 1998, 3 7, 551.

Kim, T.; Crooks, R. M.; Tsen, M.; Sun, L.; J. Am. Chem. Soc., 1995, 117, 3963.
Laibinis, P. E.; Whitesides, G. M.; Allara, D. L.; Tao, Y.-T.; Parikh, A. N.; Nuzzo,
R. G.; J. Am. Chem. Soc., 1991, 113, 7152.

Kim, Y.-T.; McCarley, R. L.; Bard, A. J .; Langmuir, 1993, 9, 743.

(a) Nakanishi, T.; Ohtani, B.; Uosaki, K.; J. Phys. Chem. B, 1998, 102, 1571. (b)
Nakanishi, T.; Ohtani, B.; Shimazu, K.; Uosaki, K.; Chem. Phys. Lett., 1997, 278,
233.

Tour, J. M.; Jones, L. 11; Pearson, D. L.; Lamba, J. J. S.; Burgin, T. P.; Whitesides,

G. M.; Allara, D. L.; Parikh, A. N.; Atre, S. V.; J. Am. Chem. Soc., 1995, 117, 9529.

140

50.

51.

52.

53.

54.

55.

56.

57.

58.

S9.

60.

61.

62.

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.

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

Katz, H. E.; 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.

Ungashe, S. B.; Wilson, W. L.; Katz, H. E.; Scheller, G. R.; Putvinski, T. M.; J. Am.
Chem. Soc., 1992, 114, 8717.

Katz, H. E.; Schilling, M. L.; Chidsey, C. E. D.; Putvinski, T. M.; Hutton, R. S.;

Chem. Mater., 1991, 3, 699.

141

63.

64.

65.

66.

67.

68.

69.

70.

71.

72.

73.

74.

75.

76.

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

Ansell, M. A.; Zeppenfield, A. C.; Yoshimoto, K.; Cogan, E. B.; Page, C. J .; Chem.
Mater., 1996, 8, 591.

Zhang, Y.; Terrill, R. H.; Tanzer, T. A.; Bohn, P. W.; J. Am. Chem. Soc., 1998, 120,
2654.

Norrod, K. L.; Rowlen, K. L.; J. Am. Chem. Soc., 1998, 120, 2656.
Schoenfisch, M. H.; Pemberton, J. E.; J. Am. Chem. Soc., 1998, 120, 4502.

Harris, J. J .; Bruening, M. L.; J. Am. Chem. Soc., in preparation.

Lowder, J. E.; Kennedy, L. A.; Sulzmann, K. G. P.; Penner, S. S. J. Quant.
Spectrosc. Radiative Transfer 1970, 10, 17.
Fyfe, W. S. J. Chem. Phys. 1953, 21, 2.

Briggs, D.; Seah, M. P. Practical Surface Analysis; Wiley: Chichester, England,
1983.

Fadley, C. S.; Baird, R. J.; Siekhaus, W. J.; Novakov.; Bergstrom, S. A. L. J.
Electron. Spectrosc. Relat. Phenom. 1974, 4, 93 — 137.
Golzhauser, A.; Panov, S.; Woll, Ch.; Surface Sci., 1994, 314, 849.

Zubragel, Ch.; Deuper, C.; Schneider, F.; Neumann, M.; Grunze, M.; Schertel, A.;
Woll, Ch.; Chem. Phys. Lett., 1995, 238, 308.

Wirde, M.; Gelius, U.; Dunbar, T.; Allara, D. L.; Nucl. Inst. Meth. Phys. Res. B,
1997, 131, 245.

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

142

77. Wagner, C. D.; Riggs, W. M.; Davis, L. E.; Moulder, J. F.; Muilenberg, G. E.;
Handbook of X — ray Photoelectron Spectroscopy; Perkin — Elmer Corporation: Eden

Praire, MN, 1979.

78. Scott, J. R.; Baker, L. S.; Everett, W. R.; Wilkins, C. L.; Fritsch, 1.; Anal. Chem,

1997, 69, 2636.

79. Kice, J. L.; Acc. Chem. Res, 1968, 1, 58.
80. Kice, J. L.; Ikura, K.; J. Am. Chem. Soc., 1968, 90, 7378.
81. Folkins, P. L.; Harpp, D. N.; J. Am. Chem. Soc., 1991, 113, 8998.

82. Folkins, P. L.; Harpp, D. N.; J. Am. Chem. Soc., 1993, 115, 3066.

143

Chapter 7

Applying Polymer Chemistry to Interfaces: Layer-by-Layer and
Spontaneous Growth of Covalently Bound Multilayers

Abstract

We report on the covalent layer-by-layer growth of multilayers of
diphenylmethane derivatives where interlayer bonding is through a urea moiety. We
demonstrate two routes to the growth of multilayer structures. The first method uses
alternating copolymerization chemistry where isocyanate-containing monomers are
reacted with arnine-containing monomers, leading to growth that is controlled layer-by-
layer. The second route uses isocyanate-flinctionalized monomers and hydrolysis to
achieve layer growth, leading to spontaneous polymerization and less well-controlled
layer growth. Both routes produce chemically identical structures as determined by 13 C

NMR, FTIR and UV-visible spectroscopies.

144

7.1 Introduction

The past decade has seen a substantial research effort focused on controlling and
manipulating the chemical properties of surfaces and interfaces, with alkanethiol-gold
self-assembled monolayers (SAMs) being perhaps the most widely studied system."2
The motivation for this effort has been that modified surfaces can, in principle, find
application in many technologically important areas such as electronics and electro-
optics3'5 and chemical sensingf”8 Of late it has become clear that the alkanethiol-gold
system is limited in a number of ways; the driving force for layer formation is modestf”IO
leading to labile structures, the thiol head groups are sensitive to oxidative degradation in

”"3 and alkanethiol/gold SAMs cannot be used to form robust

the presence of ozone,
multilayer structures without using multiple synthetic steps.'4 A chemically versatile and
robust alternative to alkanethiol-gold SAMs has been the metal-phosphate/phosphonate
chemistry demonstrated initially by Mallouk.15 "6 This family of materials has proven to
be useful in the design of a host of interesting structural motifs, with applications ranging
from controlled electron transport17 to mediation of porosity in polymer multilayer
films.‘8

The Blanchard group has incorporated or,u)-bisphosphonated thiophene oligomers
into zirconium-phosphonate (ZP) layered assemblies to understand intralayer and
interlayer optical excitation transport. ”’20 Intralayer excitation transport measurements in
conjunction with AFM data for layers bound to oxidized Si substrates revealed the
presence of island structures on these surfaces. The size and distribution of the islands

was mediated by the heterogeneous distribution of surface silanol groups.19 Interlayer

excitation transport measurements on these same systems showed that the Zr(O3PR)2

interlayer linking moiety screened dipolar coupling.20 We attributed this effect to the
polarizability of the ZP sheets between layers and have sought ways to design robust
layered materials that do not possess this functionality.

We have reported previously on the assembly of covalently-coupled disulfide
multilayers on gold.‘4 Because of the direct oxidative thiol-to-disulfide covalent bonding
scheme used in the growth of these systems, layer grth is statistical in terms of overall
assembly thickness, i. e. discrete, single layer grth is not enforceable with this
chemistry. These multilayer assemblies are also susceptible to oxidation of the disulfide
linkages to more highly oxidized species and this oxidation appears to produce lateral S-
O-S bonds parallel to the substrate.

A more general route to the growth of robust, covalent multilayer assemblies is to
take advantage of well-established polymer chemistry. The biggest challenges in
applying polymer chemistry to the controlled growth of interfacial multilayers are to
attach an initial monomer to the substrate and to devise synthetic routes where the
polymerization chemistry terminates alter the addition of a single monomer unit. We
have identified and demonstrated one system for which both conditions hold and a related
system for which only the first condition is met. Comparing the multilayers formed from
the two synthetic routes demonstrates the importance of using step-by-step alternating
copolymerization chemistry to control the growth of multilayers.

The specific chemical systems we report on here form covalent interlayer linkages
based on the reaction of isocyanates with amines to produce urea groups. A significant
benefit of using the urea moiety as an interlayer linking functionality is its ability to form

hydrogen-bonded networks parallel to the substrate. The use of urea functionalities to

146

enhance stability in supramolecular structures is not new. Recently, several groups

“'2" and Hollingsworth,27 have utilized urea

including Etter,2 1'23 Lauher and Fowler,
hydrogen-bonding in the synthesis of various assemblies. The formation of urea linkages
to bind polymer layers to silica substrates has also been reported.28 To our knowledge,
the work described here is the first report where covalent urea multilayer assemblies have
been grown layer-by—layer from dielectric and metallic substrates. We have grown up to
seven layers using alternating copolymerization chemistry without any sign of a decrease
in reactivity. The alternating c0polymerization of 4,4’-methylenedianiline (MDA) and
4,4’-methylenediphenylisocyanate (MP1) to form urea linkages between layers affords
direct control over layer thickness (scheme 7.1). Our purpose in this work is to
demonstrate the facile nature of our approach to covalent layer formation. The choice of
these monomers was based on their spectral properties as well as their commercial
availability. The lateral stability of these multilayer assemblies is enhanced through
intralayer hydrogen bonding between the urea groups. We have also demonstrated the
same urea linking chemistry for multilayers grown using only MPI, where the formation
of reactive amine at each layer is accomplished by deliberate hydrolysis of the isocyanate
to form an amine (scheme 7.2). The main difference between the chemistry presented in
scheme 7.1 and scheme 7.2 is that the grth of the discrete layers one at a time is
possible with the alternating isocyanate/amine copolymerization chemistry and
spontaneous, uncontrolled growth is achieved with the isocyanate/hydrolysis chemistry.

