..
E

I

0:1.r
Ix II

.1 if

m»

Thaw: .
:(

. .KNL ~
.. a. nu.
L WWW V M

 

. A _ .3:
lung .v . if."

. 2%

$.31 .
4m“

¢ 3 .‘

”in 9 h .
.Finnfiff u... 3:. in.“

34an , L
‘23.. 2.115
s: .

o
I

 

.. . V745
glfiwus . 1
. L,

 

. 1... :mm fivWWWTDqu
[r , I

v.2. .nfifififiwfin whmwwéfiwa
. . . . .. .x

 

 

fifiwmm. _

 

., w
.9;

   

 

.. “a ‘mfimflmw a. ._

    

Illyblnnil UL qb‘ t

i1l|x.

6.

$,-‘.
ell

THESIS
9

g)

SILTY LBI BSMRIE

||||lillll’lllllllllHill HI llHlHllHHll l|||

3129301688

HI

 

 

 

This is to certify that the

dissertation entitled

Structure and Dynamics of Layered
Molecular Assemblies
presented by

Jennifer Conrad Horne

has been accepted towards fulfillment
of the requirements for

Ph.D. degnxin Chemistry

 

 

fl ”Mg
anchard

M ajor professor

 

[yne August 14, 1998

 

MSU is an Affirmative Action/Equal Opportunity Institution ' 0—12771

 

 

LIBRARY

Michigan State
University

 

 

PLACE IN REFURN BOX
to remove this checkout from your record.
TO AVOID FINE return on or before date due.

 

MTE DUE

DATE DUE

DATE DUE

 

 

“31.270683

 

 

 

 

 

 

 

 

 

 

 

 

 

use mumps-p.14

 

STRUCTURE AND DYNAMICS OF LAYERED MOLECULAR ASSEMBLIES
By

Jennifer Conrad Home

A DISSERTATION

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

DOCTOR OF PHILOSOPHY
Department of Chemistry

1998

ABSTRACT
STRUCTURE AND DYNAMICS or LAYERED MOLECULAR ASSEMBLIES
By

Jennifer Conrad Home

This dissertation focuses on the goal of understanding and controlling layered
material prOperties from a molecular perspective. With this understanding, materials can
be synthetically tailored to exhibit predetermined bulk properties. This investigation
describes the optical response of a family of metal-phosphonate (MP) monolayers and
multilayers, materials that are potentially useful because the films are easy to synthesize
and are chemically and thermally stable. MP films have shown potential in a variety of
chemical sensing and optical applications, and in this dissertation, the suitability of MP
films for optical information storage is explored. For this application, the extent of
photonic energy transport within and between optically active layers is an important
factor in determining the stability and specificity of optical modifications made to a
material. Intralayer and interlayer energy transport processes can be studied selectively
in MP films because the composition, and thus the properties, of each layer are controlled
synthetically. It was determined by fluorescence relaxation dynamics in conjunction with
atomic force microscopy (AFM) that the substrate and layer morphologies are key factors
in determining the layer optical and physical properties. The initial MP layers in a
multilayer are structurally heterogeneous, characterized by randomly distributed islands
that are ~ 50 A in diameter. The population dynamics measured for these layers are non-

exponential, chromophore concentration-independent, and identical for two different

chromophores. The data is explained in the context of an excitation hopping model in a
system where the surface is characterized by islands of aggregated chromophores as well
as non-aggregated monomers. Within a MP monolayer, the dynamics are dominated by
intra-island excitation hopping. Forster (dipolar) energy transfer between the
energetically overlapped chromophores does not play a significant role in determining the
relaxation dynamics of these systems, as the islands are separated by distances greater
than the critical distance required for this process. Investigation of the interlayer energy
transfer behavior revealed that transport between adjacent chromophore layers is
precluded by the spatial modulation of the dielectric response of the MP film. These
results are important for understanding the morphological, structural, and optical
properties that will be useful for the incorporation of layered organic materials in future

technologies.

In Memory of Dziadzi

ACKNOWLEDGMENTS

I could not fully appreciate until now how fortunate I was to have had Gary
Blanchard as my research as my research advisor. I learned a great deal from him about
science, the professional world and life, (and how to have a smart-aleck comment about
everything). He is a mentor in every sense of the word. I would also like to thank Alice
Deckert, my research advisor at Holy Cross, for her support and encouragement through
the years. I can only hope to have the same positive effect on everyone I meet. My
committee members also deserve recognition: Marcos Dantus, Merlin Breuning, Jeff
Ledford, and especially Greg Baker, who, along with his group, helped me through the
organic syntheses for the materials used in this dissertation.

Best of luck to the Blanchard group members, past and present, and the many
other acquaintances I have made at MSU. Two group members who deserve special
recognition are Selezion Hambir and Dave Karpovich, for their generous help in the lab,
and for always having a word of encouragement (unless Dave thought 1 deserved
something a little more sarcastic). I also want to thank Greg Noonan for the interesting
times in the house on Cornell Ave. Thanks for letting me borrow your dog sometimes.
Also, summer would have been incredibly boring without all of the softball teams I have
been a part of, especially the '94 Lansing Grass Stains.

My family, especially my brothers, were real troopers for driving the twelve hours
to come out for visits. And finally, I would like to acknowledge Al Schwartz, my best
friend. Thanks for all the love and support — l'll return the favor soon. Here's to taking

chances!

TABLE OF CONTENTS

Page

List of Tables ................................................................................................................... vii

List of Figures .................................................................................................................. viii

Chapter 1. Introduction ...................................................................................................... l

1.] Literature Cited ............................................................................................... 8

Chapter 2. Synthesis and Characterization of Metal-Phosphonate Materials l l

2.1 Introduction ..................................................... . ......................................... l l

2.2 Experimental ............................................................................................. l 2

. 2.3 Results and Discussion .............................................................................. l8

2.4 Conclusions ............................................................................................... 28

2.5 Literature Cited ......................................................................................... 29
Chapter 3. Rotational lsomerization Barriers of Thiophene Oligomers in the Ground

and First Excited States. A lH NMR and Fluorescence Lifetime Study ........ 3O

3. 1 Introduction ............................................................................................... 3 l

3 .2 Experimental ............................................................................................. 3 3

3.3 Results and Discussion .......................................................... . ................... 39

3.4 Conclusions ................................ _ .............................................................. 62

3.5 Literature Cited ......................................................................................... 64

Chapter 4. Dynamics Within a Single Molecular Layer. Aggregation Relaxation and

the Absence of Motion ................................................................................... 67
4. 1 Introduction ........................................ . ...................................................... 68
4.2 Experimental ............................................................................. . ............... 68
4.3 Results and Discussion .............................................................................. 72
4.4 Conclusions ............................................................................................... 93
4.5 Literature Cited ......................................................................................... 95

Chapter 5. The Role of Substrate Identity in Determining Monolayer Motional

Relaxation Dynamics ...................................................................................... 97
5. 1 Introduction ............................................................................................... 98
5.2 Experimental .......................................................................................... 100
5.3 Results and Discussion .......................................................................... I05

 

5.4
5.5

Conclusions ............................................................................................ 126
Literature Cited .................................................. V .............................. M. 127

Chapter 6. The Effect of Layer Morphology on population Dynamics and Intralayer

Excitation Transport in Zirconium-Phosphonate Films ............................... 129

6. 1 Introduction .................................................................................... . ....... l 30

6.2 Experimental .......................................................................................... 1 32

6.3 Results and Discussion ........................................................................... 134

6.4 Conclusions ............................................................................................ l 56

6.5 Literature Cited ...................................................................................... 157
Chapter 7. Structural Mediation of lnterlayer Excitation Transport in Zirconium-

Phosphonate Multilayers ...................... . ....................... . ............................... 158

7.1 Introduction ........................................................ ,. ............. .. i59

7.2 Experimental .......................................................................................... 162

7.3 Results and Discussion ........................................................................... l65

7.4 Conclusions ............................................................................................ 177

7.5 Literature Cited ...................................................................................... 178

Chapter 8. Conclusions ................................................................................................. 179

8.1 Literature Cited ........................................................................................... 183

-vi-

Table 2.1

Table 2.2

Table 3.1

Table 4.1

Table 4.2

Table 5.1

Table 5.2

Table 6.1

Table 6.2

Table 7.1

LIST OF TABLES

1H NMR results (500 MHz) and final yields for bisphosphonated
compounds. DMSO-ds was used as the solvent for all spectra ................... 14

Calculated lengths of alkanes and oligothiophenes. The length is
measured as only the distance between the outermost oxygen atoms. ., ........ 21

Excitation and emission wavelengths used for TCSPC measurements ........ 37

Comparison of fractional BDP concentrations in deposition solution and
in the resulting monolayer, determined according to absorbance ................. 75

Dependence of infinite time anisotropy on fractional composition of
monolayer ...................................................................................................... 92

Excited state lifetimes of monolayers The time constants refer to fits of
the experimental data to the fiJnction f(t) = A.exp(-t/t.) + Azexp(-t/tz).... 109

Induced orientational anisotropy data for monolayers. For BDP, the
angle between the transition moments is 6 = 31° and for QDP, 6 = 0°,
Angles 9 were determined from 8 and R(oo) data using Equation 5.4 ....... 1 '3

Fits of excited state lifetime data and simulated lifetime data to the
function f(t) = A;exp(-t/t.) + Azexp(-t/tz) ................................................. 140

Calculated aggregate sizes for selected excitation hopping time
constants and total migration times ............................................................ 147

Descriptions of layered samples corresponding to the lifetime data
presented in Figure 7.4 ............................................................................... 169

-vii-

Figure 1.1

Figure 2.1

Figure 2.2

Figure 2.3

Figure 2.4

Figure 2.5

Figure 2.6

Figure 2.7

Figure 3.1

Figure 3.2

Figure 3.3

LIST OF FIGURES

Structures and absorbance (——) and emission (- - -) spectra (10'5 M
solutions) for 2,2'-bithiophene-5,5'-diylbis-(phosphonic acid) (top) and
2,2':5',2":5",2"'-quaterthiophene-5,5"'-diylbis(phosphonic acid) (bottom).
Note the spectral overlap between BDP emission and QDP absorption
Emission spectra have been normalized to the absorption spectra for

clarity of presentation ................................................... . .................................. 5

Synthetic scheme for quaterthiophene ........................................................... 13
Three-step synthetic scheme for phosphorylation of quaterthiophene. ( 1)
lithiation of QT; (2) formation of phosphamide; (3) hydrolysis to
phosphonate .................................................................................................. l 5

Synthetic scheme for alkane phosphonic acids. (1) Formation of
phosphonate with triethylphosphate, (2) hydroylsis to phosphonate ............ 16

 

Structures and absorption ( ) and emission (- - -) spectra for (top)
EDP; (middle) TDP; (bottom) QDP .............................................................. l9

Schematic of the buildup of two zirconium-phosphonate layers of (a) odd-
carbon chain and (b) even-carbon chain. The dashed lines indicate the
surface normal ............................................................................................... 22

Plot of ellipsometric thickness measured for lO-layer DDBPA film and
regression of data. The zero-layer thickness is that due to the native oxide
layer and the primer layer .............................................................................. 25

Reflection-absorption FTIR spectra measured for layers 4 - 11 of a Zr-
DDBPA film on silica. The dashed line is the spectrum of the silane
primer layer. Spectra were baseline corrected at 2800 cm'l ......................... 27

Time correlated single photon counting spectrometer .................................. 35

Example of excited state population decay (0) for TBSxT in n-butanol.
The instrument response (0) time is ~ 35 ps ............................................... 38

Structures and inter-ring rotations for the three thiophene Oligomers. (a).
3',4'-dibutyl-2,2':5',2"-terthiophene (DBTT), (b) 3",4"-dibutyl-
2,2':5',2":5",2"':5’", 2""-pentathiophene (DBPtT) and (c) 3',4',3"",4""-
tetrabutyl-2,2':5',2":5",2"':5"',2"“:5"",2"'"-sexithiophene (TBSxT) ................ 40

- viii -

Figure 3.4

Figure 3.5

Figure 3.6

Figure 3.7

Figure 3.8

Figure 3.9

1H NMR spectrum of DBTT taken at -20°C. Inset: Expanded aromatic
proton region with proton assignments for the aromatic protons. The
designations of the protons are indicated in Figure 3.3 ................................. 41

1H NMR spectrum of DBPtT taken at -20°C. Inset: Expanded aromatic
proton region with proton assignments for the aromatic protons. The
designations of the protons are indicated in Figure 3 .3 ................................ 42

1H NMR spectrum of TBSxT taken at -20°C. Inset: Expanded aromatic
proton region with proton assignments for the aromatic protons. The
designations of the protons are indicated in Figure 3.3 ................................. 43

Temperature dependence of the aromatic proton resonances of DBTT,
shown at 5 representative temperatures, as indicated in Figure 3.3 .............. 44

(a) Calculated isomerization barrier for DBTT, where the isomerization
coordinate is taken as the inter-ring dihedral angle. (b) Dependence of the
proton charges on inter-ring rotation angle; (0) = Ha, (I) = Hb, (A) = 11,.

(c) Dependence of the aromatic carbon charges on inter-ring rotation

angle; (0) = Ca, (I) = Cb, (A) = Cc ................................................... . ......... 47

(a) Calculated isomerization barrier for DBPtT, where the isomerization
coordinate is taken as the inter-ring dihedral angle. (b) Dependence of the
proton charges on inter-ring rotation angle; (0) = H., (I) = I-Ib, (A) = 1L,
(ale) = 114, (O) = Hc. (c) Dependence of the aromatic carbon charges on
inter-ring rotation angle; (0) = Ca, (I) = Cb, (A) = C,., ( . ) = Cd, (9) =

C6 .................................................................................................................. 48

Figure 3.10 (a) Calculated isomerization barrier for TBSxT, where the isomerization

coordinate is taken as the inter-ring dihedral angle. (b) Dependence of the
proton charges on inter-ring rotation angle; (0) = Ha, (I) = Hb, (A) = l-L,
(916) = Hg, (9) = H.. (c) Dependence of the aromatic carbon charges on
inter-ring rotation angle; (0) = C., (I) = Cb, (A) = Cc, (+) = Cd, (0) =

Ce ................................................................................................................... 49

Figure 3.11 (a) Experimental temperature-dependence of lH NMR resonances for

DBTT. (b) Temperature-dependence of aromatic proton charges,
calculated as described in the text. For both (a) and (b), (O) = Ha, (I) =
Hb, (A) = Hc .................................................................................................. 52

Figure 3.12 (a) Experimental temperature-dependence of lI-I NMR resonances for

DBPtT. (b) Temperature-dependence of aromatic proton charges,
calculated as described in the text. For both (a) and (b), (O) = 1-1., (I) =
Hb, (A) = HC, (416) = Hd, (O) = Hc ................................................................. 53

-ix-

Figure 3.13 (a) Experimental temperature-dependence of 1H NMR resonances for

TBSxT. (b) Temperature-dependence of aromatic proton charges,
calculated as described in the text. For both (a) and (b), (O) = H., (I) =

Hb, (A)=Hc, (vie) =Hd, (O)=H,. ................................................................. 54

Figure 3.14 Calculated temperature dependent proton charges (top, left) and aromatic

carbon charges (bottom, left) for DBTT, compared to experimental 'H

NMR chemical shift temperature-dependent data. The comparison is
necessarily qualitative. See text for a discussion of this point. Symbol
assignments are the same as in Figure 3.8 ..................................................... 56

Figure 3.15 (a) Viscosity dependence of fluorescence lifetimes for DBPtT. (b)

Figure 4.1

Figure 4.2

Figure 4.3

Figure 4.4

Figure 4.5

Figure 4.6

Figure 4.7

Figure 4.8

Figure 4.9

Viscosity dependence of fluorescence lifetimes for TBSxT ......................... 61

Structures of bisphosphonates used in this work. (a) 2,2'-Bithiophene-
S,S'-diylbis(phosphonic acid) (BDP) and (b) 1,6-hexane
diylbis(phosphonic acid) (HBPA) ...................................................... . .......... 69

Background corrected absorbance spectra of BDP monolayers: (—)
100%; (- - - -) 58%; (' - ' -) 34%; ( ----- )26%. Inset: Background
spectrum ......................................................................................................... 76

Dependence of monolayer absorbance on concentration of BDP in
deposition solution. Solid line: Beer’s Law predicted dependence, dashed
line: Regression of experimental data ........................................................... 77

Representative fluorescence lifetimes of (C) 10" M BDP solution and (O)
26 % BDP/HBPA monolayer collected at 390 nm ........................................ 79

(a) Fluorescence lifetimes of BDP monolayers and solution at 400 nm.

The data were fit to the fiinction f(t) = A.exp(-t/t.) + Azexp(-t/t2). (b)
Normalized prefactors of fast exponential decay component for data taken

at 400 nm ....................................................................................................... 80

Fluorescence lifetime decays of BDP monolayers at (I) 390 nm, (0) 475
nm, (0) 525 nm, and (Cl) 575 nm ................................................................. 81

Emission wavelength dependence of (a) lifetimes and (b) Normalized fast
exponential prefactors for 29% BDP monolayer ........................................... 82

Emission spectra of fresh samples (solid lines) and aged (two-months)
samples (dashed lines) for (a) aggregate-like and (b) monomer-like
samples. The vertical lines are guides for the eye to the band maxima ........ 84

Calculated rotational energy barriers for BDP: (EJ) So and (O) S. ............... 87

Figure 4.10 Reorientation data for a 26% BDP monolayer. (a) Raw transient signal

Figure 5.1

Figure 5.2

Figure 5.3

Figure 5.4

Figure 5.5

Figure 5.6

Figure 5.7

Figure 5.8

Figure 5.9

intensities for emission polarized parallel (upper) and perpendicular

(lower) to excitation pulse polarization. Included is the instrumental
response function. (b) Induced orientational anisotropy function derived
from the data presented in panel (a) and regressed fit ................................... 9O

Structures of oligobisphosphonates used in this work. (a) 2,2'-bithiophene-
5,5'-diylbis(phosphonic acid) (BDP); (b) 2,2':5',2”:5",2'"-quaterthiophene-
5,5"'-diylbis(phosphonic acid) (QDP); (c) 1,12-

dodecanediylbis(phosphonic acid) (DDBPA); (d) 1,6-
hexanediylbis(phosphonic acid) (HBPA) ...................................................... 99

Representative instrument response function (Cl), excited state
population decay for QDP in DMSO (10'5 M) (O) and excited state

 

 

population decay for QDP in a monolayer (1%) (O) .......................... . ...... 103
Dependence of monolayer absorbance on concentration of QDP in
deposition solution and regression of experimental data ............................ 106
Emission spectra of (a) QDP monolayers (- - -) 1%; ( ) 5%; (- ' ° )

100%; and (b) QDP solution in DMSO (- - -) 10'7 M; ( )10" M;

(- - - -) 10'3 M. Spectra were normalized for presentation and no
information about fluorescence quantum yield should be inferred from
these plots ................................................................................................... 108

Reorientation data for a 50% QDP monolayer on silica. (a) Emission
intensity data for parallel and perpendicular polarizations. (b) Induced
orientational anisotropy function of above data with fit (- - -) ................... 1 1 1

Emission intensities (a) and induced orientational anisotropy function
(b) for a 26% BDP monolayer on silicon ................................................... 1 l7

Emission intensities (a) and induced orientational anisotropy function
(b) for a 1% QDP monolayer on silicon ..................................................... l 18

Conformations of QDP with thiophene rings (a) all anti and (b) all syn.
Conformations of BDP with thiophene rings (c) ann‘ and (d) syn ............ 120

Calculated energy barriers for QDP inter-ring rotation for the first two
inter-ring bonds: (a) $0 and (b) S]. Twist angles indicated are for angles
with respect to their adjacent ring ............................................................... 122

Figure 5.10 Atomic force microscopy images for 300 nm x 300 nm areas of oxidized

silicon (top) and fused silica (bottom) substrates. Cursor plots show
height profiles along the line indicated in the 3-D images. Image

Figure 6.1

Figure 6.2

Figure 6.3

Figure 6.4

Figure 6.5

Figure 6.6

Figure 6.7

Figure 7.1

histograms indicate surface relief (Az) distribution of area imaged ........... 124

Structures of oligobisphosphonates used in this work. (a) The donor
2,2'-bithiophene-5,5'-diylbis(phosphonic acid) (BDP); (b) the acceptor
2,2':5',2":5",2'"-quaterthiophene-5,5"'-diylbis(phosphonic acid) (QDP);

(c) the diluent l,12-dodecanediylbis(phosphonic acid) (DDBPA) ............ 131

Representative instrument response function (0) and excited stat
population decay for QDP in a monolayer (5%) (I) ................................. 136

Dependence of donor lifetimes on acceptor concentration, for a
monolayer containing 5% donor ................................................................. 138

Kinetic model proposed for population decay dynamics in donor and
acceptor monolayers. A refers to the aggregate species and M1 and M2

 

are monomers in different chemical environments .................................... 142
Simulated population decay (0) using Equation 6.2 (see text) and fit of
data ( ) to a double exponential decay of the form f(t) = A.exp(-t/t.

