.: 1%..

. :5. ..

 

 

1..

1.3.3.11:

a“

I.

i
2:

35‘

.. Em. an" ”r

x...-

I

at» .H a.”
w... .. in...

.3324 11b:

:59)...-

3. . i.
. . lib-‘5‘! Elan-Iv
sir"). $3313.12...»
:. 3' :34?! a.

5‘ 153‘?! J1-

‘I RH. 35.5.3. 3‘;

2.....1‘

{3.5:

3.1. it»!

‘ XS.
5,...

5...).

u u :22
a:

v.7. if:
I; .. Ike-A9...- ‘4...

.... u rd...
”313:?

.1 '3‘}.
~1. 1.4.3.
. 3 ~ .—

39.8..

(1- 8; If.
2 A v . 19...... m.m.ahaunot.1murz. dmfl”: J.

: 3.5.9.»... 0!!
3.3L.“
:3

:4 2-33
33....4...X~..a.>..i.¢.
.5 ... T2» ..

Sitar}: it.

‘3‘ 1|?
‘ci ’3

.I‘:

i
i. u:.l.tlll.~
Inufi$uil.lx.3

21.5 ‘5... 5.:

gig-9.4"».

3.11:... )3. 35.3 3
a.:u..~t‘.u.|>..ux .5

3.11.1533}:

.‘oitl’fx. :31

, : .~ (5.11.7
"radiilsn!
Iii-.55?

3‘.

“in"?

.0|

. Ia\$\\$

Iii). .. «iv...

 

 

 

 

THESE}

 

.750 I LIBRARY
Mtcnugan 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.

 

i | DATE DUE DATE DUE DATE DUE

 

NI?! 31 QEJZUQZZ

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1100 m.“

USING NONLINEAR OPTICAL PROPERTIES IN CONJUNCTION
WITH CHEMICAL SYNTHESIS TO UNDERSTAND DEFECTS IN
SELF -ASSEMBLED MOLECULAR FILMS

By

Wendy Christian Flory

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

USING NONLINEAR OPTICAL PROPERTIES IN CONJUNCTION WITH
CHEMICAL SYNTHESIS TO UNDERSTAND DEFECTS IN SELF-ASSEMBLED
MOLECULAR FILMS

By

Wendy Christian Flory

The potential utility of self-assembled monolayer and multilayer (SAM) thin films
in applications such as optical information storage, photorefractives, and solid state
microelectronics is generating a great deal of attention in the chemistry and materials
science communities. The interest in these materials arises because of the ease of growth,
the mild conditions for the formation of the self-assembled layers, and the structural and
thermal stability of these assemblies, once they are formed. Because of the Zr-
bisphosphonate interlayer linking chemistry, these films are self-terminating at each
layer, allowing for the identity of individual layers to be controlled by macroscopic
means. The zirconium-phosphate/phosphonate linkage is used because it is a thermally
stable inorganic crystalline structure. The bulk properties of these films are currently
being studied, but one issue that remains to be addressed is the quantification and
characterization of defects in the formed layers. A novel approach to measuring defects
on a molecular scale is to utilize the functional chemistry for orientational control in the
grth of these films. By constructing these films with control over the orientation of
each layer, a multilayer film can be assembled with either centrosymmetric or

noncentrosymmetric bulk ordering. Theoretically, the 1(2) nonlinear response of an

opposing bilayer structure should be zero because there is a center of inversion in the
macroscopic assembly, and any surface second harmonic generation (SHG) signal
observed from this film should be the direct result of defects.

Katz and coworkers developed a nonlinear chromophore for use in SAM
assemblies. This chromophore (1) is a rigid azo dye with a large molecular
hyperpolarizability giving rise to a large second harmonic signal from each monolayer.
Functionalizing only one terminus with a phosphonate group controls the orientation of
this chromophore. When this molecule is exposed to a zirconated surface, it adsorbs with
the phosphonate group directed toward the surface. The exposed hydroxyl terminus of
the chromophore is then phosphorylated. The second chromophore (2) is identical to the
first except the functional groups at each end are exchanged. By synthesizing these
chromophores so that they are functionalized selectively at each terminus, we can control
the direction of the dipole moment for each formed layer. By assembling one layer each
of chromophores 1 and 2, we can construct a bilayer with a center of inversion.
Quantifying the intensity of the SHG signal, which is present against a small background,

we can work toward gaining an understanding of defect properties in ZP SAMs.

In Memory of Florence Wilhemina Schaus Christian

-iv-

ACKNOWLEDGMENTS

I would first and foremost like to thank my advisor, Gary J. Blanchard, for the
help and support he has given me while I was a graduate student. He has offered me his
knowledge, patience, and understanding through the years and I will always be grateful
for that. I would not have been able to learn as much as I have without Gary as an
advisor. I would also like to thank the members of my committee; Prof. Marcos Dantus,
Prof. Greg Baker, and Prof. Kris Berglund. It has been a pleasure having each of you as
part of my committee.

I would also like to thank the guys from the machine and electronic shops, who
have been a part of my research trek. Russ, Dick, Sam, Glenn, and Tom, you have all
come through in quickly machining the parts that I’ve needed in the development of the
laser systems over the years. Because of you I’ve become very good at making banana
bread! Ron and Dave, thanks for the advice when I was building the circuitry for the
system. And thanks, Ron, for the many great jokes through email!

To my mother, whose constant insistence that I go to college was the real driving
force for me to not be afraid to follow my dreams, no matter what happens, thank you.
Thanks to my father who also taught me that, even if I failed, it was okay. Both of you
have taught me many important lessons in life. To my siblings and their spouses: Pam
and Bryan, John and Sue, and Amy and Lonnie, thank you for always supporting me,
sending me care packages when I’m down, and calling me just to say hello. These
things meant a lot to me.

I would not have been the person I am today without the love and support from

my friends, who’ve always been there for me. Thanks to Jill, Dawn, Leslie, Michelle, and
Andrea. You guys have been my confidants for a very long time. Dale, you’ve been a
big part of my life here at MSU, thanks. Tom and Richard, my pseudo-parents(?), thanks
for watching out for me, and hosting a lot of great parties! Bob, thanks for always having
a positive outlook on life. To the rest of my MSU friends, Karen, Joann, Lynda, Shannon
and Jeremy, Shawn, Lee and Carina, Kristi, and Lofy, thanks for the fun times! All of
my friends will always be an important part of my life. To the many softball teams I’ve
been apart of, I’ll always remember all the great times we’ve had at Art’s. Finally, to the

past, present, and future Blanchard group members, good luck!

'Vl-

TABLE OF CONTENTS

-vii-

Page
List of Tables .................................................................................................................. ix
List of Figures ................................................................................................................ x
Chapter 1. Introduction ................................................................................................. 1
1.1 Literature Cited ............................................................................................ 9
Chapter 2. Synthesis of a Nonlinear Optical Chromophore for use in Oriented
Zirconium Phosphate/Phosphonate Multilayer Assemblies ....................... 14
2.1 Introduction ............................................................................................ . 15
2.2 Experimental .......................................................................................... . 18
2.3 Results and Discussion ........................................................................... . 24
2.4 Conclusions ............................................................................................ . 28
2.5 Literature Cited ...................................................................................... 31
Chapter 3. Laser Systems for Surface Second Harmonic Generation Measurements... 33
3.1 Introduction ............................................................................................ . 34
3.2 Experimental .......................................................................................... . 35
3.3 Results and Discussion ........................................................................... . 36
3.4 Conclusions ............................................................................................ . 52
3.5 Literature Cited ...................................................................................... 53
Chapter 4. Structural Contributions to Second-Order Optical Nonlinearities in
Oriented Interfacial Multilayers ....................................................... 54
4.1 Introduction ............................................................................................. 55
4.2 Experimental .......................................................................................... . 57
4.3 Results and Discussion ........................................................................... . 5 8
4.4 Conclusions ............................................................................................ . 76
4.5 Literature Cited ...................................................................................... 78
Chapter 5. Understanding the Spatial Dependence of the Nonlinear Response Using
Self-Assembled Multilayers ........................................................... 83
5.1 Introduction ............................................................................................. 84
5.2 Experimental ........................................................................................... 86
5.3 Results and Discussion ............................................................................. 87
5.4 Conclusions ............................................................................................ . 99

5.5 Literature Cited ..................................................................................... . 100

Chapter 6. Future Directions ........................................................................................ 104
6.1 Literature Cited ........................................................................................... 107
Appendix A ............................................................................................................... 108

- viii -

Table 2.1

Table 5.1

LIST OF TABLES

1H NMR spectra of compounds in the synthesis of chromophore 2.

Experimental susceptibility values extracted from data.

-ix-

23

89

Figure 1.1

Figure 1.2

Figure 2.1
Figure 2.2

Figure 2.3

Figure 2.4

Figure 2.5

Figure 3.1

Figure 3.2

Figure 3.3

Figure 3.4

LIST OF FIGURES

Structures of chromophores l and 2 ............................................

Bilayer structure of 1 and 2. a- bilayer has perfect symmetry about the
metal linkage and therefore cancellation of the dipole moments of each
layer; b- bilayer has a vacancy defect site and will have an overall dipole
moment and SSHG response for the assembly ................................

Chromophore 2 .....................................................................
Synthetic route to Chromophore 2 ...............................................

UV—Vis absorption spectra of (a) solution phase of 1 (solid blue) and 2
(dashed red); (b) the assemblies of 5 bilayers of l and 2 on fused silica...

Semi-empirical calculations of the linear response for chromophores 1
(left) and 2 (right). Wide lines are singlet states, and calculated
oscillator strengths for Si (— SO transitions are indicated next to the
upper state. Narrower lines indicate triplet states .............................

(a) Dependence of absorption on number of layers for both
chromophores. See text for a discussion of these data. (b) Ellipsometric
thicknesses of each layer for 15 layer assemblies of chromophores 1 (O)
and 2 (I). The slopes of the best-fit line through these data are 26.3 i
0.5 A/layer for 1 and 25.5 i 0.4 Allayer for 2. The y-intercept is a result
of the primer layer ..................................................................

Q-switched ns NszAG and mode-locked ps Nd:YAG laser systems.
Reflection geometry was used and the fimdamental light was focused
onto the sample and them recollimated. A dichroic filter was used to
remove the fundamental before the 532 nm light was detected ..............

Plot of square law dependence of the SSHG signal vs. number of
chromophore layers for the Q-switched ns Nd:YAG laser system.
Chromophore 1 is chemisorbed on a Si(100) substrate ......................

Mode-locked ps Nd:YAG laser system (a) raw data demonstrating the
small signal. (KDP reference; SSHG signal from chromophore l on Si
substrate; V=0); (b) plot of square law dependence of the SSHG signal
vs. number of chromophore layers ..............................................

Pulse envelope of a Q-switched, mode-locked pulse. The mode-locked
pulses (80 MHz repetition rate) are characterized by a 100 ps pulse

4

15
18

26

27

29

37

39

41

Figure 3.5

Figure 3.6

Figure 3.7

Figure 4.1

Figure 4.2

Figure 4.3

Figure 4.4

Figure 4.5

Figure 4.6

Figure 4.7

width .................................................................................

Q-switched, mode-locked Nd:YAG laser based system for SSHG
experiments. Transmission mode i shown ......... Peak track-and-hold
circuit schematic used for computer interfacing. The details of this
circuit are discussed in the text ............................

SSHG signal for three bilayers of chromophore 2 chemisorbed on a
fused silica substrate ..............................................................

Polarization dependence of the SSHG signal of chromophore l adsorbed
onto Si(100). The inset cartoon depicts the coupling efficiency of the
polarization to adlayers with their induced dipole moments normal to the
surface.Peak track-and-hold circuit schematic used for computer
interfacing. The details of this circuit are discussed in the
text ..............................................................

Model system of thin films on both sides of a substrate. 1 and 5 are air;
2 and 4 are interfacial thin films; 3 is the substrate ............................

Surface SHG for a clean fused-silica substrate vs. angle of sample
rotation (magenta) (0° is the condition where the incident electric field is
propagates along the surface normal axis) plotted with the model (green)
for three components of the etalon effect .......................................

(a) Calculated SSHG envelope function for a series of chromophore tilt
angles, as indicated in the plot. (b) Dependence of the envelope
function on orientational distribution width. The distribution function is
assumed to be Gaussian and the chromophore average orientation is
along the surface normal axis ....................................................

(a) Plot of the [SHG Intensity]”2 vs. the number of chromophore layers.
(b) Angle-dependent SSHG signals from chromophores l (0) and 2 (0).
These data are dominated by substrate response at small angles and
chromophore response at high angles ...........................................

(a) Comparison of experimental SSHG data to the model presented in
Eqs. 4.1-4.5 for chromophore l. (b) Comparison of experimental data
with model presented in Eqs. 4.1-4.5 for chromophore 2 ....................

Surface SHG of (a) [SiOx-l-Z] and (b) [SiOx-2-1] bilayers showing the
cancellation of SHG signal with the adsorption of the second layer. The

orange traces are data on the bilayer structures and the black circles are
SSHG data for the bare SiOx substrates ........................................

(a) Structure of [SiOx-l-Z] bilayer. (b) Structure of [SiOx-2-l] bilayer.

-xi-

43

48

45

48

60

61

65

69

71

73

Figure 5.1

Figure 5.2

Figure 5.3

Figure 5.4

Figure 5.5

Figure 5.6

Figure 1A

Note the structural difference between the interlayer linkages for the two
assemblies (boxed) ................................................................

Schematic of the assemblies of chromophores 1 and 2 spaced apart with
one x(Z)-inactive layer of 1,12-dodecanediylbisphosphonate. The
shading of the chromophore layers represents the induced dipole
moment. The experimental susceptibility values, referenced to the
known value for quartz, are located adjacent to the appropriate layer for
each step in the assembly.

Surface SHG intensities for the build up of chromophores l and 2
spaced apart with one layer of Xm-inactive 1,12-
dodecanediylbisphosphonate on fused silica substrates (a) SiOx; SiOx—l;
SiOx—l-S; SiOx—l-S-Z; SiOx—l-S-Z-P (b) SiOx; SiOx—Z; SiOx—Z-S; SiOx—
2-S-l; SiOx—Z-S-l-P. The P is the addition of a phosphonic acid
capping layer ........................................................................

Schematic of the assemblies of chromophores 1 and 2 spaced apart with
two x(Z)-inactive layers of 1,12-dodecanediylbisphosphonate. The
shading of the chromophore layers represents the induced dipole
moment. The experimental susceptibility values, referenced to the
known value for quartz, are located adjacent to the appropriate layer for
each step in the assembly .........................................................

Surface SHG intensities for the build up of chromophores 1 and 2
spaced apart with two layers of gab-inactive 1,12-
dodecanediylbisphosphonate on fused silica substrates (a) SiOx; SiOx—l;
SiOx—l-S; SiOx—l-S-S; SiOx—l-S-S-Z (b) SiOx; SiOx—Z; SiOx—2-S;
SiOx—Z-S-S; SiOx—Z-S-S-l .......................................................

Schematic of the assemblies of chromophores 1 and 2 spaced away
from the fused silica substrate with two x(2)-inactive layers of 1,12-
dodecanediylbisphosphonate. The shading of the chromophore layers
represents the induced dipole moment. The experimental susceptibility
values, referenced to the known value for quartz, are located adjacent to
the appropriate layer for each step in the assembly ...........................

Surface SHG intensities for the build up of chromophores l and 2
spaced away from the fused silica substrate with two layers of x(2)—
tvaxttme 1,12-dodecanediylbisphosphonate (a) SiOx; SiOx—S; SiOx—S-
S; SiOx—S-S-l; SiOx—S-S-1-2 (b) SiOx; SiOx—S; SiOx-S-S; SiOx—S-S-
2; SiOx—S-S-2-l ...................................................................

Front panel of the Labview® program written for the SSHG laser

75

91

92

93

94

95

96

system ............................................................................... 1 08

-xii-

Figure 2A
Figure 3A
Figure 4A
Figure 5A
Figure 6A
Figure 7A

Figure 8A

Icon based schematic of the Labview® Program ............................... 109

AIConfig SubVI details ........................................................... 110
AIStart SubVI details .............................................................. 111
AIRead SubVI details ............................................................. 112
AIClear SubVI details ............................................................. 113
Error SubVI details ................................................................ 114
Write to Spreadsheet SubVI details .............................................. 115

- xiii -

Chapter 1

INTRODUCTION

In recent years there has been a surge of research interest in processes that will
chemically modify surfaces, which can then be used in applications such as solid-state
electronics, electro-optic devices, and optical information storage."2 The thin films that
are built off of surfaces, either through physisorption or chemisorption, can be used for
these applications only if the molecular construction can be controlled. Modifications of
these surfaces have lead to the development of many clever synthetic routes aimed at
inducing molecular orientation in these layered materials.

The first monolayer assemblies studied were Langmuir-Blodgett(LB) thin films3'5
The LB technique involves the mechanical transfer of a compressed molecular monolayer
at the air-liquid interface to a substrate. The monolayer is created through the use of
amphiphilic molecules. These molecules form highly ordered monolayers when
compressed laterally at the liquid interface and can be physisorbed onto a surface to form
multilayer assemblies through sequential dipping of a substrate.6 The drawbacks of this
family of layered materials come from the relatively complex procedures for layer
formation, their sensitivity to impurities, and the weak intermolecular interactions.
Because their order derives from relatively weak van der Walls forces, they are not
thermally stable, nor can persistent order be maintained indefinitely in multilayer
assemblies.7

The most recently developed route to building thin films with molecular

orientation is using self-assembled mono- and multilayers (SAM).3'l7 SAMs are thin

-1-

films built by chemically linking the functionalization groups of a substrate with a
molecule that is in solution through covalent bonding. Noncovalent intermolecular forces
enforce in-plane ordering of the layer. An excellent review article on SAMs discusses
the various types of assemblies used, including alkanethiols on gold and organosilane and
metal phosphonate binding chemistries.18

Thiols on metals such as Au and Ag were one of the first SAM systems to be
explored as a means to control surface morphology.”"20 Many studies of these systems
have been reported.2"22 While these films are highly ordered and are structurally and
thermally stable, the thin film characteristics obtained depend sensitively on the metal
substrate used in their formation. Thiol/gold monolayers suffer from long-term chemical

stability limitationsZ3’24

and, until recently, the inability to form chemically-bound
multilayer assemblies.25

Silane multilayers from silicon and silica substrates have also been
investigated.26'29 Alkylaminosilanes, alkylchlorosilanes, and alkylalkoxysilanes require a
hydroxylated surface such as silica or alumina for the monolayer to form.18 A high
quality surface of this type is difficult to assemble because of the careful control needed
to be maintained over the amount of water present during synthesis. Too little water will
lead to incomplete layer formation, while too much water leads to polymerization of the
solution.