Both deposition schemes yield the same chemical structures as indicated by FTIR, cross-

polarization magic angle (CP/MAS) ‘3 C NMR spectrometry and X-ray diffraction. The

147

primary focus of this paper is on demonstrating the use of isocyanate/amine chemistry to

form covalent multilayer assemblies with varying degrees of control over layer growth.

M NE
.. «M o o
W 05‘ A ””2 001 N00

OH Toluene, 24 hrs, 211C 0 DMF’ 4 1118, 45K: under Al”
\ '/\/\NH- Wash wrth DMF

.1

MéMe

.\ RM 1”“: ”R,
o\ it“ :1: 31111551111110 ’
\S'Mi 11 R1

N
M51111

.1111

 

 

 

Scheme 7.1. Synthetic route for the layer-by-layer formation of urea-linked multilayer

assemblies using the alternating copolymerization of diisocyanates and diamines.
Our data show that the use of well-established polymer chemistry in the design and

growth of covalently-bonded multilayer structures is feasible, pointing the way to a

generic strategy for covalent layered material design.

148

7.2 Experimental Section

Substrates and Reagents. The substrates were made by evaporation of 2000 A of
Au on 200 A of Ti on Si(100) wafers.29 The silicon wafer substrates used were polished
7.5 mm X 15.0 mm Si(100) wafers purchased from Boston Piezo-Optics, Inc. MP1,
MDA, anhydrous and reagent grade dimethylformamide (DMF) were purchased from
Aldrich Chemical Co. and used as received. High surface area silica gel (230-400 mesh)
having a surface area of 450-550 mz/g was obtained from Spectrum Quality Products,
Inc.

Procedure for Layer Deposition: Both the gold and silicon substrates were
cleaned in piranha solution (3:1 H2SO4zH2O2) prior to deposition of multilayers. The
priming of the oxidized silicon substrates (15 A typical oxide layer thickness) was carried
out by reacting them with (3-aminopropyl)dimethylethoxysilane (APDES) in toluene for
24 hours. The priming of the oxidized silicon substrate yielded surface amino groups
which were reacted with the isocyanate containing monomer (MPI) to form the initial
layer (scheme 7.1). For external reflectance IR measurements, layers were grown on
gold-coated substrates. To fimctionalize the gold substrates with reactive hydroxyl
groups, the substrates were immersed in a 10 mM solution of 6-mercapto-1-hexanol in
ethanol at 20°C for ~18 hours. The formation of the hydroxyl-terminated self-assembled
monolayer is followed by reaction of hydroxyl groups on the surface with MP1 and,
subsequently with MDA (scheme 7.1). The polymerization reaction of diisocyanate in
the presence of a small (non-stoichiometric) amount of water yields multilayer
assemblies with the same chemical structure as multilayers produced according to

scheme 7.1. This alternative means of producing multilayer structures is shown in

149

scheme 7.2, where surface—bound terminal isocyanate groups are hydrolyzed to amines
which then react with solution phase isocyanates to generate polyurea layers. We have
also synthesized the bulk polyurea by reacting equimolar amounts of MP1 and MDA in
dry DMF for 6 hours at 40 - 50°C under an argon atmosphere. Weiner has reported the
formation of N,N’-dimethyl-N’-phenylformamidine from DMF and phenylisocyanate at
150°C,30 but we do not detect this side product for our lower temperature reaction
conditions. The procedure and conditions for growing urea multilayers on high surface
area silica are the same as those for grth on oxidized silicon or quartz substrates except

the silica gel suspension was stirred vigorously during the reaction. The resulting high

M Me
OH \
”INWSK O\Si/\/\N Ha © ©
OE‘ -' \ ‘ OCN

 

 

> Me’ Me
OH Toluene, 24 hrs, 20 0C 0\ 31:11:11, :IihthsDEIZCH O
'Si/\/\NH2 ’ 2
Me': 14c

 

 

r1 1‘
M5 Me H\ / 4
magi, © © 31.213.18.122; .0
Me" Me H H
m RRM 1R RR1
Me' Me i
m RRM 1: R1R

Scheme 7. 2. Synthetic route for spontaneous growth of urea-linked multilayer

assemblies using isocyanate/hydrolysis polymerization chemistry.

150

surface area silica with urea-linked multilayers was washed sequentially with hot DMF,
water, acetone and water to remove any physisorbed bulk polymer.

Optical Null Ellipsometry. Layer thicknesses were measured using a null
ellipsometer (Rudolph Auto-EL II) operating at 632.8 nm. The software used for data
acquisition and reduction was from Rudolph. The refractive index, u, used to calculate
the thickness of the films is taken to be 1.45 + 0i. The dependence of the film thickness
on the real part of refractive index is modest for physically realistic values of [1.

Infrared Spectroscopy. F TIR spectra of multilayer assemblies on gold substrates
were measured using a Nicolet Magna 750 FTIR spectrometer equipped with an external
reflectance accessory, operating at a beam incidence angle of ~80° with respect to the
sample surface normal. FTIR spectra of bulk polyurea were also taken in solid and
solution phase forms. The solid-state IR spectrum of polyurea was taken using a pressed
KBr pellet as the sample matrix. The solution phase spectra were acquired with the
polymer dissolved in DMSO, with a Mattson Infinity Gold FTIR spectrometer. The
spectral resolution was 4 cm'1 for all measurements.

U V- Visible Spectroscopy. The urea-linked multilayer assemblies were grown on
quartz substrates using the chemistry shown in schemes I and II. A Unicam model UV-Z
spectrometer was used to collect the spectra.

X-Ray Diffraction. The X-ray diffraction pattern of bulk polyurea was acquired

on a Rigaku Rotaflex diffractometer equipped with a rotating anode at 7» = 0.15418 nm

(Cl-11m)-
NMR Spectrometry. ‘3 C CP/MAS spectra of the bulk polymer and multilayers

grown on silica gel were acquired on a Varian VXR 400 spectrometer at 100.58 MHz.

151

The sample spinning frequencies were between 3.5 kHz and 5.1 kHz and spinning side
bands were identified by acquisition of the spectra at several sample spinning speeds. A
contact time of 2.2 ms, proton decoupling pulse of 5.0 us, recycle delay time of 2.0 s and
acquisition time of 41 .0 ms were used to collect the spectra. Typically, 6,000 scans were
required to obtain spectra with a sufficient signal-to-noise ratio.

Molecular Mechanics Calculations. Molecular mechanics calculations were
performed using the Insight 11 version 97 software package from Molecular Simulation,

Inc. on a Silicon Graphics workstation.
7.3 Results and Discussion

The primary issues we focus on in this paper are the formation of multilayer
structures using isocyanate/amine polymerization chemistry and determining the nature
and chemical speciation of the interlayer bonds. We first examine the issue of layer-by-
layer versus spontaneous grth of the multilayer assemblies using optical null
ellipsometry. The slope of best-fit line of ellipsometric thickness data yields an average
of 14 i- 1 A per bilayer for multilayers grown according to scheme 7.1 (Figure 7.1a). We
report the grth of bilayers instead of monolayers because of the sensitivity of terminal
isocyanate groups to hydrolysis by adventitious water to form a carbamic acid or an
amine. The amine-terminated surface resulting from each bilayer cycle is substantially
more stable to ambient laboratory conditions.

Synthesis of multilayers according to scheme 7.2 gives rise to a nonlinear dependence
between reaction time and multilayer thickness (Figure 7.1b). This is not an unexpected

result because of the conditions employed in this reaction. This grth behavior is an

152

unavoidable consequence of the chemistry used, in which the surface- bound terminal

isocyanate reacts with water to produce an aminated surface that can subsequently react

 

 

a
02‘ 100 L a.
7%” 80
O.) '- i
73 60 _- ‘
g a
'E 40 "' I
E _ 1
i 20 ,9
E 0 "/1 1 1 1 1 1 1 1 1
Q)

0 l 2 3 4 5 6 7 8
number of bilayer growth cycles

700- b

 

 

0:? .
g 600 ;- l/
E 500 — /
. I
o _
o 300 - '
.E .. //
E 200 — /-
o ' -/u
(D 1m '- ./
.9 -
= 0 1 1 1 1 1 1 1 1
°’ 0 1 2 3 4 5 6 7 8
number of reaction cycles

Figure 7.1. (a) Ellipsometric thickness of bilayer assemblies as a function of number of
growth cycles using the chemistry presented in Scheme 7.1. (b) Ellipsometric thickness
of multilayer assemblies as a function of number of reaction cycles using the chemistry

presented in Scheme 7.2.

153

with solution phase isocyanates. An equivalent route is that the conversion of
isocyanates to amines occurs primarily in the solution phase and the amines subsequently
react with surface-bound and bulk isocyanates. The polymerization is thus expected to
take place both on the surface and in the bulk and we do not have an experimental means
to distinguish between these cases. Regardless of which reaction pathway dominates, the
overall effect is a statistical layer growth for each cycle. We have been able to grow
multilayers with thicknesses up to 625 A with 7 reaction cycles. The average growth of
~90 A/reaction cycle far exceeds the growth achievable using anhydrous
copolymerization chemistry for layer deposition. Based on the data in Figure 7.1a, the
625 A thick surface corresponds to an average depth of ~90 monolayers. Despite the
difference in assembly thickness resulting from the differences in growth methodologies,
the same urea interlayer linking chemistry seems to be operative in the formation of
multilayers in both grth schemes (vide infra).