+ Agexp(-t/t2). Inset is a plot of the residuals fit .......... . ............................ 144

Topographic images (left panels) of the surfaces of (a) oxidized silicon
(SiOx) substrate, (b) APTES-primed substrate, (c) primed substrate with
QDP adsorbed, and (d) a mixed monolayer of 5% BDP, 50% QDP and

45% DDBPA on a primed substrate. The total scanning area is 1000 A

x 1000 A except for (a) which is 3000 A x 3000 A. The cantilever force
constant is 0.1 N/m and the imaging force used here is ~ 0.2 nN.
Corresponding cursor profiles and histograms are shown in the middle

and right panels, respectively ..................................................................... 150

Topographic image of a mixed monolayer of 5% BDP, 50% QDP, and

45% DDBPA on a primed substrate. The molecules in the central 2000

A x 2000 A area was displaced prior to this image using an AF M tip

under a force of I nN. The overall scanning area here is 4000 A x 4000

A acquired at a reduced imaging force of 0.05 nN. The cursor profile

reveals the thickness of the layer to be 40 A d: 10 A .................................. 153

Structures of organobisphosphonates used in this work. (a) 2,2'-
bithiophene-S,5'-diylbis(phosphonic acid) (BDP); (b) 2,2':5',2":5",2"'-
quaterthiophene-5,5"'-diylbis(phosphonic acid) (QDP); (c) 1,12-
dodecanediylbis(phosphonic acid) (DDBPA); ((1) 1,6-
hexanediylbis(phosphonic acid) (HBPA) ................................................. ,. 160

-xii-

Figure 7.2

Figure 7.3

Figure 7.4

Figure 7.5

Typical instrument response function (0) and excited state population
decay (I) (425 nm) for a multilayer with full donor and acceptor layers.
The emission at 425 nm is from BDP ......................................................... 164

Kinetic scheme for population dynamics in layered structures. A. is the
donor aggregate and A2 is the acceptor aggregate. M. and M2 refer to
radiative donor monomers and M3 and M4 refer to radiative acceptor
monomers. For a monolayer (Layer 1), only the species A., M. and M2

are present in the relaxed system. For a multilayer (both Layer 1 and

Layer 2), A2, M3 and M are also present in the relaxed system ................ 167

Fluorescence lifetimes measured for various layered systems. Decays
were fit to the function f(t) = A.exp(-t/t.) + A2exp(-t/t2) .......................... 172

Calculated values of K2 from Equation 7.2 as a function of donor and
acceptor orientation ......................................................................... . .......... I73

- xiii -

Chapter 1

Introduction

Molecular assemblies first captured the attention of scientists decades ago. In
1917, Irving Langmuir first studied the behavior of amphiphilic molecules at the air-
water interface“), and Katharine Blodgett later introduced the technology of depositing
the monomolecular layers on substrates”) Since then, organized molecular assemblies
have found use in numerous chemical and physical applications. This class of interfacial
materials has characteristically well defined Structures of controlled composition and,
with a small amount of material, they offer a widely tunable range of chemical, electrical,
and optical properties.

Langmuir-Blodgett (LB) films were the first examples of multilayer films with
highly ordered layers and controlled layer-by-layer growthm These films have been
used in a wide range of studies, from optical properties of chromophores to ordered
polymerizations on a substratem However, because of the delicate nature of the van der
Waals forces linking each layer and the inherent difficulty in pre-assembling the
molecules for transfer of each layer, LB films have become limited to mostly
fundamental studies of layered assemblies and their properties.

The next step in the development of this field was the discovery of self-
assembling materials, which, in contrast to LB layers, did not have to be pre-assembled
and were more chemically stable. The first detailed examinations of self-assembled
monolayers were published by Allara and Nuzzo in 1985”] Their studies showed that

certain molecules would chemisorb to a compatible substrate in situ, with crystalline-like

-1-

order within the layer. There are many possible combinations of molecules and
substrates that can be used to make self-assembled monolayers (SAMs): thiols on gold,‘51
carboxylates on aluminumm or silver,[61 and silanes on glassm. These systems have been
investigated widely, particularly the thiol/gold system, and much is known about their
formation, structure, and properties. Despite the extensive discussion in the literature on

the potential utility of alkanethiol-gold SAMSJSI recent work has demonstrated that these

[9,101 [11.12.

assemblies are labile and sensitive to ozone—mediated oxidative degradation.

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

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

[14-30]

extensively, and, like self-assembling monolayers, show potential for use in

[31.34. 135-371

practical applications such as surface modification, electronic device, nonlinear

[16.25.26] 138-431

optics, and molecular recognition applications. Phosphonic acids form
strong, sparingly soluble complexes with metal ions, giving them significant advantages
over many self-assembled monolayer systems. MP structures are comparable to SAMs in

ease of synthesis, which generally involves immersion of the functionalized substrate into

a solution of the appropriate (0am)-organobisphosphonate. Because there are two
separately deposited components required to build a layer, spontaneous multilayer growth
is precluded. MP multilayers are versatile in a chemical sense because the identity of
individual layers can be controlled selectively as the structure is assembled and, in this
way, chemical potential or optical properties can be built into the system in three
dimensions rather than two. Like other systems that exhibit mesoscopic organization,
metal phosphonate structures have been used successfully in many studies, including

[24.25]

optical second harmonic generation, artificial photosynthesis,'44' and light

harvestingml
One application for which three dimensional, layer-by-layer control of multilayer

'46] Present-day

structure could potentially be useful is in optical information storage.
storage devices use a single layer medium, and the addition of a useful third dimension
could enhance the storage density of a given volume of the surface. The use of MP
multilayers with chemically and optically distinct layers could be the foundation for such
a three-dimensional structure, but there are aspects of this scheme that could serve to
limit the utility of these materials for information storage. The density of optical
information that can be stored within a single molecular layer is limited, and the spatial
resolution or “bit size” available with optical approaches is determined by the diffraction
limit for the optics used and the wavelengths of reading and writing light. The ability to
store information in a material optically depends on making a specific, stable
modification to the optical response of a given layer of the material.

The chemical change associated with writing or erasing, while still under

investigation, will likely be a cis-trans isomerization. Regardless of the means used to

create a molecular “bit”, the writing process will necessarily be mediated by an excited
electronic state to achieve the necessary spatial selectivity and contrast. During
conversion from one logical state to another, the chromophore will be excited, probably
to its S. state, and dissipative processes from the first excited singlet state can, in
principle, serve to limit the conversion efficiency or definition of the written spot. The
conformational change must be an activated process with a sufficiently high barrier to
ensure that the operation is irreversible under ambient dark conditions. The presence of
other optically active constituents within the same layer or in neighboring layers creates
the possibility of intra- or inter-layer energy exchange processes, which serve to place
limits on the utility of these materials for information storage applications. Because of
the requisite electronic excitation and the proximity of chromophores to one another, the
issue of excitation transport must be considered in detail for these systems.

The focus of this dissertation is to understand the energy migration and excited
state population dynamics that occur within an optically active film. By using simple
chromophores that don’t “switch”, we can observe these processes without complications
associated with large scale chromophore isomerization; The investigation of MP
materials that show potential for optical switching is underway in our lab. In this
dissertation, we use oligothiophenes as chromophores because they are relatively well
understood and form ordered layers due to their rigid structure'm We chose 2,2'-
bithiophene—S,S'-diylbis-(phosphonic acid) (BDP) and 2,2':5',2":5",2'”-quaterthiophene-
5,5'"-diylbis(phosphonic acid) (QDP) because the BDP emission band overlaps the QDP
absorption band, as shown in Figure 1.1. The spectral overlap of the two chromophores

is useful for evaluating excitation transport in MP multilayers, as we explain in Chapter 6

 

 

'..ttllOH

 

normalized linear response

 

 

 

300 400 500 600

wavelength (nm)

Figure 1.1 Structures and absorbance (——) and emission (- - -) spectra (10'5 M
solutions) for 2,2'—bithiophene-5,5'-diylbis-(phosphonic acid) (top) and 2,2':5',2":5",2"'—
quaterthiophene-5,5"'-diylbis(phosphonic acid) (bottom). Note the spectral overlap
between BDP emission and QDP absorption. Emission spectra have been normalized to
the absorption spectra for clarity of presentation.

in the context of Ferster energy transfer. These chromophores can photoisomerize via
ring rotation and this property is of value in understanding their dynamical response in
layered assemblies.

This dissertation describes a systematic study of the structure and dynamics of a
family of optically active self-assembled MP monolayers and multilayers. The
microscopic structure, optical properties, and effects of local environment are all
important factors for determining the feasibility of using these materials in practical
devices. Chapter 2 describes the synthesis of the organobisphosphonate compounds and
preliminary characterization of the growth of multilayer films. Chapter 3 is a study of
related oligothiophenes in solution. lsomerization barriers measured in both ground state
and excited state gave us a baseline understanding of the conformational behavior of the
chromophores in films. Chapters 4 and 5 detail the characterization of each of the two
chromophores individually in MP monolayers. It is here that we first realize indications
of structural inhomogeneity in the layer. Chapter 5 also demonstrates the substrate-
dependence of the motional dynamics of the chromophores in a layer. Other optical
properties are not affected. Chapters 6 and 7 are investigations of the intra- and interlayer
population dynamics of mixed monolayers and multilayer structures, respectively. We
find two factors that profoundly influence the energy transport dynamics in MP films.
The morphology of the substrate and the MP layer, in which islands of chromophore are
predominant, precludes intralayer energy transfer from donor to acceptor chromophores,
but facilitates excitation transport within an island of aggregated chromophore. In
multilayer structures, interlayer transport is not observed. The spatial modulation of the

dielectric response between the metal-phosphonate and organic portions of the layers

-6-

serves to screen the dipolar coupling between chromophores in adjacent layers. Both the
structure and the dynamics components of this dissertation contribute to a broader
understanding of MP films that will be useful for further study of the feasibility of using

these materials in more applied areas.

1.1 Literature Cited

8.
9.

10

11.

12.
13.
14.
15.
l6.

17.

18.

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

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

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

Assembly Academic Press: Boston, 1991.

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

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

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

(a) Smith, E. L.; Porter, M. D. J. Phys. Chem. 1993, 97, 8032. (b) Canning, N. D. S.;
Madix, R. J. J. Phys. Chem. 1984, 88, 2437.

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

R.; Sagiv, J. J. Colloid Interface Sci. 1984, [00, 465. (c) Maoz, R.; Sagiv, J. Thin
Solid Films 1985, 132, 135.

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

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

. Schessler, H. M.; Karpovich, D. S.; Blanchard, G. J. J. Am. Chem. Soc. 1996, H8,

9645.

Zhang, Y.; Terrill, R. H.; Tanzer, T. A.; Bohn, P. W. J. Am. Chem. Soc, 1998, 120,
2654.

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

Kohli, P.; Harris, J. J.; Blanchard, G. J. J. Am. Chem. Soc, in review.

Hong, H.-G.; Sackett, D. D.; Mallouk, T. E. Chem. Mater. 1991, 3, 521.
Thompson, M. E. Chem. Mater. 1994, 6, l 168.

Katz, H. E.; 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; lverson,
B. L.; Mallouk, T. E. J. Am. Chem. Soc. 1994, 116, 4786.

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

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.
31.

32.

33.
34.

35.

36.

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.

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

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

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

Lee, H.; Kepley, L. 1.; Hong, H.-G.; Akhter, S; Mallouk, T. E. J. Phys. Chem. 1988,
92, 2597.

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

Katz, H. E. Chem. Mater. 1994, 6, 2227.
Dubois, L. H.; Zegarski, B. R.; Nuzzo, R. G. .1. Am. Chem. Soc. 1990, 112, 570.

Chidsey, C. E. D.; Bertozzi, C. R.; Putvinski, T. M.; Mujsce, A. M. J. Am. Chem.
Soc. 1990, 112, 4301.

Collard, D. M.; Fox, M. A. Langmuir 1991, 7, I 192.
Chidsey, C. E. D. Science 1991, 251, 919.

Batchelder, D. N.; Evans, S. D.; Freeman, T. L.; Haussling, L.; Ringsdorf, H; Wolf,
H. J. Am. Chem. Soc. 1994, 116, 1050.

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

-9-

37.

38.

39.

40.

41.

42.

43.

44

45

46

Wollman, E. W.; Kang, D.; Frisbie, C. D; Lorkovic, 1. M.; Wrighton, M. S. J. Am.
Chem. Soc. 1994, 116, 4395 (1994).

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

Obeng, Y. S.; Laing, M. E.; Friedli, A. C.; Yang, H. C.; Wang, D.; Thulstrup, E. W.;
Bard, A. J.; Mich], J. J. Am. Chem. Soc. 1993, 114, 9943.

Finklea, H. 0.; Hanshew, D. D. J. Am. Chem. Soc. 1992, 114, 3173.
Sun, L.; Crooks, R. M. Langmuir 1993, 9, 1775.
Sun, L.; Kepley, L. J.; Crooks, R. M. Langmuir 1992, 8, 2101.

Spinke, J.; Liley, M.; Schmitt, F.-J.; Guder, H.-J.; Angermaier, L; Knoll, W. J.
Chem. Phys. 1993, 99, 7012.

. (a) Vermeulen, L. A.; Snover, J. L.; Sapochak, L. S.; Thompson, M. E. J. Am. Chem.
Soc. 1993, 115, 11767. (b) Snover, J. L.; Byrd, H.; Suponeva, E. P.; Vicenzi, E.;
Thompson, M. E. Chem. Mater. 1996, 8, 1490.

. Kaschak, D. M.; Mallouk, T. E. J. Am. Chem. Soc. 1996, 118, 4222.
. (a) Feriga, B. L.; Jager, W. F.; de Lange, B. Tetrahedron 1993, 49, 8267. (b)

Emmelius, M.; Pawlowski, G; Vollmann, H. W. Angew. Chem. Int. Ed. Engl. 1989,
28, 1445.

-10-

Chapter 2

Synthesis and Characterization of Metal-Phosphonate Materials

2.1 Introduction

As materials science research becomes more focused, the materials required for
study become more specialized. The organophosphonate compounds needed to build MP
films are not commercially available. This Chapter summarizes the syntheses of the
compounds used for the experiments reported in Chapters 4 through 7. We chose a series
of oligothiophenes to use as chromophores. Oligomeric materials are useful as model
compounds in many areas of chemistry, and we use the oligothiophenes here because
their optical response is easily tunable by varying the length of the chain. Bithiophene
(BT), terthiophene (TT) and quaterthiophene (QT) were the oligomers chosen. QT is the
longest oligomer we use, because of solubility requirements and spectral limitations. The
syntheses as well as the layer deposition are performed in solution phase, and the
solubility of oligothiophenes longer than five rings drops drastically with increasing
oligomer length. The other limiting factor regarding Oligomeric materials is related to the
effect of conjugation length on optical response. As the chain length is increased, there is
a point at which the conjugation length can no longer increase, due to the statistical
distribution of structural defects along the chain. At this point, oligomers with longer
chain lengths exhibit approximately the same steady state spectra, as determined by the
average conjugation length of the material. The synthesis of QT and its phosphorylation

were reported by Katz, et 01.,“ and these procedures were modified slightly for the

-1]-

syntheses reported in this chapter. In general, the oligothiophenes were derivatized to the
phosphonate by lithium coupling to a phosphochloridate followed by acid hydrolysis.

We have also bisphosphonated a series of alkanes: hexane, octane, and dodecane,
to serve as optically inactive components in the MP films, in order to separate the
optically active components of the ‘film. Spatial separation of chromophores can be
accomplished by depositing alkane layers between oligothiophene layers, or by making
mixed oligothiophene-alkane layers, thereby diluting the chromophore within the layer.
The lengths of these alkanes in their all-trans conformations are approximately the same
lengths as BT, TT, and QT, respectively, in their all-anti conformations. This allows us
to dilute chromophore layers but maintain uniform height within a given layer. The
alkane bisphosphonates were synthesized by the Michaelis-Arbuzov reaction of

dibromoalkane with triethylphosphite, followed by acid hydrolysism

2.2 Experimental

Purification of Materials. All chemicals were purchased from Aldrich Chemical
Co. and used as received except where noted. All tetrahydrofiiran (THF) used in these
syntheses was distilled over CaH2 and sodium before use. Copper (11) chloride was dried
under vacuum and treated with thionyl chloride, after which excess solvent was removed
by vacuum distillation. Triethylphosphite was treated with sodium and vacuum distilled.

Synthesis of Quaterthiophene. This synthesis is shown in Figure 2.1. QT was
prepared by the reaction of 20 mmol bithiophene in THF with one equivalent of n-
butyllithium (2.5 M in hexanes) at —78 °C for 20 minutes. The mixture was transferred

by canula to a second flask containing a suspension of anhydrous CuCl2 (excess) in THF.

-12-

The clear yellow solution was allowed to warm slowly to room temperature, and was
stirred overnight. The addition of 50 mL of 1 M HCl resulted in the formation of a
yellow precipitate from a brown-green solution, which was vacuum filtered and dried

(64% yield).

/\, U My
\/\/ ———' Ll
\/ \/
W' cm. /\ /\
\/ L' \/ \/

Figure 2.1 Synthetic scheme for quaterthiophene.

Phosphorylation of oligothiophenes. The procedure described below was used for
phosphorylation of BT, TT, and QT. The synthesis of QDP is used as an example in the
text and schematically in Figure 2.2. A solution of QT (3 mmol) in THF was cooled to
—40 °C, and an excess (12.5 mmol) of n-BuLi (2.5 M in hexanes)was added. After
stirring for 2 hours, a second solution of 12.5 mmol bis(dimethylamino)
phosphochloridate and 0.625 mmol n-BuLi in THF was added to the first solution via
canula, and the resulting solution was allowed to warm to room temperature. The
reaction was stopped after three days, the product was extracted from water with ether,
and the aqueous layer was extracted twice with CH2Cl2. The product was purified
chromatograpically on silica plates (Fisher) using a mobile phase of 1 :1 CH2C12 and THF.
(For TT: 95:5 THF and CH2C12; for BT: 50:30:20 EtOAC; MeOH; THF) The most polar
component, which was yellow, was collected in 46% yield. (For TT and BT, the

component that emitted blue upon UV excitation was collected). A portion of this

-13-

tetraamide was dissolved in dioxane at 80 °C. The solution was acidified with 50 mL
each of H20 and concentrated HCl, and refluxed overnight. Upon cooling to room
temperature, bright yellow QDP precipitated from solution and was collected by vacuum
filtration in 55% yield. Final yields and lH NMR peaks are reported in Table 2. 1.
Phosphorylation of alkanes. The same synthesis was used for all alkanes, and
DDBPA is used as an example (see Figure 2.3). DDBPA was prepared by the Michaelis-
Arbuzov reaction of 1,12-dibromododecane with triethylphosphite. The mixture was
refluxed for 5 hours to allow evolution of ethyl bromide. Excess triethyl phosphite was
removed by vacuum distillation. Concentrated HCl was added to the solution, which was
then refluxed for 12 hours, and the resulting white precipitate was collected by vacuum

filtration and washed with acetonitrile. Yields are reported in Table 2.1.

Table 2.1 1H NMR results (500 MHz) and final yields for bisphosphonated compounds.
DMSO-dc was used as the solvent for all spectra.

 

 

 

Compound % Yield 1H NMR peaks
oligothiophenes BDP 10 7.3 — 7.5 ppm (m)
TDP 11 7.2 — 7.4 ppm (m)
QDP 21 7.3 - 7.4 ppm (m)
alkanes HBPA 57 1.2 — 1.6 ppm (m)
OBPA 67 1.2 — 1.5 ppm (m)
DDBPA 68 1.2 - 1.5 ppm (m)

 

 

 

 

 

 

-14-

/\ /\ ,

\/\/Li —’
\ / \ / (2equiv.)
_ / \ / \
'-' Li
\ / \ /
O
L. / \ / \ 0.446”:
I . i + l
\ / \ / L A ”'3
H3O CH3
(2equiv.)
«:Fb)2hP-fg / \ / \ )Q«3Ffifl2
(QM \ / \ / lI‘N‘CH3’2

i? / \ / \ ,NiCHah H+
(CHalzN‘F P—N(CH3)2 —’
(C 1'th \ / \ / C”)

H0413.) / \ / \ [OH
Ha \ / \ / (ii—0“

Figure 2.2 Three-step synthetic scheme for phosphorylation of quaterthiophene: (1)
lithiation of QT; (2) formation of phosphamide; (3) hydrolysis to phosphonate.

-15-

OC HzC H3

/
Br‘<CH2>-Br + CH3CH2O—P\ ———.
“ OCH2CH3
n=6,8, 12
011301120
PICHZTPPOCHzcri,
” OCH2CH3
CH30H2O H+ Ho\ ii
1140112344? \OCH2CH3 fi‘fcwi’wa
” OCH2CH3 o ” OH

Figure 2.3 Synthetic scheme for alkane bisphosphonic acids. (1) Formation of
phosphonate with triethylphosphate; (2) hydroylsis to phosphonate.

~16-

Multilayer Synthesis. Silica substrates, cut from quartz slides, or polished single
crystal (100) silicon substrates were cleaned by immersion in piranha solution (1 H2023
H2804) for 15 minutes, rinsed with distilled water, and dried under a N2 stream. Clean
substrates were used immediately. Under Ar atmosphere, 20 mL anhydrous octane was
added to a roundbottom flask containing the substrate. The octane was heated to reflux,
and 0.2 mL 3-(aminopropyl) triethoxysilane (Petrarch) was added. After ten minutes
reaction time, the primed substrate was removed and rinsed with reagent grade n-hexane
and then distilled water, and dried with N2. An alternative primer layer could be
deposited by overnight immersion in a 0.5% solution of (3-aminopropyl)
dimethylethoxysilane in dry toluene. The amine primer layer was derivatized to the
phosphonate by reaction with 100 mM POC13 and 100 mM 2,4,6-collidine in anhydrous
acetonitrile under Ar for 30 minutes and rinsed with solvent grade acetonitrile. The
phosphonic acid surface was zirconated overnight with a 5 mM solution of ZrOCl2 in
60% EtOH/I-I2O. Subsequent layers were added by sequential reaction with alkane

bisphosphonate (1.25 mM in 95% EtOH, three hours at 55 °C) and zirconium (5 mM in

60% EtOH, 30 minutes) solutions, with extensive rinsing (H2O) between each step. The
pH of the solutions was maintained between 2 and 4.

Optical null ellipsometry. The thicknesses of primed substrates and monolayer
samples were determined using a commercial optical null ellipsometer (Rudolph Auto-
EL 11) operating at 632.8 nm. Data were acquired and reduced using Rudolph software.
For both primer and ZP layers, we used a value of the refractive index of n = 1.462 + 0i

for the determination of film thickness.

-17-

Reflection-Absorption F TIR. Infrared spectra of MP films were obtained from a
Nicolet Magna 550 FTIR spectrometer. The external reflection sample mount and optics
used were obtained from Harrick Scientific Corp. and the polarization of the IR light was
controlled using a Cambridge Physical Sciences broad band IR polarizer. The light was
p-polarized and was incident at 70° with respect to the surface normal of the substrate.

Spectral resolution was 4 cm".

2.3 Results and Discussion

We have synthesized in moderate yield the bisphosphonates needed for the
monolayer and multilayer experiments in this dissertation. Most of the loss of product in
these procedures involved side-reaction of water-sensitive reagents (BuLi,
phosphochloridate) with trace amounts of moisture in the reaction vessels, and the extra
steps taken to ensure the purity of the compounds (chromatographic separations,
recrystallizations). Mechanical losses also were more significant because of the small
amounts of starting material used.