SAMs made using metal-phosphonate (MP) chemistry have been studied for the
last 12 years as a novel route to surface modified assemblies.8‘3°'34 MP SAMs have

robust physical properties with persistent structural anisotropy. These are very attractive

features for using MP thin films in materials-based applications. Zirconium phosphonate

SAMs were first demonstrated by the Mallouk group, where layered structures from one)—
alkanebisphosphonates were reported.35 A typical deposition process procedure is as
follows: a substrate is treated to produce a Zr4+-rich surface and, when exposed to
phosphonate moieties, the groups react forming an inorganic structure. The substrate is
then dipped into a zirconium solution, allowing the exposed phosphonate to react,
capping the molecular sheet and forming another zirconium rich surface. Multiple layer
growth in a two-step fashion formed films that were quickly formed and robust over
many layers. This work spurred a great deal of research effort on MP SAMs by several
groups.”48 The films initially reported by the Mallouk group were centrosymmetric
about the metal center because the direction of the organic moiety within the layer
formed could not be controlled. Either of the phosphorylated ends could react with the
zirconium surface.

Howard Katz et al. reported a scheme to incorporate molecular-level orientation
within each layer by utilizing a three-step synthetic approach to the formation of the
layers instead of the two step method reported initially.8 The same ZP inorganic structure
would still be used as the main construction, but instead of having both ends of the
organic molecule phosphonated, Katz et al. used molecules that were phosphonated at one
terminus, thereby controlling the orientation of the formed layer. The exposed terminus
of the adsorbed molecule contained a hydroxyl moiety that could be reacted with POCl3
to form a phosphate functionality. When reacted with a Zr4+ solution, the surface was
ready to be reacted through the three step cycle again. Initially, this directional order was
used to build films for use as nonlinear optical thin films. These assemblies exhibit

nonlinear responses comparable to inorganic materials, such as LiNbO3.8 We have

chosen the three step synthetic strategy to explore a novel method to characterize the
defect properties of SAMs.

With these various strategies of controlling thin films by chemical means, there is
a growing interest in characterizing the defects within these assemblies.”S3 The
techniques most commonly used are electrochemistry, atomic force microscopy (AF M),

and scanning electron microscopy (SEM). These
PO3H2 OH

methods probe a range of microscopic levels.54'60

While the results for these measurements indicate

N N

i: j E j defect sites to within the resolution of the

N N . . . . . .
instrument, little information is given about the
(C: © physical or chemical properties of the defect
NQN “LN structures. The object of this dissertation is to
characterize and quantify defects in SAMs using a
spectroscopic technique. We have chosen surface
2 2 second harmonic generation (SSHG) to achieve the
OH P03H2 characterization based on the symmetry
1 2 requirements for second order nonlinear processes.
Figure 1-1 SWCWCS Of A centrosymmetric material in the presence of a

chromophores l and 2.
strong electric field will not exhibit a second order

nonlinear response. The idea is, using the
appropriate chemistry to construct SAMs with a center of inversion within the assembly
and therefore achieve control over the nonlinear response, as long as the response is

dominated by electric dipole properties.

We have chosen to use Zr-phosphate/phosphonate (ZP) chemistry in the assembly
of our SAMs because of the structural versatility available with this system, combined
with the high degree of control that can be exercised in multilayer assemblies.”61 We
use a nonlinear optical chromophore developed by Katz and coworkers, (4-(4-(4-(4-((2-
hydroxyethyl)sulfonyl)-phenyl)azo)phenyl)piperazinyl)phenyl)phosphonic acid, 1, shown

in Fig. 1.1. The nonlinear

i it i i it y], ,L J, .l, l chromophore 1 isarigid azo dye
tittitilit a hwum.mdm.

\\\\\\\\\\\\\\

ll llllll
Tiltittti

\\\\\\\\\\\\T\\

Figure 1.2 Bilayer structure of l and 2. a- bilayer has
perfect symmetry about the metal linkage and (4-hydroxy-phenyl) piperazinyl)
therefore cancellation of the dipole moments of each
layer; b- bilayer has a vacancy defect Site and Will phenyl) azo-phenyl) sulfonyl)
have an overall dipole moment and SSHG response

for the assembly. ethylphosphonic acid, 2.64 This

 

nonlinear susceptibility, x”). The

second harmonic signal from

W b layers of this chromophore is

significant.62’63 The complement

 

to chromophore l is (2-(4-(4-(4-

molecule is identical to 1 except
that the terminal functional groups have been exchanged. By assembling one layer each
of chromophores l and 2, a bilayer film can be constructed that will have a center of
inversion around the interlayer linking metal ion. If the film is defect-free (Figure 1.2a),
the x2 response of this structure should be the same as the background because there is a
center of inversion about the metal ion. If there are defects in the bilayer assembly

(Figure 1.2b) there will not be complete cancellation of the dipole moments over two

layers and the x2 signal will nonzero. Note that the signal is only present with defects,
therefore we are trying to detect a nonzero signal against a dark background.

The first step of this dissertation project was to synthesize 2. This chromophore is
used in ZP thin films as a xZ-active molecule and its structure ensures alignment relative
to the surface. The steps used to synthesize this molecule, the resulting structural data,
and selected thin film properties, as compared to chromophore 1, are presented in
Chapter 2.

Chapter 3 details the next step, which was the design, assembly, and
troubleshooting of a surface second harmonic generation instrument for the nonlinear
measurements. We use two different thYAG lasers the source for fundamental (1064
nm) light. The lasers used operate in two different time regimes. The initial system used
pulses in the ns time regime, the bulk of this work, however, involves the use of 3 ps
Nd:YAG laser. We used both transmission and reflection configurations in the SSHG
system design, with reflection from silicon wafers being used in the us time regime and
transmission being used in the ps time regime. The reflection experiments using the ps
Nd:YAG laser with silicon wafers failed because of surface heating and ablation. For this
reason, we switched to transmission measurements using fused silica substrates.

Chapter 4 discusses our results on selected multilayer structures of l and 2 built
on silica substrates. The x‘z) nonlinear response was characterized for bare fused-silica
substrates, each chromophore chemisorbed to the substrate, and bilayers of opposing
chromophore bilayers. While the results demonstrate that there was complete cancellation
of the electric dipole moments for the bilayer assemblies, the )6” responses of the two

opposing bilayer systems are different and reveal the extent to which the “dipole

approximation”65

is valid in these layered assemblies. From the data, we elucidate the
role that quadrupolar terms play in the )6” responses of thin interfacial films. This
chapter discusses the complex and structural issues associated with the construction of
layered interfaces, and presents the theory to support the experimental data.

In Chapter 5, the dependence of the nonlinear response as a function of multilayer
structure is measured and characterized. We have shown that by spacing the individual
layers of chromophores 1 and 2 apart using l,12-dodecanediylbis(phosphonic acid), the
SSHG signal is no longer decreased due to the induced center of inversion, but increases
significantly. We have also inserted a spacer layer between the substrate and
chromophore bilayers. These results are very different to those of bilayer assemblies
adjacent to the substrate, presented in Chapter 4. The significance of these findings is
that the molecular details of interface structure can have a profound effect on the
measured response. We understand these data in the context of spatial variation in the
dielectric response of the interface. It is this spatial variation that gives rise to
nonlinearities in the induced polarization. The spatial profile of this variation influences
the propagation of the fundamental and second harmonic electric fields. Because the
chromophores 1 and 2 both have band tails at 532 nm, both the real and imaginary
components of the dielectric gradient contribute to the observed signal.

Finally, chapter 6 concludes the findings of this body of data, and future
directions of this project are discussed. There are a variety of directions that this project
can now take to further investigate ZP SAMs or other interfacial chemistries.

Modification of the SSHG system to image the surfaces is another possible route that this

project can take. SSHG imaging gives surface morphology details on the um scale.66 By

using both SSHG intensity and imaging measurements to characterize surfaces, a great
wealth of information can be gained.

A clear direction that this work must take is elucidation of the details involved in
the theory of multilayer structures. Chapter 5 discusses possible explanations to the
results of multilayer interfaces, but further work in this area is needed before concrete
conclusions can be drawn. Once the details of the multilayer structures are understood,
manipulations of thin films can then lead to selective control of the dielectric properties
for use in materials-based applications.

Another possible avenue of exploration is the study of aggregation within a
chromophore layer using SSHG as the measuring tool. It has been shown that varying
the concentration of a chromophore within a single layer leads to island formation.67’68

By using the nonlinear molecules presented on this thesis in conjunction with SSHG,

noninvasive characterization to the degree of this inhomogenaity can be assessed.

1.1 Literature Cited

1. Xia, Y. N.; Whitesides, G. M.; Annu. Rev. Mater. Sci., 1998, 28, 153.
2. Collier, C. P.; Wong, E. W.; Belohradsky, M.; Raymo, F. M.; Stoddart, J. F.; Kuekes,

P. J.; Williams, R. 8.; Heath, J. R.; Science, 1999, 285, 391.

DJ

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

A

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

5. Florsheimer, M.; Kupfer, M.; Bosshard, C.; Looser, H.; Gunter, P.; Adv. Materials,
1992, 4, 795.

6. Gaines, G. L. Insoluble Monolayers at Liquid-Gas Interfaces; Interscience: New York,
1966.

7. Ulman, A. An Introduction to Ultrathin Organic Films: From Langmuir-Blodgett to
Self-Assembly, 1St ed.; Academy Press: San Diego, 1991.

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

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

\O

. Marowsky, G.; Lupke, G.; Steinoff, R.; Chi, L. F.; Mobius, D.; Phys. Rev. B, 1990, 41,
4480.
10. Palto, S. P.; Blinov, L. M.; Yudin, S. G.; Grewer, G.; Schonholl, M.; Losche, M.;
Chem. Phys. Lett., 1993, 202, 308.
11. Kepley, L. J .; Crooks, R. M.; Ricco, A.; Anal. Chem. 1992, 64, 3191.
12. 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.

13.

14.

15.

l6.

l7.

l8.

19.

20.

21.

22.

23.

24.

25.

26.

27.

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

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

Kumar, A.; Biebuyck, H. A.; Abbott, N. L.; Whitesides, G. M.; J. Am. Chem. Soc.
1992, 114, 9188.

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

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

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

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

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

Nuzzo, R. G.; Zegarski, B. R.; Dubois, L.H.; J. Am. Chem. Soc., 1987, 109, 733.
Xia, Y. N.; Whitesides, G. M.; Annu. Rev. Mater. Sci, 1998

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.

Kohli, P.; Taylor, K. K.; Harris, J. J.; Blanchard, G. J.; J. Am. Chem. Soc, 1998, 120,
11962.

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

Gun, J .; Iscovici, R.; Sagiv, J .; J. Colloid Interface Sci 1984, 101, 201.

-10-

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

41.

42.

43.

44.

Maoz, R.; Sagiv, J .; J. Colloid Interface Sci. 1984, 100, 465.

Maoz, R.; Sagiv, J .; Thin Solid Films 1985, 132, 135.

Cao, G.; Rabenberg, L. K.; Nunn, C. M.; Mallouk, T. M. Chem. Mater. 1991, 3, 149.
Hong, H.-G.; Sackett, D. D.; Mallouk, T. E. Chem. Mater. 1991, 3, 521.

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

Byrd, H.; Snover, J. L.; Thompson, M. E.; Langmuir, 1995, 11, 4449.

Cao, 0.; Hong, H.G.; Mallouk, T. E.; Acc. Chem. Res, 1992, 25, 420.

Lee, H.; Kepley, L.J.; Hong, H.; Mallouk, T.E.; J. Am. Chem. Soc, 1988, 110, 618.
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. D.; Putvinski, T. M.; Hutton, R. S.
Chem. Mater. 1991, 3, 699.

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

-11-

45.

46.

47.

48.

49.

50.

51.

52.

53.

54.

55.

56.

57.

58.

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.

O’Brien, J. T.; Zeppenfeld, A. G; Richmond, G. L.; Page, C. J .; Langmuir, 1994, 10,
4657.

Ansell, M. A.; Zeppenfeld, A. C.; Yoshimoto, K.; Cogan, E. B.; Page, C. J .; Chem.
Mater., 1996, 8, 591.

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

Langmuir, 1.; Trans. Faraday Soc.; 1920, 15, 62.

Kleinfeld, Elaine R.; Ferguson, Gregory S.; Chem Mater., 1996, 8, 1575.

Chidsey, C. E. D., Loiacono, D. N.; Langmuir, 1990, 6, 682.

Seshadri, K.; Froyd, K.; Parikh, A. N. Allara, D. L.; Lercel, M. J .; Craighead, H. G.;
J. Phys. Chem, 1996, 100, 15900.

Patrick, David L.; Cee, Victor, J .; Purcell, Timothy, J .; Beebe, Thomas B. Jr.;
Langmuir, 1996, 12, 1830.

Sabatini, E.; Rubinstein, 1.; J. Phys. Chem, 1987, 91, 6663.

Sabatini, E.; Rubinstein, I.; Moaz, R.; Sagiv, J .; J. Electroanal. Chem, 1987, 219,
365.

F inklea, H. 0.; Snider, D. A.; Fedyk, J .; Sabatini, E.; Gafni, Y.; Rubinstein, 1.;
Langmuir, 1993, 9, 3660.

Finklea, H. O.; Snider, D. A.; Fedyk, J .; Langmuir, 1990, 6, 371.

Sabatini, E.; Cohen—Boulakia, J .; Bruening, M. L.; Rubinstein, 1.; Langmuir, 1993, 9,

-12-

59.

60.

61.

62.

63.

64.

65.

66.

67.

68.

2974.

Janek, R. P.; F awcett, W. R.; Ulman, A.; Langmuir, 1998, I4, 3011.

Zhao, Xiao-Mei; Wilbur, James L.; Whitesides, George M.; Langmuir, 1996, 12,
3257.

Lee, H.; Kepley, L.J.; Hong, H.; Sohail, A.; Mallouk, T.E.; J. Phys. Chem, 1988, 92,
2297.

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

Flory, W. C.; Blanchard, G. J .; J. Am. Chem. Soc, in review.

Flory, W. C.; Mehrens, S. M.; Blanchard, G. J .; J. Am. Chem. Soc, in review.
Guyot-Sionnest, P.; Sheri, Y. R.; Phys. Rev. B., 1987, 35, 4420.

Florsheimer, M.; Kupfer, M.; Bossard, C.; Looser, H.; Gunter, P.; Adv. Mater., 1992,
4, 795. V

Horne, 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.

-13-

Chapter 2

SYNTHESIS OF A NONLINEAR OPTICAL CHROMOPHORE FOR USE IN
ORIENTED ZIRCONIUM PHOSPHATE/PHOSPHONATE MULTILAYER
ASSEMBLIES

Summary

We report the synthesis of the nonlinear optical chromophore, (2-(4-(4-(4-(4-
hydroxyphenyl)piperazinyl)phenyl)azophenyl)sulfonyl)ethylphosphonic acid, to be used
in self-assembled multilayer (SAM) structures. This chromophore has been designed as
the structural complement of a Xm-active chromophore first reported by Katz, Wilson,
and Scheller [J. Am. Chem. Soc, 1994, 116, 6636.]. The reason for synthesizing this
structural complement is to explore the utility of structural “cancellation” of a surface are)

optical response. We report the synthesis, structural data, and preliminary surface

characterization data for the chromophore and its complement.

-14-

2.1 Introduction

There has been significant recent interest in creating nonlinear optical (NLO)
materials by incorporating organic chromophores into interfacial or polymeric support
structures. The resulting NLO materials have been reported to have x0) responses on the

same order as those characteristic of crystalline inorganic

OH
materials such as LINbO3.l The potential for self-assembling
0 multilayer (SAM) structures to be used in electro-optic device
N applications or in solid state electronics has provided the impetus
E j for the development of many clever synthetic routes aimed at
N
inducing molecular orientation in layered materials. Such systems
0 include Langmuir-Blodgett (LB) films,2 polymer films,3 and self-
N assembled mono- and multilayers.“5 Of these, SAMs appear to
’ N
be the best choice for combining robust physical properties with
O persistent structural anisotropy. Using the appropriate chemistry,
SAMs can be constructed one layer at a time, allowing control
O=S=O
over the identity and orientation of the NLO chromophore within
a given layer. An excellent review article on SAMs discusses the
PO3H2
Figure 2 1 various types of assemblies used, including alkanethiols on gold
Chromophore 2.

and organosilane and metal phosphonate binding chemistries.6

We have chosen to use Zr- phosphate/phosphonate (ZP) chemistry in the
assembly of our SAMs7’8 because of the structural versatility available with this system
combined with the high degree of control that can be exercised in their assembly. Zr-

phosphonate SAMs were first demonstrated by the Mallouk group, where layered

-15-

structures formed from or,o)-alkanebisphosphonates were reported.7 The films reported
initially by the Mallouk group were centrosymmetric about the layer center and were thus
not viable as xl2)-active interface materials. Katz et al. reported a scheme to incorporate
molecular-level orientation within each layer by utilizing a three-step synthetic approach
to the formation of the layers instead of the two step method reported initially.7 It is this
basic synthetic strategy that allows us to explore a novel way to characterize the defect
properties of these SAMs.

Our motivation for the synthesis we report here is to use the arm-active
chromophores 1 and 2 (Figure 1.1) in the characterization of ZP SAMs. A significant
question that remains to be answered for these interfaces is the quantification and
characterization of vacancy and orientational defects within the layers. Our approach to
answering this question is to construct a bilayer with two xm-active chromophores that
are identical in structure but opposed in orientation. The )6” signal of a perfectly
opposing bilayer structure should vanish because there is a center of inversion in the
macroscopic assembly and any surface second harmonic generation (SSHG) signal that is
observed will be the result of defects. The first step in this ongoing project is to
synthesize and characterize the nonlinear chromophores, and that is the focus of this
chapter.

Katz and coworkers synthesized the chromophore (4-(4-(4-(4-((2-
hydroxyethyl)sulfonyl)-phenyl)azo)phenyl)piperazinyl)phenyl)phosphonic acid, 1.9 The
nonlinear chromophore 1 is a rigid azo dye with a large second order hyperpolarizabilty,
B. The second harmonic signal from a monolayer of this chromophore is usefully large.