Molecular mechanics calculations for the copolymer predict an average length of
~19 A/bilayer. The experimental ellipsometry data, in comparison to the calculated
thickness, is of limited utility because such a comparison does not differentiate between
partial surface coverage and the chromophores being oriented at a non-zero angle relative
to the substrate normal Assuming complete surface coverage, we calculate a tilt angle of
43° with respect to the surface normal, somewhat larger than the tilt angle measured for

“'32 and ZP layers on SiOx.33 To help resolve this issue, we

alkanethiol/Au monolayers
need to consider the likely conformation of the layer constituents by examining the

conformation of the corresponding bulk polymer. For many monolayer structures with

no explicit lateral, intralayer connectivity, the correspondence

154

 

 

 

 

A 4.221
:21
.13: 4.7
E
E 11.6A
3 3.3/31
0.)
E
a
O 111'1 1' 1'11"111-1 1 '111 11
5 10 15 20. 25 30

Figure 7.2. X-ray powder diffraction data for bulk poly(MDA-MPI).

between bulk and monolayer structure is tenuous. We attempt this comparison based on
the extensive intralayer hydrogen bonding for the multilayer assembly and the fact that
this intermolecular interaction is also present in the bulk polymer.

The X-ray diffraction data on the bulk polymer are shown in Figure 7.2. There
are four relatively broad peaks corresponding to 11.6, 4.7, 4.2 and 3.3 A. The width of
the peaks suggests an amorphous nature and a lack of long range order for the bulk
polymer. This is not necessarily a surprising result because of the number of structural

degrees of freedom available to the polymer. We assign the peak at 11.6 A to the urea

155

\
.'
. -
ti“.

 

Figure 7.3. Calculated structure of two poly(MDA-MPI) oligomers indicating a helical

structure.

156

repeat unit, with the peaks in 3 - 5 A range most likely being associated with the distance
between neighboring polymer chains, as indicated in Figure 7.3. It is interesting to
compare the x-ray diffraction data with molecular mechanics calculations for urea
molecules. We consider the molecular mechanics calculations on the urea oligomers to
be instructive in a qualitative sense because they are essentially vapor phase calculations
and our data are taken on solid state material. We also recognize that molecular
mechanics calculations are prone to finding local minima with resulting structures that
can be misleading. With these caveats in mind, the calculations indicate that the lowest
energy conformation for two tetramers are parallel to each other in vapor phase. The
calculations also predict a 10 i 1 A monomer length, in reasonable agreement with the
X-ray peak at 11.6 A and an interchain spacing of 4 i 1 A, consistent with X-ray
diffraction data showing multiple peaks in the 3 to 5 A range. The calculations indicate
that, at least in the vapor phase, the oligomers adopt a-helical conformation, perhaps due
to Tt-Tt interactions between phenyl rings combined with hydrogen bonding between
adjacent urea moieties (Figure 7.3).

To this point, we have discussed the structural features of the multilayer
assemblies synthesized as indicated in schemes I and II. It is important to examine the
nature of the interlayer covalent linkages in detail. A good deal of information about the
chemical nature of the multilayers and their orientation on the surface can be obtained
with extemal—reflectance FTIR spectroscopy on multilayers grown from gold substrates.
Infrared spectroscopy has been used extensively in the examination of alkanethiol/Au and

other similar self-assembling monolayer structures, specifically regarding the

157

environment and conformation of the aliphatic chains. We apply this body of knowledge
first to understanding the properties of the 6-mercapto-l-hexanol priming layer. The
priming layer CH2 asymmetric and symmetric stretching bands were centered at 2925 and
2862 cm'1 with widths of ~ 54 and 34 cm'1 FWHM, respectively (Figure 7.4a). Both the
positions and widths of the priming layer bands indicate that the C6 alkyl chains are
disordered and exhibit a substantial fraction of gauche conformations. Figure 7.4b shows
the FTIR spectrum from 3200 cm'1 to 2600 cm'1 for polyurea multilayer assemblies
grown on the primed gold substrate. The frequencies of the 6-mercapto-1-hexanol primer
bands shift to 2918 cm"1 and 2846 cm’1 and their linewidths change to FWHNIs of 44 and
38 cm], respectively, upon reaction of the primer with diisocyanate and diamine (Figure
7.4b). The red-shift and narrowing of the asymmetric CH2 stretching vibration after
reaction with diisocyanate and diamine to form a urethane linkage with the primer
suggests that the primer layer exhibits more trans bonds after reaction with the monomer.
Overall, the presence of some residual gauche conformers in the C6 priming layer and the
distribution of MP1 and MDA conformers combines to produce CH2 stretching
resonances that are broader by a factor of ~ 2 than those seen for highly ordered long
chain alkanethiol/Au monolayers.

The IR spectra of these multilayer assemblies contain information on other
functionalities as well. A band at ~3030 cm'1 appears after reaction of primer hydroxyl
groups with MP1 and subsequent reaction with MDA (Figure 7.4b). The 3030 cm'1
resonance is the aromatic C-H stretching band of the MP1 and MDA monomer units. A
broad, unresolved band between 3400 cm'I and 3100 cm"1 (not shown) is due to the

presence of hydrogen-bonded N-H and O-H groups of urethane/urea and carbamic acid.

158

The lower energy region of the IR spectrum, between 1900 cm'1 and 1000 cm],
provides insight into the chemistry involved in the formation of these multilayers (Figure
7.5). The amide l and 11 bands appear at 1665 cm'1 and 1513 cm'l, respectively, and are
consistent with reported studies on bulk polyureas.34 The band at 1593 cm" is the C-C
stretching resonance for the phenyl ring and the band at 1411 cm‘1 is the deformation of
methylene groups. Interestingly, a new band at ~l7l7 cm'l appears in the spectra of the
urea films which can not be assigned to C=O stretching of urea linkages because of its
higher frequency position than urea C=C stretching band.“35 We offer an explanation
for the presence of this band. Carbamic acids are intermediates in the Hoffmann
rearrangement of an isocyanate to form an amine.36 The incomplete hydrolysis of
isocyanate is expected to give the corresponding carbamic acid. Recently, Yoshida et
a1.3 7‘3 8 and Aresta and coworkers”40 have reported the formation of carbamic acids from
amines and carbon dioxide. In their studies, the C=C stretching bands of the carbamic
acid appeared at 1725 cm"1 and ~1700 cm’l.39 We believe our 1717 cm'1 resonance to be
due to the presence of a small amount of carbamic acid. In an effort to verify the identity
of this band, we heated a 20 A thick polyurea film on a gold substrate in a basic solution
of K2C03 (pH 10.2) at ~45°C for 4 hours. According to the Hoffmann rearrangement,
the carbamic acid should hydrolyze to the corresponding amine in basic solution with the
evolution of C02 (scheme 7.3). Figure 7.6 shows the comparison of IR spectra of the

films grown on gold substrates before (Figure 7.6a) and afier K2C03 treatment (Figure

7.6b). The intensity of the 1717 cm'1 resonance diminishes significantly as a

159

15 a

10

 

 

absorbance x10
1

 

30200 3100 3000 2900 2800 2700 2600

frequency (cm 1)

 

 

0.006 b
".12
g: 0.004
O
t: .
w
'8
8 0.002 %I‘
.13
0.000 #

 

3200 3100 3000 2900 2800 2700 2600

-1
frequency (cm )
Figure 7.4. (a) FTIR spectrum of 6-mercapto-1-hexanol adsorbed onto gold prior to

subsequent layer growth. (b) FTIR spectra of multilayers of poly(MDA-MPI) grown on
the primer layer shown in (a). Note the grth of the bands centered at 3030 cm",

indicating the addition of phenyl C—H groups.

160

0.08 r l

.o
83
I

1 increasing thickness

 

absorbance (a.u.)
o
i

.o
8

.—_./

0.00 1
1900 1800 1100 1600 1500 1400 1300 12100 1100 1000

frequency (6111")

 

 

Figure 7.5. FTIR spectra of multilayers of poly(MDA-MPI) showing the amide I and II
band regions and higher frequency C=C stretches. See text for a discussion of band
assignments. The spectra, in order of increasing thickness are for multilayers with
ellipsometric thicknesses of 85 A, 127 A, 164 A, 209 A, 310 A and 540 A. The
multilayers used here were grown according to Scheme 7.2 but spectra for multilayers

grown according to Scheme 7.1 produced identical spectra.

161

consequence of the treatment, consistent with our assignment of this band as the
carbonyl stretching resonance of a carbamic acid. These films are also acid- stable and
resistant to organic solvent attack. Immersion in 1 M HCl at 90°C for 45 minutes
produced no change in either the ellipsometric response or the IR spectrum of the amide

region of the spectrum. Immersion in strong acid appeared to produce oxidation of some

 

0.012
A a
:5 0009 I
3;
8
C:
,8 0.006
§
'8 b
0.003
0.000

1800 1700 1600 1500 1400 1300

frequency (cml)
Figure 7.6. FTIR spectra of the 1800 cm' to 1250 cm'1 spectral region (a) before and (b)

after reaction with K2C03 to establish the chemical origin of the 1717 cm'1 resonance

indicated by an arrow.

fraction of the thiol sulfur, and this is an expected result.14 Immersion of the polyurea

multilayers in cyclohexane and DMF yielded no change in either the ellipsometric

162

thickness or IR spectrum. The multilayer films we report here are resistant to a wide

variety of chemical attacks.