The property of interest for the chromophores in this study is that their spectra are
overlapped, to allow efficient energy transfer from one to another. As illustrated in
Figure 2.4, this condition applies. The spectra of the three oligothiophenes are highly
overlapped. In order to distinguish between the optical responses of multiple
chromophores in a given sample, the excitation wavelengths of each and the emission
maximum of each must be sufficiently resolved, ideally by 50 nm or more. We note that
TDP is not well resolved from either BDP or QDP. However, the emission spectra of

BDP and QDP are sufficiently resolved, and the emission band of BDP and the

-13-

 

 

 

 

 

 

0
II
a Two“
.5 OH
1‘:
B
.E
"U
Q)
.53
E
C
=
O OH

” S /\ S /\ 7,0...

”0;? \/ s \/ s it"

 

 

L l _1 L

400 500 600

wavelength (nm)

 

Figure 2.4 Structures and absorption ( ) and emission (- - -) spectra for (top) BDP;

(middle) TDP; (bottom) QDP.

-19-

absorption band of QDP are still well overlapped. Clearly, the best choice is to use BDP
as the donor chromophore and QDP as the acceptor chromophore in the energy transfer
experiments. We can excite a film containing both BDP and QDP at 315 nm, which is a
maximum in the BDP absorbance and a minimum for QDP absorbance. Emission can be
collected for BDP and QDP at 400 nm and 525 nm, respectively, without significant
interference.

The alkanes synthesized for this study serve two purposes: dilution of
chromophore layers and spacing of chromophore layers. The separation of chromophore
layers within a film with optically inactive material is expected to be important for
controlling energy transfer that occurs between layers. We can control sensitively the
interlayer spacing using alkanes of various lengths. We have confirmed by absorbance
(Figure 4.2) and fluorescence (not shown) that the alkane phosphonates do not have
visible spectra that would potentially interfere with the chromOphores. Each alkane was
purposely chosen to have a length matching closely the length of a corresponding
oligothiophene. We determined the lengths of the molecules using molecular modeling
software (Hyperchem 4.0). The dimensions of the molecules were determined for the
most extended conformations: all-trans for the alkanes and all-anti for the thiophenes.

These theoretical lengths are recorded in Table 2.2.

-20-

Table 2.2 Calculated lengths of alkanes and oligothiophenes. The length is measured as
the distance between the outermost oxygen atoms.

 

 

Length (A) Length (A)
HBPA 12.0 BDP l 1.3
OBPA 14.5 TDP 14.9
DDBPA 19.6 QDP 18.5

 

 

 

 

 

 

We note that the calculated length for undecane-BPA (C..) is 18.3 A, which
corresponds more closely to the length of QDP than C .2 does. We chose to use DDBPA
for this work because of the possibility of odd-even effects in the ordering of the layers,
as has been seen in other types of monolayers. In the case of bisphosphonates, the active
phosphonate groups of even-carbon chains are oriented more favorably for multilayer
growth than they are on odd-carbon chains, as illustrated in Figure 2.5. Using even-
carbon chains is more beneficial for the packing of individual layers and for the stability
and uniform buildup of additional layers in the film.

Prior to starting the energy transfer work, we needed verification that we could
reproduce the ordered MP films that others had synthesized. In this Chapter, we study
the growth of individual layers and of MP multilayer films by building multilayer films
of alkanes. Throughout this work we use substrates with silanol active sites, namely
fused silica and single crystal silicon (100) with its native oxide layer. To anchor the film
to the substrate, we treat the substrate with a short aminoalkylsilane primer. The identity
of the substrate and the primer has been shown to have an effect on the optical properties

of chromophore monolayers, as discussed in later chapters. However, these effects are

-21-

,0
P\O Zr

 

Figure 2.5 Schematic of the buildup of two zirconium-phosphonate layers of (a) odd-
carbon chain and (b) even-carbon chain. The dashed lines indicate the surface normal.

-22-

not relevant in the preliminary characterization of the alkane films discussed in this
Chapter.

The terminal amino groups on the surface are derivatized to aminophosphonates
to provide the active sites necessary for MP growth. The next step is to deposit a layer of
metal ions. MP films have been synthesized using a wide variety of metals, including
Zn2+, Cu“, Y1”, Zr“, and HELP] Zirconium is the most common, because the structure
of the ZP lattice has been studied extensively, and because the complex that Zr forms
with phosphonates is one of the least soluble of the MP complexes. This property allows
fast (5 minute) formation of the interlayer linkages and makes the films especially robust.
Longer times are required for organophosphonate depositions to allow for the ordering of
organic portion of the layer. Byrd, et a1. [4] have demonstrated by atomic force
microscopy that an alkane bisphosphonate layer is well ordered within four hours.

Part of the annealing process within a layer involves conformational relaxation of
the alkanes or thiophenes. One concern associated with this process is the possibility that
a bisphosphonate could form “loops,” connecting both phosphonate ends to the previous
layer and thus terminating layer growth at those sites. This is more of a concern for the
longer alkanes. The oligothiophenes are energetically unlikely to loop to the same extent
because of the sterics involved with the more bulky ring structure and the substantial
electron density associated with the sulfur in each ring. We address the reproducibility of
MP multilayer synthesis by monitoring the growth of films layer-by-layer.

One method we use to monitor the buildup of multilayer films is ellipsometry. In
this technique, changes in polarization of a light beam are measured as it passes through a

film and is reflected off the substrate. The polarization angles are directly related to the

-23-

real and imaginary parts of the refractive indices of the substrate and the film, and the
thickness of the film. There are two factors that afiect how we take these measurements.
One is that the optical constants of the film must differ substantially from those of the
substrate. If the indices are matched, the instrument is unable to distinguish the film from
the substrate, resulting in meaningless film thicknesses. To eliminate this problem, we
use only silicon (n = 3.858 + 0.018i) substrates for ellipsometry and not silica (n = 1.46 +
01'), as the estimated refractive index of MP films is n z 1.51 + 0i.[5'3d] The native oxide
layer on silicon has constant thickness, and is easily corrected for with the software. The
other limitation is in knowing the exact indices of the film. For very thin films, (< 100
A), 11mm and km... do not affect the polarization measurably. These parameters can be
estimated from bulk measurements, but many studies have shown that these values are
not necessarily accurate for the same material in films. Because of the uncertainty in n.~...n
and because our films are thin (20 - 200 A), we use it = 1.462 + 0i, which is equivalent to
that of the native oxide layer. In this case, we measure the MP film and the oxide
together, and manually subtract the oxide thickness for each measurement.

Figure 2.6 shows the results from the synthesis of a ten layer MP film of DDBPA.
We observe linear growth of layers with an average thickness of 19 A. The fact that the
thickness increases monotonically for all layers indicates that looping of the alkanes is
not a factor in these films. If looping was occurring, we would observe a leveling-off in
the ellipsometric data as more active sites were terminated. Assuming the input indices
are approximately accurate, we can geometrically determine the average tilt angle of the
alkanes. The theoretical length of a Zr-DDBPA layer is 3.8 A“1 + 19 A = 21.8 A. This

produces an average tilt angle of 30°, which agrees well with molecular tilts found for

-24-

 

220

200

180

160

l—I t—a
NJ $3
CD CD

thickness (A)

 

 

 

47 1.4 1 4 l 1 1 l 1 1 l 1 J4,. l 1 l 1 1 I

0 l 2 3 4 5 6 7 8 9 10
# Zr-DDBPA layers

 

Figure 2.6 Plot of ellipsometric thickness measured for lO-layer DDBPA film and
regression of data. The zero-layer thickness is that due to the native oxide layer and the
primer layer.

-25-

many other layered systems.

Another technique that is especially sensitive in monitoring thin film buildup is
reflection-absorption FTIR spectroscopy (RAIRS). RAIRS stands in contrast to a typical
transmission experiment, where a spectrum is obtained of the film on both sides of the
substrate and includes potentially large interference from the substrate, which can be on
the order of 105 times thicker than the film itself. In the RAIRS configuration, the IR
beam is reflected off the substrate at a grazing angle, typically 70 — 85°, so that the
absorbance of the film is the main component of the spectrum. At grazing angle, most of
the incident radiation is reflected through the film, in accordance with the Fresnel
equations. The substrate still contributes to the overall spectrum, but not to as great an
extent as in transmission experiments. RAIRS spectra were collected for films on both
silica and silicon substrates, with similar results. The spectra for a representative
DDBPA film on silica are shown in Figure 2.7.

The region of interest in these measurements is the — CH2 stretching region. The
CH2 stretches can provide two important pieces of information about these films. One is
the uniform growth of layers, as evidenced by the linear increases in peak height and area
as layers are added. Another aspect is the relative order of the film overall. The peak
positions of the CH2 stretches provide a measure of the order of the alkane portions of the
film. Using this criterion, many researchers have determined that the order of alkane-
thiol monolayers on gold is relatively high, even superior to that of the same material in
crystalline form. The peak positions for the alkane bisphosphonate layers are v(CH2)a =

2925 cm'1 and v(CH2), = 2854 cm". In comparison to the low frequencies measured for

alkanethiols on gold (v(CH2)a = 2916 cm"; v(CH2)s = 2848 cm") and bulk crystalline

-26-

0.06

 

 

 

 

 

F
0.05 L
=3
3 0.04 -
3?:
'7.
c .-
93
.E
8 0.03 —
E
e l-
O
B
9 0.02 ~
0.01 —
0.00
l l 1 l l L L
2800 2850 2900 2950 3000

frequency (cm?!)

Figure 2.7 Reflection-absorption FTIR spectra measured for layers 4 — 11 of a Zr-
DDBPA film on silica. The dashed line is the spectrum of the silane primer layer.
Spectra were baseline corrected at 2800 cm".

-27-

alkanes (v(CH2). = 2918 cm'1 and v(CH2), = 2849 cm”), the alkanes in MP films are not
well-ordered. It has been proposed that the disorder in alkane MP films is due to lattice
mismatch of the organic and inorganic portions of the layers. The ZP lattice produces
active sites that can accommodate molecules with molecular area up to 24 A2,[6‘7' which
is about the size of a benzene ring. That area is much greater than that required for a
simple alkane chain, allowing for a greater extent of conformational freedom within each

site, and thus less overall order in each layer.

2.4 Conclusions

The compounds required for making some simple metal-phosphonate films have
been synthesized. These compounds and subsequently synthesized films were
characterized by 1H NMR, ellipsometry, FTIR, and steady-state optical spectroscopy.
We found that alkane bisphosphonate films grow monotonically, as evidenced by both
ellipsometry and FTIR. We determined from the ellipsometric thicknesses that the
average tilt angle of the alkanes in a film is 30°, which is common for self-assembled
layers. The FTIR data indicate that, while the layers can be grown uniformly, there is
some structural heterogeneity that decreases the average order within a layer. Between
the uniform layer growth and the favorable overlap of the emission band of BDP and
absorption band of QDP, these materials will be useful for future energy transport

studies.

-23-

25 Literature Cited

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

2. Bhattacharya, A. K.; Thyagarajan, G. Chem. Rev. 1981, 81, 415.

3. (a) Lee, H.; Kepley, L. J.; Hong, H.-G.; Akhter, S.; Mallouk, T. E. J. Phys. Chem.
1988, 92, 2597. (b) Akhter, S; Lee, H.; Hong, H.—G.; Mallouk, T. E.; White, J. M. J.
Vac. Sci. T echnol. A 1989, 7, 1608. (c) Yang, H. C.; Aoki, A.; Hong, H.-G.; Sacket,
D. G.; Arendt, M. F.; Yau, S.-L.; Bell, C. M.; Mallouk, T. E. J. Am. Chem. Soc. 1993,
115, 11855. (d) Zeppenfeld, A. C.; Fiddler, S. L.; Ham, W. K.; Klopfenstein, B. J.,
Page, C. J. J. Am. Chem. Soc. 1994, 116, 9158.

4. Byrd, H.; Snover, J. L.; Thompson, M. E. Langmuir 1995, 11, 4449.
5. Hanken, D. 6.; Corn, R. M.; Anal. Chem. 1995, 67, 3767.
6. Thompson, M. E. Chem. Mater. 1994, 6, 1168.

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

-29-

Chapter 3

Rotational Isomerization Barriers of Thiophene Oligomers in the
Ground and First Excited Electronic States. A 1H NMR and
Fluorescence Lifetime Investigation

' Summary
We report on the isomerization dynamics of a family of substituted thiophene oligomers
designed for the synthesis of soluble poly(alkylthiophene)s possessing a minimal coupling
defect density. The data we report for 3',4'-dibutyl-2,2‘:5',2"-terthiophene, 3",4"-dibutyl-
2,2':5',2":5",2"':5"', 2""-pentathiophene and 3',4',3"",4""-tetrabutyl-
2,2':5',2":5",2"':5"',2"":5"",2"'"-sexithiophene indicate that the ground state isomerization
barrier for all of these oligomers is on the order of 8 kcal/mol, with no discernible
dependence on the length of the oligomer. For oligomer isomerization in the S. electronic
state, we find that the isomerization barrier for both the pentathiophene and sexithiophene
oligomer is ~52 kcal/mol, slightly larger than the value we reported previously for S.
3',4'-dibutyl-2,2:5',2"-terthiophene. Experimental data for both electronic states in these
oligomers are in excellent agreement with semi-empirical calculations using the PM3

parameterization.

-30-

31 Innoducfion

This Chapter focuses on the rotational isomerization barriers of thiophene
oligomers in the context of determining the intrinsic, structural limits to the synthesis of
conjugated polymers possessing a minimum defect density. This is a central issue in the
development of practical organic optical and electronic devices. Understanding and
controlling the extent of structural disorder in a material is also critical to studies aimed at
optimizing the nonlinear optical response of strongly coupled three level systemsml
There are a limited number of conjugated polymers for which it is possible to synthesize
crystalline materials possessing a very low structural defect density. For the majority of
conjugated polymer systems, however, syntheses are by electrochemical or, less
commonly, photochemical means and are often carried out in solution. For this reason,
any amount of structural heterogeneity intrinsic to the monomer units will be expressed in
the resulting polymer. We focus here on poly(alkylthiophene) because of the importance

-8 . . . . . -
'3 1 and nonlinear optical lnformatlon processmg'9

of this family of materials for conducting
'31 applications.

Polythiophenes and poly(alkylthiophene)s have received a great deal of recent
attention for both filndamental and practical reasons. This family of polymers has found
application in the fabrication of organic light emitting diodes and for this reason, there has
been a significant amount of research effort aimed at producing processable materials with
a minimal density of structural defects. The first polythiophenes were synthesized from
thiophene and the dominant defect in this material was from inter-ring linkages between

positions at and B to the S heteroatom on adjacent thiophene ringsml These materials are

also insoluble and, as a result, were of limited utility for device applications requiring

-31-

material processability. The next advance in the synthesis of polythiophenes was to use a
3-a1kylthiophene as the monomernmol While the solubility of the resulting polymer was
improved, head-to-head and tail-to-tail coupling defects contributed significantly to
molecular scale disorder in the resultant material. There have been several successful
attempts to minimize these coupling defects synthetically by using asymmetric monomers,
and it is now possible to make poly(3-alkylthiophene)s where the inter-ring coupling is
~95% “pure”, i.e. head-to-tail coupling.'6'2"23’ An alternative approach to such syntheses
is to use oligomeric forms of thiophenes where head-to—tail coupling defects are precluded
for structural reasons. Recent work by Wang et all” has shown this to be a successful
approach, yielding corresponding polymer material that exhibits conductivities higher than
those reported for poly(thiophene)s synthesized from 3-alkylthiophene monomers. Even
with the development of synthetic routes aimed at minimizing the extent of inter-ring
coupling defects, rotational isomerization between individual thiophene rings in oligomeric
and polymeric thiophenes can give rise to a different class of structural defects. For this
reason it is important to consider the intrinsic inter-ring isomerization characteristics of
thiophene oligomers and to understand how the energetics of isomerization in this family
of oligomers are affected by oligomer length. We have examined the rotational
isomerization characteristics of three substituted thiophene oligomers in order to address
these questions. In previous work in the Blanchard group"' it was indicated that the
ground state isomerization barrier of 3',4'-dibutyl-2,2':5',2"-terthiophene (DBTT) was ~20
kcal/mol based on variable temperature lH NMR measurements. Further investigation of
this system with a higher field (500 MHz) NMR spectrometer has revealed that the data

reported previously were limited in resolution, causing the estimate of the isomerization

-32-

' barrier to be too large. The data we report here possess the resolution required to
establish the ground state isomerization barrier of DBTT, and two larger oligomers, to be
~8 kcal/mol. Measurements of the excited state isomerization characteristics of thiophene
oligomers indicate this energy to be ~5.2 kcal/mol for both the substituted pentathiophene
and sexithiophene compounds. In an earlier report the isomerization energy determined

1‘” We discuss below the

for the substituted terthiophene oligomer was 4.2 i 0.4 kcal/mol.
means used to acquire the experimental data, the form and temperature dependence of the
1H NMR data and the way in which we relate the NMR chemical shift results to the
dynamical response of the molecule, and the viscosity-dependence of the fluorescence
lifetime data for the pentathiophene and sexithiophene oligomers. In addition, we present
semi-empirical calculations'u'zs' for these molecules using the PM3 parameterization. The
PM3 parameterization offers a significant improvement over the AMI parameterization

for these oligomers. The calculated predictions are in excellent agreement with

experimental data.

3 .2 Experimental

Chemicals. The thiophene oligomers used in this study were provided by Prof. E.
LeGoff. A brief summary of the syntheses follows. 3',4'-Dibutyl-2,2':5',2"-terthiophene
(DBTT), was synthesized according to a published procedurem' To synthesize the 3",4"-
dibutyl-2,2':5',2":5",2'":5'", 2""-pentathiophene oligomer (DBPtT), DBTT was
dibrominated at the bi positions of the terminal thiophene rings using tetramethyl

ammonium tribromide and subsequently coupled on both ends to thiophene magnesium

-33-

bromide in the presence of a catalyst. The 3',4',3"",4""-tetrabutyl-
2,2':5',2":5",2"':5"',2"":5"",2""'-sexithiophene oligomer (TBSxT) was synthesized by
coupling two equivalents of DBTT which had been treated sequentially with n-butyl
lithium and copper chloride. A detailed description of these oligothiophenes has been

[29’ For the fluorescence lifetime measurements, all solvents were

reported elsewhere.
purchased from the Aldrich Chemical Company as either reagent grade or spectrosc0pic
grade and were used without further purification. For the 1H NMR measurements, d..-
toluene (99.6 atom% D) was purchased from Isotech, Inc. and used as received.

Steady State Optical Spectroscopy. The absorption spectra of DBPtT and TBSxT
were measured using a Hitachi U-4001 UV-visible spectrophotometer. The emission
spectra for these oligomers were determined using a Hitachi F-4500 fluorescence
spectrophotometer. These data were used to determine the appropriate excitation and
observation wavelengths for the fluorescence lifetime measurements.

NMR Spectroscopy. 'H NMR spectra of DBTT, DBPtT, and TBSxT in d...
toluene, referenced to tetramethylsilane (TMS), were measured using a 500 MHz Varian
(VXR—SOO) NMR spectrometer. The temperature of the samples was controlled with an
FTS Systems air-jet and temperature control unit, and was varied between -40°C and

+90°C for DBPtT and TBSxT measurements and between -40°C and +105°C for DBTT

measurements. Uncertainty in the sample temperature was determined to be 05°C.

Time Correlated Single Photon Counting Spectroscopy. The spectrometer we
used for the fluorescence lifetime measurements of the oligomers DBPtT and TBSxT is
shown schematically in Figure 3.1. A mode-locked CW Nd:YAG laser (Quantronix 416)

produces ~ 11 W average power at 1064 nm, with 100 ps pulses at 80 MHz. An angle-

-34-

 

 

 

 

    

U»;

 

  

bio—u OLD 255

 

>22.
6523.3
830 Loam

 

cuss-966% wasp-:8 cos-3 EwEm “6822.8 SE... _.n 9.5!..—

q 0286.30
Em: 6288:6-

honEoEoocoE
:55 .62 EV

 

BEE-com
so 65:8
938.30g

Es-a

finch-#1-
v...

   

E: coo . E: com

     

.ll

 

vvvv

     

lll

 

ll"

      

‘.

 

(

   

8:8:me
8:6 ~82 05w 3:02-282:

-35-

tuned Type II KTP crystal is used to frequency double the pulse train to 532 nm (~ 750
mW). The 532 nm light excites a cavity-dumped, synchronously pumped dye laser
(Coherent 702-2) which is dumped at a repetition rate of 4 W12, producing ~ 5 ps pulses.
The output from the dye laser is divided using a 90/10 beam splitter into two paths: 90%
of the light is used to excite the sample, and 10% is used for the reference. The beam sent
to the sample is frequency doubled with a Type I LiIO3 crystal. The fundamental is
separated from the second harmonic using color filters, and the excitation polarization is
controlled using a filsed silica half-wave rhomb. (CV1) Emission from the sample is
collected at 90° with respect to the incident light, and the polarization of the emission is
selected using a Gian-Taylor prism. The reflective microscope objective shown in Figure
3.1 is used for low-signal samples, such as the monolayers in the following Chapters. The
polarization of this light is subsequently scrambled, and collection wavelengths are
selected using a subtractive double monochromator. (American Holographic DB-IO) The
detection electronics have been described in detail previously,” and we briefly review
them here. The signal from the detector (MCP-PMT, Hammamatsu R2809U-07) is sent
to a constant fraction discriminator (CFD, Tennelec TC454), where it is processed and
sent to the time-to-amplitude converter (TAC, Tennelec TC864), which we operate in
I reverse mode. The reference consists of the remainder of the divided dye laser beam
(10%), which is sent through a fiber optic delay line, where we control the time offset
between signal and reference channels. The TAC output counts are observed using an
oscilloscope (Tektronix 2230) and are sent to a multichannel analyzer (MCA, Tennelec
PCA-II) for collection. DBPtT was excited at 390 nm (LDS 751, Exciton), and TBSxT

was excited at 406 nm (LDS 821, Exciton). Fluorescence was collected at 547° with

-36-

respect to the polarization of the excitation pulse at the wavelengths indicated in Table

3.1. A representative fluorescence lifetime scan, together with the instrumental response

filnction (typically ~ 35 ps FWHM) is shown in Figure 3.2.