The orientational complement to chromophore l is (2-(4-(4-(4-(4-

-16-

hydroxyphenyl)piperazinyl)phenyl)-azophenyl)sulfonyl)ethylphosphonic acid, 2, shown
in Figure 2.1. This molecule is identical to 1 except that the terminal functional groups
have been exchanged. By assembling a bilayer with one layer each of chromophores 1
and 2, we can construct a nominally centrosymmetric bilayer with opposing dipolar
orientation of the layers. We report on the details of the )6” response for this system in
chapter 4.

The focus of this chapter is on the synthesis of nonlinear chromophore 2, the
characterization of its linear optical properties and the ellipsometric determination of its
layer thickness in a homogeneous multilayer structure. The synthetic route to
chromophore 2 is shown in Figure 2.2. The optical and layer properties are compared to
the response of 1. The pair of complementary chromophores 1 and 2 were chosen
because of their rigid structure, a characteristic that will minimize SSHG signal from
intralayer sterically-mediated defects. Isomerization about the azo bond could, in
principle, create cis conformers, but using 1064 nm light we do not excite these layers on
resonance, so isomerization about the azo functionality is not likely to be a significant

problem.

2.2 Experimental

Chemicals. All chemicals used were obtained in the highest purity grade
available. (4-Acetlyaminobenzene)sulfonyl chloride, 2-chloroethanol, 2,3,4-collidine,
POCl3, ZrOClz-8H20, triisopropyl phosphite, p-anisidine, bromotrimethylsilane, and
CDC13 were obtained from Aldrich Chemical Co. p-Toluene sulfonyl chloride, sodium

sulfite, and sodium bisulfite were obtained from Spectrum Chemicals. N,N-Bis-2-

-17-

i... 1
es <3
.12

NH
502
H3 >
O
\
SO

iNH fiNH
3020 $0211

3

2

4
CH3
5
802 S02 80;
S ..
PO3[iPr]2 OsllPl’lz l OCH:
. 7 . (<5
N N
U m
N
II J/CI OCH; N
~ . © __. N] __. (C?
E N
(5 N... N \ ~
9
€02 $02
PO3[1PI]2 P03H2
10 2

Figure 2.2 Synthetic route to Chromophore 2.

-13-

chloroethanol was obtained from Pfaltz & Bauer, d6—DMSO was purchased from Isotec
Inc., and 3-aminopropyltrimethoxysilane was purchased from United Chemical
Technologies, Inc. All water used was distilled in-house. All chemicals were used as
received except for triisopropyl phosphite, which was purified by drying over sodium
under an inert atmosphere, followed by vacuum distillation.lo

Measurements. lH NMR spectra of all compounds were taken with a Varian
Gemini-300 MHz NMR spectrometer. Optical absorption measurements were made
using a Hitachi U-4001 UV-visible absorption spectrometer. The thickness of
chromophore layers bound to Si substrates was measured using an optical null
ellipsometer (Rudolph AutoEL-II) operating at A = 632.8 nm. The software used to
calculate layer thickness was supplied by Rudolph. The complex refractive index of the
individual layers was taken to be 11 = 1.61 + Oi'll

Calculations. Semi-empirical and molecular mechanics calculations were
performed on chromophores l and 2 using Hyperchem® v. 4.5 on a PC. Energy level
calculations were performed on ground—state optimized structures using the PM3
parameterization.

Synthesis. The synthetic route is shown in Figure 2.2. For the compounds listed
in the following section, the 1H NMR spectral properties are summarized in Table 2.1.
The syntheses described below are typical results for individual experiments. The
absolute amounts of compounds used in each section are not limited by that reported
here.

(4-acetylaminophenyl)sulfinic acid12(3) 50.1 g Na2803 (anh.) (400 mmol) was

dissolved in 250 mL H20. (4-Acetylaminophenyl)sulfonyl chloride (47.7 g, 204 mmol)

-19-

was added slowly to the solution. The pH of the solution was maintained above 9 using
50% NaOH. After 2 hours, 60% H2804 was added dropwise until a white precipitate
formed. The white precipitate was collected by vacuum filtration. The yield was 48.0 g
(97%). Pure by NMR, FTIR, and MS.

(4-acetylaminophenyl)-B-hydroxyethyl sulfonel3 (4) 56.5 g of (4-
acetylaminophenyl)sulfinic acid (284 mmol) and 21.9 g of NaOH (548 mmol) were
added to 230 mL H20. Once dissolved, 45.0 mL of 2-chloroethanol (671 mmol) was
added and the solution was refluxed for 2 hours. The solution was cooled gradually then
placed in an ice bath and the resulting white crystals were collected by vacuum filtration.
The filtrate was made slightly alkaline and was refluxed for an additional 2 hours with
10% NaOH added at intervals. One-third of the solvent was then removed by rotary
evaporation, and the remaining solution was cooled in an ice bath. The crystals were
collected by vacuum filtration and added to the first batch. A total of 62.8 g of product
was obtained (91%). Pure by NMR, MS, and F TIR.

4-acetylaminophenyI-B-(p-toluenesuIfonyloxy)ethyl sulfonel3 (5). p-Toluene
sulfonylchloride (2.03 g, 10.6 mmol) was added to 2.25 g of (4-acetylaminophenyl)-B-
hydroxyethylsulfone (9.25 mmol) in 9.5 mL of pyridine at 0°C. The solution was stirred
for 2 hours at 0°C then poured into 30 mL of ice water. The precipitate was collected by
vacuum filtration and then recrystallized from ethanol. A total of 3.08 g (83%) of white
crystalline product was recovered. Pure by NMR.

4-acetylaminophenyl-B-iodoethyl sulfonel3 (6). 2.34 g NaI (15.4 mmol) was
added to 3.08 g 4-acetylaminophenyl-B-(p-toluenesulfonyloxy)ethyl sulfone (7.72 mmol)

in 23.4 mL of acetone. The solution was refluxed for 15 hours then poured into a

-20-

solution of 3.03 g of sodium bisulfite in 93 mL of H20. The crystalline product was
collected by vacuum filtration. Total product obtained was 2.26 g (82%). Pure by NMR
and MS.

(4-acetylaminophenyl)sulfonylethyldiisopropyl phosphonate (7). 27.0 mL of
purified triisopropyl phosphite (109 mmol) was added to 25.7g of 4-acetylaminophenyl-
B-iodoethyl sulfone (73.1 mmol) under Ar. The solution was refluxed for 5 hours at
145°C. The solution was cooled and excess triisopropyl phosphite and isopropyl iodide
was removed by vacuum distillation. The remaining crystals were dissolved in ethanol
and recrystallized. The product was collected by vacuum filtration and dried, yielding
22.6 g (79%) of white product. Pure by NMR and MS.

2-(4-aminophenyl)-sulfonylethyldiisopropyl phosphonate (8). 7.90 g of (4-
acetylaminophenyl)sulfonylethyldiisopropylphosphonate (20.2 mmol) was refluxed at
75°C in 75 mL of 0.5M HCl for 48 hours. The solution was cooled and the precipitate
collected by vacuum filtration. The precipitate was recrystallized from ethanol yielding
3.57 g (51%). Pure by NMR.

4-(4-methoxyphenyl)phenyl piperazinel4 (9). 7.27 g of N,N-bis-2-
chloroethylaniline (33.5 mmol) was combined with 8.77 g of p-anisidine (71.1 mmol) in
25 mL of waterzacetone (1:1), and refluxed for 2.5 hours. The solution was cooled in an
ice bath and the crystals were collected by vacuum filtration. The product was
recrystallized from ethanol, yielding 3.09 g of lustrous white platelets (35%). Pure by
NMR and MS.

(2-(4-(4-(4-(4-methoxyphenyl)piperazinyl)phenyl)azophenyl)sulfonyl)

ethyldiisopropyl phosphonate (10). The compound (10) was synthesized by formation

-21-

of an azo bond between (8) and (9). These functionalities were reacted by first forming
the diazonium salt. This was done by dissolving 0.260 g of (8) (0.745 mmol) in 57 mL
of 2M HCl and then slowly adding a solution of 150 mg of NaN02 in 13 mL of water at
0°C. After 15 minutes, the diazonium salt was added to 0.199 g of (9) (0.742 mmol)
dissolved in 30 mL of propanoic acidzH20 (1:5) at 0°C, producing an orange solution.
Sodium acetate was added to increase the pH to ~ 4. The solution was stirred for 24
hours at 0°C, then slowly brought to room temperature. The liquid was decanted and the
remaining precipitate was dissolved in 50 mL of CH2Cl2. This solvent layer was washed
3 times with 25 mL of 0.5 M HCl. After the third washing, 25 mL water was added and
the CH2Cl2 layer was boiled off, leaving a red-orange precipitate. This solid was washed
several times with 5:1 waterzisopropanol to remove starting material. This solid was then
purified by chromatography on 100 g of silica gel, eluting with acetoneztoluene (1:1).
The yield is 72.2 mg (15%). Pure by NMR.

(2-(4-(4-(4-(4-hydroxyphenyl)piperazinyl)phenyl)azophenyl)sulfonyl) ethyl
phosphonic acid (2). 67.2 mg of (10) (0.011 mmol) was placed in a flask under Ar. To
this, 8.0 mL of dried CH2C12 and 130 uL of dried pyridine were added. 620 uL
Bromotrimethylsilane was added and the solution was allowed to stir at ambient
temperature. Afier 24 hours, 10 mL of H20 was added to hydrolyze the silanes. The
CH2C12 layer was washed with 50 mL water and 50 mL 0.5M HCl. The CH2Cl2
evaporated and 53.0 mg of red-orange solid was collected (93 %). Pure by NMR and
UV-Vis.

Surface Preparation. Si(100) wafers (Multi Crystal Optics, Inc.) and silica

substrates were used. Si substrates were cleaned in piranha solution (3:1 H2S04zH202 —

-22-

Table 2.1 1H NMR spectra of compounds in the synthesis of chromophore 2.

 

Molecule (solvent) Resonances

 

 

 

 

 

 

 

3 (dg—DMSO) 2.02(s) ; 7.48(s) ; 7.52(d) ; 7.69(d) ; 10.16(s)
4 (d6—DMSO) 2.08(s) ; 3.35(t) ; 3.62(m) ; 4.85(t) ; 7.79(s) ;10.38(s)
5 (dd—DMSO) 2.08(s) ; 3.37(t);3.61(t) ; 7.79(s) ; 10.38(s)
6 (dd-DMSO) 3.15(t) ; 3.78(t) ; 7.8s(s) ;10.42(s)
“7 (d6—DMSO) 1.22(t); 1.88(m) ; 2.18(s) ; 3.22(m) ; 4.54(m) ; 7.84(s) ; 10.4l(s)
8 (dg—DMSO) l.18(t) ; 1.84(m) ; 3.09(m) ; 4.52(m) ; 6.22(s) ; 6.64(d) ;
7.48(d)
9 (dg—DMSO) 3.12(t) ; 3.24(t) ; 3.68(s) ; 6.82(m) ; 6.96(m) ; 7.22(t)

 

10 (d6—DMSO) 1.19(t) ; 1.95(m) ; 3.40(m) ; 3.79(s) ; 3.81(t) ; 3.87(d) ; 4.53(m)
; 6.97(d) ; 7.17(d) ; 7.25(d ofd) ; 7.42(d of d) ; 7.85(d) ;

7.98(d) ; 8.07(d)

 

2 (d6-DMSO) 1.77(m) ; 3.88(m) ; 6.97(d) ; 7.17(d) ; 7.25(d of d) ; 7.42(d of

d) ; 7.85(d) ; 7.98(d) ; 8.07(d)

 

 

 

 

caution- strong oxidizer!) for 10 minutes, rinsed with distilled water, placed in 2 M HCl
for 10 minutes, and rinsed with water. The substrates were dried in a dry N2 stream.
Under an Ar atmosphere, 20 mL anhydrous octane was added to a vessel containing the
substrate. The octane was heated to reflux and 3-aminopropyltrimethoxysilane was
added to make a 1% v/v solution. After 10 minutes of reaction the substrate was
removed and rinsed with warm hexane, then water. A N2 stream was used to dry the

surface. Silica surfaces were cleaned in piranha solution, dried, but were not primed.

-23-

Both surfaces were then phosphorylated. This was done under Ar, using a solution of 20
mM 2,3,4-collidine and 20 mM P0Cl3 in anhydrous acetonitrile at ambient temperature.
After 10 minutes, the substrates were removed, rinsed with acetonitrile and water and
dried with N2. The surfaces were then zirconated by immersing in aqueous 5 mM Zr0C12
solution for 10 minutes. The surfaces were rinsed with water and dried.

Chromophore Deposition. The chromophores were deposited on the substrates
from a saturated solution of each dissolved in 1:4 DMFzEtOH. For Si wafers, the
solution was kept acidic by adding dropwise 4 M HCl in dioxane. The temperature of
each solution was held just below boiling while the surfaces were immersed for 10
minutes followed by rinsing with warm ethanol and drying with N2. The phosphorylation
and zirconation of the surfaces is repeated as described previously. To monitor layer
adsorption, ellipsometric thickness was used for Si wafers and UV-Vis absorption was

used for silica substrates.

2.3 Results and Discussion

We have synthesized and performed initial characterization of (2-(4-(4-(4-(4-
hydroxyphenyl)piperazinyl)phenyl)azophenyl)sulfonyl)ethyl phosphonic acid, 2, a new
nonlinear chromophore developed for use in zirconium phosphate/phosphonate SAMs.
As discussed in the Introduction, our motivation for this work is to construct multilayer
assemblies with controlled orientations and, using surface second harmonic generation
measurements, evaluate interface defect properties.7 In addition to the synthesis of the
chromophore, it is important to perform some basic characterization steps to ensure that

appropriate layered structures can be made. We need to determine that the substitution of

-24-

the terminal groups in 2 relative to 1 does not alter the electronic spectroscopic response
of the chromophore seriously. This is an important issue because unwanted resonance
effects in nonlinear spectroscopic measurements can cause difficulty in interpretation.
We also need to consider the ability of 2 to form layered structures in the same manner as
1. We have synthesized l and report a comparison of the linear optical responses and
ellipsometric thicknesses per layer for SAMs formed from each of these two compounds.
The linear optical responses of chromophores l and 2 in ethanol are shown in Figure
2.3a. The absorption maxima for l and 2 in solution are 429 nm and 415 nm,
respectively. The slight difference between the two chromophores is due to the position
of the terminal functional groups and the influence each has on the TC system of the
chromophores. We have assembled layers of 1 and 2 separately on SiOx substrates. In
Figure 2.3b, we show the absorption spectra of layers of chromophores l and 2, revealing
very similar electronic structures. Both chromophores are dominated by a broad
absorption band centered at 417 nm, with only very slight differences in the red edge of
the spectra. The higher energy bands appear to be somewhat different for the two
chromophores and the reason for this effect is not obvious by inspection. Semi-empirical
calculations of the linear response for these two chromophores show the ordering and
energies of the first three excited singlet states to be slightly different, especially in the
region of the most prominent band(s) (Figure 2.4). These calculations provide some
justification for the differences in the absorption data, but given the spectral widths of the
absorption bands, making the comparison between experiment and calculation at any
meaningfiil level is not feasible. By plotting the absorbance at 417 nm of each

chromophore against the number of deposited layers, we recover slopes of 0.0567 $0.001

-25-

0.06

Absorbance Intensity (a.u.)

0.05

0.04

0.03

0.02

0.01

0.00

 

 

 

300 400 500 600

 

 

l I l l I

 

300 400 500 600

Wavelength (nm)

Figure 2.3 UV-Vis absorption spectra of (a) solution phase of 1 (solid blue) and 2 (dashed
red); (b) the assemblies of 5 bilayers of 1 and 2 on fused silica.

-26-

and 0.0546 :1: 0.0007 for 1 and 2, respectively as shown in Figure 2.5a. These data
demonstrate that the same concentration of chromophores is deposited in each layer, an
important result for future uses of these chromophores in nonlinear applications. Because
consistent build up is seen for each layer, we also conclude that priming the surface prior

to chromophore deposition does not yield a more uniform surface coverage.

 

 

 

 

 

 

 

 

 

f= 0.02
40000 -
l- f: 0.12
f: 1.2 f= 0.81
30000 -
f= 0.001 F 0002

 

 

 

 

 

energy (cm")
8
l

 

 

 

Figure 2.4 Semi-empirical calculations of the linear response for chromophores
1 (left) and 2 (right). Wide lines are singlet states, and calculated oscillator
strengths for St (— So transitions are indicated next to the upper state. Narrower
lines indicate triplet states.

From the solution phase spectra, we extract smax = 207,600 L/mol-cm for 1 and given the

same linear dependence of A on layer growth for 2, we infer the same value of 8. Using

the experimental data shown in Figure 2.5a, we estimate the surface loading density to be

-27-

1.6 x 1014 cm’z—layer'l.

From optical null ellipsometry, we recover a best—fit slope of 26.3 i 0.5 A/layer
for chromophore 1 and 25.5 i 0.4 A/layer for chromophore 2. (Figure 2.5b) These values
are the same to within the experimental uncertainty. Our results are in good agreement
with Katz and coworkers data on the layered growth of 1.15 The length of each molecule
in its most stable conformation is calculated by molecular mechanics to be 27 A. It has
been reported previously that the thickness of a Zr-P03- linkage is 3.75 A.16 An average
layer thickness of 25.6 A indicates that the chromophores are tilted at an angle of ~20°
from the surface normal, presuming full surface coverage. Katz et al. also reported this
result for 1. Because both chromophores tilt to similar angles when each is adsorbed into
a layer, it can be assumed that assemblies of alternating layers of 1 and 2 will result in
similar increases in the layer thickness. This finding also suggests that our approach to
the measurement of defect densities in opposing bilayer structures is feasible with these
complementary chromophores.

These chromophores are characterized by the same layer density, electronic
structure, and tilt angle based on the linear response and ellipsometry data. We conclude
that multilayers of 1 and 2 are structurally similar making them a good choice for
investigations of the chemical structural contributions to the x0) response of oriented ZP

systems.

2.4 Conclusions

Chromophore 2 has been synthesized and characterized using 1H NMR, UV-

visible spectroscopy, and optical null ellipsometry. Layers of both 1 and 2 have been

-23-

Absorbance

 

 

 

5 400} 3

a 350- .i b
. i

g 300r i:

g 250.” ii

:5 200- .1 *

g 150" j I

a 128' . o

'63 T i:
O.- l 1 l A l l l n l l

 

 

-2‘ 0 2 4 6 8 ‘1'0‘112’114‘16
NumberofIayers

Figure 2.5 (a) Dependence of absorption on number of layers for both chromophores. See
text for a discussion of these data. (b) Ellipsometric thicknesses of each layer for 15 layer
assemblies of chromophores 1 (O) and 2 (I). The slopes of the best-fit line through these
data are 26.3 i 0.5 A/layer for 1 and 25.5 i 0.4 A/layer for 2. The y-intercept is a result
of the primer layer.