R—NH
NAOH —*——* 2 + C02

a carbamic acid
Scheme 7.3. Hydrolysis of an isocyanate to form a carbamic acid, with subsequent

decomposition to produce an amine.

O
2 \NAOH ‘ \N/\ O/\N/
I +H20 I I
H H H

 

Scheme 7.4. Dehydration of carbamic acids to form an anhydride.

163

e,d g= .

 

 

 

 

 

 

 

 

Tfi' j Y Y I I I I T I I I I V I I l I Y T Y

250 200 150 100 50 0 -50 -100 -150 ppm

Chemical Shift, ppm

Figure 7.7. CP/MAS ‘3 C NMR spectrum of multilayers of poly(MDA-MPI) grown of
high surface area silica. Band assignments are indicated and referenced to the inset
structure. Bands marked with an asterisk are spinning side bands, identified by their

positional dependence on sample spinning speed.

Another interesting feature in the IR data is the presence of a weak band at ~178O
cm"l (Figure 7.5). This resonance is seen for both polyurea films and polymers
synthesized in the bulk. This band remains constant with the addition of polymer layers.
A carbonyl stretching this high in energy is typically assigned to a carbonyl group
incorporated in five- or six-membered rings or in an anhydride. Our explanation for the
1780 cm'1 resonance is a carbonyl stretching resonance for an anhydride formed by the

reaction of two carbamic acids, as shown in scheme 7.4. However, it should be noted

164

that the possibility of this reaction is very low because of the sensitivity of anhydrides to
hydrolysis. In an effort to better understand the chemical structure and identity of
multilayers, we have acquired 13 C CP/MAS NMR spectra of urea films grown on high
surface silica (Figure 7.7). The resonances in the region of O - 50 ppm are the aliphatic
carbons of APDES primer layer and methylene groups between the phenyl rings in the
polyurea layers. There are two resonances in 150 - 170 ppm region associated with
carbonyl groups from the several possible structures discussed above. Although our
ability to characterize the multilayer films is limited by the linewidths of the resonances,
the bands in the 150 - 180 ppm window are fully consistent with the formation of urea-
linked multilayers synthesized according to either scheme 7.1 or scheme 7.2.

The optical spectroscopy of the polymer multilayers we grow provides some
insight into their structure. A point relating to the identity of the chromophore requires
consideration first, however. We deposit substituted diphenylmethane monomers, whose
absorption spectra are characterized by bands at ~240 nm (e = 3,100 M'l cm“) and 290
nm (e = 22,000 M'1 cm'l).41 The formation of the interlayer linkage produces an N,N’-
biphenylurea moiety which can also act as a chromophore. The absorption spectra of
diphenylureas are characterized by a strong absorption near 260 nm (a = 37,200 M’1 cm’
1),42 in agreement with our experimental spectra (Figure 7.8). The data presented in
Figure 7.8 are for a multilayer assembly grown according to scheme 7.1, with the
spectrum of each successive bilayer shown separately. The absorbance for successive
bilayers does not increase linearly. Specifically, at layers 4, 6 and 9, the absorbance is
less than expected based on linear extrapolation. This result is not an artifact. Repeated

growth of multilayers gives similar results; the absorbance of the layers is not constant

165

for each layer but varies with number of layers. It is therefore incumbent upon us to offer
an explanation for this finding. Because the ellipsometric thickness measurements of the
multilayers do not reveal deficits in density that correspond with the absorbance data, the
only explanation for this finding lies in the structure of the oligomers we grow. The
absorbance data point to the fact that the layers are not simply linear extensions of the

layer beneath them. The molecular mechanics calculations we presented in Figure 7.3

0-22 " 0.20: -
0.20 .

0.18
0.16
0.14
0.12
0.10? ,
0.08 f. .
0.06 "
0.04
0.02
0.00

 

 

absorbance (a.u.)

 

 

 

 

wavelength (nm)

Figure 7.8. Absorbance of multilayers as a function of number of bilayers added. The
band at 260 nm is assigned to the diphenylurea chromophore. Inset: Absorbance as a

function of number of bilayers for both resonances.

166

suggest a helical structure for the oligomers we grow and, depending on the turn rate of
the helix, the orientation of some of the biphenylurea chromophores may be such that
they do not couple efficiently to the incident electric field for absorbance measurements.

Clearly more detailed data will be required to resolve this complex structural issue.

7.4 Conclusion

We have demonstrated layer-by-layer covalent grth of multilayers using urea
interlayer linking chemistry. We have shown two routes to polyurea multilayers.
Alternate exposure of the surface to diisocyanates and diamines produces a multilayer
structure where there is discrete control over layer growth. Using only the isocyanate
monomer in the presence of a small amount of water results in spontaneous multilayer
growth. The structural properties of the urea multilayer assemblies grown using scheme
7.1 and 7 .2 is found to be the same despite the difference in extent of control that can be
exerted using the two growth methods. The layers are hydrogen bonded laterally through
the urea functionalities, as indicated by the IR spectra. UV-visible spectroscopy indicates
the orientation of layer varies with the number of layers, consistent with molecular
mechanics calculations. The data we present here serve to demonstrate the general
principle that alternating copolymerization chemistry can be applied effective in the
design of novel covalently-linked interfaces. Depending on the specific polymerization
reaction used, we can achieve layer-by-layer control over multilayer growth. We
anticipate that other, similar chemical approaches will find use in the controlled growth

of interfacial materials.

167

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. Batchelder, D. N., Evans, S. D., Freeman, T. L., Haussling, L., Ringsdorf, H., Wolf,
H. J. Am. Chem. Soc. 1994, 116, 1050.

4. Tarlov, M. J., Burgess, D. R. F., Gillen, G. J. Am. Chem. Soc., 1993, 115, 5305.

5. Katz, H. E., Wilson, W. L., Scheller, G. R. J. Am. Chem. Soc., 1994, 116, 6636.

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

7. Yang, H. C., Dermody, D. L., Xu, C., Ricco, A. J., Crooks, R. M. Langmuir, 1996,
12,726.

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

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

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

11. Zhang, Y.; Terrill, R. H.; Tanzer, T. A.; Bohn, P. W.; J. Am. Chem. Soc., 1998, 120,
2654.

12. Norrod, K. L.; Rowlen, K. L.; J. Am. Chem. Soc., 1998, 120, 2656.

13. Schoenfisch, M. H.; Pemberton, J. E.; J. Am. Chem. Soc., 1998, 120, 4502.

14. Kohli, P.; Taylor, K. K.; Harris, J. J.; Blanchard, G. J.; J. Am. Chem. Soc., 1998, 120,

11962.

168

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

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.

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

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

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

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

Etter, M. C.; Panunto, T. W.; J. Am. Chem. Soc., 1988, 110, 5896.

Etter, M. C.; Urbanczyk-Lipkowska, Z.; Zia-Ebrahimi, M.; Panunto, T. W.; J. Am.
Chem. Soc., 1990, 112, 8415.

Etter, M. C.; Acc. Chem. Res., 1990, 23, 120.

Zhao, X.; Chang, Y.-L.; Fowler, F. L.; Lauher, J. W.; J. Am. Chem. Soc., 1990, 112,
6627.

Kane, J. J.; Liao, R.-F.; Lauher, J. W.; Fowler, F. W.; J. Am. Chem. Soc., 1995, 117,
12003.

Schauer, C. L.; Matwey, E.; Fowler, F. W.; Lauher, J. W.; J. Am. Chem. Soc., 1997,
119, 10245.

Hollingsworth, M. D.; Brown, M. E.; Santarsiero, B. D.; Huffinan, J. C.; Goss, C. R.;

Chem. Mater., 1994, 6, 1227.

Beyer, D.; Bohanon, T. M.; Knoll, W.; Ringsdorf, H.; Blender, G.; Sackman, E.;

Langmuir, 1996, 12, 2514.

169

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

41.

42.

Harris, J. J.; DeRose, P. M.; Bruening, M. L.; J. Am. Chem. Soc., 1999, 121, 1978.
Weiner, M. L.; J. Org. Chem, 1960, 25, 2245.
Bain, C. D.; Troughton, E. B.; Tao, Y.-T.; Evall, J .; Whitesides, G. M.; Nuzzo, R. G.;
J. Am. Chem. Soc., 1989,11], 321.
Allara, D. L.; Nuzzo, R. G.; Langmuir, 1985, 1, 52.
Home, J. C.; Blanchard, G. J.; J. Am. Chem. Soc., 1998, 120, 6336.

Vien, D. L.; Colthup, N. B.; Fateley, W. G.; Grasselli, J. G. Ihe_fland_bqu_gf
J r 1;! 2-... {2.01-1 re- -. ‘r' f! .‘r ' o, C no. u o ‘r . ‘ Academic
Press, London, 1991.

Pouchert, C. J. The Aldrich Library for Infrared Spectra, The Aldrich Chemical Co.:
Milwaukee, WI, 1970.
Ege, S. N. W, D.C. Heath and Company: Lexington, 1989.
Yoshida, Y.; Ishi, S.; Watanabe, M.; Yamashita, T.; Bull. Chem. Soc. Jpn., 1989, 62,
1534.