Table 3.1 Excitation and emission wavelengths used for TCSPC measurements.

 

 

 

 

 

 

 

 

DBPtT TBSxT
solvent Am (nm) A.,... (10 nm bandwidth) Am (nm) Ac,“ (10 nm bandwidth)
methanol 3 90 48 1 406 504
l-butanol 390 484 406 507
l-heptanol 390 485 406 508
l-decanol 390 485 406 508

 

 

Data Analysis. The lifetimes we report here are derived from data deconvoluted
from the instrumental response fitnction using software written by Snyder and DemasPO'
The lifetimes we report are averages and standard deviations of at least twelve individual
determinations for each solvent.

[24-28]

Calculations. Semiempirical calculations were performed using HyperchemP

Release 4.0 on an IBM compatible PC (Gateway 2000 486-66V). The PM3

parameterization, used for these calculations, is a modification of the AMI
parameterization that treats molecules containing polar heteroatoms, such as sulfur, more
accurately than previous parameterizations. The calculation strategy for the three
oligomers was to perform an initial optimization of the structures using a molecular

mechanics routine (MM+)[3" followed by a geometry optimization at the semiempirical

-37-

 

1 000

100

Normalized Counts

 

 

 

 

 

 

Time (ns)

Figure 3.2 Example of excited state population decay (I) for TBSxT in n-butanol. The
instrument response (0) time is ~ 35 ps.

-33-

level using an SCF algorithm. Semiempirical optimization was performed until the lowest
energy conformation for each molecule was attained. Heats of formation and electronic
charges of individual protons and carbons were calculated for the geometrically optimized

molecules.

3.3 Results and Discussion

The primary focus of this paper is to obtain information on the isomerization
characteristics of the thiophene oligomers DBTT, DBPtT and TBSxT (Figure 3.3), in both
their ground and first excited singlet electronic states. The motivation for this work is to
understand the intrinsic limits to conformational disorder in polymers synthesized from
these oligomers and, based on this information, determine whether there is a preferred
synthetic strategy for alkylated thiophene polymers. Because the measurement strategy
for the isomerization of a molecule in its ground state differs significantly from that for
excited state measurements, we will treat measurements of rotational isomerization in
these two states separately.

Ground State Measurements. Representative lH NMR spectra are shown for each
of the three oligomers in Figures 3.4 — 3.6. The proton assignments are as indicated in
Figure 3.3. The chemical shift of each proton exhibits a unique temperature dependence,
from 0.0001 - 0.002 ppm/K, as shown for a series of DBTT spectra in Figure 3.7.
Analogous protons on different oligomers exhibit almost identical temperature
dependencies. Measurements of proton resonances that appear furthest downfield were
limited, due to the presence of solvent peaks in the same chemical shift range. In a

previous examination of DBTT in the Blanchard group, it was observed, using a 300 MHz

-39-

/\ s /\

Ha S \/ S a
Hd He
Ha S ‘/\ S /\ S b
\/ s l\/ s \/
Hb HC
Hb Hc Hd He

Figure 3.3 Structures and inter-ring rotations for the three thiophene oligomers. (a).
3',4'-dibutyl-2,2':5',2"-terthiophene (DBTT), (b) 3",4"-dibuty1-2,2':5',2":5",2”':5"', 2""-
pentathiophene (DBPtT) and (c) 3',4',3"",4""-tetrabutyl-2,2':5',2":5",2'":5"',2"":5"",2""'-
sexithiophene (TB SxT).

-40-

.m.m Emmi E 3:865 08 9.968 2: mo mcomamcflmov 2:. .2888 £3865 05
.5; BEE—aims :22; £3, :2on .885 £8863 novcmaxm :85 doom- a :83 Ema no 83.58% 52 Z I. WM 953...—

EAE o _ m m w m c x.
_ _ _ _ _ _ _ _

a- 11

 

 

2% 2e 23 0% of o. 2.
_ _ _ p _

a . é?

a: or «I

 

 

 

 

 

 

 

.41-

E-E o

.m.m 08w?— 5 .8822: 08 2888 2: .8 828—888 2:. .8088 68:88 05

8.. 388-868 888 5:5 8&8 .888 28:88 tonnage-m 88— .erm- 8 28.8 ,3me 8 83828 M22 1. Wm 0.5”:—

 

 

 

 

 

 

 

 

m m e m c N.
_ , _ E- _ _ _ _ L-
4. J5 31.1
_ as. 88 2e 98 88 2:.
_ _ _ A _
D l - 41-
D. x.
..-- L
e: e: e: A

 

 

 

 

c:

 

 

0:

 

 

 

 

 

 

 

 

-42-

 

.m.m 2:»:— E 88685 8.8 2888 2.: .8 8282388 2:. .8688 28:88 of
8.2 388888 888 53> 882 888 68:88 uovcaaxm 88. .erN- 8 88.5 Exmmh .8 828on M22 I. 9m 9...»:—
Eggs

 

 

 

 

 

_ N m e m c e
_ _ _ _ _ _ _ _
ll. fil < \i
2% Se 98 Se 2: S s
_ _ _ _ _
B: 8:

 

 

 

 

c:

 

 

 

0:

 

 

e:

 

 

 

 

-43-

.m.m 083m 5 88685 8
owd 88 2:.
_ _ _

woe-288883 828.8882 o>c 8 =32? .Emn .8 828888 888 63888.8 65 .8 38888 8888883. EM 2:»...—
88 2.8

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6:

 

 

 

-44-

NMR spectrometer, that the several doublet resonances split by ~1.5 Hz appeared to
coalesce and, based on this information, an isomerization barrier of ~20 kcal/mol was
reported for ground state DBTTM Our subsequent measurements using a higher field
spectrometer have shown that the apparent coalescence we reported earlier is actually only
a loss of resolution. The split resonances for a given individual proton do not coalesce,
because the splittings arise from J-coupling between non-adjacent protons. Based on the
higher resolution data we report here for DBTT, DBPtT and TBSxT, we conclude that
ring rotation is much faster than the time scale of the NMR experiment. That is,
coalescence occurs at a temperature much lower than -40°C for these oligomers. The
rotational isomerization barriers for these oligomers are, in fact, significantly smaller than
20 kcal/mol. Despite the absence of a coalescence in these data, the temperature-
dependence of the proton chemical shifts for these oligomers can be used to infer the
approximate energy of the isomerization barrier. We understand the temperature
dependence of the individual proton resonances in these oligomers based on the
temperature-dependent conformational distribution of the rotamers.

It is well established that the chemical shift of a proton bound to an aromatic
system is related to the effective charge on the carbon atom to which the proton is
boundm'm
A6,, =aAqC [3.1]
The effective charge (qc) on the carbon of interest is determined by its electron density,
and the 1H NMR resonance energies (chemical shifls, 6) are also affected by secondary,
inductive effects such as “ring current”.l3“’36' We expect the effective charge on the proton

to play a central role in determining the chemical shift, although we are not aware of any

-45-

explicit treatments of the relationship between proton charge and chemical shifi. We find,
in fact, that the proton charges play a critical role in determining the proton NMR
resonance frequency, and the carbon 7t electron density is important as well (vide infra).

[32‘3” but is not

The proportionality constant a can be determined for a given system,
expected to be constant for all molecules because of secondary, structure-dependent
shielding effects. We understand and model the temperature-dependent proton resonance
energies for the thiophene oligomers as follows. The charge of each proton is influenced
by the inter-ring dihedral angle between individual thiophene rings. Because there is a
finite energy barrier to inter-ring rotation, (Figures 3.8a, 3.9a, 3.10a) the extent to which
each rotamer will contribute depends on the temperature of the system. We have
calculated the conformation dependence of the proton and adjacent carbon charge,
(Figures 3.8b,c; 3.9b,c; 3.10b,c) as well as the isomerization barrier for inter-ring rotation
(Figures 3.8a, 3.9a, 3.10a) for each oligomer using semiempirical molecular orbital
calculations with the PM3 parameterization. We calculated the charge for each aromatic
proton and carbon at 5 degree increments between 0 and 180 degrees. We assume in our
treatment that the there are no anomalies on this energy surface that would give rise to
discontinuities (e.g. sudden polarization phenomenamd”). The average charge of a given
proton or a given carbon for an oligomer isomerizing in solution, then, depends on the
amount of time that the oligomer exists in each rotational conformation. This treatment
presumes that ring rotation is fast compared to the time scale of the NMR measurement
and that we sample a thermally-weighted portion of the isomerization surface. Changes in

temperature thus alter the contribution of high energy portions of the isomerization

surface to the average charges sampled in the NMR measurements. To determine the

-46-

 

 

 

 

VCR
O
8 P \
g o
a £5 ' \
Q) C)
g g 4 l- \ 000000
B ‘5 \0 00000000
& O 1- 1 1 OQOOQOOQOO? 1 1 1 1 1 A
0.16 - b
- ‘\
0.15 ~ \-
.p-I-I-Ikil-I-I-I-l-I-I-I-I-I-I-I-I-I-I-I-I-I-I-l-I-I-I-I-I-I-.-.--_-,.,.
A
O. 14 *- \kk
,_ k -‘-A—A-A-A-A-A-A
L‘L‘. AJAvA‘A‘.‘-‘
% 0.13 .. A-A-A-A—A-A-A-
g L11~1~1111
L) C:
-0.09

‘A-“‘-‘.‘-‘-‘-‘-‘-‘-L‘~LL

'0.12 I 12:313.”:*

A. _ , -.-._
.".‘.-.-.-.-.-.-.-O-.-.-O-O'.“.‘.:‘:t-{... . .
A

I-l‘.‘.'.'.-.-.-.-.-".'.‘-'-"‘.-l-I-I-I-I-I-I’

 

-l-I'
-O,3O . --.’.’.-.'I.
r -
'0'32 1 1 1 1 1 1 1 1 1 1 1 1
O 30 60 9O l20 150 180
Twist angle

Figure 3.8 (a) Calculated isomerization barrier for DBTT, where the isomerization
coordinate is taken as the inter-ring dihedral angle. (b) Dependence of the proton charges
on inter-ring rotation angle; (0) = H., (I) = Hb, (A) = He. (c) Dependence of the
aromatic carbon charges on inter-ring rotation angle; (0) = C., (I) = C1,, (A) = Cc.

-47-

<3'°-o\ a
36 3 ' °\
33 c 0
s: o - \
3 g 0
> R 4 _ \ 0.00000
‘3 .32 o 00' i
——I b O
0 - 1 1 O 100Q009001 1 1 1 1 1 1

 

 

b

0'15 [I-I-I-I-I-I-I-l-I-|-I-I-l-l-I-l-I-I-l-I-I-I-I-I-l-l-I-l—l-l-I-I'l""

 

 

0. 14 ‘M.‘.
\"o-
.- - - .‘Wx:xzx:fi:x:x:x:x:*:x:xzx
0 13 firigztztemwaaaa
Q) +—o-o-o-o-o-o-o-o-o-o-o-o-o—o-o-o-o-oa-o-o-o-o—o-o-o-o-o.o-.-.-.-.,.,.
bag) 1 I J l 4 l 1 I J J 1 I
.C
U i «WAN, C
wwwf'gw-r 111g,» ..
x x—H" kk , VIE-xix? -
‘X—X' A
"“"‘OA: k; .4-0 .-

 

0 30 6O 90 120 150 180
Twist angle

Figure 3.9 (a) Calculated isomerization barrier for DBPtT, where the isomerization
coordinate is taken as the inter-ring dihedral angle. (b) Dependence of the proton charges
on inter-ring rotation angle; (0) = H., (I) = Hb, (A) = He, (1) = H.., (O) = H-. (c)
Dependence of the aromatic carbon charges on inter-ring rotation angle; (0) = C., (I) =
C1,, (A) = Ce, (9*) = Ca, (0) = C1.

-43-

 

8 r- a
36 “K
33 :2? *CK
a: o o
3 g 4 - \
.5 g o\
:6 x 0000
T) V . O\ 0.00000

0. 0000000
0 “' 1 1 09009003 1 4 1 1 1 1 1

 

b

O. 147 q;--I-I-I-I-I-I---I~--I-I-.-I-I-I-I-I-I-I-I-I-I'--I-I-I-I-I-I-I-.-.-._._-

0.140 -

 

 

 

  
  

W
~ Ozapwwik,x-x-x-x-x-x-%-X-x-x-x-x
O. 133 l L‘.“_‘_§-fii:‘_‘_‘_‘.‘-‘-‘-A-A-A-A-A-A-A-A-A-A
°’ 0 126 _
go ' M444-Q-O-O-Q-O-O-Q-O-O-Q-O—O-Q-O-O-Q-O-O-Q-O-o-1-o-.-t.-.-,
.fl
0 -o.110 C

ixwmwmttx: 1 - .

- 1:13-

      

a¢oaa4fi4"'

-0. 120

- ‘I-T-I-I-T-I-I-’-I-I-T-I-I-f-I-I-T-I-I-E-I-I-T-I-I-=-I-I-T-I-I-’-I-

0 30 60 90 120 150 180

 

Twist angle

Figure 3.10 (a) Calculated isomerization barrier for TBSxT, where the isomerization
coordinate is taken as the inter-ring dihedral angle. (b) Dependence of the proton charges
on inter-ring rotation angle; (0) = FL, (I) = Hb, (A) = He, (+) = Hd, (O) = FL. (c)
Dependence of the aromatic carbon charges on inter-ring rotation angle; (0) = C., (I) =
C1,, (A) = Cc, (at) = Cd, (9) = C1.

-49-

TE

\l’

in:

3‘15

fractional contribution of a given conformer to the average charge of a given proton or
carbon, the energy of the oligomer was calculated as a function of inter-ring dihedral
angle. For DBTT, there is one unique isomerization to consider. For DBPtT there are
two distinct dihedral angles relevant to the isomerization surface and for TBSxT there are
three relevant dihedral angles, as shown in Figure 3.3. In this treatment of DBPtT and
TBSxT we assume that the individual isomerization coordinates are nominally uncoupled
(this is a valid assumption because the barriers are small), and thus the average of the two
isomerization barriers provides a usefiJl one-dimensional potential well representation for
these molecular motions on the time scale of the NMR measurement. This assumption
allows us to calculate ring rotation—dependent atomic charges without accounting for
inductive, multiple-ring effects. From these energy surfaces, reduced to one-dimensional
representations of the overall molecular motion, we can determine the contribution
(fractional population) of a given rotamer based on the Boltzmann distribution because the

isomerization is driven thermally.

N
i = cxp(— (£0 - EO)/kBT) [3.2]

Where N9 is the population of the rotamer with dihedral angle 9, No is the population of
the lowest energy conformation, and the energies E9 and E0 are the relative energies of the
two conformers. We assume that the degeneracy of each conformer is the same, and

include no degeneracy weighting factors in Equation 3.2. For each atom of interest, the

average charge density at a given temperature T, (77, weighted for the appropriate

rotamer distribution, can be calculated from

-50-

21"an

__9____
2M)
9

9T:
T

The temperature dependence of (77 results from the Boltzmann terms. We obtain (77 for
each relevant proton and carbon .in these oligomers accordingly. A plot of the
experimental temperature dependencies of the aromatic proton resonances for DBTT is

shown in Figure 3.1 la and a corresponding plot of (77 vs. T for these same protons is

shown in Figure 3.11b. We present the same comparison for DBPtT and TBSxT in
Figures 3.12 and 3.13, respectively. The temperature-dependence of the average charge
on a given proton correlates well with the experimental chemical shift temperature-
dependence. We note that the poorest agreement between experiment and calculation is
for DBPtT, where the approximation of an average isomerization coordinate is expected
to be most limiting. We recognize that the magnitude of the chemical shifi(s) in Hz are
not calibrated, necessarily, with the temperature-dependent charge variation on each
proton, due in part to the inherently qualitative nature of the molecular orbital calculations
we used to estimate the charges. Registration of these two scales (charge and chemical

3 .37 - '
' 2 ' Thus our qualitauve

shift) would be empirical even with the appropriate standards.
agreement between experiment and model is encouraging in and of itself. Previous
models, relating electronic charge to lH NMR chemical shifi have focused on the carbon
atom to which the proton is bound. We observe that the charge on the proton is a
dominant factor in determining the deshielding and thus the NMR chemical shift. The

relationship between the charges on each proton and carbon and the proton chemical shift

will be determined by proportionality constants that are different for each species. Thus

-51-

a

-_— .— -——- —
I I I
'—

3500 -

3450 e /A/‘

3400 — /

Chemical shift (Hz)

3350 1/ ././

 

0.1474 1 b

0.1472 lI—l—I—I—I-—I-I—I—I—I—I—I—I—I-I-I

0.)

CD

a 0.1314 — ‘AA‘

0 ,3

cu A/‘A

on 0.1312 - ,AA

‘13 A)

3 at"

< 1
0.1248 l—

lD—O—O—O—O—O—O—O—O—O—O—O—O—O-O-O

0.1246 -

 

l J l l l l l l l j J J- l l l

240 260 280 300 320 340 360 380
Temperature (K)

Figure 3.11 (a) Experimental temperature-dependence of lH NMR resonances for
DBTT. (b) Temperature-dependence of aromatic proton charges, calculated as described
in the text. For both (a) and (b), (O) = H., (I) = 111,, (A) = He.

-52-

 

 

 

3480 - . u ciao—0‘0
9"“.1:::.. *_*-—-¥—X
A 3440 —"r “##1##”
v X/*/ .
SE 3400 - AA/‘
'5) 1- A/
A
73‘ 3360 r A‘A
5:) 3320 - AA‘ .o-O'"
U ‘ coo-0‘"
3280 .1 1 1 1 1 1 1 1 l 1 1 g 1 1 1 1
b
0.1480tlllllllllllllll
0.1460
3001380 00990000999000.
:3 ..
5 0.1330 —
a) I-
313“ 0.1325 m—A—A—A—A—A—A—A—A—A—A—A—A—;—;——;
a) x x x x x a! 3K 3*
3: 0.13201lE " " "‘ ’K
1
0.1315 .-
0.1260

01240 J 1 l 1 l 1 l 1 l 1 l 1 I 1 l 1
240 260 280 300 320 340 360 380

 

Temperature (K)

Figure 3.12 (a) Experimental temperature-dependence of 'H NMR resonances for
DBPtT. (b) Temperature-dependence of aromatic proton charges, calculated as described
in the text. For both (a) and (b), (O) = H., (I) = Hb, (A) = 1L, (+) = Hd, (O) = H1.

-53-

a

 

 

 

3550 FI—I—I—I—I—I—l—-—I——I—-I—I-—-I
E 3500 -
' *0 ———O—-O-—.
LE 3450 to" */,1/x/ A
U) */*/ /‘
“a " */*/
.0 /alt/ /‘/‘
a 3400 -* /A
<1.) /A __,...-——-.""".
6 b /‘/A 0 "’°’./.
. 1 1 1 1 1 1 1 1 1 1 1 1 1
- b
O_1475 i/I—I—I—I—I—I—I-—I—I—I—I—I
“-1474 r . ./.____.,.._..-+4
y’.’ /
“DQ001350 ,A/O”
.1: _
0 0.1342 ’*’*
.4g-——3K~*‘¥
& . __-_*-———¥""*'"'*
E x—-—**"‘"*
3 0.1323
< __ ____‘__-_‘——-A

0.1250 i
e—o—o——o——o—o——o—o—o—o——o—o—o
1 1 1 1 1 1 1 1 1 1 1 1 A

240 260 280 300 320 340 360
Temperature (K)

Figure 3.13 (a) Experimental temperature-dependence of 'H NMR resonances for
TBSxT. (b) Temperature-dependence of aromatic proton charges, calculated as described
in the text. For both (a) and (b), (C) = H1, (I) = H1, (A) = H1, ( t) = H1, (0) = H...

-54-

Equation 3.1 can be modified to include these dependencies explicitly,

A511 = aHAqH +aCAqC [3.4]
where a” and ac are different. Consideration of either of these charges individually does
not provide the optimum agreement, although it is clear from the experimental data that
the dominant contribution to A6” comes from aHAqH. Consideration of aCAqC would lead
to an improved correlation between experiment and model, as suggested by Figure 3.14.
While we can calculate the terms qc and q”, the determination of each coefficient a is
problematic because of the limited experimental information available. That is, we do not
have sufficient experimental information to determine either ac or an separately. Despite
this limitation, we recognize that the trends presented in the experimental data are
consistent with contributions from the electron densities of both the proton and carbon.
Proton a of DBTT will exhibit relatively little temperature-dependence because the charge
density on the carbon to which it is attached is comparatively rotation-independent. The
temperature-dependence of the proton b NMR resonance will increase in slope due to the
inductive effect of the adjacent carbon, and proton c will exhibit a slight decrease in its
temperature-dependence. The comparisons between experimental NMR data and our
calculated charge temperature dependencies are in excellent predictive agreement with the
relationship between electronic charges and chemical shift indicated in Equation 3.4. The
agreement between the experimental data and our model also provides justification for our
estimation of an isomerization barrier on the order of 8 kcal/mol for ground state
thiophene oligomers, as calculated. A few words are in order at this point on the
meaning(s) of the isomerization barriers we report in Figures 3.8 - 3.10. For DBTT, there

is only one unique dihedral angle, and the barrier height of ~ 1 O kcal/mol calculated for this

-55-

.w,m Eswi E 3 0E3 2: 2m 04:08:96.8 BDE-Am ESQ 25.6 5.338.: a :8 :8. com 025:3:
3:388: m: :Omcrd E8 of. Sm:05:53:12.:quan cEm 78:55 122 I. _S:oE:oaxo 2 coEQEOo .1E1m0
:8 82 .E288 8,353 :088 use-£018 cc“ 32 do; 8920 :905 E2053: SEEKEB 38.330 :6 FEB:

:ossncéoo :onao

 

Av: 23809:“;
com omm owm ovm
- 1 A 1 _ a q a q < u .u\ I:\ In\ 0
e\ I\ Q\ |\ I\
\\ 1
4\4\1\1\ 1

(‘4

I I I IIIII‘I‘I‘I‘I‘I‘I‘.‘

1

I11 I11 I‘ I

 

Ommm

oovm

omvm

comm

:o:=£b:oo :88:

(2H) 131118 19011119113

 

 

 

 

. 488-

mmomd-
VNQ .o-

same/xv

o

CNN—d-
owe—d-

aqu

08—6-

 

 

 

CC 2323 Eek
oom omm owm gm
lJfldrldllltldrldllcrld11¢
com omm owm ovm
1
111111\C\i 1

 

ovm _ .o
wvmfio

N33.
32.