-29-

built up on Si and silica wafers and each has an average thickness of 25.6 A and an
average tilt angle of ~20° from the surface normal assuming full coverage. Absorption
data indicate a surface coverage of 1.6 x 1014 cm'z-layer‘l. The chromophores exhibit
very similar x0) signals on silica substrates, as we detail in the next paper. The similar
absorption and layer-forming characteristics of both chromophores make them ideal
choices for surface second harmonic generation investigations aimed at characterizing

and quantifying defect density and type in this family of SAMs.

-30-

2.5 Literature Cited

1. Katz, H. E.; Scheller, 0.; Putvinski, T. M.; Schilling, M. L.; Wilson, W. L.; Chidsey,
C. E. D.; Science, 1991, 254, 1485.

2. Florsheimer, M.; Kupfer, M.; Bosshard, C.; Looser, H.; Gunter, P.; Adv. Materials,
1992, 4, 795.

3. Zhu, X.; Chen, Y. M.; Li, L.; Jeng, R. J.; Mandal, B. K.; Kumar, J.; Tripathy, S. K.;
Opt. Commun, 1992, 88, 77.

4. Marowsky, G.; Lupke, G.; Steinoff, R.; Chi, L. F.; Mobius, D.; Phys. Rev. B, 1990, 41,
4480.

5. Palto, S. P.; Blinov, L. M.; Yudin, S. G.; Grewer, G.; Schonholl, M.; Losche, M.;
Chem. Phys. Lett., 1993, 202, 308.

6. Ulman, Abraham; Chem. Rev., 1996, 96, 1533.

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

8. Lee, H.; Kepley, L.J.; Hong, H.; Sohail, A.; Mallouk, T.E.; J. Phys. Chem, 1988, 92,
597.

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

10. Perrin, D. D.; Armarego, W. L. F.; Perrin, B. R.; Purification of Individual Organic
Chemicals, Pergmon Press, Oxford, 1980, 449.

ll. Hanken, Dennis G.; Corn, Robert M.; Anal. Chem, 1995, 67, 3767

12. Smiles,S.; Bere, C.M.; Organic Syntheses, Wiley, New York, 1941, Collect. Vol.1, 7.

13. Baker, B. R.; Querry, M. V.; J. Org. Chem, 1950, 15, 413.

14. Popov, D.; C. R. Acad. Bulg. Sci, 1966, 19, 1163.

-31-

15. Flory, W. C.; Blanchard, G. J.; J. Am. Chem. Soc, in review.
16. Katz, H. E.; Schilling, M. L.; Chidsey, C. E. D.; Putvinski, T. M.; Hutton, R. S.

Chem. Mater. 1991, 3, 699.

-32-

Chapter 3

LASER SYSTEMS FOR SURFACE SECOND HARMONIC GENERATION
MEASUREMENTS

Summary

We report on the laser systems used for the surface second harmonic generation
(SSHG) measurements. We use Nd:YAG lasers (1064 nm) as our fundamental light
source and detect the second harmonic signal (532 nm). We present the system design,
characteristics, and some example data from each of the systems used. Two time regimes
were used as well as two configurations, reflection and transmission geometries. The
details of these systems as well as preliminary work to interface the laser system to a

personal computer are presented.

-33-

3.1 Introduction
With the discovery of masers and then lasersI came the ability to generate

harmonic frequencies from materials P= me + x(2)EtE + meE’E +___ (bulk)
because of the high electric field I" = “E + BE‘EJ' YEE‘E + ---(molecular)
intensities that these light sources produce. Equation 3.1

Equation 3.1 describes the induced polarization for a medium in an electric field on
microscopic (u) and macroscopic (P) scales. For a low intensity electric field (<102
W/cmz), the higher order terms can be neglected and the response is described using only
the linear portion of the equation (1“), or). For high intensity illumination (>105 W/cmz),
the electrons within the system can no longer oscillate with the electric field, and the
expanded form of the equation needs to be used to describe the induced polarization of
the system. One result of the expanded form of this equation is that moieties containing a
center of inversion will not exhibit even harmonic responses. This is because the x0)
tensor elements of a material with a center of inversion are zero in the dipole
approximation.

(2)

The first nonlinear term is x , the second order nonlinear susceptibility, is related

to B, the first hyperpolarizability. An assembly of B-active molecules aligned
appropriately will give an area with an overall dipole moment, yielding a x0) signal. If
this same molecule is dissolved in solution, there will be no x0) response because there is
orientational averaging in the bulk.

The driving force for using second harmonic generation (SHG) in these
experiments is the symmetry restriction described above. Our intent is to build bilayer

systems purposefully incorporating a center of inversion and thereby turning off any SHG

-34-

from defect-free portions of the thin film. In principle, any residual response arises from
defects within the bilayer system.

The focus of this chapter is to describe the development of the laser systems used
for the surface SHG measurements. We use the 1064 nm light generated by Nd:YAG
lasers as our fundamental light source. Two different time regimes, ns and ps, have been
used to generate the SHG signal. We have also used reflection geometry for films built
on Si and transmission geometry for films built on fused silica substrates. A Labview®

program was written and a peak track-and-hold circuit was built to interface the laser

system to a personal computer. These items are detailed in this chapter.

3.2 Experimental

Two different Nd:YAG lasers were used for this thesis. A Q-switched ns
Nd:YAG (Quanta-Ray DCR) and Q-switched, mode-locked ps Nd:YAG (Quantronix
416) were used as sources of fundamental (1064 nm) light. The details of these laser
systems are presented in section 3.3. The data presented characterizes each of the
systems, but the majority of the results were acquired using the Q-switched, mode-locked
Nd:YAG system.

Surface Preparation. Si(100) wafers (Multi Crystal Optics, Inc.) and silica
substrates were used. Si substrates were cleaned in piranha solution (3:1 H2S04:H202 —
caution- strong oxidizer!) for 10 minutes, rinsed with distilled water, placed in 2 M HCl
for 10 minutes, and rinsed with water. The substrates were dried in a dry N2 stream.
Under an Ar atmosphere, 20 mL anhydrous octane was added to a vessel containing the

substrate. The octane was heated to reflux and 3-aminopropyltrimethoxysilane was

-35-

added to make a 1% v/v solution. After 10 minutes of reaction the substrate was
removed and rinsed with warm hexane, then water. A N2 stream was used to dry the
surface. Silica surfaces were cleaned in piranha solution, dried, but were not primed.
Both surfaces were then phosphorylated. This was done under Ar, using a solution of 20
mM 2,3,4-collidine and 20 mM P0Cl3 in anhydrous acetonitrile at ambient temperature.
After 10 minutes, the substrates were removed, rinsed with acetonitrile and water and
dried with N2. The surfaces were then zirconated by immersing in aqueous 5 mM Zr0C12
solution for 10 minutes. The surfaces were rinsed with water and dried.

Chromophore Deposition. The chromophores were deposited on the substrates
from a saturated solution of each dissolved in 1:4 DMFzEtOH. For Si wafers, the
solution was kept acidic by adding dropwise 4 M HCl in dioxane. The temperature of
each solution was held just below boiling while the surfaces were immersed for 10
minutes followed by rinsing with warm ethanol and drying with N2. The phosphorylation
and zirconation of the surfaces is repeated as described previously. To monitor layer
adsorption, ellipsometric thickness was used for Si wafers and UV-Vis absorption was

used for silica substrates.

3.3 Results and Discussion

ns Q—switched Nd: YAG laser system
The first laser system used for this work was a ns Q-switched Nd:YAG laser
(Quanta-Ray DCR). A schematic of the ns SHG system in reflection mode is shown in

Figure 3.1. The laser produces ~14W average power at 1064 nm with 5ns pulses at 10

-36-

 

 

 

 

 

   

Figure 3.1 Q-switched ns Nd:YAG and mode-locked ps Nd:YAG laser systems.
Reflection geometry was used and the fundamental light was focused onto the sample and
them recollimated. A dichroic filter was used to remove the fundamental before the 532
nm light was detected.

-37-

Hz. The 1064 nm pulse train is directed through a 1 mm spatial filter. The beam, having
an average power of 13 mW (intensity: 1.65x107W/cm2), is passed through a color filter
(RG 630) to remove any SHG generated by the optics, is focused onto the sample using a
dichroic focusing mirror (100 um focal diameter),and is recollimated with a dichroic
mirror. A dichroic filter is used to remove the residual fundamental, and the 532 nm
beam is directed through a monochromator (Spex) and then to a photomultiplier tube
(Hammatsu 1106 PMT). The detector output is buffered using a voltage follower, and is
then read and saved using a digital signal analyzer (Tektronix DSA 602A). An incident
angle of 61° with respect to the surface was used for all measurements. The plotted data
are the average of three readings, each being an average of 1024 single scans. The ratio
of the signal vs. bare reference substrate is plotted to normalize the signal.

Katz et a1. has reported previously on the second order nonlinear reSponse of
chromophore 1.2 Our data are consistent with his and we observe a square-law
dependence of the second harmonic intensity with increasing number of layers for
chromophore 1 built off of Si(100) substrates. This is an expected result for a system
where the nonlinear medium is significantly thinner than the coherence length of the
incident light source. The data we present (Figure 3.2) demonstrate this relationship and
show that the system is sufficiently ordered to allow the square-law relationship to be
observed for this chromophore. The drawback to using this system for this-body of work
was the heating of the surface and surface mount. Much of the incident energy was
dissipated thermally within the substrate. It was thought that this extra energy could
create defect contributions to the SSHG signal not originally related to the surface

synthesis. The use of a ps laser to reduce the amount of energy per pulse was the next

-33-

16

 

 

 

 

 

i \’ i ‘
l .\ J _N a
l l p—l 0.;
C6
,4
i FA . a
O
"E.
W ~00 O
8
i=3
i U
q...
0
‘fi' 5
.0
E
q :3
Z
CD
ltltliltltliltJiltltlili
In In In to tn tn
\0 ln <1“ m N v-i
coins QIdLIIBS
Z/I[(00Z) We 1]

Figure 3.2 Plot of square law dependence of the SSHG signal vs. number of chromophore
layers for the Q-switched ns Nd:YAG laser system. Chromophore l is chemisorbed on a
Si(100) substrate.

-39-

step in the building of the SSHG system.

ps mode-locked Nd: YAG laser system

The Nd:YAG laser (Quantronix 416) produces ~10 W average power at 1064 nm
when mode locked. The mode-locked pulse train produces a 100 ps pulse at a repetition
rate of 80 MHz. The 1064 nm beam is attenuated using a polarizing beamsplitter cube
(CVI) in conjunction with a polarization rotator. The beam, with an average power of
200 mW, is then directed through the same reflection mode optical train as described for
the ns Nd:YAG laser system (Figure 3.1).

When a silicon substrate was at the focal point of the focusing mirror, the damage
threshold was exceeded and the surface was ablated. When the beam remained
collimated and directed onto the surface using a flat mirror, no signal was detected
because the incident intensity was not sufficient to generate a detectable SSHG signal.
An intermediate, defocused arrangement allowed signal detection without destruction of
the substrate. This finding is consistent with the ns system and the intensities needed to
observe a SSHG signal. The first data from this system came from layers of
chromophore 1 built on silicon substrates and the raw data for the SSHG response of
eight layers are plotted in Figure 3.3a. The reference SHG signal was taken
simultaneously from a KDP crystal. A zero-signal line is drawn in red for ease of
reading. The linear relationship between the square- root of the SHG intensity vs.
number of layers was seen (Figure 3.3b), demonstrating that the system is sufficiently
ordered, but the offset of the signal from zero remains an issue. The downfall of this

configuration is the small signal obtained, even for many chromophore layers. In order

-40-

0.25

0.20

0.15

0.10

Voltage (mV)

0.05

0.00

0.05

0.04

(20) )1 1/2

KDP

0.03

(2w)/1

0.02

[ sample

0.01

0.00

 

 

 

 

 

 

0.001 '
WWW
~ 0.000 ’”‘
. a
-0.001F . - A - . _ l
-0010 -0005 0.000 0.005
r
I 1 l I I l 1
-0.010 -0005 0.000 0.005
Time (s)

 

l 1 l 1 l 1 l 1 l 1 l

 

0 2 4 6 8 10
Number of Chromophore Layers

Figure 3.3 Mode-locked ps Nd:YAG laser system (a) raw data demonstrating the small
signal. (KDP reference; SSHG signal from chromophore 1 on Si substrate; V=0); (b) plot
of square law dependence of the SSHG signal vs. number of chromophore layers

-41-

for the system to be sensitive to a small SHG signal generated from defects alone, the )6”
signal from individual chromophore layers must be large. Since the peak power on the
sample at the focal point damaged the surface, a modification to the system to reduce the
duty cycle and increase the peak intensity without damage to the surface was needed. We

report two alternatives below.

ps mode-locked Nd: YAG laser system with pulse compression

Optical pulse compression is a technique used to achieve ultrashort pulses by
using self-phase modulation to expand the pulse frequency width, then manipulation of
the “chirp” on the pulse using either a grating and a prism or a grating pair.3'7 By
decreasing the pulse width in time, very high peak powers can be achieved. We
compressed the 100 ps pulses to ~10 ps pulses, but could not sustain the pulse integrity
because of the instability of the system. The input pulse would heat the tip of the fiber
and eventually burn the cladding, distorting the tip, and ruining the end. Stable
compressed pulses, before the fiber damage could only be maintained for one hour. We

abandoned the idea of compression to achieve high peak powers.

Q-switched, mode-Iockedps Nd: YAG laser system

We installed a Q-switch on the mode-locked Nd:YAG laser to give an envelope
of mode-locked pulses with high peak power. The pulse profile of the Q-switched mode-
locked pulse is shown in Figure 3.4. A Q-switched, mode-locked Nd:YAG laser

(Quantronix model 416) produces ~1.4W average power at 1064 nm with a 2 us envelope

at 500 Hz. The mode-locked pulses (80 MHz repetition rate) within the 2 us envelope are

-42-

characterized by a 100 ps pulse width. Each 2 us envelope contains 2.8 mJ total energy,
with the maximum ps pulse in the envelope having ~125 u] pulse energy.8 A schematic

of the surface SHG system is shown in Figure 3.5. The 1064 nm pulse train is directed

Intensity (a.u.)

0.00 ~_——-——~ WM»

 

 

A l . . l . l . I
—4.0xio" -2.0xl0'7 00 2.0x10" 40x10" some"
Time (s)

Figure 3.4 Pulse envelope of a Q-switched, mode-locked pulse. The mode-
locked pulses (80 MHz repetition rate) are characterized by a 100 ps pulse width.

through a 2 mm spatial filter and a polarization rotator, then through a focusing lens and a
color filter (RG 610) to remove any 532 nm light generated by the optics. The 1064 nm
beam is focused onto the sample (~100 um focal diameter), and is then recollimated.
After the recollimating lens, three dichroic mirrors are used to separate the fundamental
from the second harmonic light. The second harmonic light is directed into a 0.25 m

monochromator (Fisher) and is detected witha PMT (Hammatsu 466). The

-43-

 

 

 

 

Figure 3.5 Q-switched, mode-locked Nd:YAG laser based system for SSHG experiments.
Transmission mode i shown.

resulting transient is amplified (Ortec Model 451) and then stored using a digital signal
analyzer (Tektronix DSA 602A). The sample is located at the focal point of the
fundamental laser beam and is rotated manually about its vertical (y) axis from 0° to 70°
relative to the incident beam using a precision rotation stage (Newport 481 A). The
angular dependence of the second harmonic intensity is shown in Figure 3.6. Each

experimental datum point is the result of three individual readings, with each reading

 

 

 

SSHG signal (a.u.)
A
l

 

 

 

 

 

 

 

 

 

 

3 — ,l' ‘
. ll"I ;
2 - .llm‘h-‘l I
P 911' it ‘
1 '- i.‘ I
0 »— '
1 l 1 l 1 l 1 I 1 I 1 l 1 l
O 10 20 30 40 50 60 ' 70
Sample Rotation Angle (deg)

Figure 3.6 SSHG signal for three bilayers of chromophore 2 chemisorbed on a
fused silica substrate.

-45-

being an average of 532 scans of the amplified transient recorded by the digital signal
analyzer. The complete scan is normalized by referencing it to a scan of a reference bare
fused silica substrate.

Data from the transmission mode geometry scanned from 0°-70° angle of

incidence are shown in Figure 3.6. These data are from a quartz substrate with 3 bilayers

 

 

0.16-
0.14-
3 0.12-
EL .
.§‘ 0.10-
g .
E 0.08-
% 0.06-
m n
0.04-

l

0.02 J

 

90 120 150 180 210 240 270
Incident Polarization

Figure 3. 7 Polarization dependence of the SSHG signal of chromophore 1 adsorbed
onto Si(100). The inset cartoon depicts the coupling efficiency of the polarization to
adlayers with their induced dipole moments normal to the surface.

of 2 chemisorbed to the surface. A complete interpretation of these data are presented in
chapter 4. The data were taken with the 1064 nm light polarized at 90° with respect to
the surface. A plot of the polarization dependence is shown in Figure 3.7 for

chromophore 1 chemisorbed on silicon. The polarization dependence of the SSHG signal

for chromophore 1 is shown in Figure 3.7 inset.

~46-

The next step in this project was to interface the laser system directly to a
personal computer using Labview®. It was quickly determined that the data acquisition
(DAQ) board that we had was slow compared to the temporal width of the detector
output pulse. The DAQ board could only detect pulses with durations greater than 20 us
while the detector output pulse was 2 us. I built a circuit to hold this peak value in time

so that Labview® could detect the pulse.

Peak Track-and-Hold Circuit.

A peak track and hold circuit originally designed by John Rugis (MSU)
electronics shop, was built to track and hold the voltage output amplitude of a PMT for
an amount of time that allows the AT—MIO-16-L9 board sufficient time to read this value.
The schematic of a modified version of the original circuit is shown in Figure 3.8. The
holding time was chosen to be 88 us, so that two channels could be sampled
simultaneously with 8-10 points forming each waveform depending on the scanning rate
of the Labview® program.

Figure 3.8 (inset) shows the relative timing between the trigger, input pulse, and
the output pulse. A TTL input is needed to trigger the circuit. We used the analog output
of the laser Q-switch driver as the trigger after converting it to a 5V signal with an
inverting amplifier (LF357N). The duration of the trigger signal is 1 us. The trigger is
the signal input to a retriggerable monostable multivibrator (74LSl23). The output is a
square wave with a time dependence related to the resistor (R) and capacitor (C) used.
We used RT = 267 k0 and CEXT = 1000 pF to give a holding time of 88 us as

calculated

-47-

 

 

«15.95:.