Yoshida, Y.; Ishi, S.; Kawato, A.; Tadataka, Y.; Yamashita, T.; Yano, M.; Inque, S.;
Bull. Chem. Soc. Jpn., 1988, 62, 2913.
Aresta, M.; Quaranta, E.; Tetrahedron, 1992, 48, 1515.
Aresta, M.; Quaranta, E.; J. Org. Chem. 1988, 53, 4153.

Simons, W. M.; W, Heyden and Son, Ltd.,
London, England, 1979.

Schroeder, W. A.; Wilcox, P. E.; Trueblood, K. N.; Dekker, A. 0.; Anal. Chem,

1951, 23, 1740.

170

Chapter 8

Design and Demonstration of Hybrid Multilayer Structures.
Layer-by-Layer Mixed Covalent and Ionic Interlayer Linking
Chemistry

Abstract

We report on the growth of layered molecular assemblies where the interlayer
attachment chemistry is controlled layer-by-layer. We demonstrate the compatibility of
chemistry where the layers are connected by ionic coordination chemistry [ROP032’-Zr4+-
'OzCR]+OH' and by the formation of covalent urea moieties. A maleimide-vinyl ether
(MVE) copolymer containing pendant benzoic acid and butyl alcohol functionalities is
used for ionic layer growth. Coupling covalently-bonded adlayers to the MVE polymer
is achieved by attachment of poly(ethylene imine), PEI, to the MVE surface. Subsequent
reaction of the PEI surface with diisocyanates and diamines produces urea-linked
covalent multilayers. The covalent multilayers can be converted to ionic growth
chemistry by treatment of the aminated terminal surface with POCl3 and water, followed
by further reaction with MVE polymer and Zr4+ ions. We report the reaction schemes for
these hybrid layer structures and the characterization of these novel materials by optical
ellipsometry, FTIR and UV-visible spectroscopy, XPS and X-ray diffraction. The data
show the formation of robust multilayer assemblies characterized by limited order within

each layer.

171

8.1 Introduction

The organization of mono- and multilayer molecular assemblies at solid surfaces
provides a rational approach for the fabrication of interfaces with well-defined structure,
composition, and thickness. Layered assemblies may ultimately find application in areas

2 chemical sensing,3 surface passivation,4

such as nonlinear optical devices,"
photoreactivity,5'7 and chemical separations?“9 The ability to control interfacial processes
has important implications for both fundamental scientific and technological advances.
Of particular significance is the ability to grow layered materials where there is
substantial control over the layer thickness and uniformity. Layer-by-layer deposition of
films provides spatial resolution over composition and molecular orientation relative to
the substrate. Both of these properties are often critical to the macroscopic properties of
the resulting system. Metal-bisphosphonate chemistry has been used extensively to
produce robust multilayer interfacial materials because of its simplicity and
versatility."'0'27

We have reported previously on polymeric multilayer assemblies using
zirconium-phosphonate (ZP) interlayer linking chemistry.28'3° The maleimide-vinyl ether
polymer layers we used in that work are resistant to both thermal and solvent attack and
these materials can be of use in the design of a wide variety of multilayer assemblies.
While the ZP interlayer linking chemistry is robust and allows for exquisite control over
layered material assembly, this chemistry is not well suited for all applications. Our work
on multilayer assemblies of cam-bisphosphonated thiophene oligomers indicated that the

ZP interlayer linking moiety screens dipolar excitation transport efficiently, owing to the

polarizability of the ZP sheets between layers and the resultant spatial variation in the

172

dielectric response of the multilayer stack.3 "32 To avoid this structural limitation to facile
interlayer excitation transport, we reported recently on a synthetic route to multilayer
interfacial assemblies using covalent interlayer linking chemistry. Controlled growth of
multilayers is achievable using diisocyanate and diamine alternating copolymerization
chemistry.33 The resulting urea-linked multilayers possess structural stability both
transverse and lateral to the substrate because of the propensity of urea to hydrogen-bond.

With both ionic and covalent multilayer chemistry demonstrated independently,
we report here on the combination of these chemistries to produce multilayers where the
interlayer linking chemistry can be changed between ionic and covalent in a well-
controlled manner. We report in this paper on the synthesis and characterization of
layered hybrid interfaces formed using zirconium-phosphate/carboxylate (ionic) and
diisocyanate/diamine (covalent) chemistries. We have characterized these hybrid
multilayer assemblies using optical ellipsometry, X-ray diffraction and FTIR, UV-visible

and X-ray photoelectron spectroscopies.

8.2 Experimental Section

Substrates and Reagents. The substrates used in this work were made by
evaporation of 2000 A of Au on 200 A of Ti on Si( 100) wafers.34 The substrates were
polished 7.5 mm X 15.0 mm Si( 100) wafers purchased from Boston Piezo-Optics, Inc.
Maleic acid, 4-aminobenzoic acid, 4,4’-methylenedianiline, 4,4’—
methylenediphenyleneisocyanate, phosphorous oxychloride, collidine, the sodium salt of
4-styrene sulfonic acid hydrate, branched polyethyleneimine (MW ~ 10,000),
dimethylformamide (anhydrous and reagent grade) were all purchased from Aldrich

Chemical Co. Potassium persulfate was obtained from Fisher Scientific Company, (3-

173

aminopropyl)-dimethylethoxysilane was purchased from United Technologies, high
surface area silica gel (230-400 mesh, surface area 450-550 mz/g) was obtained from
Spectrum Quality Products and maleic anhydride is purchased from Arco. All
compounds were used as received.

Synthesis of monomers and polymers: N-(4-carboxyphenyl)maleimide (NCPM) is
synthesized in a two step scheme as reported previously.”36 In the first step, equimolar
amounts of 4-aminobenzoic acid (ABA) and maleic anhydride were reacted to produce
the corresponding amic acid quantitatively. In the second step, the cyclization of amic
acid is accomplished by reaction with acetic acid and sodium acetate at ~ 80°C for 2
hours. 1H NMR (do-DMSO, ppm): 7.2 (s, 2H of vinylic group on the imide), 7.48 (d,
2H, phenyl adjacent to carboxylic acid), 8.4 (d, 2H, phenyl group adjacent to imide). The
alternating copolymer of N-(4-carboxyphenyl)maleimide and 4-vinyl ether butanol
(poly(NCPM-VEB» was synthesized by radical copolymerization in chloroform using
AIBN as initiator.28’30’37 ‘H NMR(d6-DMSO, ppm): 1.0-1.8 (2H); 1.8-2.4 (2H), 3.0-3.8
(6 H), 4.2-4.6 (1H), 7.0- 7.6 (2H) and 7.8-8.2 (2H).

Layer Deposition: Both the gold and silicon substrates were cleaned in piranha
solution (3:1 HZSO4zH202) prior to layer deposition. Oxidized silicon and silica
substrates were phosphorylated using POC13 in dry acetonitrile, followed by zirconation
using a 5 mM solution of ZrOClz (60:40 ethanol:water)”30 For external reflectance IR
measurements, layers were grown on gold-coated substrates. To functionalize the gold
substrates with reactive hydroxyl groups, the substrates were immersed in a 10 mM
ethanolic solution of 6-mercapto—1-hexanol at 20°C for ~18 hours. The formation of the

hydroxyl-terminated self-assembled monolayer is followed by reaction of the surface

174

hydroxyl groups with POCl3 in dry acetonitrile. The zirconation of the layers is
accomplished by immersion in the ZrOClz solution for 20 minutes. This procedure yields
a zirconium-rich surface which is used to form [ROPO32'-Zr4+-’02CR]' multilayers. We
have used the chemistry demonstrated by Bakiamoh and Blanchard38 and Alsten et (11.,39
in which they have shown that, in addition to phosphates and phosphonates, zirconium
also complexes with carboxylates and sulfonates.

Covalent multilayers can be grown on ionic multilayers by a reaction scheme
where surface-bound terminal amino groups are reacted with diisocyanates, then with
diamines, repetitively, to generate polyurea multilayers. The terminal amine can be
converted readily to a phosphate by reaction with POCl3/HZO followed by complexation
with Zr4+ and complexation with carboxylate pendant groups on the polymer.

Optical Null Ellipsometry. Layer thicknesses were measured using an optical null
ellipsometer (Rudolph Auto-EL 11) operating at 632.8 nm. The software used for data
acquisition and reduction was from Rudolph (DAF IBM). The refractive index, u, used to
calculate the thickness of the films is taken to be u = 1.462 + 0i. While the index of the
covalent and ionic layers may differ slightly, for the film thicknesses reported here,
modest changes in Re{n} give rise to negligible changes in the calculated film thickness.

Infrared Spectroscopy. FTIR spectra of multilayer assemblies on gold substrates
were measured using a Nicolet Magna 750 FTIR spectrometer equipped with an external
reflectance accessory, operating at a beam incidence angle of ~80° with respect to the

sample surface normal. The spectral resolution was 4 cm'1 for all measurements.

175

U V- Visible Spectroscopy. A Unicam model UV-2 spectrometer was used to
collect absorption spectra of the ionic, covalent and hybrid multilayer structures. Spectral
resolution was 2 nm for all measurements.

X-Ray Diffraction. The X-ray powder diffraction pattern of bulk polyurea was

acquired on a Rigaku Rotaflex diffractometer equipped with a rotating anode at it =

0.15418 nrn (Cum).

X-ray Photoelectron Spectroscopy. XPS measurements were made on multilayer
samples with a PHI 5400 X-ray spectrometer. The X-ray source is the Al KCl line and all
the values reported are referenced to the Cl 5 line at 284.6 eV.