CD CD
9312110 afietoAV

N516
E: .o

-56-

isomerization is determined almost exclusively by steric constraints imposed by the n-butyl
moieties at the 3' and 4' positions of the molecule. Our calculation did not take into
account the ability of these aliphatic groups to move in such a way as to minimize steric
hindrance to the ring rotation, and we thus believe that the calculated barrier is a slight
overestimate of the actual experimental barrier. For the larger oligomers, DBPtT and
TBSxT, the barriers we report in Figures 3.9 and 3.10 are averages of the isomerization
barriers for the 2 (DBPtT) or 3 (TBSxT) dihedral angles involved. We recognize that the
~8 kcal/mol barrier is an over-estimate for the DBPtT terminal ring because steric
hindrance by the n-butyl groups will not play a role. For rotation of only the DBPtT
terminal ring, we calculate an isomerization barrier of ~l kcal/mol. The relatively poorer
agreement between experiment and calculation for this oligomer (Figure 3.12) is therefore
an expected result. For TBSxT, the rotation of each thiophene ring will be hindered to
some extent by the presence of the n-butyl groups on an adjacent ring, and thus the
average of the barriers will provide a reasonably accurate one-dimensional representation
of the isomerization surface of this molecule. Because of the qualitative nature of the
calculations, and the limitations intrinsic to our reduction of these complex, large
amplitude molecular motions to a one-dimensional coordinate, we do not imply that 8
kcal/mol is an explicit upper or lower bound on the isomerization barrier energy for these
molecules. We assert, however, that the experimental barriers must be close to those we
calculate. If the barriers to ring rotation were much higher than 8 kcal/mol, the
temperature-dependence of the NMR resonances would be much weaker because the
oligomers would remain in energetically favored conformations even at elevated

temperatures, i.e. a small portion of the multidimensional isomerization surface would be

-57-

sampled. Conversely, if the isomerization barrier were negligible compared to kBT, then
the isomerization surface would be sampled uniformly for all temperatures and we would
observe a temperature-independent 1H NMR response. The correlation between the
semiempirical calculations and the experimental NMR data also indicates that the electron
density on the protons is at least as important as the charge on the carbons in determining
the proton chemical shift for aromatic systems.

Excited State lsomerization. Dilute (~10"‘ _M_) solutions of DBPtT and TBSxT in
methanol, n-butanol, n-heptanol, and n-decanol were excited at 390 and 406 nm,
respectively, and fluorescence was collected at the wavelengths indicated in Table 3.1.
The lifetimes were observed by collecting fluorescence polarized at 547° with respect to
the exciting electric field to eliminate any rotational diffusion contributions to the data. In
a previous report, the lifetime of DBTT showed a linear dependence on viscosity, which
can be understood in the context of the existence of an excited state rotamer that is
coupled efficiently to the ground state surface. For this model, there is essentially ~90° of
uncertainty in the inter-ring dihedral angle at which efficient relaxation occurs. Many
organic molecules exhibit a minimum energy in their S. isomerization surfaces at an inter-
ring dihedral angle of 90°, because of the decoupling of the 1: system (break in
conjugation) at this dihedral angle.“2"‘3' Relaxation from this conformation is thought to
be rapid, and the limiting rate in the relaxation of the excited electronic state in this model
is therefore mediated by the time required to execute an activated isomerization along the
Si surface. The other possibility is that there is no 90° conformer that relaxes efficiently,
but rather, one conformer is significantly shorter lived than the other conformer. These

two conformers are expected to differ by an inter-ring dihedral angle of ~l80°. In other

-53-

words, the ~syn conformer excited state lifetime is expected to be significantly different
than that of the ~anti conformer. The obvious difference between these conformers is the
extent to which S atoms on adjacent rings can interact. We speculate that such an
interaction, enhanced in the ~syn conformer, would serve to decrease the excited state
lifetime, although there is no direct data to support this speculation. Semiempirical
calculations of the S ; thiophene oligomers (not shown) predict isomerization surfaces that
do not exhibit energetic minima at 90° dihedral angles, but rather indicate the presence of
energetic maxima at 90°. Regardless of the model used, the time required for an excited
thiophene oligomer to relax is mediated by two barriers to the ring rotation; one intrinsic
to the molecule and one imposed by the surrounding solvent medium. The extrinsic
portion of the isomerization barrier, contributed by frictional interactions between the
oligomer and the surrounding solvent, can be controlled by changing the solvent viscosity.
In this model, a high viscosity solvent will impede the intramolecular ring rotation
significantly, producing a comparatively long excited state lifetime. For a low viscosity
solvent, where intramolecular motion is comparatively unhindered, the lifetime is expected
to be shorter due to the greater rotational mobility of the rings. In order to estimate the
intrinsic barrier to ring rotation for the thiophene oligomers, the viscosity dependence of
the S; lifetime was extrapolated to zero viscosity. The experimental data for DBTT were
regressed to a 123 ps zero-viscosity fluorescence lifetime. This lifetime can be interpreted

in the context of an activated process,
_ E. + E ,.
kisom = A , exp( ( ISO"! ”5ka 7‘) [3 ' 5]

using a preexponential factor of IO”, which is appropriate for the n-alcohols. From this

-59-

time constant, EM", was determined to be 4.2 i 0.4 kcal/mol, with the uncertainty in this
value being determined by uncertainty in the value of A. For the longer thiophene
oligomers DBPtT and TBSxT, we find that there is no discernible viscosity dependence, as
shown in Figure 3.15. Despite the outward difference in the appearance of the DBPtT and
TBSxT data compared to the data for DBTT, we believe that the same model applies
because of the structural similarity of these molecules. We are confident in this assertion
based on the similarity of the ground state barriers for all of these oligomers, as well as the
qualitative similarities in their linear optical responses. If the intrinsic barrier to rotation is
greater than the extrinsic isomerization barrier, due to solvent frictional forces, then little
or no viscosity dependence will be observed. The lifetimes for DBPtT and TBSxT are
676 i 4 ps and 779 i 8 ps, from which rotational barriers of 5.1 i 0.4 and 5.2 i 0.4
kcal/mol were calculated. These barriers are approximately I kcal/mol higher than that of
DBTTJ” and we can rationalize this behavior in the context of a more delocalized II
system for the longer oligomers. We expect that the excited state of polythiophenes
should be qualitatively similar to that of other conjugated polymers, such as polyaniline or
poly(phenylenevinylene). For such polymers, the excited electronic state is expected to
increase the bond order of the inter-ring bond(s). The longer the oligomer, the more
capable it will be of supporting a “quinoid-like” resonance structure,'9' and the larger will
be the expected isomerization barrier. The experimental data and intuitive reasoning we
describe above are consistent with semiempirical calculations of the excited thiophene

oligomers, which indicate an S. barrier height of ~5 kcal/mol for both DBPtT and TBSxT.

-60-

720

700

680

660

640

1,, (1)8)

800

780

760

740

 

 

 

1 l 1 l 1 l 1 l 1 l L 1

O 2 4 6 8 10 12

Solvent Viscosity (c?)

Figure 3.15 (a) Viscosity dependence of fluorescence lifetimes for DBPtT. (b) Viscosity
dependence of fluorescence lifetimes for TBSxT.

-6]-

3.4 Conclusions

We have investigated the isomerization characteristics of several thiophene
oligomers in order to understand the dependence of oligomer length on the intrinsic
disorder in resulting polythiophenes. We find that the inter-ring isomerization barrier is ~8
kcal/mol for rotation in the 8.. state of all oligomers we have examined. This estimation is
based on the temperature-dependence of the 1H NMR resonances for the aromatic
protons, and is not derived from a coalescence measurement. For the S. DBPtT and
TBSxT oligomers, we have determined isomerization barriers of 5.1 i 0.4 and 5.2 i 0.4
kcal/mol respectively. These barrier heights are ~1 kcal/mol larger than that found for
DBTT, and we rationalize this difference in the context of the longer oligomers being able
to support a more quinoid-like excited electronic state. Our findings imply that the small
barrier height values we have determined preclude the existence of corresponding long-
term order in the polymers synthesized from these materials. The similarities of the
isomerization barriers in these two states indicates that there is no significant basis for
favoring a photopolymerization scheme over an electrochemical route. We speculate,
based on these data that the ground state inter-ring isomerization barrier for
poly(thiophene)s are on the order of ~8 kcal/mol plus any additional frictional contribution
associated with the motion of associated aliphatic moieties. Thus rotational defects in
poly(thiophene)s do not likely contribute significantly to the material properties because
any mesoscopic or macroscopic material response is averaged over a large number of
rotations of each thiophene ring. If the ground state ring rotations in the polymer can be
treated as an activated process, then the time constant for a ring rotation should be on the

order of 60 ns. In this limit, extrinsic frictional effects in the solid state polymer will likely

-62-

dominate the speed of ring rotation in these materials, but in any event such rotation will

be relatively fast.

-63-

3 .5 Literature Cited

8.

9.

IO.

11.

12.

13.

14.

15.

l6.

17.

18.

. Hambir, S. A.; Blanchard, G. J.; Baker, G. L. J. Chem. Phys. 1995, 102, 2295.

. Hambir, S. A.; Wolfe, D.; Blanchard, G. J.; Baker, G. L. J. Am. Chem. Soc. 1997, 119,

7367.

. Wang, C; Benz, M. E.; LeGoff, E.; Schindler, J. L.; AIlbritton-Thomas, J .; Kannewurf,

C. R.; Kanatzidis, M. G. Chem. Mater. 1994, 6, 401.

. DeWitt, L., Blanchard, G. J.; LeGolf, E.; Benz, M. E; Liao, J .-H.; Kanatzidis, M. G. J.

Am. Chem. Soc. 1993, 115, 12158.

. Wang, C.; Benz, M. E.; LeGolf, E; Schindler, J. L.; Kannewurf, C. R; Kanatzidis, M.

G. Polym. Preprints 1993, 34, 422.

. McCullough, R. D.; Lowe, R. D. J. Chem. Soc, Chem. Commun. 1992, 70.

Guay, J.; Kasai, P.; Diaz, A.; Wu, R.; Tour, J. M.; Dao, L. H. Chem. Mater. 1992, 4,
1097.

Roncali, J. Chem. Rev. 1992, 92, 71 l.

Nunzi, J .-M.; Pfeffer, N; Charra, F. Chem. Phys. Lett. 1993, 215, 1 l4.
Robitaille, L.; Leclerc, M.; Callender, C. L. Chem. Mater. 1993, 5, 1755.
Beljonne, D.; Shuai, Z.; Bredas, J. L. J. Chem. Phys.l993, 98, 8819.

Thienpont, H; Rikken, G. L. J. A., Meijer, E. W.; ten Hoeve, W.; Wynberg, H. Phys.
Rev. Lett. 1990, 65, 2141.

Zhao, M.-T.; Singh, B. P.; Prasad, P. N. J. Chem. Phys. 1988, 89, 5535.
Gamaggi, G. M.; Deluca, G.; Tundo, A. J. Chem. Soc. Perkin II 1972, 1594.

Jen, K. Y.; Miller, G. G.; Elsenbaumer, R. L. J. Chem. Soc. Chem. ('ommzm. 1986,
1346.

Sato, M.; Tanaka, S.; Kaeriyama, K. J. Chem. Soc, Chem. Commun. 1986, 873.

Hotta, S.; Rughooputh, S. D. D. V.; Heeger, A. J.; Wudl, F. Macromolecules 1987,
20, 212.

Nowak, M.; Rughooputh, S. D. D. V.; Hotta, S.; Heeger, A. J. Macromolecules
1987, 20, 965.

-64-

19.

20.

2].

22.
23.

24.

25.
26.
27.
28.
29.
30.
31.

32.

33.

34.

35.

36.

37.

38.

39.

Rughooputh, S. D. D. V.; Nowak, M.; Hotta, S.; Heeger, A. J.; Wudl, F. Synth. Met.
1987, 21, 41.

Rugooputh, S. D. D. V.; Hotta, S.; Heeger, A. J.; Wudl, F. .l. Polym. Sci. Polym.
Phys. 1987, 25, 1071.

McCullough, R. D.; Tristam-Nagle, S.; Williams, S. P.; Lowe, R. D; Jayaraman, M.
J. Am. Chem. Soc. 1993, 115, 4910.

Tour, J. M.; Wu, R. Macromolecules 1992, 25, 1901.
Benz, M. E.; PhD. Thesis, Michigan State University, East Lansing, MI 1992.

Dewar, M. J. S.; Zoebisch, E. G.; Healy, E. F.; Stewart, J. J. P. J. Am. Chem. Soc.
1985, 107, 3902.

Dewar, M. J. S.; Dieter, K. M. J. Am. Chem. Soc. 1986, [08, 8075.

Stewart, J. J. P. Comput.-AidedMol. Des. 1990, 4, l.

Dewar, M. J. S.; Thiel, W. J. Am. Chem. Soc. 1977, 99, 4899.

Dewar, M. J. S.; Thiel, W. J. Am. Chem. Soc. 1977, 99, 4907.

Liao, J.-H.; Benz, M. E; LeGofl‘, E; Kanatzidis, M. G. Adv. Mater. 1994, 6, 135.
Snyder, S.; Demas, J. N; Private Communication.

Allinger, N. L. J. Am. Chem. Soc. 1977, 99, 8127.

Fraenkel, G.; Carter, R. E.; McLachlan, A.; Richards, J. H. J. Am. Chem. Soc. 1960,
82, 5846.

Cobb, T. 8.; Memory, J. D. J. Chem. Phys. 1969, 50, 4262.

Memory, J. D; Wilson, N. K. NMR of A romatlc Compounds, John Wiley and Sons,
New York, 1982.

Pauling, L. J. Chem. Phys. 1936, 4, 673.

Bernstein, H. J.; Schneider, W. G.; Pople, J. A. Proc. Royal. Soc. 1956, A236, 515.
Musher, J. I. J. Chem. Phys. 1962, 37, 34.

Bonacic-Koutecky, V. J. Am. Chem. Soc. 1978, 100, 396.

Salem, L. Acc. Chem. Res. 1979, [2, 87.

-65-

40. Malrieu, J. P. Theo. Chim. Acta 1981, 59, 251.
41. Lam, B.; Johnson, R. P. J. Am. Chem. Soc. 1983, 105, 7479.
42. Rulliere, C. Chem. Phys. Lett. 1976, 43, 303.

43. Awad, M. M.; McCarthy, P. K.; Blanchard, G. J. J. Phys. Chem. 1994, 98, 1454.

-66-

Chapter 4

Dynamics Within a Single Molecular Layer. Aggregation, Relaxation
and the Absence of Motion

Summary

We report on the transient and steady state optical responses of the chromophore 2,2'-
bithiophene-S,5'-diylbis(phosphonic acid) (BDP) incorporated within a single zirconium
phosphonate layer as a function of chromophore density. While the dilute solution optical
response of BDP reveals no anomalous behavior, its characteristics are substantially
different when confined within a monolayer. We vary the concentrations of layer
constituents to determine the extent of interaction between BDP moieties within a single
monolayer. We observe limited initial aggregation of BDP, the extent of which is
determined largely by the conditions under which the monolayer is formed. Over time, the
fractional contribution of BDP aggregates to the total optical response decreases to a
limiting value, implicating surface adsorption site density as the dominant factor in
determining the morphology of the organobisphosphonate layer. Motional relaxation
measurements of BDP within the layer show that the chromophores are immobile on the

hundreds-of-picoseconds time scale of our experiments.

-67-

4.1 Introduction

In this chapter, we consider intra-layer relaxation processes of MP monolayers
from a fundamental perspective using 2,2'-bithiophene-5,5'-diylbis(phosphonic acid) (BDP,
Figure 4.1) as the chromophore. To gain the information of interest here, we vary the
concentration of the chromophore within a Zr-phosphonate (ZP) monolayer. An optically
inactive component can be used either for spacing between chromophore layers or for
dilution of the chromophore within layers. In these experiments, we dilute the
chromophore within a single layer using 1,6-hexane diylbis(phosphonic acid). (HBPA)
This spacer molecule, in its all-trans form, is approximately the same length as the
chromophore, so that the thickness of the monolayer is uniform. (Figure 4.1). We discuss
below the extent of synthetic control we have over the composition of an individual layer,
the steady-state optical response of BDP, and how it is related to the measured dynamical
relaxation processes operating within the layer. Our data indicate that both aggregated
and non-aggregated forms of the chromophore exist within a monolayer, and that the
relative amount of each is determined primarily by the silanized site density of the surface

and not simply by the chromophore density.

4.2 Experimental

Synthesis of Bisphosphonates. 2,2'-Bithiophene-5,S'-diylbis(phosphonic acid)
(BDP) was synthesized by modification of a published procedure on a longer (cum)-
bisphosphonated thiophene oligomer, as reported in Chapter 2. Briefly, bithiophene was
phosphorylated at both termini by lithium coupling to a phosphochloridate, followed by

acid hydrolysis. The synthesis of I,6-hexanediylbis(phosphonic acid) (HBPA) is also

-68-

Figure 4.1 Structures of bisphosphonates used in this work. (a) 2,2'-Bithiophene-5,5'-
diylbis(phosphonic acid) (BDP) and (b) 1,6-hexane diylbis(phosphonic acid) (HBPA).

~69-

reported in Chapter 2. HBPA was synthesized from 1,6-dibromohexane by the Michaelis-
Arbuzov reaction, followed by acid hydrolysis.

Metal-Phosphonate Multilayer Synthesis. Silica substrates, cut from quartz slides,
were cleaned by immersion in piranha solution (I H20213 H2801) for 15 minutes, rinsed
with distilled water, and dried under a N2 stream. The substrate surface was silanized by
refluxing in 1% v/v (3-aminopropyl) triethoxysilane in anhydrous octane under Ar
atmosphere for 10 minutes, followed by thorough rinsing with reagent grade n-hexane.
This amine primer layer was derivatized to the phosphonate by reaction with 100 mM
POCI; and 100 um collidine in anhydrous acetonitrile under Ar for 1 hour and rinsed with
solvent grade acetonitrile. The phosphonic acid surface was zirconated overnight with a 5
mM solution of ZrOClz in 60% EtOH/HzO. Subsequent layers were added by sequential
reaction with bisphosphonate (1.25 mM in 95% EtOH, three hours at 55 °C) and
zirconium (5 mM in 60% EtOH, 30 minutes) solutions, with extensive rinsing (H20)
between each step. The pH of the solutions was maintained between 2 and 4.

Steady-State Optical Spectroscopy. The absorption spectra of BDP monolayers
were measured using a Hitachi U-4001 UV-visible spectrophotometer. The samples were
held vertically in place at 45° with respect to the incident beam and were collected with 5
nm spectral resolution. The emission spectra of the monolayers and solutions were
collected using a Hitachi F-4500 fluorescence spectrophotometer. Monolayer samples
were excited at 320 nm through a 10 nm slit. The emission bandpass was adjusted
according to spectral intensity. The samples were held vertically, at 45° with respect to
the excitation beam propagation in such a way as to minimize reflection of excitation

radiation into the emission collection optics.

-70-

Time Correlated Single Photon Counting Spectroscopy. The spectrometer we
used for the fluorescence dynamics measurements of BDP solutions and monolayers is
described in detail in Chapter 3, and we present here only a brief overview of the system.
The light pulses used to excite the sample are generated with a cavity-dumped,
synchronously pumped dye laser (Coherent 702-2) excited by the second harmonic of the
output of a mode-locked CW Nd:YAG laser (Quantronix 416). All samples were excited
at 320 nm (Kiton Red, Exciton; LiIO; Type I SHG). Monolayer samples were held
approximately horizontally, with 5° tilts away from horizontal in two directions: toward
the excitation beam and toward the detector. Fluorescence from the sample was imaged
through a reflecting microscope objective. Lifetimes were collected across the emission
bands at 547° with respect to the excitation polarization for solutions, and without
polarization bias for monolayers. Fluorescence was collected at 390 nm (5 nm FWHM
bandwidth for solution, 20 nm for monolayers) with polarizations of 0° and 90° for
rotational diffiision dynamics measurements on all samples.

Data Analysis. The lifetimes we report here were fit to sums of exponentials using
Microcal Origin v. 4.1 software , and are reported as the values and uncertainties of the fit
to individual decays. Rotational diffusion information for solution measurements is
reported as the average of six data sets with their associated 95% confidence limit
uncertainty. For monolayers, eight pairs of alternating parallel and perpendicular scans
were used to produce an average anisotropy fiJnction. R(0) was determined to i 0.01 by
regression of data at times after the instrument response, and several data sets were

averaged for each sample.

-7]-

ll-5l

Calculations. Semiempirical calculations were performed using Hyperchem

Release 4.0 (Hypercube, Inc.) on an IBM compatible PC (Gateway 2000 PS-120). The
PM3 parametrization, used for these calculations, is a modification of the AMI
parametrization that treats molecules containing heteroatoms, such as sulfur, more
accurately than previous parametrizations. An initial optimization of the structure was
performed using a molecular mechanics routine (MM+)'6' followed by geometry
optimization at the semiempirical level using an SCF algorithm. The torsion of the 2,5'
BDP 0 bond was set at 10° increments, and semiempirical optimization was performed
until the lowest energy conformation for the fixed inter-ring bond torsion was attained.
The heat of formation and So <—> S. transition energy were calculated for each

geometrically optimized conformation.

4.3 Results and Discussion

We are concerned with understanding the optical response of (0am)-
bisphosphonated chromophores, specifically oligomeric thiophene derivatives, as
constituents of metal-phosphonate multilayer structures. As discussed above, the
structural regularity attainable in these systems allows for the selective examination of
intermolecular interactions between layers and within layers. Our initial work in this area
focuses on the simplest case, that of a single layer containing a variable chromophore
density. We want to discern, ultimately, how intermolecular interactions and
photophysical processes within one layer will affect the optical response of, and relaxation

dynamics within, more complex layered structures.