00— 0w 8 3 ON 0

— EDP—.30

 

 

 

 

FDA. 2.

 

_. :5:me 9%: D

[II

— «NOD—m...

 

amt—3.

 

 

 

 

 

 

 

b N
m
him—mm:— 91
GNJN
If- N >
QZO
08v
38—
020

 

 

 

 

 

m~_mnvs
:40
IE 0 a 1111:;
I. a m .Inmnmmm>
my m.
m_ mg n. (.nvuquuuulnlul.
u u
T_n
“coco.
Co:
na>
m~_mnv~
an“.
dos- .- lJl-e
m m ruaau¢Wm>
O M 3
W m w (UlwullllvllnmmOO—dh
unocc.
fluxpeN
~_7r2 .uzmv

 

 

 

 

 

755.5:— _ 1 D20
22> H
u noon
H30 .129. D o + l—Iu
0:: a - :28
t Bogus—h

NE>

 

 

 

 

 

Figure 3.8 Peak track-and-hold circuit schematic used for computer interfacing. The

details of this circuit are discussed in the text.

-43-

using equation 3.2:

Tw = 0.33*RT* CEXT ( 1+ 0.7/RT) Equation 3.2

This is the timing window that starts and resets the peak track and hold circuitry. The
output Q from the multivibrator is the input to a bilateral switch (4066). When Q is high
(H) the switch is on or tracking the pulse, and turns off or resets when Q is low (L). The
time between the Q—>H and Q—>L is Tw.

The input pulse is sent into a high-speed inverting amplifier (LT1363) with a gain
of one. The experimental signal travels through a fast diode where it charges a capacitor.
This capacitor is isolated and therefore holds this charge while the bilateral switch is in
the on position. The charged capacitor is isolated from the AT-MIO-l6-L9 board
through a voltage follower. This is a buffer that allows the held voltage to be read by the
board without the capacitor losing its charge. When the bilateral switch turns off, the
capacitor is shunted to ground, and discharged. This process resets the circuit, which

waits for the next trigger/pulse input.

Labview® Code

A program was written that would read the data fi'om the AT-MIO-16-L9 board,
and plot this array as a waveform. This program was written using the icon-based
programming software, Labview® from National Instruments. The program collects a
scan of data in time and saves it as an array, saves the array, and resets itself and waits for
the next trigger/pulse input. As it collects the arrays of data, it averages them. It then

saves the array as an ASCII file, which can then be processed by a graphing program.

-49-

The program can be found in Appendix A.

The first step is to initialize the program. This is done with the AI CONFIG
subVI, one of several subVIs included in Labview®’s software package. This subVl
configures the hardware and allocates a buffer for the input. Here the device number and
the buffer size are determined.

The AI START subVI sets the parameters such as tn'gger type, scan rate, and the
number of scans to acquire. The program is externally triggered by the signal from the
Q-switch. It is triggered by the falling edge of the square wave. The scan rate and
number of scans are variable parameters that are connected to the front panel of the
program so that they can be selected by the user.

The AI READ subVI is where the data from the AT-MIO-16-L9 channel 0 or 1 is
read by the program. It reads the channel values and fills the allocated buffer at a rate
determined by the set parameters. The array has a number of columns depending on how
many channels are scanned. (i. e. if two channels are scanned, there will be two columns
of data in the array). This array is sent to a bundle function, which plots the data points
of the array depending on the time, set by the inverse of the scan rate. This data is also
sent to a shift register at the edge of the for loop, where it is saved and then added to the
next collected array. The summed array of all of the scans is divided by the number of
scans giving a final array that is the average waveform of the scans. This array is plotted
for the user, and is also sent to out of a second for loop. This array is broken down into
individual columns, depending on the number of channels read, transposed, and saved
with a second column of time values. The data is saved as an ASCII file, with values for

X and Y. This is to make it very simple to call this data into a plotting program and

-50-

graph it.

Interface Troubleshooting

Each of the components works independently, but there were several problems
encountered with interfacing the peak track-and-hold circuit to the Labview® program.
The details and possible solutions to each are presented. The interfacing software and
hardware are not currently being used due to these problems, but are presented for future
reference.

One problem with the circuit is that when there is no input signal, the 390 pF
holding capacitor charges to a voltage of 150 mV. This signal could be present because
there may be an offset in this first op amp that is charging the capacitor in time. If this is
the problem, then a voltage follower can be inserted between the first op amp and the
diode to compensate for this offset.

Another problem between the interfacing of the laser system with the personal
computer, is that the PULSE IN signal arrives so quickly after the trigger, Labview®
misses the first two points of the waveform. One way to delay the read time is to trigger
a TTL channel on the AT-MIO board and delay the TTL output of this channel by ~10
us. Then the TTL out of this channel can be sent into the external trigger input that
triggers Labview® to read.

One problem with the Labview® code is that it sometimes gives an error code
when running the program. The error is an “overFlowErr” with a description of “at a
time of the update clock for the input channel, the device-resident memory was unable to

accept additional data- one or more data points may have been lost”. This error could be

-51-

the result of the data not being transferred to Labview® before the next data is available
to be read. The reason for this situation is not completely clear, but the overall
conclusion of this interfacing using the current laser system is to obtain a faster DAQ

board for fast data acquisition.

3.4 Conclusions

The systems developed for surface second harmonic generation measurements
have been described. Two different time regimes were used and also two different
geometries. The laser system used for measurements reported in the following chapters
is a Q-switched, mode-locked Nd:YAG laser and measurements are made in transmission
mode. It has been characterized and data obtained from this system are used for the study
of defects and layered structure effects in self-assembled multilayer zirconium
phosphate/phosphonate films. A peak track-and-hold circuit was built and a program
using Labview® software was written in order to interface the laser system to a personal
computer. Several problems with these components need to be solved before they can be

used with this system.

-52-

35 Literature Cited

1. Franken, P. A.; Hill, A. E. Peters, C. W.. Weinreich, 0.; Phys. Rev. Lett., 1961, 7(4),
118.

2. Katz, H. B.; Wilson, W. L.; Scheller, G.; .1. Am. Chem. Soc, 1994, 116, 6636.

3. Shank, C.V.; Fork, R.L.; Yen,R.; Stolen, R. N.; Tomlinson, W. J .; Appl. Phys. Lett.,
1982, 40(9), 761.

4. Nikolaus, B.; Grischkowsky, D.; Appl.Phys.Lett., 1983, 42(1),1.

5. Choi, Kee-Ju; Gustafson, Terry L.; IEEE J. Quantum Electron, 1989, 25(12), 2441.

6. Tomlinson, W. J .; Stolen, R. N.; Shank, C.V.; J. Opt. Soc. Am. B., 1984, 1(2), 139.

7. Nikolaus, B.; Grischkowsky, D.; Appl.Phys.Lett., 1983, 43(3),228.

8. Quantronix, Model 416 Lasers, Technical Manual; Quantronix Corporation,

Smithtown, NY, pp. 1-3.

-53-

Chapter 4

STRUCTURAL CONTRIBUTIONS TO SECOND-ORDER OPTICAL
NONLINEARITIES IN ORIENTED INTERFACIAL MULTILAYERS

Summary

We report on the characterization of Zr-phosphate/phosphonate (ZP) self-assembled
multilayer structures using surface second harmonic generation measurements. Two
structurally complementary Xm-active chromophores that can be deposited with
orientational control relative to the substrate were used. These chromophores produce
multilayer xm responses identical in magnitude but of Opposite sign. O\opposing
bilayers formed with these chromophores were used to produce two different structural
motifs, each with a local center of inversion about the ZP interlayer bonding plane. The
x”) responses of these two bilayer systems are different and reveal the extent to which the
dipole approximation is valid in these layered assemblies. Our data elucidate the role that
quadrupolar terms play in the x0) responses of thin interfacial films. This work illustrates
the complex and subtle structural issues associated with the design and construction of

layered interfaces, and provides a means to evaluate the vacancy and orientational defect

density in layered materials.

-54-

4.1 Introduction

Self-assembled mono- and multilayer structures (SAMs) have received a great
deal of attention in the materials community for fundamental as well as more near-term
reasons."9 An overarching goal of this effort has been to make the connection between
molecular-scale organization, interface chemical identity, and macroscopic materials
properties. A tremendous amount has been learned about interfacial monolayers and the
associated measurement technology using the alkanethiol/gold system.'°'12 While studies
of this archetypal system have led to profound advances in our understanding of organic-
modified interfaces, thiol/gold monolayers suffer from long-term chemical stability

13,14

limitations and, until recently, the inability to form chemically-bound multilayer

5

assemblies.I These limitations have been a driving force for research to identify

2,16-18

alternative layer growth strategies, with both Silane and Zr-phosphate/phosphonate

(ZP) chemistrym'35

proving successful and versatile. The interest in SAMs based on ZP
interlayer linking chemistry arises from their ease of assembly, the mild conditions used
in the formation of these layers, and their structural and thermal stability, once formed.

Recent work on ZP-based SAMs has pointed to limitations in their properties (e. g.
mesoscopic ordering, optical response) that are mediated by structural defects.”37 The
characterization of defects in mono- and multilayer interfaces is central to understanding
SAM properties and is an area of interface and materials science that remains to be
explored more fiilly. For alkanethiol/gold monolayers, cyclic voltammetry and
impedance measurements have proven to be valuable tools in detecting the presence of

38-43

vacancy defects. For multilayers, electrochemical methods are not generally as

useful owing to the short-range nature on the electron tunneling process responsible for

-55-

the experimental signal. We are investigating a means to study defects in multilayer
assemblies using nonlinear spectroscopic methods. For many optical methods, the
characteristic response of surface defects is small compared to the bulk material
contribution, limiting the utility of spectroscopy for such investigations. We are
interested in measuring vacancy and substitution defects in layered interfaces by utilizing
the functional chemistry of selected second order nonlinear chromophores to control their
orientation during layer growth. We have chosen x0) measurements because of their
intrinsic surface-selectivity.44 By constructing interfaces with specific, predetermined
orientation of the chromophore nonlinear transition moment within each layer, multilayer
films can be assembled to produce either centrosymmetric or non-centrosymmetric bulk
ordering to first approximation. The 1(2) nonlinear response of a centrosymmetric bilayer
structure will be null to within the validity of the electric dipole approximation. This
dipolar cancellation allows examination of the residual x0) response in the context of
vacancy and substitution defects and higher order multipole contributions to the
experimental signal against a nominally dark background. The resolution of the
contributions to the )6” response of opposing bilayers is an important first step in
establishing this means of defect characterization in layered materials.

We use two compounds to explore the utility and practical limitations of
chemically-based cancellation of the second order nonlinear response. The first is the
Xm-chromophore (4-(4-(4-(4-((2-hydroxyethyl)sulfonyl)phenyl)azo)phenyl)piperazinyl)
phenyl)phosphonic acid, (1), reported by Katz and coworkers and designed for its
combined rigid structure and large first hyperpolarizability, [3.33 The second compound,

(2—(4-(4-(4-(4-hydroxyphenyl)piperazinyl)-phenyl)azophenyl)sulfonyl)ethylphosphonic

-56-

acid, (2), is the structural complement of 1, where the functional groups at the termini
have been exchanged. The synthetic route to 2 is reported in the preceding paper,45 and
the structures of both molecules are shown in Figure 1.1. We report on the surface
second harmonic generation (SSHG) response of multilayer assemblies of these
chromophores for several different bilayer structural arrangements. Following a brief
discussion of the experimental details, we consider the results of our experiments on
these complementary chromophores in the context of the physical and chemical origins of
the nonlinear response. We discuss the form of the SSHG signal and how that signal is
related to the several contributions to the second order nonlinear susceptibility, )6”. We
find direct evidence for the role of electric quadrupolar contributions to our data, with the
magnitude of the quadrupole contribution depending sensitively on the chemical

functionality in the vicinity of the ZP linking moieties.

4.2 Experimental

Synthesis. The synthesis of chromophore 2 is detailed in chapter 2 and the
synthesis of chromophore l was reported by Katz et al.33 All chemicals used were
obtained in the highest purity grade available and the procedures reported in the literature
were followed.

Surface Preparation. Fused silica substrates were cleaned in piranha solution (3:1
H2S04:H202 — caution - strong oxidizer!) for 10 minutes and rinsed with distilled water.
The substrates were then dried in a N2 stream and introduced to a reaction vessel. The
surfaces were phosphorylated under an Ar atmosphere using a solution of 20 mM 2,3,4-

collidine and 20 mM POC13 in anhydrous acetonitrile at 20°C. After 10 minutes the

-57-

substrates were removed, rinsed with acetonitrile and water and dried with N2. The
surfaces were then zirconated by immersion in an aqueous 5 mM ZrOClz solution for 10
minutes. The zirconated substrates were rinsed with water and dried.

Chromophore Deposition. Chromophores 1 and 2 were deposited on zirconated
substrates from a saturated solution (~ 0.2 mM) of the appropriate chromophore
dissolved in 1:4 DMFzEtOH. Deposition of each layer was from a solution containing
either 1 or 2, but not both. The temperature of each deposition solution was maintained
slightly below boiling while the zirconated substrates were immersed for 10 minutes,
followed by rinsing with warm ethanol and drying with N2. Subsequent phosphorylation
and zirconation of the chromophore-containing surfaces was performed as described
above. Before measurement of adlayers for x0) activity, terminal OH groups were
phosphorylated. To monitor layer adsorption, the UV-visible absorbance spectrum of the
sample was measured after each deposition cycle, and in all cases linear growth was
observed.

Surface Second Harmonic Generation Laser System The details of the SSHG

laser system are given in chapter 3.

4.3 Results and Discussion

The purpose of this chapter is to evaluate the use of structural cancellation at
interfaces as a means of creating a nominally null )6” background, making possible the
characterization of interface vacancy and orientational defect density. We provide an
overview of the relevant theory and discuss our experimental findings in the context of

the contributions to the angle—dependent surface SHG data. We consider the nonlinear

-53-

response of the bare fused silica substrate first, then the same surface with layers of each
chromophore adsorbed. With that information in hand, we present our data on two
opposing bilayer structures; [SiOx-l-Z] and [SiOx-2-1]. Their SSHG responses are not
identical, pointing to the role and chemical identity of an adlayer quadrupole contribution
to the experimental signal.

Many surface SHG experiments are performed in reflection mode,46419 with the
angle of incidence of the fundamental electric field being the independent experimental
variable. The form of the signal in these experiments is an incidence-angle dependent
intensity of the reflected second harmonic light. The specific form of the data can, under
favorable circumstances, be used to estimate the average tilt angle of the nonlinear
chromophore relative to the surface normal and the magnitude of the experimental signal
can be used to determine the value(s) of selected x0) tensor elements. For SSHG

transmission measurements on transparent substrates,”54

such as those we report here,
the form of the experimental signal is somewhat more complicated and optical
interference and transmission/reflection effects must be accounted for. We detail these
effects below and discuss how information on interface properties can be obtained from
our data.

We consider the system shown in Figure 4.1, where the incident electric field at
frequency 00 and polarized along the x-axis and propagates along the z-axis. The sample,
composed of a transparent substrate, has interfacial layers present on both the front and

back surfaces. The sample is rotated about the y-axis and the resultant second harmonic

signal intensity is recorded as a function of rotation angle. Four effects contribute to the

-59-

data. The first is the angle-dependence of the reflection and transmission properties of
the sample at both the fundamental and second harmonic frequencies. The second effect

is optical interference between second harmonic light generated at the front and back

X

 

 

polarization

T >
i

 

 

 

 

 

 

Figure 4. 1 Model system of thin films on both sides of a substrate. 1 and 5 are air; 2 and
4 are interfacial thin films; 3 is the substrate.

faces of the sample. Both the first and second effects are understood in the context of
simple optical phenomena. The third and fourth features are the magnitude and shape of
the envelope function of the SHG data. These quantities are related to the second order

nonlinear susceptibility, x0), and the orientation and angular distribution of the dominant

x0) tensor element(s) for the sample under consideration.

Optical eflects. The angular dependence of the SHG signal we recover is
oscillatory with respect to the angle of incidence of the fundamental electric field (Figure
54-56

4.2). This pattern appears to be outwardly similar to that produced by Maker fringes,

but it cannot be accounted for using Maker’s treatment because the coherence length of

-60-

70

 

 

 

 

 

 

 

 

 

 

 

 

 

O
"‘ \O
O
"‘ V3
'l d.)
a
O d.)
--—t
0
8:1
- I—t
Q—4
O O
23‘ M 2
in
O
" N
O
— 1—1
0 00 \O V N C
—t O O Q 0 O
O C O O O o

(W) l<11911911119HSS

Figure 4.2 Surface SHG for a clean fused-silica substrate vs. angle of sample rotation
(magenta) (0° is the condition where the incident electric field is propagates along the
surface normal axis) plotted with the model (green) for three components of the etalon
effect.

-61-

the light source we use is long relative to the thickness of the portion of the sample that
generates the second harmonic signal. The origin of the signal shown in Figure 4.2 is
interference between the second harmonic light generated at the front and back interfaces
of the transparent substrateso'sz'53 For the model system shown in Figure 4.1, we can
decompose the induced polarization at 20) into the components that arise from nonlinear

interactions at each interface ( layers 2 and 4, Figure 4.1),57’58

Ez‘” = E,” + E42“ Equation 4.1
For E2” generated at each interface, we must account for the angular dependence of

reflection and transmission, dispersion in each of the media and the thickness of the

interface layer.