Molecular Mechanics Calculations. Molecular mechanics calculations were performed

using Hyperchem software (v. 4.0) on an IBM-compatible PC.

8.3 Results and Discussion

Controlling the properties of ultrathin films and coatings on metallic and
dielectric surfaces has been a challenge to the chemistry and materials communities for
years. Many methods have been devised for the formation of thin films, including

4 . - . - , ,2
l covalent,33 and romc11028313 layer growth and

Langmuir—Blodgett deposition,40’
electrostatic binding.”43 We report on the growth of hybrid multilayers in this paper,
where interlayer attachment chemistry can be determined layer-by-layer to be either
ionic, [ROPO32'-Zr4+-'02CR]+OH’, or covalent, [RNHCONHR].44“18 We expect these
hybrid multilayer assemblies to have properties that may be difficult to obtain using only
a single type of interlayer linking chemistry within the assembly. The focus of this paper

is to demonstrate our ability to synthesize multilayer structures and control the interlayer

linking chemistry at each step to produce interfaces containing “bands” of ionically and

176

covalently linked multilayer stacks. We have used the chemistry reported by Bakiarnoh

and Blanchard,38 Alsten,39 and Schwartzfoso

in which zirconium ions can complex with
sulfonate, carboxylate and phosphate. Because of the ability of Zr4+ to complex with
these fimctionalities, robust multilayer fihns using mixed complexation chemistry have
been demonstrated. A key point in the use of mixed complexation chemistry is that such
a structural arrangement requires the use of free ionic species to achieve macroscopic
charge neutrality. Such mixed complex layers should, therefore, be conductive.

Ionic Multilayers. To form an ionically-bound multilayer assembly, we deposit a

layer of poly(NCPM-VEB), on the primed and zirconated substrate. The bonding of the

3;.
3

ONOONOONO

, g 0‘3.

 

Zr Zr Zr Zr Zr Zr
0&0 (x80 09,0 0.0 , 00180 0. 02,0 (I) P00. 063,0 0.0,0 069,0
t) t) l) Poly(NCPM-VEB) l) l (l; (2) Zr‘*(aq) l, p l
> L m
I Poly(NCPM-VEB) l

 

(l ) POCI3 additional layers
(2) Zr“*(aq)
Scheme 8.1. Surface functionalization and adsorption scheme for poly(NCPM-VEB)

multilayers.

177

polymer to the zirconated substrate is accomplished by complexation of the pendant
carboxylate on the phenylsuccinimide monomer units of the polymer backbone, resulting
in a mixed complex structure [ROPO32'-Zr4+-'02CR]+, with charge neutrality being
enforced by the presence of either OH' or Cl’ (vide infra). The first step in bonding the
next polymer layer to the interface is to convert the VEB vinyl ether butanol functionality
to a butylphosphate (using POC13) followed by complexation with Zr“, then adsorption
of another poly(NCPM-VEB) layer. This process is shown in the scheme 8.1.

We observe a linear increase in the ellipsometric thickness of the poly(NCPM-
VEB) adlayer with number of growth cycles (Figure 8.1). We have grown up to 8 layers
using this chemistry and we see no evidence of decrease in the reactivity with increasing
number of layers. The slope of the best fitting line is ~ 23i2 A/layer and each data point
in Figure 8.1 represents an average of 18 individual ellipsometric measurements.
Molecular mechanics calculations predict the length of the repeat unit of poly(NCPM-
VEB) to be ~21 A. The apparent agreement between the measured ellipsometric
thickness and molecular mechanical calculation prediction is likely fortuitous because of
the substantial disorder that exists in the polymer layers. The similarity between
calculation and experiment suggests, but does not prove, the absence of significant inter-

layer penetration in these systems.

178

 

 

5003

A400-

°$

3

153300?

3

.0

@200-

O

a

%100-
0111111..lr....11..lr
O 5 10 15 20

number of layers

Figure 8.1. Ellipsometric thickness as a function of reaction cycle. The first eight layers
are poly(NCPM-VEB), bound together with Zr“. The next 7 data points are for growth
of poly(MPI-MDA), bound together with urea linkages. The ~100 A step between the
ionic and covalent layers is due to the adsorption of the PEI layer. The four data points

for layers 16 - 19 are for poly(NCPM-VEB) grown on the covalent layers.

179

To this point, we have considered only the growth of ionic multilayers. We can
make use of either reactive carboxylic acid or hydroxyl functionalities on poly(NCPM-
VEB) to convert from ionic to covalent layer growth. The hydroxyl groups can be

- - 51,52
reacted wrth isocyanates or

 

I 1. soc1,/ CH3CN

A» \\~//— N>. ‘ 2. ‘ xiN y
9::- / R
.l;; 0 .

O
”O\/\/\\O // NH:
/\ 4{
l N
\/ \
\\
O

 

 

   

Scheme 8.2. Reaction of poly(NCPM-VEB) layer with SOClz and PEI to produce an

amine-rich surface.

180

acid chlorides to form urethane or ester interlayer linkages, respectively. The
carboxylates can be converted to the corresponding acid chloride53 or acid anhydride,54
followed by reaction with an alcohol or amine to form ester or amide linkages,
respectively. To activate the surface of the 8-ionic layer assembly, we reacted the
polymer pendant benzoic acid functionality with thionyl chloride in the presence of

collidine. The resulting surface is reacted with branched poly(ethyleneimine), PEI. We

I
l

(3%

f0

3:
\/z\/\2/\/z\/\Z/\/z\/\Z/\/z\
\?O
E
2
8

1

if"

5.

o
9.

x?

 

H...

1

I

‘5’ R
0 he
‘33
:30
0%
Q 0

O

m-

DMF, 45°C, 4 hrs.

I
l

E

2:

\/z\/\z/\/z\/\z/\/Z\/»/\/z\

Scheme 8.3. Reaction of PEI surface with MP1 and MDA for form covalent multilayer

structures.

181

chose to use branched PEI because the amine groups that do not react to form an amide
linkage with the poly(NCPM-VEB) underlayer provide a new reactive surface having
primary and secondary amine groups available for subsequent layer growth. Treatment
with PEI yields a relatively high density of reactive amine groups on the surface and
provides a facile means for altering the density and distribution of surface reactive sites.
This surface functionalization chemistry is presented in scheme 8.2.

In this reaction, we believe that there are a relatively small number of free
carboxylic acid groups remaining on the poly(NCPM-VEB) surface after complexation
with zirconium, and will return to a discussion of this point later. A result of the low
density of reactive attachment points for the PEI is the formation of a relatively open PEI
layer. This prediction is borne out experimentally, as is seen in Figure 8.1, where there is
a substantial jump in adlayer thickness upon formation of the PEI layer.

Once the amine-rich surface is formed, we react it with 4,4’-methylenediphenyl-
isocyanate, MP1, to form a urea interlayer bond (scheme 8.3).33’53’54 Because we use an
excess of diisocyanate and maintain anhydrous conditions, the terminal isocyanate groups
on the new adlayer are available to react with 4,4’-methylenedianiline, MDA, producing
another urea linkage and an amine-terminated surface. Repeating this sequence of

33’53’54 We observe

reactions with MP1 and MDA produces covalently-bound multilayers.
a linear growth in the thickness of the covalent multilayers with a slope of 12 A/layer.
This value is in agreement with our previous work33 and with molecular mechanics
calculations, which predict the repeat unit to be ~11 A.

With the grth of covalent layers established, we consider the conversion back

to ionic layer growth. Reaction of the amine-rich surface with POCl3/water yields a

182

phosphated surface, capable of [ROPO32'-Zr4+-'O;;CR]+ complexation chemistry to form
ionic multilayers (scheme 8.4). On top of the silicon substrate with 8 layers of
poly(NCPM-VEB) and 6 urea layers already grown, we have deposited 4 more
poly(NCPM-VEB) layers (Figure 8.1). We have thus “sandwiched” 6 urea covalent
multilayers between ionic multilayers. The thickness of the top 4 layers is 29 i 3
A/layer, close to but larger than the 8 poly(NCPM-VEB) layers grown previously on
oxidized silicon (Figure 8.1). We attribute the larger slope of this second poly(NCPM-
VEB) region to the change in active site density induced by the PEI and covalent layer
chemistry. As noted above, the application of PEI, in effect, changes the density and
distribution of chemically reactive sites on the surface.

Using external reflectance FTIR spectroscopy, we can obtain a good deal of
information about the interlayer bonding nature of multilayers and their orientation
relative to the substrate. For FTIR measurements, we form the layers on a gold surface
using 6-mercapto-l-hexanol as the priming layer. Figure 8.2 show the 3600 cm'1 to 2600
cm'1 and 2000 cm’1 to 1000 cm'1 regions of the FTIR spectra of poly(NCPM-VEB)
multilayers as a function of number of deposition cycles. As the thickness of
poly(NCPM-VEB) increases from 30 A to 180 A, the CH stretching bands of the 3100
cm'1 to 2800 cm’l spectral window go from showing resonances dominated by 6-
mercapto-l-hexanol to a larger contribution from the polymer layers. The most obvious
manifestation is the grth of the band near 2963 cm'1 in the top spectrum (Figure 8.2a),
indicating a contribution from the benzoic acid pendant side groups. Also important is
the increasing contribution from H-bonded hydroxyl groups centered at ~ 3300 cm'l,

indicating interactions between the polymer butanol side groups. The bands in the 1750

183

 

cm’l to 1650 cm'1 region are characteristic of the benzoic acid functionality, with the top-
most spectrum in Figure 8.2b, showing an additionalband at ~1650 cm"l due to metal
ion-complexed -COO' groups. The resonance between 1200 cm'1 and 1100 cm'1 are

characteristic of the PO group, which exhibits a linear dependence on number of growth

cycles.