-72-

/1_—

 

There are several factors that can, in principle, contribute to the optical response of
this system. One factor is the stoichiometry of the monolayer deposition, which we can
determine using UV-visible absorption spectrometry. The spontaneous emission response
of the monolayer is also a rich source of information on how the components of the
monolayer are assembled and how they interact. Fluorescence lifetime and reorientation
dynamics data obtained using transient emission spectroscopy provide insight into the
local environment(s) of the chromophores within a monolayer. Semiempirical calculations
performed on the chromophore aid in our interpretation of the experimental data and thus
our understanding of monolayer properties. We consider in the following sections each of
these aspects in the characterization of mixed ZP monolayers.

Stoichiometry of Monolayer Formation. Little is known about the competition
between different bisphosphonates for active metal ion sites. There is some evidence that
the formation of different mixed monolayers does not follow the stoichiometry of the
solution from which they are formed precisely,m and thus we need to determine whether
or not such is the case for the BDP/HBPA system. In addition to complications arising
from competition for binding sites, the stoichiometry of a multi-component deposition
solution may not be reflective of the resulting monolayer concentration if solution phase
aggregation of one or more of the constituents is occurring. For these reasons, it is
important to establish a direct means of calibration for this particular system. To study the
deposition stoichiometry of mixed monolayers relative to the solution from which they are
formed, we varied the proportions of the chromophore (BDP) and the optically inactive
component (HBPA) in ethanol solution and compared these data to absorbance

measurements of the resulting monolayers. The total concentration of bisphosphonates

-73-

(BDP + I-IBPA) in the deposition solutions was 1.25 m_M_, with the concentration of BDP
fixed in increments from 1.25 m1\_4 (100%) to 0.0125 mM (1%). The absorption spectra
of the resulting monolayers are shown in Figure 4.2. These data are background
subtracted using a separate primed and zirconated substrate as the background (Figure 4.2
inset). This background spectrum is identical to that of a pure HPBA monolayer. The
raw spectra exhibit a slight sample-to-sample variation in absorbance baseline, which can
be attributed to slight absorptive differences between individual quartz substrates and
positioning of the sample in the instrument. To correct for this baseline variation, the
absorbances reported in Figure 4.3 are the differences between the absorbance at A...“ and
the interpolated baseline. A linear regression fits the data well, indicating that there is a
direct relationship between the deposition stoichiometry and the fractional concentrations
of the components present in the deposition solution. These data reveal that the
correspondence between solution composition and monolayer composition is not one-to-
one. Beer’s law requires there to be no residual sample absorption at [BDP] = 0, and the
experimental background corrected data produce a linear regression to a measurable
absorbance. Clearly for [BDP] = 0, AB...) = 0, and the regression to a nonzero intercept
implies that monolayer composition is offset from solution composition. In other words,
for low fractional BDP concentrations (< 1%), the relationship between solution and
monolayer composition is curvilinear. From these data we can establishing a working
relationship between solution and layer composition that is straightforward. For
monolayer formation where BDP is the only phosphonate containing constituent in the
deposition solution, we assert that a limiting layer of the chromophore is present, and we

know experimentally that a pure HBPA monolayer represents a zero-absorbance

-74-

condition. We further assert that Beer’s Law holds for this system so that the absorbance
of monolayers formed from mixed solutions is indicative of the chromophore
concentration. For high fractional chromophore coverage, a direct correspondence
between solution and monolayer composition is recovered, but for lower fractional
coverage, we find significant preference for adsorption of BDP compared to I-{BPA (Table
4.1).

Table 4.1 Comparison of fractional BDP concentrations in deposition solution and in the
resulting monolayer, determined according to absorbance.

 

 

% BDP in % BDP in
deposition solution monolayer
100 100
50 58
10 34
S 29
l 26

 

 

 

 

This is not an unexpected result. For many ionic association processes, the formation
constant of the complex depends sensitively on the identity of the species involved, and
there is no reason to expect that not to be the case here. This result is observed, at least in
part, because there is a ~25 fold stoichiometric excess of BDP relative to the number of
substrate active sites (~15 x 10”) for the 1% BDP deposition solution. This ratio
correlates remarkably well with the observed monolayer concentration for this solvent
system. We expect that it would be possible to achieve lower fractional coverages of the
SiOx surface by using smaller absolute amounts of BDP in the deposition solution, and we

recognize that the relative efficiency of adsorption of both species will be determined by

-75-

()136

 

 

 

 

 

 

0.06'
0.04L
- (1025\\\\\‘\\_-,.ffinu\vflf"
. l
0'00 A 360 C 357) 460 A 450
6 ‘-
3 0.04 - ,/ ‘.
V , \
a) I ‘,
c) ,I ,
5 . ~,
—e I x
8 l I, ‘
.ID ‘, I \
CU \ I \
’o \ s I] \
Q) \ , \
'EE , , .-q_ '.
E 3- 1" \'-r- '-
C) ()112 .\.._.. .5 fl. \\K
\ - - .
C -- \ . . .l. \_\1 \
\ ‘ \.’./ \:. ..._.. I \
ik-cc. ,
\-_,- ', ..'- I
b \k}, \1. \
ta.”
'1 \~ . \
\1“;\... . ‘
-/~\.: .1 I: ‘
o oo . 1 1 I . I 4
250 300 350 400 450
wavelength (nm)
Figure 4.2 Background corrected absorbance spectra of BDP monolayers: (—) 100%;
(- - - -) 58%; (- - - °) 34%; ( ----- )26%. Inset: Background spectrum.

-76-

0.05

 

 

 

 

0.04 - [I
a h I I
:3. , ’
3 , I
Q) 0.03 '- I
0 I I
a I, . .....
"E ,
O ' I
.8 x
N /
“U , ’
g 0.02 - ’1
'EE ,,’
E .13. ............
O .
S: ,4 ..............

L1. ...............
0.01 -
0.00 l l I .1 . I l :1 1 4L 1
0 20 4O 6O 80 100

% BDP in monolayer

Figure 4.3 Dependence of monolayer absorbance on concentration of BDP in deposition

solution. Solid line:
experimental data.

Beer’s Law predicted dependence, dashed line: Regression of

-77-

the solvent system.

The monolayer emission intensities varied with each sample, but did not follow any
trend, suggesting that the quantum efficiency of the monolayer samples is related to some
factor(s) in addition to the amount of chromophore present. Part of the variation, again,
may be due to inconsistency in the positioning of the sample. The relatively complicated
behavior of the emission response for this system is also not a surprising result, and the
details of this complexity contain useful chemical information as we will discuss below.

Monolayer Composition. The dilute solution behavior of BDP is typical of a
simple chromophore. Its absorption and emission spectra have well-defined bands, and it
exhibits a single exponential fluorescence intensity decay in time. With monolayer
samples, a double exponential decay is observed, with a fast component that is the same as
the solution lifetime, ~200 ps, and a slow component, ~1 ns, as shown in Figure 4.4. This
decay functionality is observed for all monolayer samples, and the time constants
recovered for each component are the same over the entire concentration range studied, as
shown in Figures 4.5a and b. We observe that, as we collect lifetime data at several
wavelengths within the BDP emission band, the form of the decay changes, as indicated
Figure 4.6. Fitting these data recovers the same two time constants, but the fractional
contribution of each decay component varies as a function of the wavelength at which the
data are collected. Specifically, the relative contribution of the longer lifetime increases on
the red side of the emission band for all samples. (See Figures 4.7a and b).

Such behavior has been observed previously for different chromophores imbedded
in ZP layers, although its origin was not determinedm One explanation for the existence

of two lifetime components in these monolayers is the interaction of the chromophore with

-73-

 

1000

 

 

 

 

Z 8
I 5%

A .
if:
c:
8
3 100 :—
.é‘ I
(I)
a .
8 r
.E .
t: 8
.2 . 0
.2 8
g o
B 10 To?
.2! :2
"g 38'
o l
:

If

I 1 J 1 l
0.0 0.5 1.0 1.5 2.0
time (ns)

Figure 4.4 Representative fluorescence lifetimes of (O) 10" M BDP solution and (O) 26
% BDP/HBPA monolayer collected at 390 nm.

-79-

2000 l—

1600 -

lifetimes (ps)

p—A
N
O
O
1

800g 5

240 -

 

 

hid

200 B—.-—-_--i ------------------------

 

160 -

1.00

0.95

A1/(Al+ A2)

0.85

 

0.80

J

soln

0

20 40 60 80 100

% BDP in monolayer

Figure 4.5 (a) Fluorescence lifetimes of BDP monolayers and solution at 400 nm. The
data were fit to the function f(t) = A.exp(-t/t.) + Azexp(-t/t2). (b) Normalized prefactors
Offast exponential decay component for data taken at 400 nm.

-80-

 

 

 

 

OUDDWnWlWDQqukM OO QWOO “I. III
0 00 I In I I
wwwosMss so 09%- Nu..." ._.._
WW.- 3. o 00 h...
@ 3a.... a... .
0 mm wo‘o\o o In“
DfluDMHWMn- AWmfi hru“.13- 1
m o
no chum MWOR on D Slot!
B o 00 in.-
plmfim .I©I_II‘. I ‘- - —. . p P p . p p l-
m m

@553 33:85 coma-Eu 3538.5:

1.0

0.5

0.0

time (ns)

Figure 4.6 Fluorescence lifetime decays of BDP monolayers at (I) 390 nm, (0) 475 nm,

(0) 525 nm, and (D) 575 nm.

-81-

 

 

 

 

? fir—
1600 - .
a 1200 "" __l_. E: E
a.
.E 800::
E D
T 240 l-
’ E
200 l- I I I I
’ I
160b1l1111111l111
IHOO -
0.96 --E E E
<6 .92 - E
+—
3 ' I
45' (188 -
(184 P
080 1 I 1 J 1 l 1 l 1 l 1 l 1

 

400 420 440 460 480 500

emission wavelength (nm)

Figure 4.7 Emission wavelength dependence of (a) lifetimes and (b) Normalized fast
exponential prefactors for 29% BDP monolayer.

-32-

active hydroxyl sites on the silica surface. Wang and Harris found similar lifetime behavior
for pyrene on silica, and determined that it was due to the existence of two distinguishable

active silanol sites on the surfacem

To explore the possible role of chromophore-
substrate interactions, we synthesized a multilayer structure with a ~26% BDP layer
spaced away from the SiOx surface by three HBPA layers. This structural motif prevents
the chromophore from interacting with the surface. The same lifetime measurements were
made, and the results were identical to those of the monolayers that were not isolated
from the substrate, indicating that the presence of the second lifetime component is not the
result of a single chromophore interacting with different silica sites on the same surface.
We consider next the possibility that we are observing emission from an aggregate
or excimer moiety that forms at short intermolecular spacings, such as are found in a
monolayer. Aggregation has been observed for several molecules, such as pyrene, which
is known to form excimers in concentrated solutions and at solid interfaces.llol The steady
state emission spectra of the monolayer samples are consonant with this explanation, as
we find two distinct classes of spectra; one which has been observed at all chromophore
concentrations and is similar to the dilute solution spectrum, and one seen for
chromophore monolayer concentrations above 34%. (See Figures 4.8a and b). This latter
feature, which we designate as the aggregate spectrum, is broader and red shifted by about
50 nm from the dilute solution, or monomer, spectrum. The intensity of the monomer
emission spectrum is several times that of the aggregate spectrum, although quantitative
intensity comparisons are difficult. We observe both monomer and aggregate peaks in the

100% BDP (1.25 mM) deposition solution, implying that the resulting monolayer will be

comprised of both monomer and aggregate, with the ratio of the two forms being

-33-

1200

1000

800

 

600

400 _

 

200

 

 

4000

3000

emission intensity

2000

1000

 

 

 

 

wavelength (nm)

Figure 4.8 Emission spectra of fresh samples (solid lines) and aged (two-months) samples
(dashed lines) for (a) aggregate-like and (b) monomer-like samples. The vertical lines are
guides for the eye to the band maxima.

-34-

determined, at least initially, by their deposition kinetics. We note that, occasionally, a
monolayer synthesized from concentrated BDP solution yields a monomer-like spectrum,
which we believe is due to incomplete monolayer formation.

The monolayer aggregate spectra anneal over time to more monomer-like spectra,
but samples that were characterized initially by a monomer-like emission response stayed
the same except for diminution of intensity, as seen in Figure 4.8. There is evidence in the
literature that the packing density of the organic portions of the ZP structures is mediated
by the lateral spacing of the inorganic inter-layer linking fiinctionalities.""'3' The ZP
lattice has been shown to cause disorder in layers because the spacing it enforces is larger

“4"” This means that the spacing

than the molecular area of most organic moieties.
between individual chromophores may be larger than the intermolecular distance that 'is-
optimum for aggregate formation, so that, even if the chromophores are deposited onto
the surface as pairs, they are likely to separate over time, giving rise to the spectral
“annealing” we observe. This finding is fully consistent with our FTIR data on HBPA
monolayers on SiOx (see Chapter 2) indicating some structural heterogeneity within the
layer. Also, we detect an apparent photodegradation of the chromophores within the layer
(Figure 4.8a) which may indicate that the aggregates, as well as the monomer, are being
destroyed, resulting in a sample with a higher proportion of monomer and a spectrum
reflecting that change.

We expect that, based on the ionic nature of the metal-phosphonate bond(s),
surface dissociation and diffusion will play a role in achieving the annealed condition for

these layers. Our linear response data, however, do not provide significant insight into the

role of diffiision for this system. If bound BDP aggregation was mediated by surface

-35-

diffusion, we would observe aggregate-like absorption spectra in proportion to the surface
concentration of the BDP for the annealed layers. Instead, we observe that the stable
long-term condition for these films is in the form of monomeric chromophores, regardless
of the initial BDP concentration. While the implications of these data remain to be
explored more completely, it is clear that, for this system, surface diffusion is not the only
process at work in determining the absorption properties of these films.

The existence of two measurably different spectra based on two chromophore
species can be understood from a structural perspective with the aid of semiempirical
calculations. We calculated the inter-ring rotational energy barriers for both the ground
state and the first singlet excited state of BDP, as shown in Figure 4.9. The 5.. potential
energy surface is essentially independent of thiophene inter-ring torsion angle, with
rotational barriers of less than 1 kcal/mol. However, the S. surface has 12 kcal/mol
barriers at 90° and 270° (rings perpendicular). The energetic minima occur at 0°, the anti
conformer, and 180°, the syn conformer, where the molecule can adopt a quinoid-like S.
resonance structure. In a system where there is freedom of motion, such as in solution,
the conformational distribution of the chromophore is unbiased with respect to inter-ring
rotation prior to excitation. Upon excitation, the conformational distribution of BDP is
biased toward the S. minima, where the molecules take on the planar, quinoidal structure.
Planar BDP chromophore pairs are more likely to aggregate than non-planar species if
they are in close proximity, by a dipole coupling mechanism. In a monolayer, we expect
different behavior, because the chromophores are not able to rotate freely (vide infra). 1n
the monomer limit within a monolayer, chromophores are spaced such that the probability

of intermolecular interactions between chromophores is low, so the chromophores are

-86-

 

 

-79506 -
. . , .o-o
_ /0000 o O\
O 0
-79509 - o/ / \
o o
’ / \ / °
-79512 '— /O O O \
' \ / 00
o 0
-79515 - /O/ \O o/ \o
'5 r o \ / \o
E l 'O/ 0\o 0’0 \O‘o
% -79518 ‘0'
o
.3:
V "I
I
EB «r
a)
5 79602 l-
23.
2 .-
‘3
73 -79605 —
o
1-
-79608 -
~79611 r-
3.0.0.0.0.0-U'U'U'D‘D°D-D-D.D.O.DID.D.D.D.DODID.U,D-D-D-D-D-D-D.D.D.D.D.D
-79614 -
1 1 1 1 1 J 1 l 1 l 1 l
0 60 120 180 240 300 360

torsion angle

Figure 4.9 Calculated rotational energy barriers for BDP: (D) So and (O) S..

-37-

excited from an average of all conformations, resulting in an emission spectrum with a
distribution of energies skewed to high energies by S. emission contributions from twisted
conformers. In more concentrated monolayers, chromophores are likely to be close
enough to allow interaction which would serve to fix the molecules in more planar
conformations, and resulting in the lower energy spectrum. In addition, aggregation is
expected to yield a spectral red shift, as is seen for the pyrene excimer.“°'

Rotational Dynamics. The rotational dynamics of the chromophores within a
monolayer are an information-rich component of the optical response. Molecular
reorientation measurements are well established as sensitive indicators of intermolecular

“6'2” Most rotational diffusion measurements are performed on probe

interactions.
molecules in solution, where the orientational distribution of the chromophore can be
completely random while for a (partial or complete) monolayer of the same molecule, we
expect its motional freedom to be constrained both by surface attachment and by
interactions with neighboring molecules. For such a system, the hindered rotor model is

mm In this model, the surface-bound probe

appropriate for interpreting anisotropy data.
molecule is confined to a cone of rotation, where the base of the cone is defined by the
attachment point of the probe to the surface and the cone semiangle is designated as 00.
The optical and geometric properties of the tethered probe molecule also affect the form
of the experimental signal. We select a non-random orientational distribution of
chromophores within the monolayer with a laser pulse and monitor the change in emission

intensity as a function of time following excitation. This measurement is made for

emission signals polarized parallel and perpendicular to the excitation polarization and

-33-

these data are combined to produce the induced orientational anisotropy function, R(t),

 

(Figure 4.10)
_ IMO—11(1)
R(t)" I"(t)+21i(t) [4"]

The function R(t) is related to the properties of theprobe molecule through‘23 '

R(t)=%a(cos6)(a[y(0)-p‘(z)]> [4.21

where 6 is the angle between 1.1 and u“, the excited and emitting transition moments and P2
is the second order Legendre polynomial. For the hindered rotor model, it is important to
consider two limiting cases, t = 0 and t = 00. At t = 0, where the second term in Equation
4.2 is unity, the information available from the measurement is determined by the optical
properties of the probe molecule and thus we can determine 6 directly. For BDP in
solution we measure R(0) = 0.24, yielding 5 = 31° i 2°. At t = 00, any re-randomization
of the chromophores will be complete and the information available from the measurement
is related only to the average tilt angle of the chromophore ensemble within the layer

according to the hindered rotor modelml

2 2
R(oo) : 313(cos 19,, )P2 (cos 6e",)(P2 (cos 19)) [4.3]
In Equation 4.3, 0... is the angle of the excited transition moment with respect to the cone
center axis, 0,... is the emitting transition moment angle with respect to this same axis and
0 is the average tilt angle of the chromophore within the cone. The terms for 0... and 0....

are, together, representative of the angle 5, and for simplicity, we take 0... = 0 and 0.... = 5.

From these quantities we can extract the average tilt angle, 9, from the experimental data.

-39-

4000

2000

emission intensity

 

 

 

0.2 — a":

R(t)

 

 

 

 

 

 

 

0.5 1.0 1.5

time (ns)

Figure 4.10 Reorientation data for a 26% BDP monolayer. (a) Raw transient signal
intensities for emission polarized parallel (upper) and perpendicular (lower) to excitation
pulse polarization. Included is the instrumental response function. (b) Induced
orientational anisotropy function derived from the data presented in panel (a) and
regressed fit.

-90-

In the hindered rotor model, there can also be a measurable relaxation associated
with the randomization of chromophore orientations within the confining cone, with the
time constant of this relaxation being related to the motion of the chromophore about its
tether and the semi-angle of the cone. It is evident from the experimental data presented
in Figure 4.10 that we do not detect a transient relaxation of the chromophore within the
monolayer, once contributions from the instrumental response function are accounted for.
This finding is important in and of itself because it implies the rigidity of these systems on
a ~nanosecond timescale, and from the infinite time response, we can derive USCfiJl
information on the average tilt of the chromophore in the cone.

It is fair to question whether the absence of a decay in R(t) is reflective of the
absence of dynamics or the presence of unresolved, fast motion. We expect that the
motion of BDP in a low viscosity solvent will place a qualitative lower bound on the time
constant for any such motion in the monolayer. For BDP in 95% ethanol, the
reorientation time we measure, 10R = 299 i 41 ps, is on the same order of magnitude as
the monomer lifetime. This reorientation time is significantly slower than would be
predicted by the Debye—Stokes-Einstein model‘z‘” in the stick limitlzs' (60 ps) and we
attribute this difference to partial deprotonation of the terminal phosphonate moieties.
Ionic charge will give rise to strong association with the solvent, slowing rotation of the
chromophore. In a monolayer, the dynamics are expected to be mediated by attachment
of the chromophore to the substrate as well as by dipole-dipole interactions between
molecules and by the proximity of its neighbors. We expect that, if the chromophore were
to exhibit a dynamical response within the monolayer, it would occur on a timescale at

least similar to that seen for BDP in ethanol, if not longer. We believe that the absence of

-91-

a measurable decay in R(t) is indicative of a highly rigid environment. We recover the
same value for the steady-state anisotropy for all monolayer concentrations studied, as
shown in Table 4.2, another indication that the chromophores are unable to randomize on

the timescale of the BDP excited state lifetime.

Table 4.2 Dependence of infinite time anisotropy on fractional composition of monolayer.

 

 

 

 

 

% BDP R(oo)
100 0.06 1 0.01
58 0.05 i 0.01
34 0.07 1 0.01
26» 0.05 x 0.01

 

From the monolayer data, R(oo) = 0.06 i 0.01, yielding 0 = 35° : 2°. This value for the
tilt angle is similar to that recovered for alkane bisphosphonates'Z"I but, for the BDP
chromophores, indicates that organization within the layer is not determined fully by the
structure of the (presumably all-anti) chromophore, which would suggest a tilt angle of
~l7°. It appears that there is either a contribution to the measured tilt angle of BDP from
the metal-phosphonate coordination or from intermolecular interactions between the
organic functionalities of the bisphosphonate species. Unfortunately, we do not have
sufficient knowledge of the surface roughness or active site spacing to resolve this
question, especially in light of the several possible coordination numbers of Zr" .‘27'

We can use the rotational dynamics results to clarify some ambiguity about the

origin of the aggregates. One explanation for the presence of aggregate is that all of the

-92-

chromophore within the layer is initially monomer. Upon excitation, BDP relaxes to its S;
conformational minima, where it is planar, and if other planar chromophores are in close
proximity, they will aggregate. The second possibility is that the chromophore is pre-
aligned as aggregates in solution, and is deposited onto the surface as such. We found no
evidence of motion in the dynamical data, likely ruling out the possibility that aggregate
formation in BDP monolayers is driven by optical excitation. We note also that, if
aggregation were mediated by excitation of BDP, then the measured population relaxation
dynamics would be more complex than the double exponential decay functionality we
recover, and we would expect a chromophore concentration dependence, which we fail to
observe. We believe that the deposition of pre-formed aggregates is more likely, based on
both the dynamics data and the observation of monomer and aggregate bands in

concentrated BDP deposition solution spectra.