47rx(2)

 

E2“ = E‘” 2 TWR‘” 2T2° 2“"T2°"T2“’ x
2 ( )(12 2) 23 2 34 45 (n2)2_(n}2‘°)2 Equation 4.2

. . 20) w
exp(zq’)f2 )exp[z€f Thu}; cost2 — 17%;” costf)—l

47rx(2)

 

E2“) = E” 2 TwaTwTwa 27’2sz“)
4 l l l '2 2 23 34 4) 45 4 (n74)2_(n}j’)2 x Equation 4.3

. . 2w 2 2
ex 1 ex 11? —-— “’cosOw— ”c056“ —1
FUN.) P[ f c ]("A f4 '71. f.)
where If is the film thickness, the terms 11 are the refractive indices of the films 2 and 4 at
the frequencies indicated in the superscripts, the angles 6 are the angles of refraction for

the layer at the frequencies indicated, and the terms (1) are the phase angles of the light at

200 relative to the fundamental at each interface. The R and T terms are the Fresnel

-62-

transmittance and reflectance coefficients for a TM-polarized electric field at the

interfaces indicated.59

 

 

TM _ 277m cos6,,, TTM _ 77,, c036,, TM 2
tmn _ mn — tmn
77,, cos 6,,, + 77,, cos 6,, 77,, cos 6,,,
Equation 4.4
TM _ 77” 6059”; — 77m 0089,, RTM _ TM 2
rmn _ mn — rmn
17,, cos 6,,, + 17", cos 6,,

The second harmonic light generated at interfaces 2 and 4 is initially in phase with the
fundamental electric field at the same interfaces. The term that dominates the observed
angular-dependence of the signal is the phase mismatch between the second harmonic
light generated at interfaces 2 and 4 resulting from dispersion in the substrate. This phase

relationship is given by52
13¢ = (dfz — (pf, ) = 23d (7752“ cos 632‘” — 17:." cos 6;”) Equation 4.5
c

where d is the substrate thickness. The oscillatory nature of the signal can be accounted
for quantitatively using Eq. 4.5. With this information, I2‘0 = (Em)2 can be calculated as a
function of the incidence angle of the fundamental electric field. We show the
experimental second order response of a bare fused silica substrate in Figure 4.2 along
with the calculated signal based on Eqs. 4.1-4.5. We note the presence of several
different oscillating components and account for the presence of these components by
considering that the substrate will act as an etalon. This is an expected phenomenon.
The incident electric field makes multiple passes in the substrate, with the relative
contribution from each odd-numbered pass (for a transmission measurement) depending
on the angle of incidence in a manner dependent on the flatness of the SiOx substrate and

the Fresnel factors. The data shown in Figure 4.2 provide important insight into the

-63-

dominant contributions to the xa) response for this system, and we will return to a
discussion of this point after considering the chemical contributions to the nonlinear
response.

Chemical effects. The chemical properties of the system that we sense with
surface SHG measurements are the relative magnitudes of the several contributions to the
first hyperpolarizability of the system and the distribution of orientations of the x‘2)-active

species. Recent work by Simpson and Rowlen has treated the issue of x0)

-chromophore
orientation in surface second harmonic generation experiments.”62 Among the
important findings of their work is that the chromophore tilt angle recovered
experimentally converges to a “magic angle” of 392° with respect to the surface normal
as the orientational distribution broadens. The actual treatment of the distribution can be
complex, depending on its functional form, and it is typically assumed that the
distribution is relatively narrow. In this limit, the dependence of the SHG response on
chromophore tilt angle will scale with cosz<0>, where <0> is the average angle between
the incident electric field polarization vector and the chromophore nonlinear transition
moment. For a molecule with its transition moment oriented along the surface normal,
the maximum signal will occur when the incident electric field polarization vector is
aligned with the substrate normal. The minimum signal for this condition will occur for
the electric field propagating along the surface normal axis. The orientation-dependence
of our data is manifested most prominently in the envelope function of the experimental
data, as schematized in Figure 5.3a. It is clear from these calculations, based on Eqs. 4.1

and 4.4 and neglecting the Optical interference effects, that the nonlinear chromOphore

orientation will have a significant effect on the form of our signal. In order to understand

-64-

 

LW tilt=0°

SHG intensity (a.u.)

 

 

 

 

 

_2 1 r . r .4 1 r . r . r . r
0 10 20 30 40 50 60 70
angle of incidence

10 F

2?

3

.§‘

.5

CD

I

U)
- 10 distribution width

0 1 l 1 l 1 1 1 I 1 1 1 1 1 l
0 10 20 30 40 50 60 7O

angle of incidence

Figure 4.3 (a) Calculated SSHG envelope function for a series of chromophore tilt
angles, as indicated in the plot. (b) Dependence of the envelope function on orientational
distribution width. The distribution function is assumed to be Gaussian and the
chromophore average orientation is along the surface normal axis.

-65-

the limits inherent to the interpretation of our data, we need also to consider the effect of
the chromophore orientational distribution width on the envelope function. We show the
effect of increasing orientational distribution width on the calculated envelope function in
Figure 4.3b. This calculation is for a chromophore orientational distribution centered
around the substrate normal. The broadening of the function near 0° tilt angle is
reflective of the change in orientational distribution while the invariance of the signal
near 70° demonstrates that, at high angles, the SSHG signal is dominated by optical
considerations as described in Eqs. 4.4. Based on these calculations, it is clear that the
unambiguous resolution of chromophore tilt angle and distribution width is limited by the
S/N ratio of our data.

In the interpretation of many second harmonic generation measurements, it is
common practice to assume that the electric dipole term in )6” is much larger than the
electric quadrupole and higher order terms. While this approximation holds for many
experimental conditions, it is not universally the case. For the data we report here, the
electric quadrupole term plays a significant role in determining the form of the
experimental signal. Guyot-Sionnest and Shen have investigated the contributions of
dipolar and quadrupolar terms to the second order nonlinear susceptibility of surfaces and

63 Their work focused on the distinct contributions from the structural and

interfaces.
electric field discontinuities that exist where the incident electric field propagates across
the interface. In their model, the electric dipole contribution to the xm response is
determined by the structural properties of the interface and they term it a local, intrinsic

response. The electric quadrupole term results from the discontinuity experienced by the

electric field as it propagates through the interface, and it is termed a nonlocal response

-66-

because it depends explicitly on the bulk properties of the two phases of matter
comprising the interface. Using Guyot-Sionnest and Shen’s terminology, the second

order nonlinear susceptibility for an interface is given by63

2’12) =01 [i (Z)S(Z)-§-ll?m(2)5(2)l+1,1.) (Z)§-S(Z)]dz Equation 4.6
yzy 1 W 62 , , ,y‘y 62 _

where s(z) = 1/81, i=l,2, and e is the optical frequency dielectric constant. Before the
electric field is incident upon the interface (2 < 0), s(z)=1/el and after the interface (2 >
0), s(z)=l/ez. Between these two limits, s(z) varies continuously from 1/81 -* 1/82
through the interface. )6”, is the total second order susceptibility and the integration is
over the thickness on the interfacial region. The first term in Eq. 4.6 is the electric dipole
term associated with the structural properties of the interface, the second is the nonlocal
term associated with the induced electric quadrupole moment as the electric field
propagates through the interfacial dielectric gradient, and the third term arises from the
magnetic dipole moment. The theory is based on the simplest case; an interface between
two transparent, nonmagnetic, isotropic media, 1 and 2. For these conditions, the first
two terms in Eq. 4.6 will contribute most significantly to the observed xms.

Because the interface electric dipole contribution to the nonlinear susceptibility is
an intrinsic property of the material, it can be assumed constant over a range of
experimental conditions. The discontinuity in the electric field as it propagates through
the interface will depend on the difference in a between the two bulk media and this
quantity can be varied systematically. Guyot-Sionnest and Shen performed a series of
surface SHG measurements using a fused silica substrate immersed in bulk liquids with

3

varying dielectric constants.6 In this way, the term 6/6z(s(z)) could be made small,

-67-

revealing the role of the electric dipole term, x0, for this substrate. They determined
from those experiments that the susceptibility of the air-fused silica interface is x“), ~
2.7x10'” esu/cmz, with xD ~ 5.7x10'l8 esu/cmz.63 Although they did not explicitly
perform the subtraction owing to the presence of the magnetic dipole term in Eq. 4.6, we
estimate from their findings is that x0 ~ 2.1x10’l7 esu/cmz.

With the theoretical framework established by Guyot-Sionnest and Shen in place,
we can see from the data in Figure 4.2 that the envelope function points to a signal
maximum at normal incidence. For the fused silica substrate, the electric quadrupole
term is dominant and the envelope function maximum near normal incidence suggests
that the quadrupolar component lies in the plane of the substrate, and not perpendicular to
it. We expect the electric dipole contribution to x0) to be aligned with the surface
normal. This finding can be understood in the context of the greatest electric field
gradient at the interface being parallel to the interface. Thus a discontinuity in the
incident electric field at the interface will be most pronounced in the plane of the
interface. We can estimate the resulting “tilt” angle for the bare substrate if we assume
that 20% of the total x0) response is from the electric dipole term, normal to the surface
and 80% of the response is from the electric quadrupole term, in the surface plane. The
“tilt” angle recovered experimentally should be ~72° with respect to the surface normal
for fused silica. We extract a best fit tilt angle of 63° from our experimental data. As
discussed above, the extraction of tilt angle and orientational distribution information
from our data is S/N limited, so to within the accuracy of these determinations, we view

the agreement as good.

-68-

ram)"2 (a.u.)

SSHG response (a.u.)

 

1.0 L
0.8 _
0.6
0.4 P
0.2

 

I‘ll-l

F.-+—-.—-1

I l I 1 I

2 3 £1 5
nunber of layers

p—s

 

 

 

 

 

 

 

Ar; 0 p
/\ .
m 9; '1 r
.. I 1. o é
C: 1‘. S" "’ l .:
'- V;‘ (‘1 J I i. .1
'.‘- l I. l 1! i
P ill 1. O (i O
9.22., .1 .,
i : 1" 1‘. t
‘ ,- 1 l > z .1
_»1 . ,1 (1“,) . :1 k; ‘ , i
o ‘ \O ‘ l “
_. (l o t 12“ ll 1" ’1 a"
c- .".O ., .1 l\ a .11
j'». I p i 9 F ”i. .4 . " 7’
, gym}? I f“ if; 0,1 i. ’9' .1 l e
I) 1. r5; v'.|‘ ,- i . f
0 CC’ ‘3 ,‘l, 0', O’?\ .1331 ' '.
,_ T) K'- t t l ,
.- .~ ~ 10. I. 1 l <
’- w“ '01 - .21.! MI ' 0’
o o ’% o 7 "a" C? 7 \l "O 3““ 1. l 0
90> ‘- .’ c.» 9 \O. I. t o. 1“ V ‘9" ¢ 3?
’ 7“ ‘ (x! . 1 03 0 gr, ~. 0 a o
O Ottzrrmg'C‘IIf-KQ’ )4 1' 1 ‘1‘ ' L" 1 1 , l
' 20 3O 4O 50 60

0

 

sample rotation angle

Figure 4.4 (a) Plot of the [SHG Intensity]”2 vs. the number of chromophore layers. (b)
Angle-dependent SSHG signals from chromophores l (0) and 2 (0). These data are
dominated by substrate response at small angles and chromophore response at high

angles.

-69-

We consider next the nonlinear optical properties of multilayers of each of the
chromophores. Katz et al. has reported previously on the second order nonlinear response

of chromophore 1.33

Our data are consistent with his and we observe a square-law
dependence of the second harmonic intensity with increasing number Of layers for both
chromophores. This is an expected result for a system where the nonlinear medium is
significantly thinner than the coherence length Of the incident light source. The data we
present in Figure 4.4a demonstrate this relationship and show that the system is
sufficiently ordered to allow the square-law relationship to be manifest for both
chromophores. We estimate xm = 4.0 x 10"6 esu/cmz-layer for 1 based on the
experimental signal relative to that of the quartz substrate. Katz et al. reported a value Of
B = 150 x 10'30 esu for a 4-layer stack of 1 characterized by an order parameter Of 0.2.33

'2 45 and the same layer order, our data

Assuming a layer density of 1.6 x 1014 cm
correspond to a value of 200 x10'30 esu.

There are many possible molecular and bulk contributions to the nonlinear
response of the interfaces we study here and, based on the subtle differences in the linear
response of 1 and 2, it is possible that the magnitude of their xm responses could be
different. From the data shown in Figure 4.4a, it is clear that the magnitudes of the xm
responses for chromophores 1 and 2 are the same to within the experimental uncertainty.
While the magnitudes are equal, the sign of the nonlinear response for these two
chromophores is Opposite owing to their complementary orientations, and this condition
is apparent in the phase relationship of the angle-dependent SSHG data (Figure 4.4b).52

The experimental angle-dependent SSHG data for each chromophore can be

modeled using Eqs. 4.1 — 4.5. We present the correspondence between the calculated

-70-

 

 

 

   

 
     
 

’7
=1
CC
V
I—
<6
Eh
(I)
(D
I
m
V)
D. -
(U
V
_‘ :—
CV3
3
OD
. 1—1
W _.
(D
m
m _ I,
m '
4'34 trfifififiymz? ,5; t:

 

 

01050‘30‘40‘50
samplerotation angle

Figure 4.5 (a) Comparison of experimental SSHG data to the model presented in Eqs.
4.1-4.5 for chromophore l. (b) Comparison of experimental data with model presented in

Eqs. 4.1-4.5 for chromophore 2.

-71-

response and the experimental response for a single layer of chromophore l in Figure
4.5a and for a single layer of chromophore 2 in Figure 4.5b. For both chromophores, the
agreement between experiment and model is reasonable. For small sample rotation
angles, the signal has a measurable contribution from the SiO" substrate while at higher
rotation angles the chromophore response dominates. For the chromophores, it is likely
that the electric dipole contribution to their )6” response is dominant, owing to their
structures and the fact that their response as a function of angle shows directly the
orientation dependence. As discussed above, the manner in which we acquire SSHG data
is not amenable to precise orientation angle or distribution determinations, but it is clear
from the experimental signals that the largest x0) susceptibility terms for chromophore
layers lie close to the substrate normal.

We now turn to the issue of assessing the nonlinear response of multilayer
assemblies that contain both chromophores. For a bilayer system, there are two possible
structural permutations; [SiOx-l-Z] and [SiOx-2-l]. Before comparing their nonlinear
responses, it is important to ensure that the formation of these bilayers does not give rise
to anomalous changes in their linear response. The absorption spectra of both bilayers
are identical to the linear responses of the individual chromophores. We present the )6”
data for these two bilayer structures in Figures 4.6a and 4.6b. The data include )6”
responses of the bare substrate before deposition for comparison. The most striking
feature of these data is that the responses of the two complete bilayers are not the same.
This result is reproducible and cannot be accounted for in the context of simple additivity
of the constituent electric dipole contributions. If electric dipole or any other

contribution(s) intrinsic to the chromophores and the substrate accounted for the overall

-72-

SSHG signal (a.u.)

SSHG signal (a.u.)

 

 

 

 

0.8 —
0.7 - a
0.6 —
’ 2
0.5 —
0.4 -
0.3 : 1
0.2 ~
0.1 . p
v ' fl :‘s‘.:'~1a)‘| 1... 9..
0.0 Ligr'ff :11” 971331“ ‘1': A?” 1
Sample Rotation Angle (deg.)
0.8 - b
0.7 —
' 1
0.6 —
0.5 —
’ 2
0.4 —
0.3 ~ "m ,
0.2 —
01 — fl. "0 31“}.0‘0;'Oo
0.0‘Kfiv’ 5' s“.- '-".°v "o'- “7°." hr fiw"\

Sample Rotation Angle (deg.)

Figure 4.6 Surface SHG of (a) [SiOx-l-Z] and (b) [SiOx-Z-l] bilayers showing the
cancellation of SHG signal with the adsorption of the second layer. The orange traces are
data on the bilayer structures and the black circles are SSHG data for the bare SiOx
substrates.

-73-

x0) response, the two bilayer structures would necessarily produce identical nonlinear

responses, with any difference between the bilayer and bare substrate responses being
attributable to vacancy and/or orientational defects. There may be some hint of this
effect in the data in Figure 4.6a for high rotation angles, but this issue remains under
investigation. We note that the cancellation we observe in these data demonstrates the
feasibility of our approach to x0") background nulling. Because the interlayer linking
chemistry is the same for both bilayer structural permutations and the formation constant
for ZP materials is characteristically so large, there is no reason to expect a difference for
the bilayer responses based on differences in the efficiency Of layer formation.

The fact that the bilayer data in Figures 4.6 are not identical demonstrates the
importance Of subtle structural contributions to the overall nonlinear response of the
system. The only structural difference between the two bilayers lies in the region near
where the two layers are connected through ZP linkages. It is known that the zirconium
bisphosphonate solid state structure possesses a center of inversion about the metal ion,
ruling out an electric dipole allowed contribution to the signal from this moiety. At first
glance, one could invoke the phosphate/phosphonate asymmetry to account for a xa)
response, but a more detailed examination of the bilayer structure reveals that this
asymmetry is oriented in the same direction for both bilayer structures. The only
structural difference between the two bilayers is the polarizability, and thus the
hyperpolarizability, of the organic functionalities attached to the phosphate and
phosphonate moieties. For [SiOx-l-Z], the interlayer 1-2 connection is Of the form
indicated in Figure 4.7a and for [SiOx-Z-l] the analogous structure is shown in Figure

4.7b. As indicated above, the net dipolar contributions should be the same for both

-74-

EN]
2 (‘2 1

I

I

02 I

I -------- I |
I , I
I I I

P.

i O/ \00 ' I
Q2: 0 I |

1 O\‘ / I l
I f , I
' 0 I |
I | |
I

I

I

 

[N]
N
-..P\ P

 

 

ATV I
.a'“

A, _ ‘U .. l- 0., ., . watt, 0, 50.2w,
I ~ Q ‘* / SrOx substrate E5--- 0 7

Figure 4. 7 (a) Structure of [SiOx-l-Z] bilayer. (b) Structure of [SiOx-Z-l] bilayer. Note
the structural difference between the interlayer linkages for the two assemblies (boxed).

-75-

structures, but the quadrupole moment for [SiOx-l-Z] should be substantially less than
that for [SiOx-Z-l]. The electric quadrupole moment for [SiOx-Z-l] should be oriented
along the chromophore tilt axis and both of these structurally-based predictions are
consistent with the experimental data. Unfortunately, there is no reliable means to
estimate the magnitude of the nonlinear responses associated with these interlayer
connecting structures. The central point is, however, that the nonlinear response of these
bilayers cannot be accounted for simply by adding the dipolar contributions of the

component parts.