 

Scheme 8.4. Reaction of MDA-terminated surface with POClg/ZrOC12(aq) and

poly(NCPM-VEB) to convert from covalent to ionic interlayer linking chemistry.

184

 

 

 

 

   

 

0.024 -
180 A
0.020 —
3
3 0.016 -
Q)
g 0.012 - 96 A
"3
g (1008 r"'7flfl-—_—7—777“‘--~_/c«~\~“_“--
.0
m 0 004 —
. 30 A
0000 r 1 . r . L . r . J
3600 3400 3200 3000 2800 2600
frequency (cm")
0.03 - b
2?
3
o 0.02 -
O
5
e 180 A
8 (101 -
‘2 96 A
30 A W
0.00

 

 

 

2000 i 1800 i 1600 A 1400 ‘ 1200 i 1000
frequency (cm")
Figure 8.2. (a). FTIR spectrum of poly(NCPM-VEP) layers grown on a gold substrate
and primed with 6-mercapto-1-hexanol. The growth of the band at 2963 cm'1 and the
broad feature around 3300 cm'1 are indicative of layer growth. The 3300 cm'1 band is
characteristic of H-bonded polymer side groups. (b) F TIR spectrum of the same system
in the 2000 cm'I to 1000 cm'1 region. The band at 1650 cm'1 is characteristic of the
presence of uncomplexed COOH groups. The bands in the 1200 cm‘1 to 1100 cm'1

region are associated with the phosphate group.

185

 

Following the reaction of reactive hydroxyl groups of poly(NCPM-VEB) on the surface
with diisocyanate and diamine yielded the urethane and urea linkages respectively. The

evidence of formation of urea linkages comes from sharp bands in the

0.25

1

0.20

I

.9

p—t

LII
1

 

 

 

Absorbance (AU)
o
S

.O

O

LII
T

 

0.00 M J‘ANLLAM
A

36001340013200.3000. 1180011600‘140011200‘1000

frequency (0111")

 

 

Figure 8.3. FTIR spectra of poly(MPI-MDA) covalent layers grown on poly(NCPM-
VEB) ionic multilayers. The bottom spectrum is of the poly(NCPM-VEB) layers and the
top spectrum is with covalent multilayers added. The amide I and amide 11 bands are
seen for the covalent multilayers at 1643 cm’] and 1563 cm", respectively. The
resonance in the top spectrum at ~ 3300 cm'1 is characteristic of H-bonding of the urea

nitrogens.

186

1700-1500 cm’1 regions (Figure 8.3). We observe amide I and amide II vibration bands
at ~ 1643 cm"1 and ~ 1563 cm", respectively, having ~100 A thick urea-linked layers and
~96 A of poly(NCPM-VEB) and these bands are characteristic of urethane and urea
linkages (Figure 8.3). The increase in the urea multilayer thickness from 100 A to 810 A
leads to an increase in the intensities of stretching resonance peaks of amide I and II,

C=C and C-H which is also consistent with a previous report.”

 

 

2468101214161820

 

20

Figure 8.4. X-ray powder diffraction pattern for a sample of an 18 layer hybrid
multilayer stack. The absence of sharp features demonstrates the absence of crystalline

structure in these layers.

187

We have also taken the X-ray diffraction spectrum of 18 layered multilayers assembly
and is shown in the Figure 8.4. We do not observe any sharp diffraction features but we
observed a very broad response between 20 z 6° to 18° with a maximum at 20 ~12°,
consistent with amorphous polymeric multilayers. Although the peak of X-ray
diffraction pattern is very broad, we can get a rough estimate of the thickness of these
layers from this spectrum. Assuming 20 = 12°, we recover the distance between two

layers of ~1 0 A, in qualitative agreement with the ellipsometric data.

25000 - \Auger C O ls
M

g
$515000 W 1 N15 /Cls
8

,3

Zr3d

5000 -
P29

0 1 r 1 1 r l
1200 1000 800 600 400 200 0
Binding Energy (eV)

 

Figure 8.5. XPS survey spectrum of an 18 layer hybrid multilayer stack. The data show

the presence of the expected elements (see text) and the absence of Cl.

188

X-ray photoelectron spectroscopy (XPS) is useful in understanding the layer
constituents and their relative amounts. Such data provide a valuable check on the
chemistry we apply to these surfaces. We show in Figure 8.5 a survey scan of oxidized
silicon substrate modified with 18 hybrid multilayers. We find the expected elements (0,
Si, C, N, P, and Zr) on the SiOx substrate following deposition of multilayers. All the
peaks are referenced to Cls peak at 284.6 eV. We assign the peaks as follows:
Elemental Si;p at 98.7 eV; SiOx (of Sizp) at 102.5 eV; P2, at 191.5 eV and P2p at 135 eV;
Zr3d5/2 at 183.5 eV and Zr3d7/2 at 185.8 eV; N], at 401.5 eV; and O], at 532.5 eV. Also,
the Cls at 284.6 eV and a low intensity peak at 289 eV corresponding to the carbonyl of
the amide and urea moieties. There are also Auger resonances for O at ~ 750 eV and C at
~l keV. Since silica is an insulator and silicon a semiconductor, a gold contact was used
to avoid charging of the sample. One open issue for these materials is the chemical
identity of the counterion used to enforce charge neutrality in the ionic interlayer
lamellae. Because we use ZrOClz(aq) as the zirconating solution, it is not immediately
clear whether C1' or OH" is the dominant counterion. The XPS data shown in Figure 8.5
do not exhibit any Cl resonances, suggesting OH’ to be the dominant counterion.

ur ellipsometry and FTIR data suggest relatively uniform layer growth. We can
also examine this issue using UV-visible absorption spectroscopy (Figure 8.6). For
[ROPO32'-Zr4+-'OZCR]+OH' multilayers, the 4-succinimido-benzoic acid constituent is
the chromophore and for the covalent layers, N,N’-biphenylurea is the chromophore. In
our calculations we have used a value of ~ 11,900 L/mol cm for extinction coefficient (8)
of benzoic acid at 7mm = 243 nm.55 The observed linear increase in absorbance of

poly(NCPM-VEB) with number of growth cycles yields ~ 1.25 x1014 chromophores/cm2

189

 

per layer. Assuming a surface density of silanol groups on silicon of 4 X 1014 sites/cmz,2

8

the polymer is present at a density of ~1/3 of the possible surface coverage. The most

likely reason for this apparent partial surface coverage is the steric limitations imposed on

the system by the structures of the substrate and the polymer side groups.

absorbance (AU)

0.16

0.12

0.08

 

 

0.10 '

99
£883

absorbance (AU)
9:

p
O
N

0.00

 

wavelength (nm)

covalent layers (260 n’r:)(./r
I/é ionic layers (243 nm)

L l l l l l l 1

0123456789101112
numberofbilayers

 

Figure 8.6. Absorbance spectra of multilayer stack as a function of layers added. For

the ionic layers, the absorbance maximum is at 243 nm and for the covalent multilayers,

the absorbance maximum is at 260 nm. Inset: Absorbance at 243 nm (ionic) and 260 nm

(covalent) for the multilayer stack showing linear grth with each type of chemistry.

190

The central point is that we observe approximately the same loading density of
poly(NCPM-VEB) for each layer. We note that the Beer’s law slope of the poly(NCPM-
VEB) layers is the same, both before and after the addition of covalent layers. This result
stands in contrast to the ellipsometric data and argues for a change in the structural
properties of the polymer that depend on substrate identity.

Following the formation of urea multilayers on poly(NCPM-VEB) films, the peak
maxima of the UV-Vis absorption is red shifted and the absorbance per layer has also

increased. The N,N’-biphenylurea chromophore has a Xmax at ~ 260 nm and 2 ~ 37,200

L/mol cm.56 As can be seen from Figure 8.6, the Arm of poly(NCPM-VEB) is ~ 243 nm,
and the observed band shifts to ~ 249 nm after the formation of the first urea-linked layer.
Following the deposition of two more biphenylurea layers, the absorption maximum
shifts to 260 nm, consistent with the presence of the biphenylurea chromophores.”55
The extinction coefficient of biphenylurea is approximately twice that for poly(NCPM-
VEB), resulting a slope for the biphenylurea absorbance data that is double that for the
poly(NCPM-VEB) multilayers. Once ionic layer chemistry using poly(NCPM-VEB) was
reinstituted, the slope of the absorbance data reverts to ~0.004/layer at km, ~ 245 nm.

These absorbance data also confirm that we have ability to change from [ROP032‘-Zr4+-’

OzCR]+ interlayer linkages to and from covalent linkages.