4.4 Conclusions

Our initial investigation of the optical response of a chromophore-containing ZP
monolayer indicates that these assemblies are ideal for the selective examination of inter-
layer and intra-layer energy relaxation phenomena. We have demonstrated control over
the concentration of chromophore in the layer, which is limited by the higher affinity of
BDP for the metal binding sites. Some degree of aggregation occurs in the BDP
monolayers, as verified by the presence of the two species detected with both the steady-
state and fluorescence lifetime spectroscopies. The presence of aggregated chromophore
does affect the population relaxation dynamics within the monolayer, but in a way that can

be monitored. The time constants for the fluorescence decays remain concentration

-93-

independent, ~200 ps for the monomer and ~l ns for the aggregate, although the
contribution of each species’ lifetime varies across the emission band. The rotational
motion data are constant within the chromophore concentration range of 26% to 100%
BDP. We find a steady-state anisotropy of ~0.06, consistent with the restrictions of
monolayer attachments and interactions, from which we determine an average tilt angle of
35°. The absence of rotational motion demonstrates that the chromophores are in a highly
rigid environment on the hundreds of picoseconds timescale of our experiments. This
finding will simplify the interpretation of results on interlayer transport within multilayer
structures, because the chromophores should be even more restricted in an environment

with attachments at both phosphonate functionalities.

-94-

4.5 Literature Cited

9.

. Dewar, M. J. S.; Zoebisch, E. G.; Healy, E. F.; Stewart, J. J. P. J. Am. Chem. Soc.

1985, 107, 3902.

Dewar, M. J. S.; Dieter, K. M. J. Am. Chem. Soc. 1986, 108, 8075.
Stewart, J. J. P. Comput.-AidedMoI. Des. 1990, 4, 1.

Dewar, M. J. S.; Thiel, W. J. Am. Chem. Soc. 1977, 99, 4899.
Dewar, M. J. S.; Thiel, W. J. Am. Chem. Soc. 1977, 99, 4907.
Allinger, N. L. J. Am. Chem. Soc. 1977, 99, 8127.

Rabolt, J. F.; private communication.

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

Wang, H.; Harris, J. M. J. Phys. Chem. 1995, 99, 16999.

10. Berlman, I. B. Handbook of Fluorescence Spectra of Aromatic Molecules; Academic

ll.

12.

13.
14.

15.

l6.
17.
18.
19.

20.

Press: New York, 1971.
Frey, B. L.; Hanken, D. G.; Corn, R. M. Langmuir 1993, 9, 1815.

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

Cao, G.; Hong, H.-G.; Mallouk, T. E. Acc. Chem. Res. 1992, 25, 420.
Thompson, M. E. Chem. Mater. 1994, 6, 1168.

Bent, S. F.; Schilling, M. L.; Wilson, W. L.; Katz, H. E.; Harris, A. L. Chem. Mater.
1994,6,122.

Jiang, Y.; Blanchard, G. J. J. Phys. Chem. 1994, 98, 6436.
Blanchard, G. J. J. Phys. Chem. 1991, 95, 5293.
Blanchard, G. J. J. Phys. Chem. 1989, 93, 4315.
Blanchard, G. J. Anal. Chem. 1989, 61, 2394.

Blanchard, G. J. J. Phys. Chem. 1988, 92, 6303.

-95-

21.

22.

23.

24.

25.

26.

27.

Blanchard, G. J.; Cihal, C. A. J. Phys. Chem. 1988, 92, 5950.

Szabo, A. J. Chem. Phys. 1984, 81, 150.

Lipari, G.; Szabo, A. Biophys. J. 1980, 30, 489.

Debye, P. Polar Molecules; Chemical Catalog Co.: New York, 1929.
Hu, C. M.; Zwanzig, R. J. Chem. Phys. 1974, 60, 4354.

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

Fay, R. C.; Pinnavaia, T. J. Inorg. Chem. 1968, 7, 508.

-96-

Chapter 5

The Role of Substrate Identity in Determining Monolayer Motional
Relaxation Dynamics

Summary

We report on the lifetime and motional dynamics of zirconium phosphonate (ZP)
monolayers containing oligothiophene chromophores in a range of concentrations.
Monolayers were formed on fiised silica substrates and on a 15 A oxide layer formed on
crystalline Si(lOO) substrates. For both interfaces, the fluorescence lifetime behavior of
the chromophores is identical and does not depend on chromophore concentration within
the monolayer. Transient anisotropy measurements reveal that, for both substrates, the
chromophores are oriented at ~3 5° with respect to the surface normal. For monolayers
formed on silica, there is no evidence for chromophore motion, while motion is seen for
monolayers formed on silicon. Despite the substantial similarity between the two
families of monolayers, the surface roughness of the primed silicon substrate allows for
greater motional freedom of the chromophores in the monolayers. We discuss these
findings in the context of the differences in substrate surface roughness and domain sizes

as measured by atomic force microscopy (AFM).

-97-

5 .1 Introduction

Our focus in this Chapter is to understand the energy migration and motional
dynamics that occur within an optically active layer. By using simple chromophores, we
can observe the presence or absence of excitation transport and motional dynamics
without complications associated with large scalechromophore isomerization. We use
2,2'-bithiophene-S,5'-diylbis(phosphonic acid) (BDP) and 2,2':5',2":5",2"'-
quaterthiophene-S,5"'-diylbis(phosphonic acid) (QDP) (Figure 5.1) because the BDP
emission band overlaps the QDP absorption band, as shown in Figure 1.1. The spectral
overlap of the two chromophores is useful for excitation transport studies. These
chromophores can photoisomerize via ring rotation and this property is of value in
understanding their dynamical response in layered assemblies (vide infra).

In these initial studies, we concentrate on the intralayer motional relaxation
properties of each chromophore by varying the chromophore concentration in a series of
BDP and QDP monolayers. The chromophore concentration is controlled by
incorporating optically inactive, alkanebisphosphonate diluents into the monolayer. We
chose 1,6-hexanediylbis(phosphonic acid) and l,l2—dodecanediylbis(phosphonic acid) as
diluents because, in their all-trans conformations, they are approximately the same
lengths as BDP and QDP, respectively, enabling us to make the monolayer thickness as
uniform as possible, at least on molecular length scales.

We have characterized the chromophore dynamics of a single layer of BDP on
fused silica substrates in Chapter 4. In this work, we compare the behavior of QDP to

BDP on fused silica and oxidized Si( 100) substrates. For BDP on silica, we do not detect

-93-

P .....
HOV: P 3
H5 \ / S u
o
0
OH
HOV'IDI S / \ S / \ P uIOH b
H5 \ / S \ / S u
o
0
II OH
HO';P\/\/\/\/\/\/\P’...1IIOH C
H6 II
0
II OH
HOVEPWpr-HIOH d
H6 C")

Figure 5.1 Structures of oligobisphosphonates used in this work. (a) 2,2'-bithiophene-
S,5'-diylbis(phosphonic acid) (BDP); (b) 2,2':5',2":5”,2"'-quaterthiophene-5,5'"-
diylbis(phosphonic acid) (QDP); (c) 1,12-dodecanediylbis(phosphonic acid) (DDBPA);
(d) 1,6-hexanediylbis(phosphonic acid) (HBPA).

-99-

any chromophore rotational motion. The fluorescence dynamics of QDP on silica are
essentially the same, as we describe below in more detail.

The identity of the substrate affects the dynamical behavior of the chromophores.
The 15 A oxide layer on Si(lOO) possesses the same active sites (-OH) as silica, but there
are fimdamental differences in site density and surface roughness on oxidized silicon that
affect the primer density and organization and thus influence the adlayers. The steady-
state optical response and the fluorescence lifetime behavior were identical on both
substrates, for a given chromophore. We detect chromophore motion on silicon, and this
finding suggests substantially different organization and structural freedom in the two
systems. The active sites on silica appear to allow a higher chromophore density on the
initial phosphonate primer layer than is seen for silicon. giving rise to an improvement in
the ordering of the first layers. We present AFM images of the SiO-\ substrates that
complement the dynamic spectroscopic measurements and demonstrate the differences in

surface roughness of the substrates.

5.2 Experimental

Synthesis of 2,2':5',2":5".2"'-Qualerthiophene-5,5"’-dIy/bis(phosphomc acid)
(QDP); 2,2'—bithiophene-5,5 diylbis(phosphonic acid) (BDP); 1,12-dodecanediylbis
@hosphonic acid) (DDBPA) and I, 6-hexanediylbis(phosphonic acid) (HBPA). The
syntheses of these compounds were described in more detail in Chapter 2. Briefly,
bithiophene and quaterthiophene were derivatized to the phosphonic acid by lithium

coupling to a phosphochloridate, followed by acid hydrolysis. DDBPA and HBPA were

-100-

synthesized by the Michaelis-Arbuzov reaction'” of the corresponding dibromoalkane
with triethylphosphite, followed by acid hydrolysis.

Metal-phosphonate multilayer synthesis. Fused silica and polished Si(lOO)
(MultiScanning Plus) substrates were cleaned by piranha solution for 10 minutes and
rinsed with flowing distilled water. Samples were dried with a N2 stream and used
shortly afier for deposition. The substrates were primed by reflux in a 1% v/v solution of
3-(aminopropyl) triethoxysilane in anhydrous octane for 10 minutes, followed by rinses
with hexane and water. Afler drying with N2, the amine surface was derivatized to the
phosphonate in a solution of 0.1M POC13 and 0.1M collidine in anhydrous CH3CN for
one hour, followed by rinses with CH3CN and water. The resulting surface was
functionalized with Zr4+ by immersion in a 5 m_M_ solution of ZrOClz in 60% aqueous
ethanol solution for 10 minutes. Monolayers were deposited from bisphosphonate
solutions with a total 1 mM concentration ([chromophore] + [alkane] = 1 m_M_).
Alkanebisphosphonate, BDP, and mixed alkanebisphosphonate/BDP solutions were
made in 90% EtOH (aq); QDP and mixed alkanebisphosphonate/QDP were made in 80%
DMSO, 20% (90% EtOH (aq)). The substrates were immersed in the appropriate
bisphosphonate solution for 4 hours at 55°C.

Steady-state optical spectroscopy. The absorbance spectra of oligothiophene
chromophores in solution and on surfaces were measured using a Hitachi U-4001 UV-
visible spectrophotometer. Monolayer samples were held vertically in place at a 45°
angle with respect to the incident beam. Spectra were collected with 5 nm resolution.

Fluorescence spectra of solutions were measured on a Hitachi F-4500 fluorescence

spectrophotometer. Excitation and emission slits were adjusted according to spectral

-101-

intensity. Monolayer emission spectra were collected on a SPEX Fluorolog 2
spectrophotometer. Samples were mounted vertically and at a 45° angle with respect to
excitation, to minimize reflection of excitation light into the emission collection optics.
Slits were adjusted according to spectral intensity.

Time correlated single photon counting spectroscopy ( T CSPC ). TC SPC was used
to measure excited state fluorescence lifetimes and rotational dynamics of the
chromophores. This system has been described in detail previously in Chapter 3, and we
review it briefly here. The light pulses used to excite the sample are generated with a
cavity dumped, synchronously pumped dye laser (Coherent 702-2) pumped by the second
harmonic output of a mode-locked CW Nd:YAG laser (Quantronix 416). The excitation
beam was frequency doubled (LiIO3 Type 1 SHG) to excite BDP samples at 320 nm (640
nm fundamental, Kiton Red, Exciton) and QDP samples at 390 nm (780 nm fundamental.
LDS 821, Exciton). Monolayer samples were held approximately horizontally, with 5°
tilts away from the horizontal in two directions: toward the excitation beam and toward
the detector. Fluorescence from the sample was imaged through a reflecting microscope
objective. Lifetimes were collected across the emission bands at 547° with respect to the
excitation polarization for solutions and without polarization bias for monolayers. While
such collection can, in principle, lead to small contributions to the measured relaxation
from motional dynamics, for our experimental conditions this is not a problem.“3|
Fluorescence was collected at 390 nm and 465 nm (5 nm FWHM bandwidth for solution,
30 nm for monolayers) for BDP and QDP, respectively, with polarizations of 0° and 90°
for orientational anisotropy measurements on all samples. A representative lifetime

decay and instrument response function (~ 35 ps FWHM) are shown in Figure 5.2.

-102-

 

intensity (counts)

 

1 -o-c)-o~.oo-- .. .0.

mama-Inn
rm: mummdou fi'm‘.’- “° .
a. o o -o_--

 

 

 

I I . 1 q. .0,- o _ b
0 1 2 3 4
time (ns)

Figure 5.2 Representative instrument response function (El). excited state population
decay for QDP in DMSO (10'5 M) (O) and excited state population decay for QDP in a
monolayer (1%) (0).

-lO3-

Data analysis. The lifetimes we report here were fit to sums of exponentials
using commercial software (Microcal Origin v. 4.1) and are reported as the averages and
standard deviations of 4-6 individual decays. Rotational dynamics parameters were
determined in two ways. For monolayers on silica, eight pairs of alternating parallel and
perpendicular scans were used to produce an average anisotropy function. Each scan pair
was recorded for a different physical location on the substrate. R(O) was determined to i
0.01 by regression of data at times after the instrumental response, and several data sets
were averaged for each sample. For monolayers on oxidized Si(lOO). the rotational
diflitsion information is reported as the averages and standard deviations of data from
several decays. Four pairs of alternating parallel and perpendicular scans were used to
produce an average anisotropy function, which was fit with Origin to determine the
parameters R(O), TMR, and R(oo).

Calculations. Semiempirical calculations were performed using Hyperchem
Release 4.0 (Hypercube, Inc). The PM3 parameterization used for these calculations is a
modification of the AMI parameterization that treats molecules containing heteroatoms,
such as sulfur, more accurately than previous parameterizations. An initial optimization
of the structure was performed using a molecular mechanics routine (MM+) followed by
geometry optimization at the semiempirical level using an SCF algorithm. The torsions
of the QDP 2-2' and 5'-5" 0 bonds were set at 10° increments, and semiempirical
optimization was performed until the lowest energy conformation for the fixed interring
bond torsions was reached. The heat of formation and So <—> 81 transition energy were

calculated for each geometrically optimized conformation.

-lO4-

5.3 Results and Discussion

In this section we compare the steady-state and transient optical responses of four
optically active chemical systems using two chromophores and two substrates. A basic
understanding of the optical response and the intralayer relaxation effects of each
chromophore is necessary for future investigation of interlayer energy transport between
the two chromophores in complex multilayer structures. In addition to the determination
of the optical properties of these layers, the effect of substrate must also be considered.
Several measurements contribute to our understanding of the optical response, including
UV-visible absorption, fluorescence emission, excited state population decays, and
induced orientational anisotropy. The results from the anisotropy measurements are
critical for comparing the systems. We detail the subtle differences between the systems
below.

Monolayers on fused silica

Steady state spectroscopy. BDP monolayers on silica were characterized in
Chapter 4. For BDP and HBPA mixed monolayers, we found a linear relationship
between the proportion of chromophore in the deposition solution and the proportion of
chromophore in the resulting monolayer. The calibration did not regress through the
origin, however. This is a simple partitioning effect and can be understood in terms of
preferential solvation of HBPA over BDP in ethanol. For QDP and DDBPA mixed
monolayers, we also observe a calibration that is linear in QDP concentration, shown in
Figure 5.3. In this case, the solvent system used for the monolayer deposition solution

gives rise to a calibration curve that passes through the origin. The implication of these

-105-

 

0.025 -

0.020

0.015

0.010

corrected absorbance

0.005

 

 

 

 

0.000
0 20 40 60 80 100

% QDP in monolayer

Figure 5.3 Dependence of monolayer absorbance on concentration of QDP in deposition
solution and regression of experimental data.

-106-

data is that solution concentration ratios will result in a monolayer with the same
constituent ratio, if the solVent system is chosen correctly. The proportions of
constituents that are incorporated into a mixed monolayer depend mostly on their
solubilities in the solvent used.

The steady-state emission spectra for both BDP and QDP are similar in the sense
that they reveal more than one form of the chromophore in the monolayer. The BDP
spectrum has two features, a narrow peak at 390 nm that we see for dilute (< 35% BDP)
monolayers and a broader, less intense peak at 425 nm that we observe for concentrated
monolayers. These bands correspond to the two peaks present in concentrated solution
spectra. QDP exhibits four unresolved bands in solution and typically three are present in
monolayer spectra (see Figure 5.4). For the monolayers, the 445 nm band is not
observed. The QDP spectra also depend on concentration in the monolayer. Blue-shified
peaks are prominent in the dilute monolayer spectra, and red-shifted features dominate
the concentrated monolayer spectra. Over time, the BDP spectra anneal to spectrum
dominated by the blue band, (Figure 4.8) but it is not clear from our data that QDP
monolayers exhibit this same evolution. The complexity of these monolayer spectra is
also manifested in the temporal evolution of emission intensity. We will discuss this
behavior in more detail in Chapter 6.

Transient spectroscopies. The dilute solution behavior of both BDP and QDP is
that of a simple chromophore. The absorption and emission bands of the steady-state
spectra are well-defined, and dilute solution lifetimes are fit to a single exponential decay
for both chromophores. The lifetime for dilute solutions is approximately 200 ps for

BDP and 450 ps for QDP. The population decays of monolayers approximate a double

-107-

 

 

normalized emission intensity

 

l‘|'~.§'..

r 5*

n .
-- -
\~5v-\ ‘~-.
‘- ______________

 

i l 1 l i l 1 I

450 500 550 600
emission wavelength (nm)

 

Figure 5.4 Emission spectra of (a) QDP monolayers (- - -) 1%; ( ) 5%; (' - ' -) 100%;
and (b) QDP solution in DMSO (- - -) 10'7 M; ( )104’ M; (- - - -) 10'3 M. Spectra were
normalized for presentation and no information about fluorescence quantum yield should
be inferred from these plots.

 

-108-

exponential, as shown in Figure 5.2. The fast component of the monolayer lifetimes is ~

200 ps for both chromophores and the longer lifetime is ~ 1.2 ns, as shown in Table 5.1.

Table 5.1 Excited state lifetimes of monolayers. The time constants refer to fits of the
experimental data to the fimction f(t) = Alexp(-t/11)+ Azexp(-t/rz).

 

 

TI (P5) 12035)
BDP on silica 187i ll ll40i337
QDP on silica 232 $16 1265 i 138
BDP on silicon 213 i 12 1277 :l: 96
QDP on silicon 243 :t 23 1115 i: 156

 

 

 

 

 

We interpret the fast monolayer lifetime as that due primarily to an isolated chromophore
species in an alkane-like environment, owing to its similarity to the solution phase value,
and the long lifetime primarily to a chromophore within an aggregated island. These
assignments are, of course, only approximations and the actual contributions to the
measured decay are not as clear-cut. This relaxation behavior is well understood and we
will present a detailed treatment in Chapter 6. The lifetimes do not vary with respect to
either chromophore density in the monolayer or emission wavelength collected for either
chromophore. For this reason, the lifetimes reported in Table 5.1 are averages over all
concentrations and emission wavelengths used for chromophore monolayers. However,
the fractional contribution of each lifetime component varies with chromophore
concentration and emission wavelength. For BDP, as the chromophore concentration is

increased or as emission is collected on the red edge of the emission band, the relative

—109-

contribution of the long lifetime increases. This trend is not observed with QDP, as the
contributions of the different species to the total signal varied with no simple pattern.
This effect may be due to the greater complexity of the QDP emission spectrum
compared to BDP.

Rotational motion measurements are not complicated by population decay
functionality effects and can provide insight into how the chromophores move within
their environment, offering an indication of how ordered the layers are. Rotational
diffusion measurements are most typically made on probe molecules in solution, where
the chromophores can relax into all possible orientations. We have measured the
dynamics of BDP and QDP in monolayers. Because one end of each chromophore is
attached to a substrate, the motion of the chromophore is confined to a conic volume.
The cone dimensions are described by Lipari and Szabo’s hindered rotor model,”' where
the base of the cone is the point of attachment of the chromophore to the substrate, and
the semiangle of the cone is designated 00. A nonrandom orientational distribution of the
chromophore ensemble in the monolayer is selected with a polarized excitation pulse, and
components of the emission that are polarized parallel and perpendicular to the excitation
polarization are collected. These data (Figure 5.5a, for example) are combined to

construct the induced orientational anisotropy function, R(t):

_ [Kn-[1(1)

— [5.1]
ln(t)+21l (t)

R(t)

 

The decay of R(t) (see Figure 5.5b) is fit to:

R(t) .-. R(oo) +[R(O) — R(oo)]exp(’%m) [5.2]

-110-

1500 -

fl
O
O
O
I

  
  

500 +- parallel

 

 

emission intensity (counts)

 

 

 

 

 

 

r perpendicular .-
0 . - . ‘ ' “"‘ "I
b
0.4 —
0.2 L
9 ' _i ' '1 " lr 1’,"|‘ ll 1[iii -' i, If i' _
m i r ‘ I]
00 'i I illl
- l ll l' t
F
_0.2 I i 1 l 1 I l
0.0 0.5 1.0 1.5

Figure 5.5 Reorientation data for a 50% QDP monolayer on silica. (a) Emission
intensity data for parallel and perpendicular polarizations. (b) Induced orientational
anisotropy function of above data with fit (- - -).