4.4 Conclusions

We have characterized Zr-phosphate/phosphonate (ZP) self-assembled multilayer
structures using surface second harmonic generation measurements. We have used two
Xm-active chromophores with complementary structures and have assembled bilayers
with controlled orientation relative to the substrate. The nonlinear optical responses of
the SiOx substrate and of multilayers of each of the chromophores have been
characterized. For the substrate we find substantial contributions to the nonlinear
response from the electric quadrupole contributions to the total x0) term, in agreement

with the work of Guyot-Sionnest and Shen.63

For each chromophore, multilayer
structures provide the expected dependence of second harmonic signal on number of
layers and, based on the magnitudes of these signals, the electric dipole term likely
dominates the chromophore x0) responses. Using these same chromophores, we have

formed bilayers to produce two different, canceling structural motifs, each with a local

center of inversion about the ZP interlayer bonding plane. The xlz) responses of the two

-76-

bilayer systems are measurably different, revealing the limitations of accounting for
nonlinear optical responses simply in terms of additive contributions from the
constituents. The differences in the nonlinear responses of the two systems can be
accounted for through cancellation of the electric dipole contribution to )6”, with the
residual difference arising from the electric quadrupole contribution. This higher order
response is associated with the region centered around the interlayer linking group. Our
data underscore the complex and non-additive issues associated with the design and
construction of layered interfaces and, in principle, provide a means to measure the
vacancy defect density in layered materials. Clearly the chromophores we have used in
this report are not ideal candidates for vacancy defect density measurements, but simple
synthetic means can be used to improve their utility for this application. These data also
raise the issue of the thickness of interface that SSHG measurements are sensitive to, and

we anticipate future work in this area to shed new light on this matter.

-77-

4.5 Literature Cited

1. Katz, H. B.; 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.

5. Calvert, J. M.; Georger, J. H., Jr.; Peckerer, M. C.; Perhsson, P. B.; Schnur, J. M.;
Scheon, P.; B.; 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, I.; Vansant, E. F.; Grobet, P.;
J. Chem. Soc, Faraday Trans. 1992, 88, 3197.

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

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

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

12. Xia, Y.; Whitesides, G. M.; Angew. Chem. Int. Ed. 1998, 3 7, 551.

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

-78-

14

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

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

Kohli, P.; Taylor, K. K.; Harris, J. J.; Blanchard, G. J .; J. Am. Chem. Soc., 1998, 120,
11962.

Gun, J .; Iscovici, R.; Sagiv, J .; J. Colloid Interface Sci 1984, 101, 201.

Maoz, R.; Sagiv, J .; J. Colloid Interface Sci. 1984, 100, 465.

Maoz, R.; Sagiv, J .; Thin Solid Films 1985, 132, 135.

Hong, H.-G.; Sackett, D. D.; Mallouk, T. B.; Chem. Mater. 1991, 3, 521.
Thompson, M. B.; Chem. Mater. 1994, 6, 1168.

Katz, H. B.; Wilson, W. L.; Scheller, G.; J. Am. Chem. Soc. 1994, 116, 6636.
Yonemoto, E. H.; Saupe, G. B.; Schmehl, R. H.; Hubig, S. M.; Riley, R. L.; Iverson,
B. L.; Mallouk, T. B.; J. Am. Chem. Soc. 1994, 116, 4786.

Katz, H. B.; 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. B.; J. Am. Chem. Soc. 1993, 115, 11855.

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

Ungashe, S. B.; Wilson, W. L.; Katz, H. B.; 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.

-79-

29

3O

31

32

33

34

35

36

37

38

39

40

41

. Katz, H. B.; Schilling, M. L.; Chidsey, C. E. D.; Putvinski, T. M.; Hutton, R. 8.;
Chem. Mater. 1991, 3, 699.

. Katz, H. B.; 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. B.; 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. B.; Mallouk, T. B.;
Coord. Chem. Rev. 1990, 97, 237.

. Lee, H.; Kepley, L. J.; Hong, H.-G.; Akhter, S.; Mallouk, T. B.; 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. B.; Chem. Mater. 1994, 6, 2227.

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

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

. Sabatini, B.; Rubinstein, 1.; J. Phys. Chem, 1987, 91, 6663.

. Sabatini, B.; Rubinstein, 1.; Moaz, R.; Sagiv, J.; J. Electroanal. Chem, 1987, 219,
365. l

. Finklea, H. 0.; Snider, D. A.; Fedyk, J.; Sabatini, B.; Gafni, Y.; Rubinstein, I.;
Langmuir, 1993, 9, 3660.

. Finklea, H. 0.; Snider, D. A.; Fedyk, J.; Langmuir, 1990, 6, 371.

-80-

42. Sabatini, B.; Cohen-Boulakia, J .; Bruening, M. L.; Rubinstein, 1.; Langmuir, 1993, 9,
2974.

43. Janek, R. P.; Fawcett, W. R.; Ulman, A.; Langmuir, 1998, I4, 3011.

44. Shen, Y. R.; The Principles of Nonlinear Optics, John Wiley and Sons, Inc., 1984.

45. Flory, W. C.; Mehrens, S. M.; Blanchard, G. J .; J. Am. Chem. Soc., preceding article.

46. Heinz, T. F .; Tom, H. W. K.; Shen, Y. R.; Phys. Rev. A., 1983, 28, 1883.

47. Bloembergen, N.; Chang, R. K.; Jha, S. S.; Lee, C. H.; Phys. Rev, 1968, 174, 813.

48. Naujok, R. R.; Higgins, D. A.; Hanken, D. G.; Corn, R. M.; J. Chem. Soc. Faraday
Trans, 1995, 91, 1411.

49. Daschbach, J. L.; Fischer, P. R.; Gragson, D. E.; Demarest, D.; Richmond, G. L.; J.

Phys. Chem, 1995, 99, 10690.

50. Lupo, D.; Prass, W.; Scheunemann, U.; Laschewsky, A.; Ringsdorf, H.; Ledoux, 1.;
J. Opt. Soc. Am. B, 1988, 5, 300.

51. Berkovic, G.; Shen, Y. R.; Marowsky, G.; Steinhoff, R.; J. Opt. Soc. Am. B, 1989, 6,
205.

52. Sato, 0.; Baba, R.; Hashimoto, K.; Fujishima; A.; J. Electroanal. Chem, 1991, 306,
291.

53. Kajikawa, K.; Takezoe, H.; Fukuda, A.; Chem. Phys. Lett., 1993, 205, 225.

54. Herman, W.; Hayden, L. M.; J. Opt. Soc. Am. B, 1995, 12, 416.

55. Maker, P. D.; Terhune, R. W.; Nisenhoff, M.; Savage, C. M.; Phys. Rev. Lett., 1962,
8, 21.

56. Jerphagnon, J.; Kurtz, S. K.; J. Appl. Phys, 1970, 41, 1667.

-31-

57. Bloembergen, N.; Pershan, P. 8.; Phys. Rev, 1962, 128, 606.
58. Kajzar, F.; Messier, J.; Zyss, J.; Ledoux, 1.; Opt. Comm, 1983, 45, 133.

59. Pedrotti, F. L.; Pedrotti, L. 8.; Introduction to Optics, Prentice-Hall, Inc., 1987, pp.

 

472-487.
60. Simpson, G. J.; Rowlen, K. L.; J. Am. Chem. Soc., 1999, 121, 2635.
61. Simpson, G. J.; Rowlen, K. L.; J. Phys. Chem. B, 1999, 103, 1525.
62. Simpson, G. J.; Rowlen, K. L.; J. Phys. Chem. B, 1999, 103, 3811.

63. Guyot-Sionnest, P.; Shen, Y. R.; Phys. Rev. B., 1987, 35, 4420.

-32-

Chapter 5

UNDERSTANDING THE SPATIAL DEPENDENCE OF THE NONLINEAR
RESPONSE USING SELF-ASSEMBLED MULTILAYERS

Summary

We report on the structure-dependence of second harmonic generation (SHG) in Zr-
phosphate/phosphonate (ZP) self-assembled multilayer structures using two Xm-active
chromophores separated by xl2)-inactive layers. The chromophores are identical save for

their terminal functional groups, allowing for control over nonlinear chromophore

orientation with respect to the surface to produce bilayer structures with inversion
symmetry. We have shown previously that the x9) response of such structures can cancel
if the interlayer linking functionality does not produce an explicit x0) response. We
report here on the dependence of the x0) response of structurally opposing multilayers
spaced apart by pre-determined distances using centrosymmetric alkanebisphosphonates.
Our data show that the nonlinear response of multilayer stacks, even when each layer is

of molecular dimension, is more complex than that of a single Xm-active layer.

-33-

5.1 Introduction

Self-assembled monolayer and multilayer (SAM) thin films continue to receive a
great deal of attention in the materials community because of their potential relevance to

27 The reason for

a variety of applications, including nonlinear optical device design."
this interest is that SAMs are easy to synthesize, require mild conditions for formation,
and are structurally and thermally robust once formed. In addition, it is possible to grow
highly oriented multilayer assemblies, allowing for the possibility of efficient second
harmonic generation and other phenomena that depend on )6”. For SAMs to be used in
applications that are sensitive to surface conditions, the nature of structural defects
inherent in layer synthesis needs to be understood.

28'“ There is a rich literature on

The interest in defect determination is not new
the characterization of defects in self-assembled alkanethiol monolayers on gold. In that
work, the primary means of defect characterization is electrochemical.”37 For multilayer
films and films deposited on non-conducting substrates, electrochemistry is not a viable
technique and we are left to devise alternative means for the examination of defects.
Whitesides has used chemical modifications to detect vacancy defects.38 Beebe and
coworkers use “molecular corrals” to elucidate information defects from surface
structures.39 A major goal of our work has been to develop ways to quantify defects on a
molecular scale in multilayer films. Our approach is to utilize surface second harmonic
generation (SSHG) measurements in conjunction with oriented layer growth. For the xm
chromophores we use, the functional chemistry of their termini allow for orientational

control in the growth of these films. By constructing films with specific orientation

relative to the substrate in each layer, a bilayer film can be assembled such that a center

-34-

of symmetry is enforced across the interlayer linking functionality. This structural
arrangement effectively nulls the second order nonlinear response from portions of the
assembly where the bilayer structure exists and, in principle, the recovered SSHG signal
exists as a consequence of vacancy or orientational defects in the bilayer structure. We
have demonstrated this effect in the previous chapter.

The chromophore used for our SSHG measurements, (4-(4-(4-(4-((2-
hydroxyethyl)sulfonyl)-phenyl)azo)phenyl)piperazinyl)phenyl)phosphonic acid, 1, was
developed by Katz et al. for its rigid structure and large nonlinear response.40 We have
detailed the synthesis and characterization of the complementary chromophore, 2, (2-(4-
(4-(4-(4-hydroxyphenyl)piperazinyl)phenyl)azophenyl)sulfonyl)-ethylphosphonic acid in
Chapter 2.41 The molecular structure of these chromophores is shown in Figure 1.1. The
similar absorption and layer-forming characteristics of both chromophores make them
ideal candidates for SSHG investigations. The results presented in Chapter 4 have shown
that the SSHG responses of bilayer structures of l and 2 grown on fused silica are
decreased substantially due to symmetry about the ZP linkage with the residual x0)
response arising from the structure of the interlayer linking moiety.42 For bilayers of the
form [SiOx-l-Z], complete cancellation was observed while for bilayers [SiOx-Z-l] there
was a residual SSHG signal. The origin of this residual signal is the formation of a
polarizable interlayer linking moiety that is absent in the [SiOx-l-Z] layer structure.
Subsequent to that study we have investigated the dependence of the SSHG signal on the
structure of the multilayer assembly, and it is that work we report in this chapter.

The x”) cancellation effects we see for relatively simple bilayer structures are not

obtained for structurally more complex systems. Despite the fact that each layer in these

-35-

interfacial assemblies is of molecular dimension, achieving an understanding of the
experimental xlz) response requires understanding of the structural complexity of the
interface. The focus of this chapter is on reporting and understanding the experimental

SSHG response for several multilayer assemblies.

5.2 Experimental

Chemicals. All chemicals used were obtained in the highest purity grade
available and were used as received except for triisopropyl phosphite, which was purified
by drying over sodium under an inert atmosphere, followed by vacuum distillation.43

Synthesis. 1,12-dodecanediylbisphosphonic acid44 20.0 mL of purified
triisopropyl phosphite (81.1 mmol) was added to 11.41 g of 1,12-dibromododecane (34.8
mmol) under Ar. The solution was refluxed for 18 hours at 150 °C, then cooled, and
excess triisopropyl phosphite and isopropyl bromide were removed by vacuum
distillation. 50 mL 121_VI_ HCl was added to the vessel and the solution was refluxed for 5
hours at 150 °C. Upon cooling, crystals precipitated from solution and were collected by
vacuum filtration. The crystals were dissolved in ethanol and recrystallized. The product
was collected by vacuum filtration and dried, yielding 9.45 g (82%) of white product.
Pure by 'H NMR.

Chromophore Deposition. The chromophores were synthesized according to the
literature procedure (1)40 or as described in Chapter 2 of this thesis (2), and were
deposited on zirconated substrates from a saturated solution (~ 0.2 m1\_/I) of the

appropriate chromophore dissolved in 1:4 DMFzEtOH. Deposition of each layer was

from a solution containing either 1 or 2, but not both. The temperature of each deposition

-86-

solution was maintained slightly below boiling while the zirconated substrates were
immersed for 10 minutes, followed by rinsing with warm ethanol and drying with N2.
Subsequent phosphorylation and zirconation of the chromophore-containing surfaces was
performed as described above. Before measurement of adlayers for x0) activity, terminal
OH groups were phosphorylated. To monitor layer adsorption, the UV-visible
absorbance spectrum of the sample was measured after each deposition cycle, and in all
cases linear growth was observed.

Spacer Layer Deposition. The alkyl bisphosphonate was deposited on zirconated
substrates from a ~61mM solution using 1:4 DMFzEtOH as before. The temperature of
the deposition solution was maintained slightly below boiling while the zirconated
substrates were immersed for 10 minutes, followed by rinsing with warm ethanol and
drying with N2. Zirconation of the surfaces was performed as described above.
Measurements. 1H NMR spectra were taken with a Varian Gemini-300 MHz NMR
spectrometer. Optical absorption measurements were made using a Hitachi U-4001 UV-
visible absorption spectrometer.

Surface Second Harmonic Generation Laser System. The laser system used for

SSHG has been described in detail in Chapter 3.

5.3 Results and Discussion

The purpose of this chapter is to investigate the surface SHG response of
multilayer assemblies greater structural complexity than those reported in the Chapter 4.
In that work, structural cancellation of the surface SHG signal was demonstrated for a

specific molecular configuration. The SSHG signal could be treated and understood in

-37-

(2)

the context of a single thin layer of 1 -active material at the interface between two bulk ,

Xm-inactive materials. We demonstrated that the chemical identity of the xlz)-active
interface layer can play a critical role in determining the form and magnitude of the
experimental signal, as expected. In this work we have investigated structurally more
complex interfaces and report that their SSHG responses cannot be understood in the
same context as those reported in the previous chapter. The initial intention of this work
was to establish the persistent length of layer organization by spacing the chromophores l
and 2 away from one another and the substrate by varying distances using Xm-inactive
alkylbisphosphonate layers. We will first consider the experimental data and then discuss
the methods by which it can be interpreted.

We have investigated six structural permutations of multilayer assemblies and
have recorded the SSHG response of each layer at each step in the assembly procedure.
The structures we have investigated are shown schematically in Figures 5.1, 5.3, and 5.5
and their integrated SSHG responses are reported in Table 5.1. The corresponding
angular dependent experimental data are shown in Figures 5.2, 5.4, and 5.6. For these
measurements there is no explicit attempt to separate quadrupolar from dipolar
contributions to the experimental signal. There are several unexpected results contained
in these data. The first is that the structurally-based SSHG cancellation seen for bilayers
of chromophores 1 and 2 is not seen for multilayer assemblies containing more than two
molecular layers. The second unexpected finding is that the magnitude of the SSHG
response for a given chromophore is found to depend on its spacing from the dielectric
substrate. There are two qualitative approaches to the interpretation of these data and we

outline them below.

-33-

Table 5.1. Experimental susceptibility values extracted from our data.

 

 

 

 

 

 

 

 

esuxlO‘” SiOx- 132 SiOx- lSSZ SiOx- 281 SiOx- ZSSI SiOx- SSlZ SiOx- SSZl
SiOx 2.70 2.70 2.70 2.70 2.70 2.70
Layer 1 12.7 29.1 16.8 32.0 0.39 3.33
Layer 2 11.2 26.2 2.15 2.70 2.72 0.99
Layer 3 206 73.8 197 1.50 26.4 4.44
Layer 4 - 252 - 204 77.2 28.9
cap 23.8 166

 

 

 

 

 

 

 

 

In their work on understanding SSHG experiments, Guyot-Sionnest and Shen
demonstrated that the SSHG response of a blank fused silica substrate contained
contributions from both structural discontinuities and from variations in the dielectric
response of the media at the interface which led to nonlinearities in the electric field as it
traversed the interface.45 They expressed their findings in an equation that takes into

account both dipolar and quadrupolar contributions to the experimental signal.

11” = [ Ix” (2)s(2) - 311.9. (2)3(2)] + 1.? m (2)3s(z)]dz Equation 5.1
”Y I W)" 62 y'y ., y 62

where s(z) = Na (i=1,2, e = the optical frequency dielectric constant). Before the electric
field is incident upon the interface (2 < 0), S(Z)=l/81 and after the interface (2 > O),
S(Z)=l/82. Between these two limits, s(z) varies continuously from 1/81 -* US; through
the interface. xms is the total second order susceptibility and the integration is over the

thickness on the interfacial region. The first term in Eq. 5.1 is the electric dipole term

associated with the structural properties of the interface, the second is the nonlocal term

-39-

 

 

associated with the induced electric quadrupole moment as the electric field propagates
through the interfacial dielectric gradient, and the third term arises from the magnetic
dipole moment. The theory is based on the simplest case; an interface between two
transparent, nonmagnetic, isotropic media, 1 and 2.

An assumption inherent in this treatment is that the xl2)-active layer is
characterized by a single (complex) dielectric constant, and the nonlinearity experienced
by the electric field is the result of a continuous or simple step-wise variation in the
dielectric response of the medium from one boundary condition to the other. Indeed, this
treatment is implicit in essentially all of the current explanations of SSHG phenomena
and, in Eq. 5.1, the term s(z) describes this variation in the dielectric across the interface.
There are thus two ways to treat the structurally more complex systems we report on
here. We can either consider the function s(z) to be a non-monotonic function across the
interface or we can consider these systems to be multilayer stacks, with each layer
possessing its own s(z) functionality. The difference between these two approaches lies
in the treatment of the Fresnel factors at each interface. If we consider the xm active
portion of the interface to be a single layer with a complicated s(z) profile, there is no
need to explicitly consider the Fresnel factors at each layer. For a multilayer assembly,
however, the Fresnel factors come into play at each interface.