8.4 Conclusions

We have reported the growth of layered molecular assemblies where the
interlayer attachment chemistry is controlled layer-by-layer. We demonstrate the

compatibility of layered growth chemistry where the layers are connected by ionic

191

coordination chemistry of the form [ROPO32'-Zr4+-’OZCR]+OH'. A maleimide-vinyl
ether (MVE) c0polymer containing pendant benzoic acid and alcohol functionalities is
used for ionic layer growth. Coupling covalently-bonded adlayers to the MVE polymer
is achieved by attachment of poly(ethyleneimine), PEI, to the MVE surface. Subsequent
reaction of the PEI surface with diisocyanates and diamines produces covalent
multilayers. The covalent multilayers can be converted to ionic growth chemistry by
treatment of the aminated terminal surface with POCl3 and water followed by further
reaction with MVE polymer and Zr“ ions. The optical ellipsometry, FTIR and UV-
visible spectroscopy, XPS and X-ray diffraction data show the grth of robust layers
with limited organization within each layer. We anticipate that these layered structures
will find use in applications where the formation of ultrathin conducting and insulating

layers are important, such as the formation of organic semiconductor structures.

8.5 Literature Cited

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

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

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

4. Swalen, J. P.; Allara, D. L.; Andrade, J. P.; Chandross, E. A.; Garoff, S.; Israelachvilli,
J.; McCarthy, T. J.; Murray, R.; Pease, R. F.; Rabolt, J. F.; Wynne, K. J.; Tu, H.

Langmuir 1987, 3, 932.

192

5. Calvert, J. M.; Georger, J. H., Jr.; Peckerer, M. C.; Perhsson, P. E.; Schnur, J. M.;
Scheon, P.; E. Thin Films 1992, 114, 9188.

6. Dulcey, C. S.; Georger, J. H.; Krauthamer, V.; Stenger, D. A.; Fare, T. L.; Calvert, J.
M. Science 1991, 252, 551.

7. Kumar, A.; Biebuyck, H. A.; Abbott, N. L.; Whitesides, G. M. J. Am. Chem. Soc.
1992, 114, 9188.

8. Vrancken, K. C.; Van Der Voort, P.; Gillis-D’Hammers, 1.; Vansant, E. F .; Grobet, P.;
J. Chem. Soc., Faraday Trans. 1992, 88, 3197.

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

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

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

12. 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.

13. 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.

14. Cao, G.; Rabenberg, L. K.; Nunn, C. M.; Mallouk, T. E. Chem. Mater. 1991, 3, 149.

15. Rong, D.; Hong, H. — G.; Kim, Y. — 1.; Krueger, J. S.; Mayer, J. E.; Mallouk, T. E.
Coord. Chem. Rev. 1990, 97, 237.

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

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

193

 

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

l9. Snover, J. L.; Byrd, H.; Suponeva, E. P.; Vicenzi, E.; Thompson, M. E. Chem.
Mater. 1996, 8, 1490.

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

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

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

23. Ungashe, S. B.; Wilson, W. L.; Katz, H. E.; Scheller, G. R.; Putvinski, T. M. J. Am.
Chem. Soc. 1992, 114, 8717.

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

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

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

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

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

29. Kohli, P.; Blanchard, G. J. Langmuir 2000, I6, 695.

30. Kohli, P.; Rini, M.C.; Major, J. S.; Blanchard, G. J. Langmuir, in review.

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

4419.

194

 

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

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

. Harris, J. J.; DeRose, P. M.; Bruening, M. L.; J. Am. Chem. Soc. 1999, 121, 1978.

. Liu, F. J.; Munukutia, S.; Levon, K.; Tesoro, G. J. Polym. Sci. Part A: Polym. Chem.

1992, 30, 157.

36.

37.

38.

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

Roa, B. S. J. Polym. Sci. Part C: Polym. Lett. 1988, 26, 3.
Kohli, P.; Scranton, A. B.; Blanchard, G. J. Macromolecules 1998, 31, 5681.
Bakiamoh, S. B.; Blanchard, G. J. Langmuir 1999, I5, 6379.

Alsten, J. G. V. Langmuir 1999, 15, 7605.

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

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

Keller, S. W.; Kim, H. N.; Mallouk, T. B.; J. Am. Chem. Soc., 1994, 116, 8817.

Keller, S. W.; Johnson, S. A.; Brigham, E. S.; Yonemoto, E. H.; Mallouk, T. E.; J.

Am. Chem. Soc., 1995, 117, 12879.

Chen, Y.; Iroh, J. 0. Chem. Mater. 1999, 11, 1218.

Arkels, B. CHEMTECH 1999, 12, 7.

Colvin, V. L.; Schlarnp, M. C.; Allvisatos, A. P. Nature 1994, 370, 354.
Loy, D. A.; Shea, K. J. Chem. Rev. 1995, 95, 1431 and reference therein.
Caruso, F.; Caruso, R. A.; Méhwald, H. Science 1998, 282, 1111.

Lu, G.; Purvis, K. L.; Schwartz, J .; Bemasek, S. Langmuir 1997, I3, 5791.

195

 

50. Amoff, Y. G.; Chen, B.; Lu, G.; Seto, C.; Schwartz, J .; Bemasek, S. L. J. Am. Chem.
Soc.l997,119,259.

51. Beyer, D.; Bohanon, T. M.; Knoll, W.; Ringsdorf, H.; Elender, G.; Sackman, E.;
Langnudr1996,12,2514.

52. Marx-Tibbon, S.; Ben—Dov, 1.; Willner, I. J. Am. Chem. Soc. 1996, 118, 4717.

53. Duevel, R. V.; Corn, R. M. Anal. Chem. 1992, 64, 337.

54. Yan, L.; Huck, W. T. S.; Zhao, X.-M.; Whitesides, G. M. Langmuir 1999, 15, 1208
and references therein.

55. The Sadtler handbook of Ultraviolet Spectra, page 719, The Sadtler Research
Laboratories, PA, 1979.

56. Schroeder, W. A.; Wilcox, P. E.; Trueblood, K. N.; Dekker, A. 0. Anal. Chem. 1951,

23, 1740.

196

 

Chapter 9

Future Prognosis

The information and understanding presented here is only a starting point to a
new field of devising new nano -materials and -chemistry. We think that there are many
more opportunities available to be carried out in both in basic understanding and
applications of these materials. Below are tasks either in progress or we intend to carry
out in the future which will further enhance our ability to understand these layered

materials for both the basic understanding and their applications.

OH OH

   

Figure 9.1. The structure of MVE copolymers having two different chromophores R1
and R2 to get better understanding of rate of complexation of a metal (e. g. Zr“, Ca2+, Y3+

etc.) with different protonic acids using UV-Vis spectroscopy.

197

 

7;\\ 72>h

l.
1.

\‘;‘\\\_>’I\/“\ l ’
\/ F V\ I
coon

PO3H2

Figure 9.2. The structure of MVE copolymers having two different chromophores R1
and R2 to get better understanding of rate of complexation of a metal (e. g. Zr“, Ca“, Y“
etc.) with different protonic acids using XPS.

Since phosphonate, the carboxylate and sulfonate functionalities can complex
with zirconium, it would be interesting to see the rate of complexation of these acids with
zirconium on various flat substrates. This can easily be checked spectroscopically by
depositing layers from a mixture of MVE solutions having R1 and R2 in various
proportions (Figure 9.1). One of most widely used methods is UV-Visible spectroscopy.
In our group, we have expertise in the synthesis of specialty MVE monomers and
polymers and also in the use of absorbance spectroscopy. We intend to choose the
pendant chromophores R1 and R2 on the succinimide group such that their absorption
maxima are resolvable. By knowing the absobance and their extinction coefficients, we
can calculate the relative concentration of each chromophore and thus the MVE polymer
concentration. XPS can also be used to get the relative concentration of two species in

quantitative manner by using “labeled” R1 and R2 functionalities (Figure 9.2). By

198

“labeling” we mean that RI- and R2-bearing functional groups which are distinguishable.
For example, R1 bears fluorine while R2 contains iOdine, making two chromophores
distinguishable.

OH

I

/P_———O
O—(CH2)lgOH

 

Figure 9.3. The structure of MVE copolymer where the length of vinyl ether

comomoner is greater than the length of the azobenzene chromophore.

In Chapter 6, we observed a change of trans-to-cis conformer ratio as the number
of layers was increased. We intend to perform future following experiments to confirm
the argument that trans conformers isomerize to cis conformers due to steric constraints.
First we intend to increase the length of the vinyl ether monomer bearing a phosphoester
to be longer than azobenzene moiety (Figure 9.3). In this case, we expect to have a little

or no affect on the trans-to-cis conformation ratio. Secondly, by selecting the

199

 

azobenzene chromophores (Figure 9.4) having very high isomerization barrier so that the
trans-to-cis isomerization energy is more than the zirconium-phosphonate complexation
energy. This experiment has a direct impact on the formation of ZP complexation since
the azobenzene So isomerization barrier is expected to be greater than the ZP

complexation energy. Thus, using a combination of chromophores such as shown in

 

Figure 9.4. The structure of MVE polymer where the trans—to-cis isomerization is

expected to be higher than ZP complexation energy.

Figure 9.4 and an unsubstituted azobenzene and applying the soft-lithography technique
pioneered by Whitesides, we can devise “write-once” information storage devices.
We have demonstrated that it is possible to change the identity of each polymer

layer in the multilayer assembly, the next obvious step is to make the multilayer

200

assemblies where we can create a gradient. This gradient could be a surface energy,
porosity, chemical, optical (refractive indices) or electrical conductivity. Some work is
already is in progress.

Furthermore, it would be very interesting to carry experiments where the vinyl
ether having phosphonate group is copolymerized with two or more different maleimides
having pendant fiinction groups with different surface energies. We expect these types of
copolymers may give an amphiphilic films and coatings.

It is clear that there is much work needed to acquire a complete understanding of

the structural and dynamical properties of these systems.

201