The fundamental difference between these two approaches lies in the fact that the
explicit multilayer treatment considers the )6” response of the interface to be a simple
linear superposition of the nonlinear response of each layer, with the Fresnel factors for

each interface determining the transmission of the fundamental and second harmonic

-90-

23.8 166

   

 

 

 

 

 

 

 

 

P P

206 197
2 1

11.2 2.15
S S

12.6 16.8
1 2

/ 2.70 ‘

 

 

 

 

 

 

 

 

(x10'17 esu/cmz)

Figure 5.1 Schematic of the assemblies of chromophores l and 2 spaced apart with one
x(Z)-inactive layer of 1,12-dodecanediylbisphosphonate. The shading of the chromophore
layers represents the induced dipole moment. The experimental susceptibility values,
referenced to the known value for quartz, are located adjacent to the appropriate layer for

each step in the assembly.

-91-

 

 

 

 

 

 

 

a

’7
=2
C6
V
.3?
m
C
Q)
E ‘
H 18 —
CD .
E 16 _
m r-
CI) 14 _

12 —

10 — b

8 _.

6 P— i;

4 — ‘ l \

2 — , “

J \
0 . ’W . ..::2llllll$i:.;, aka-‘11“ .
t 4 l . . I
0 10 20 30 4o

 

Sample Rotation Angle (deg)

Figure 5.2 Surface SHG intensities for the build up of chromophores l and 2 spaced apart
with one layer of xm-inactive 1,12-dodecanediylbisphosphonate on fused silica substrates
(a) SiOx; SiOx—l; SiOx—l-S; SiOx—l-S-Z; SiOx—l-S-Z-P (b) SiOx; SiOx—Z; SiOx—Z-S;
SiOx—Z-S-l; SiOx—2-S-1-P. The P is the addition of a phosphonic acid capping layer.

-92-

 

 

 

 

 

 

 

 

 

 

 

 

252 .m_, 204
2 1
73.8 1.5
S s
26.2 2.70
s s
29.1 32.0
1 ' l 2

 

 

 

 

 

 

 

 

(x10'17 esu/cmz)

Figure 5.3 Schematic of the assemblies of chromophores 1 and 2 spaced apart with two
xm-inactive layers of l,12-dodecanediylbisphosphonate. The shading of the chromophore
layers represents the induced dipole moment. The experimental susceptibility values,
referenced to the known value for quartz, are located adjacent to the appropriate layer for

each step in the assembly.

-93-

_s .A .A .1
o —* N (A)
F' I

IIIIfi—[rllllllllllllllll

(O

   

O—‘NOOLU'IQVG

 

SSHG Intensity (a.u.)

 

 

 

0 1O 20 ' 30 40 50 60 70
Sample Rotation Angle (deg)

Figure 5. 4 Surface SHG (intensities for the build up of chromophores 1 and 2 spaced apart

with two layers of x( 2)-inactive 1,12-4- ‘ "at? on fused silica

substrates (a) SiOx; SiO)( —;1 SiOx—l- S, SiO x—-1-S S, SiOx —1-S- S-2 (b) SiOx; SiOX —2;
SiOx—Z-S; SiO x—2-S S, SiOx—2-S- S- l.

-94-

 

 

77.2

 

 

 

 

 

 

 

 

 

 

 

 

28.9

4.44

0.99

3.33

 

2 I 1
f 26.4
1
2.72
s s
0.39
S s

 

 

 

(x10'17 esu/cmz)

.......
......

 

 

 

Figure 5.5 Schematic of the assemblies of chromophores 1 and 2 spaced away from the

fused silica substrate with two xlz)

-inactive layers of 1,12-dodecanediylbisphosphonate.

The shading of the chromophore layers represents the induced dipole moment. The
experimental susceptibility values, referenced to the known value for quartz, are located
adjacent to the appropriate layer for each step in the assembly.

-95-

1.a _
1.6 —
1.4
1.2

1.0 -

 

 

 

 

SSHG Intensity (a.u.)

0.3 —
0.2 '— I '. o‘ ' +9 0 ‘

O Q. .
, ‘ .. ‘
0 1 ' . ‘
— . “i
. -
o .0
. .
9 O

 

 

 

Sample Rotation Angle (deg.)

Figure 5.6 Surface SHG intensities for the build up of chromophores 1 and 2 spaced away
from the fused silica substrate with two layers of x(Z)-inactive 1,12-
dodecanediylbisphosphonate (a) SiOx; SiOx—S; SiOx—S-S; SiOx—S-S-l; SiOx—S-S-l-Z (b)
SiOx; SiOx—S; SiOx—S-S; SiOx-S-S-Z; SiOx—S-S-Z-l.

-96-

electric fields through the assembly. Implicit in this treatment is that there are no
substantial inter-layer effects. This approximation is suspect, as we demonstrated
experimentally in the preceding chapter. Treating the multilayer assembly in the context
of a single layer with a non-monotonic dielectric response function, s(z), is potentially
appealing in the sense that it eliminates the use of the Fresnel factors that can produce
misleading results if the complex dielectric response of each layer is not known with
certainty. This approach is also attractive in the sense that it offers the connection
between the depth profile of the interface and its net nonlinear response. While it would
be appealing to model s(z) from the experimental data, this is not a tractable task because
we cannot determine a priori the relative contributions of the dipolar and quadrupolar
terms in the effective xm response. Because these two contributions depend differently
on the function 5(2), and we measure experimentally the single quantity X12), the
unambiguous resolution of s(z) is not possible. Notwithstanding this limitation, we can
gain some insight into the factors that contribute to s(z) from the experimental data.

The first experiments in the series we report here are for three layer systems,
where the two opposing bilayer structures are each modified by the insertion of a )6”-
inactive alkane bisphosphonate spacer. Because the length of the alkane bisphosphonate
spacer layer (20 A)44 is substantially less than the coherence length of the laser light
source (2 3cm), we expect to recover cancellation of the chromophore dipolar
contributions to the xm response. We do not observe this condition experimentally. As
shown it Table 5.1 and schematized in Figures 5.1,5.3, and 5.5, there are two notable
features regarding these data. The first is that cancellation is not obtained between

chromophores 1 and 2 and the second is that the magnitude of the 1(2) response for the

-97-

 

chromophores is different by a factor of ~10, depending on their position with the
interface. For either chromophore in close proximity to the surface, we recover a SSHG
signal that is ~ 7 times larger than that of the fused silica substrate. For these same
chromophores spaced away from the substrate by ~ 50 A, we recover SSHG signals that
are ~ 70 times larger than the substrate. The difference in these responses must not be
due to simple chemical structural effects. We recover essentially the same layer-
dependent responses for the case where the chromophore layers are separated by two
spacer layers, ~ 40 A, demonstrating that the anomalous response we observe is not
associated with a simple optical interference phenomenon.

In an effort to resolve whether our experimental data are the result of
chromophore separation, we constructed the two canceling chromophore bilayer
structures, with chromophores in adjacent layers, on a substrate where two spacer layers
had been chemisorbed to isolate the chromophores from the fused silica. The results of
these experiments are similar to those for the preceding ones; we do not observe
structural cancellation, and the x0) responses of the chromophores appear to correlate
with the spacing from the silica substrate.

As discussed above, we need to consider the function s(z) in Eq. 5.1 to understand
these data. For an interface of single composition, the dielectric response is considered to
vary smoothly between its two boundary conditions. Guyot-Sionnest and Shen indicate
that s(z)i = l/si, and that s(z) varies smoothly across the interface. In their treatment it
was implicit that e was real, but for our systems, where the second harmonic electric field
is in resonance with the red edge of the absorption spectra of the chromophores, s will

necessarily be complex. In addition, the dielectric response of the ZP interlayer linking

-93-

 

1!-

moiety is not the same as that of the organic layer constituents. We are thus left with a
function s(z) which is, in general, complex and does not vary in a simple manner across
the interface. The role of the ZP functionality is underscored by changes in the x”)
response of multilayers capped with phosphoric acid, as indicated in Table 5.1. While we
can estimate the spatial variation of s(z) for our layered assemblies, there is still
significant ambiguity remaining in the modeling of the experimental phenomena. This
ambiguity remains because of the undetermined relative contributions from the various

(2)

contributions to x (Eq. 5.1), and the uncertainty that exists in the experimental values of

n and k for the individual layer constituents.

5.4 Conclusions

We have determined that there is a strong interface structural dependence on the
experimental SSHG signal recovered from systems expected to exhibit cancellation. By
spacing two complementary chromophores apart using one or two layers of 1,12-
dodecanediylbisphosphonic acid, the SSHG intensity is seen to deviate from expectations
and increase anomalously for each structural motif. The different SSHG signal responses
of each layer before and after phosphorylation indicates a contribution to the overall
response from the Zr-phosphate/phosphonate linkage adjacent to different functional
groups. The immediate horizons in this work are the development of models to account
for complex spatial variations in 8(2) and to treat these interfaces as explicit multilayer
assemblies. Comparing the predictions of each of these models will allow us to

determine the most appropriate way to treat our data.

-99-

 

 

5.5 Literature Cited

fl

. Nelson, Bryce P.; Frutos, Anthony G.; Brockman, Jennifer M.; Corn, Robert M.; Anal.
Chem, 1999, 71, 3928.

2. Frutos, Anthony G.; Weibel, Stephen C.; Corn, Robert M.; Anal. Chem, 1999, 71,

3935.

3. Xia, Y. N.; Whitesides, G. M.; Annu. Rev. Mater. Sci, 1998, 28, 153.

A

. Collier, C. P.; Wong, E. W.; Belohradsky, M.; Raymo, F. M.; Stoddart, J. F .; Kuekes,

P. J.; Williams, R. 8.; Heath, J. R.; Science, 1999, 285, 391.

 

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

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

O\

. Marowsky, G.; Lupke, G.; Steinoff, R.; Chi, L. F.; Mobius, D.; Phys. Rev. B, 1990, 41,
4480.
7. Palto, S. P.; Blinov, L. M.; Yudin, S. G.; Grewer, G.; Schonholl, M.; Losche, M.;

Chem. Phys. Lett., 1993, 202, 308.

00

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

9. 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.

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

l 1. Dulcey, C. S.; Georger, J. H.; Krauthamer, V.; Stenger, D. A.; Fare, T. L.; Calvert, J.

M.; Science 1991, 252, 551.

~100-

12.

13.

14.

15.

l6.

17.

18.

19.

20.

21.

22.

23.

24.

Kumar, A.; Biebuyck, H. A.; Abbott, N. L.; Whitesides, G. M.; J. Am. Chem. Soc.
1992, 114, 9188.

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

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

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

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

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

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

-101-

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

O’Brien, J. T.; Zeppenfeld, A. G; Richmond, G. L.; Page, C. J .; Langmuir, 1994, 10,
4657.

Ansell, M. A.; Zeppenfeld, A. C.; Yoshimoto, K.; Cogan, E. B.; Page, C. J .; Chem.
Mater., 1996, 8, 591.

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

Langmuir, 1.; Trans. Faraday Soc.; 1920, 15, 62.

Kleinfeld, Elaine R.; Ferguson, Gregory 8.; Chem Mater., 1996, 8, 1575.

Chidsey, C. E. D., Loiacono, D. N.; Langmuir, 1990, 6, 682.

Seshadri, K.; Froyd, K.; Parikh, A. N. Allara, D. L.; Lercel, M. J.; Craighead, H. G.;
J. Phys. Chem, 1996, 100, 15900.

Sabatini, B.; Rubinstein, 1.; J. Phys. Chem, 1987, 91, 6663.

Sabatini, B.; Rubinstein, 1.; Moaz, R.; Sagiv, J .; J. Electroanal. Chem, 1987, 219,
365.

F inklea, H. O.; Snider, D. A.; F edyk, J .; Sabatini, B.; Gafni, Y.; Rubinstein, I.;
Langmuir, 1993, 9, 3660.

Finklea, H. O.; Snider, D. A.; Fedyk, J .; Langmuir, 1990, 6, 371.

Sabatini, E.; Cohen-Boulakia, J .; Bruening, M. L.; Rubinstein, 1.; Langmuir, 1993, 9,
2974.

Janek, R. P.; Fawcett, W. R.; Ulman, A.; Langmuir, 1998, 14, 3011.

Zhao, Xiao-Mei; Wilbur, James L.; Whitesides, George M.; Langmuir, 1996, 12,
3257.

Patrick, David L.; Cee, Victor, J .; Purcell, Timothy, J .; Beebe, Thomas B. Jr.;

-102-

7‘. " “5"— :“"111

40.

41.

42.

43.

44.

45.

Langmuir, 1996, 12, 1830.

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

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

Perrin, D. D.; Armarego, W. L. F.; Perrin, B. R.; Purification of Individual Organic
Chemicals, Pergmon Press, Oxford, 1980, 449.

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

Guyot-Sionnest, P.; Shen, Y. R.; Phys. Rev. B., 1987, 35, 4420.

~103-

 

 

Chapter 6

Future Directions

This dissertation has demonstrated that there are a minimal amount of defects
within zirconium phosphate/phosphonate multilayer assemblies. With the development
of the SHG laser system and the synthesis of chromophores 1 and 2, there are a variety of
directions that this work can take. The system can be modified to image surfaces. A
variety of layer motifs can yield a large body of data further investigating these systems
and the spatial effects that the two chromophore layers can have with respect to the
SSHG signal obtained.

Modification of the SSHG system to image the surfaces would improve the
amount of knowledge obtainable from ZP systems. Nonlinear surface imaging is a
technique that has recently been employed for use as a measuring tool of interfaces."3
SSHG imaging measurements can yield information about the surface morphology on a
pm scale. Images are acquired using two fundamental beams temporally intersecting at
the sample. The SHG signal from the sample is collected using CDD detectors, and the
sample is then translated normal through the beam thereby forming an SHG images of the
area scanned. The SHG images can be collected using either transmission or reflection
modes. Control over the input polarization of each of the beams will allow selective
control of the electric field polarization with respect to the chromophores nonlinear
transition moment, and more specifically, access to the tensor elements of the x0) optical
susceptibility.

By using both SSHG intensity and SSHG imaging measurements in

characterizing SAMs, a wealth of information can be acquired. Insight into the

-104-

 

 

chromophore layer’s nonlinear susceptibility, tilt angle, and the angular distribution can
be ascertained from the direct SHG intensity data, while information relating to surface
morphology on a pm scale can be obtained using SSHG imaging experiments.

One possible avenue of exploration using these systems in tandem is the study of
aggregation within a chromophore layer using SSHG as the measuring tool. It has been
shown that varying the concentration of a chromophore within a single layer leads to

island forrnation.4’5

By using these nonlinear molecules in conjunction with SSHG,
noninvasive characterization to the degree of this inhomogenaity can be assessed. By
choosing an alkane bisphosphonate of the appropriate length and using this to ‘dilute’ the
chromophore in solution, layers containing aggregated islands of the chromophore can be
formed. The SSHG intensity and image from this assembly will be from these island
defects and information about their properties can be elucidated in a more detailed
manner then previously reported.

One may also be able to draw information about the assembly of the
chromophores by mixing the chromophores in the same solution. Using a 1:1 solution of
chromophores 1 and 2, one would expect an even distribution of each on the surface since
the formation properties are identical, based on the linear response presented in Chapter
2. The monolayer SSHG signal should be only the response of the substrate, because the
distribution of signs of the nonlinear response within the 100 um beam area. If a bilayer
is synthesized from sequential deposition from this solution, the response should yield a
signal that contains cancellation of the electric dipole moments in the bilayer as well as

the monolayer, and half the signal intensity observed for [SiOx-Z-l] (Figure 4.6b) is

expected due to the quadrupole moment contribution.

-105-

 

Based on the discussion from Chapter 5, a greater understanding of the response
of the spatially modified surfaces needs to be achieved. Because the chromophores 1 and
2 both have absorption tails at 532 nm, both the real and imaginary components of the
dielectric gradient contribute to the observed signal. It is this gradient of the dielectric
properties that leads to the observed SSHG signal. It is believed that this is the first work
that studies dielectric variance in thins films. By controlling the dielectric properties of a

film, one can begin to understand the effects the imaginary component has in thin films.

-106-

 

 

6.1 Literature Cited

 

l. Florsheimer, M.; Kupfer, M.; Bossard, C.; Looser, H.; Gunter, P.; Adv. Mater., 1992,
4, 795.

2. Florsheimer, M.; Jundt, D. H.; Looser, H.; Sutter, K.; Gunter, P.; J. Phys. Chem,
1994, 98, 6399.

3. Florsheimer, M.; Phys. Stat. Sol. A, 1994, 173, 15.

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

4419.

 

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

 

-107-

APPENDIX A

 

1.1.:
1“ ‘

 
 

   
  
    
    

[Impact 59.5313

2000 3000 Numberotl
Scans
10000 >_4000 9399191291
/‘ . ‘ ,
0 mm 5'“ l

 

39911 later] ' . -' '--.;..;-. .. _ .7 .. .- .

W]

LII).

...I'_iQ.99' I
591

L111.

channel:

ISLE

 

 

 

Figure 1A Front panel of the Labview® program written for the SSHG laser system.

-108-

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 2A Icon based schematic of the Labview® Program.

-109-

interchannel delay [secs] 1...
measurement mode structure
coupling ix input contig Inc...

Input limits [no change]

 
    
 

 

device [1] — 15:le
channels [0) 4 1 number of channels
bu‘fer size [IMO scans] ‘ DSP handle structure out

[group] [Cl] error 011
error in [no error]

[number at butters] [1]
allocation mode [no changefl]
[number of AM LIX boards] [no...

 

Figure 3A AIConfig SubVI details.

-110-

edge or slope [no change]
pretrrgger scams [0]

trigger type [no "1910]

IMHO '11

number of scan: to acquie
scan rate [1000 scans/sec]
number of buffers to acquir...
error in [no error]

scan clock source [no Chang...
analog Chan [-1 a level [0.0]

l additional trio oaramsl

 

-.-W_-.......-._..l

 

Figure 4A AIStart SubVI details.

-111-

scan backlog

conditional retrieval [off] number read
D 'n tasle or!
number of scans to read [-1... scaled date
time limit in sec [no chang... binary data
output units [scale-c121] retrieval complete
error in [no error] error out

read/“search positIon [from ..
DSP handle structure

 

 

Figure 5A AIRead SubVI details.

-112-

taskID 'n E 133le out
error it [no error] 2 error out

 

Figure 6A AIClear SubVI details.

-113-

error ’11 [no error)

 

[-
I
I
I
§

1
I .__-..-

 

Figure 7A Error SubVI details.

-114-

newfiepatthotA Path L.

   
   
   

1D data .

append to tie? [new flezF] -------- é
um”? [m:F] ............_...._....4

delimiter lTabl

1"”
i
l

 

Figure 8A. Write to Spreadsheet SubVI details.

-115-