-lll-

‘tMR is the motional relaxation time constant and the quantity R(O) is related to the angle

between the chromophore absorbing and emitting transition moments, 5,
2
R(O) = ~5-P2 (0056) [5.3]

where P; is the second order Legendre polynomial. For BDP in solution, R(O) = 0.24,
which corresponds to an angle of 6 = 31°. For QDP in solution, R(O) = 0.40, indicating
parallel transition moments. We consider that the QDP transition moments lie along the
long molecular axis. We use these values for 5 in the treatment of monolayer data
because this quantity is intrinsic to the chromophore and, for the solution data, represents
an average over all conformers. We note that R(O) does depend to a certain extent on
environment because the intramolecular rotational freedom available to the chromophore
depends on the limitations imposed by its surroundings (vide infra). Specifically, syn and
anti conformers are likely to exhibit different values of 5.

For chromophores in solution, R(oo) = 0 because they can completely re-
randomize at infinite time. For chromophores in monolayers, |R(oo)| > 0 because of the
geometric restriction on their motional freedom. R(oo), the steady-state anisotropy, is

related to 0, the average tilt angle of the chromophore within the cone by'4'

R(oo) = %P2(cosl9u )P2 (c056? )(P2(cosl9))2 [5.4]

cm

The angles Bax and 0m are the transition moment angles with respect to the cone center
axis. The calculation of 0 is simplified by setting 0,,- = 0° and 0m, = 6. In our previous

work we reported that BDP on SiOx exhibited a concentration-independent steady-state

anisotropy of 0.06 i 0.01, which yields an average tilt angle of 35° :1: 2°. We report here

-112-

that QDP on silica has a similar, concentration-independent steady-state anisotropy of

0.07 d: 0.05, and an average tilt angle of39° :l: 7° (Table 5.2).

Table 5.2 Induced orientational anisotropy data for monolayers. For BDP, the angle
between the transition moments is .5 = 31° and for QDP, 6 = 0°. Angles 0 were
determined from 5 and R(oo) data using Equation 5.4.

 

 

 

 

 

 

 

 

R(O) IMR (ps) R(oo) 0 (degrees)
BDP on silica - - 0.06 :t 0.01 35 :t 2
QDP on silica - - 0.07 :t 0.05 39 :l: 7
BDP on silicon 0.403: 0.15 393 :t 119 0.18:1:008 19: ll
QDP on silicon 0.32 i 0.06 479 :t 153 0.13 :t 0.06 32 :l: 7

 

Our measurement of the same tilt angles for both chromophores implies that the
oligothiophene/alkane environment is not affected by the lengths of the constituent
molecules. We note that the angles we report here for layer tilt are quite similar to the tilt
angle of ~30° seen for alkanethiol monolayers on goldlm In the case of alkanethiol/gold
SAMs, the tilt angle is determined by aliphatic inter-chain interactions and the registry of
the thiol head groups on the crystalline gold substrate surface. For ZP layers. it is not as
obvious that intermolecular interactions dominate the tilt angle because, for the substrates
we use, the spacing between primed sites is significantly less well determined than for a
crystalline metal surface. In the case of ZP layers, it may be the bonding geometry of the
metal-phosphonate connecting layers that determines the range of structural properties

available to the system.

-ll3-

We do not detect any chromophore motion in either BDP or QDP monolayers on
silica, as indicated by the absence of a decay in R(t). At times after the instmment
response function, regression of R(t) yields a line of zero slope. We can evaluate the
possibility that the chromophores are reorienting significantly faster than our detection
system time resolution (~ 35 ps). The time constants of reorientation in solution are 299
:l: 15 ps for BDP in ethanol and 1.2 i 0.2 n5 for QDP in DMSO. Based on viscosity
considerations, we believe that the chromophore imbedded in the monolayer will not
reorient more rapidly than in low viscosity solvents such as ethanol or DMSO. If there
was any motion of the chromophore in the monolayer, it would occur on a time scale we
can determine because of the proximity of neighboring molecules and dipolar and van der
Waals interactions between them.

Monolayers on silicon

The substrate on which self-assembling structures are synthesized affects the
ordering of the adlayersw” This effect is seen for alkanethiol layers on various metal
surfaces, because the infrared-active symmetric and asymmetric CH2 stretching modes
are sensitive indicators of aliphatic chain ordering. We compare the optical response of
the chromophore monolayers on silica to that on oxidized Si(lOO) substrates to evaluate
the role of substrate-induced chromophore ordering in determining the dynamical
response of the system. Frey and coworkers have shown effects of the substrate on the
order of ZP layers,“2] but this effect is seen in the inorganic portion of the IR spectrum
(v. (PO32')) rather than in the aliphatic CH2 stretching region. We chose silicon

substrates for the comparison because the surface priming chemistry is identical to that of

-114-

silica, due to the 15 A thick layer of native oxide that is present on silicon under our
experimental conditions.
It is fair to consider that vibrational spectroscopy could address the question of

“2' Aside from the different

monolayer order for ZP layers on these two substrates.
dielectric responses of the two substrates, especially in the IR spectral region, vibrational
spectroscopy will not be as sensitive to order for our samples as they are for pure
alkanebisphosphonate layers. Contributions from the thiophene rings combined with the
structural perturbations they may induce in adjacent aliphatic domains will serve to
obscure structural information for our samples. The strong absorbance of SiOx at low
energies obscures information about the inorganic portion of the structures. We therefore
rely on lifetime and anisotropy measurements that are sensitive to the local environment
of the oligothiophene chromophores.

The concentration dependence of the lifetime data is important in understanding
excitation transport in these monolayers. lf excitation transport does play a significant
role in the transient response of these systems, then we would expect a highly

“3' The fact that we observe concentration

concentration-dependent population decay.
independent lifetime behavior indicates that either excitation transport in this system does
not play an important role in determining the observed motional dynamics, excitation
transport dynamics are faster than can be resolved for all of the systems we examined, or
the double exponential decay is itself characteristic of the excitation transport dynamics.
We consider this issue in detail in Chapter 6. The chromophores we have chosen possess

moderately high fluorescence quantum yields and large absorption cross sections (emax =

27,200 Mlcm'l for BDP and 27,300 M"em" for QDP). Because the oligothiophene

-llS-

chromophores relax primarily by radiative decay and they possess a large Stokes shifi
(Figure 1.1), the oligothiophene chromophores are ideal candidates for energy transport
studies. We are presently characterizing intralayer excitation transport in mixed
monolayers of BDP and QDP. Multilayer assemblies of these chromophores will be
useful in investigating interlayer excitation transport phenomena.

While the lifetime data for the chromophores reveal no discernible dependence on
the identity of the substrate, the rotational dynamics of BDP and QDP on silicon are
fundamentally different than those on silica. As we show in Figures 5.6 and 5.7, a
measurable decay of the anisotropy is present for BDP and QDP, with an average
reorientation time of ~ 435 ps for both, that was not seen for these chromophores bound
to silica. These data were taken over a broad spectral range, with emission collected at
465 nm i 15 nm, and the values reported in Table 5.2 are averaged over all
concentrations because there was no statistically meaningful concentration dependence.
The average tilt angles, 0, calculated from R(oo) using Equation 5.4, are approximately
the same for both QDP and BDP monolayers, ~ 35°, (Table 5.2) which correlates well
with tilt angles determined for other MP monolayers'n‘m We note that the average tilt
angles of the chromophores on silica are similar, so measurement of 0 cannot explain the
existence of motion on the oxidized silicon substrate and not silica.

Origins of Chromophore Motion. A necessary first step is to determine the nature
of the molecular motion we sense in a single layer. In particular, we need to distinguish
between inter-ring isomerization and large amplitude precessional motion. The hindered
rotor model describes the physics of a chromophore imbedded in a monolayer. It is

typically assumed that a decay in R(t) is related to large amplitude motion of a

-l16-

O\
O
O
O
I

 
 
   

A
O
O
O
1

parallel

 

 

 

emission intensity (counts)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2000 -
_ perpendicular
O _
l l L
b
0.4 —
02 ~ ‘ - j
A I 4* k-
5?
0.0 . (I [ll
_02 l i l I l‘ . I
0.0 0.5 1.0 1.5

time (ns)

Figure 5.6 Emission intensities (a) and induced orientational anisotropy function (b) for
a 26% BDP monolayer on silicon.

-117-

2000 -

 
 
 
  

1500 —
parallel

1000 —

500 perpendicular

 

emission intensity (counts)

 

0.2- H 'IlJ“ . ”ll ‘ i l

R(t)

00 .....

 

 

 

 

 

0.0 0.5 1.0 1.5
time (ns)

Figure 5.7 Emission intensities (a) and induced orientational anisotropy function (b) for
a 1% QDP monolayer on silicon.

-118-

chromophore in its environment about its tether to the surface, which is one possible
explanation for the dynamics we detect. Ifthis is the dominant motion in our systems, a
larger or longer chromophore would experience more hindered movement, and would
exhibit a correspondingly longer relaxation time. This is essentially a viscous drag
argument, borrowing from the underlying principles of the Debye-Stokes-Einstein model
for rotational diffusion in liquids.“5‘l6]

If large amplitude motion dominates our experimental data then we would also
expect to see a concentration dependence, since the thiophene chromophores occupy a
larger “footprint” in the lattice sites than alkanes do. A chromophore with chromophore
neighbors will move more slowly in its environment than if it were surrounded by
aliphatic constituents because of geometric constraints and dipolar coupling. The average
motional relaxation time, IMR, (Table 5.2) for BDP and QDP on silicon are the same to
within the experimental uncertainty, arguing that the relaxation we measure is not large
amplitude precessional motion within a conic volume.

Another type of motion that is important in these chromophores is intramolecular
relaxation, specifically the rotation of the thiophene rings about their connecting bonds.
This motion is intrinsic to the molecule, and is not limited to a hindered environment. As
seen in Figure 5.8, a transition from an all-anti to an all-syn conformation of QDP
represents a substantial change in the orientation of the transition moments relative to the
substrate. We believe that rotation of the ring(s) is more reasonable and it can account
for our data (vide infra). The Optical anisotropy decay is a correlation function for the

transition moment relaxation, so the dynamics we observe can be accounted for by inter-

ring conformational changes in QDP and large amplitude motion does not need to be

-119-

 

o" "

e

a b
9Hon—i
?/
\ P
\ \\
s O
\ S
‘O

 

Figure 5.8 Conformations of QDP with thiophene rings (a) all anti and (b) all syn.
Conformations of BDP with thiophene rings (c) anti and (d) syn.

-120-

invoked. The anisotropy decay due to isomerization is not as pronounced for BDP as it is
for QDP because BDP possesses fewer conformational degrees of freedom, but there is
still a discernable change in the transition moment angle with respect to the substrate
normal (Figure 5.8). The BDP and QDP measured parameters of R(O), rm, and R(oo) are
likely similar because they are constrained to the same molecular area, as dictated by the
substrate, primer, and metal phosphonate lattice structure. Within the constraints of a
well-packed, uniform monolayer, BDP and QDP may actually have the same net change
in transition moment direction, because QDP will not be able to access all of the degrees
of conformational freedom available to an isolated molecule (Figure 5.8).

“7'2” rotation of the thiophene rings of an isolated

Based on semiempirical calculations,
ground state chromophore is a low energy activated process, and certainly occurs at room
temperature. We reported in Chapter 4 that the BDP chromophore possesses low barriers
to rotation between conformations in the ground state (< l kcal/mol). This finding
implies that there will be a statistical distribution of conformers that are excited, with any
bias in their conformational distribution being induced by environmental, not structural
constraints. In Figure 5.9, we show results from semiempirical calculations on QDP for
twists between two end rings and between the two center rings. These plots are
representative of most energetically favorable conformations, with the exclusion of
contributions from the twist between the other two end rings, which were held at 0° in the
anti conformation. 1n the first excited singlet state, energetic minima are present at
approximately 0° and 180° ring torsion (anti and syn), while a maximum of ~ 10 kcal/mol

is calculated for an inter-ring angle of 90°. This maximum is the result of the tendency of

thiophenes to form a planar quinoid structure in the excited state. The syn and anti

-121-

-lll435.0

     

4114355 0
a $5,
g -111436.0 gtf’ég 43%;?
3 fig? _ Quests
354114365 figs 1? w giggle,
er
204114370 §§§£€f¢5§§ @ 494a
:- “‘9'
9 4114375

-lll339

-lll342

aUnKfl)

-lll345

-lll348

energy (kc

-lll35|

-111354

Figure 5.9 Calculated energy barriers for QDP inter-ring rotation for the first two inter-
ring bonds: (a) So and (b) 5.. Twist angles indicated are for angles with respect to their
adjacent ring.

-122-

conformations correspond to the lowest energy difference between $0 and 8;, so
transitions associated with these conformations are expected to produce the red shifted
features we see. From these calculations we can see that QDP exhibits facile
interconversion between all conformations in the ground state and the excited state
possesses definite minima. These calculations are, of course, qualitative and are limited
by the utility of the PM3 parameterization in the calculation of S. surfaces. Despite these
limitations, they do demonstrate the likelihood for the chromophores to exhibit ring
rotation in the S1, consistent with intramolecular relaxation being the dominant
phenomenon we detect in the dynamical measurements.

The semiempirical calculations serve to limit the possible explanations for why
we observe motional relaxation for monolayers on oxidized Si(100) and not on fused
silica. The observation of motion on the silicon substrate could be taken to imply that we
are observing motion in this system because the rotation of the chromophore rings is not
hindered. It is clear based on the data for the Si(lOO) substrate and comparison to bulk
liquid data that motion of the thiophene rings on silica would occur within a time window
accessible to our measurements. We believe that thiophene ring rotation on the silica
substrate must be hindered in some way. A simple difference in the dielectric responses
of the two substrates cannot account for the differences in motional behavior, because the
population dynamics, which depend sensitively on the dielectric response of the
substrate, do not vary.

The surface morphology of the substrates is more likely to provide an explanation
for our observations. As shown in atomic force microscopy (AFM) imagesl22| (Figure

5.10), the two substrates are characterized by very different surface features. The silica

-123-

.3925 «Be mo :o:=nt%% fis

“0:2 “data 8865 mEEwoamE 03E. .momSE Q-m 2: E 3:865 0:: 2: mass 8an £32.. 32.-e. 32m .850 33:33
€832: 8:? came use as: 50:: vufiExo me 325 E: 8m x E: can .8 8mg: 30822:. 09.8 2E9< 36 9...»...—

 

C

 

325.5
4

l'/.ul) thumb»;

 

 

02

.65 K
8.

(mun

 

 

 

 

 

 

 

 

 

(mull

 

 

 

 

 

424-

has a granular, irregular surface, and the oxidized silicon exhibits striations, probably a
result of polishing. The grooves in the oxidized Si surface are 200 — 500 A apart, and 10
— 30 A deep. The roughness of the two surfaces was determined from the FWHM of a
histogram of the imaged area, which is slightly narrower for the silica substrate. On a

23' the variations in

length scale more relevant to our spectroscopic measurements (50 A),l
height across the surfaces reveal much flatter domains on silica (i 2 A) than on silicon (i
10 A). Because of the greater variation in surface relief, chromophores on silicon should
have more motional freedom than on a corresponding monolayer formed on silica. An
additional piece of information that corroborates this explanation is that for 810,, x,...,.,,., <
xsmca, which could result in a less dense layer on silicon. Haller has shown that. for this
priming chemistry on silicon, under all except the most stringent conditions, the
triethoxysilane primer polymerizes easily, resulting in large islands on the surface instead

[2“] It is possible that the higher density of hydroxyl active sites on the

of a single layer.
silica surface precludes much of the polymerization because the tighter packing does not
allow as much multilayer buildup. For silicon, with fewer active sites on which to grow
the bisphosphonate monolayer, the chromophores will be spaced further apart and
therefore ring rotation is less hindered. The relative active site density for the two
substrates does not explain, however, why we observe a larger fluorescence signal from
the monolayers on silicon than on silica. One factor is that the polished silicon substrate
is more reflective, so that a larger fraction of emitted light is collected. A more important
factor is that, in the more dense monolayers on silica, aggregation of the chromophores is
likely. Aggregates typically exhibit lower emission quantum yields than isolated

[25]

chromophores, and the less densely packed monolayers on silicon will, on average,

~125-

have more non-aggregated chromophores. Regardless of these qualitative observations
relating to emission intensity, our motional relaxation data can be understood in the

context of substrate roughness, as characterized by the AFM data.

5 .4 Conclusions

The lifetime and motional dynamics of oligothiophene chromophores within ZP
monolayers reveal a subtle difference in the monolayers formed on different substrates.
For both silica and oxidized Si(lOO) substrates, the fluorescence lifetime behavior of each
chromophore is identical and is concentration independent, pointing to the limited
contribution of intralayer excitation transport in these systems. Transient anisotropy
measurements reveal that, for both substrates, the chromophores are oriented at ~3 5° with
respect to the surface normal. For monolayers formed on silica, there is no evidence for
chromophore motion, while motion is seen for monolayers formed on the silicon
substrate. Despite the substantial similarity between the two families of monolayers, the
surface roughness of the polished silicon substrate allows for greater motional freedom of
the chromophores in the monolayers. The comparatively lower site density for active
hydroxyl groups on the oxidized Si(lOO) surface likely also plays a role in determining
the observed dynamics. These data underscore the important role of substrate
morphology in determining the properties of molecular interfaces and provide a

foundation for the examination of inter-layer excitation transport studies.

-126-

5 .5 Literature Cited

1. Bhattacharya, A. K.; Thyagarajan, G. Chem. Rev. 1981, 81, 415.

2. Flory, W. C.; Blanchard, G. J. Appl. Spectrosc. 1998, 52, 82.

3. Spencer, R. D.; Weber, G. J. Chem. Phys. 1970, 52, 1654.

4. Lipari, G.; Szabo, A. Bi0phyng. 1980, 30, 489.

5

6

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

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

7. Smith, E. L.; Porter, M. D. J. Phys. Chem. 1993, 97, 8032.

00

. Fenter, P.; Eisenberger, R; Li, J .; Camillone, N. 111; Bernasek, S.; Scoles, G,
Ramanarayanan, T. A., Liang, K. S. Langmuir 1991, 7, 2013.

9. Chau, L.-K.; Porter, M. D. Chem. Phys. Lett. 1990, I6 7, 198.

10.
ll.
12.
l3.

14.

15.

l6.

l7.

l8.
19.
20.

21.

Chen, S. H.; Frank, C. W. Langmuir 1989, 5, 978.

Sondag, A. H. M.; Raas, M. C. J. Chem. Phys. 1989, 91, 4926.

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

Gochanour, C. R.; Andersen, H. C; Fayer, M. D. .1. Chem. Phys. 1979, 70, 4254.

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

Debye, P. Polar Molecules Chemical Catalog Co.: New York, 1929.

Fleming, G. R. Chemical Applications of Ultrafast Spectroscopy Academic Press:
New York, 1986.

Dewar, M. J. S.; Zoebisch, E. G.; Healy, E. F.; Stewart, J. J. P. .I. Am. Chem. Soc.
1985, 107, 3902.

Dewar, M. J. S.; Dieter, K. M. J. Am. Chem. Soc. 1986, [08, 8075.
Stewart, J. J. P. Comput.-AidedMol. Des. 1990, 4, 1.
Dewar, M. J. S.; Thiel, W. J. Am. Chem. Soc. 1977, 99, 4899.

Dewar, M. J. S.; Thiel, W. J. Am. Chem. Soc. 1977, 99, 4907.

-127-

22.

23.

24.

25.

We are grateful to Prof. Gang-Yu Liu and Yan Huang (Wayne State University) for
providing the AFM images shown in Figure 5.10.

We estimate the F 6rster critical transfer radius for these chromophores to be 25 A.
For rotational diffusion measurements, the relevant distance is probably shorter, on
the order of the nearest neighbor spacing. Due to the amorphous nature of these
surfaces, we can only estimate this distance to be s 10 A.

Haller, I. J. Am. Chem. Soc. 1978, 100, 8050.

Song, Q.; Bohn, P. W.; Blanchard, G. J. J. Phys. Chem. B. 1997, 10!, 8865.

-128-

Chapter 6

The Effect of Layer Morphology on Population Dynamics and Intralayer
Excitation Transport in Zirconium-Phosphonate Monolayers

Summary

We use fluorescence relaxation dynamics in conjunction with atomic force microscopy
(AFM) to understand the morphology of zirconium-phosphonate monolayers containing
oligothiophene chromophores. The population relaxation dynamics we measure are non-
exponential, chromophore concentration-independent and identical for two different
chromophores. We account for these data in the context of an excitation hopping model
in a system where the surface is characterized by islands of aggregated chromophores as
well as non-aggregated monomers. From the spectroscopic data we infer the sizes of the
aggregated islands to be between ~50 A and 100 A in diameter, in excellent agreement
with the size range of 40 A to 100 A measured directly using atomic force microscopy
(AFM). The AFM images also reveal that the nearest neighbor spacing of the
chromophore islands varies from 80 A to 200 A, a separation sufficient to preclude
efficient inter-island excitation transport. The data, taken collectively, indicate that the
silanol group distribution on the surface determines the heterogeneity of the initial primer
layer, and chromophore aggregation plays an additional role in mediating the size of the

aggregated islands.

-129-

6.1 Introduction

In this work, we are interested in elucidating the details of excitation transport
within a given molecular layer. The bisphosphonated oligothiophene chromophores we
use are not candidates for optical information storage applications but, because of their
spectroscopic properties, they allow us to determine the worst-case effects of excitation
transport in a molecular layer. lsomerization about the inter-ring bond in oligothiophenes
is facile in both the So and S 1 states, as determined in Chapter 3, and this process does not
interfere with the measured population dynamics. We can thus study excitation transport
in the absence of complications associated with large amplitude molecular motion. We
use 2,2'-bithiophene-5,5'-diylbis(phosphonic acid) (BDP) as the donor chromophore and
2,2':5',2":5",2'"-quaterthiophene-5,5"'-diylbis(phosphonic acid) (QDP) for the acceptor
chromOphore. The structures of these molecules are shown in Figure 6.1.

We have characterized the optical response of each chromophore separately in
monolayer environments as well as in mixed monolayer systems where both BDP and
QDP are present in varying concentrations. These two chromophores represent a limiting
case in the context of information storage applications because the BDP (donor) emission
band overlaps the QDP (acceptor) absorption band, making radiative energy transfer
favorable (Figure 1.1). For all of the monolayer assemblies, we recover the same non-
exponential population decay, independent of chromophore concentration. These data are
not consistent with Forster excitation transport in a homogeneous system. The fact that
the population relaxation dynamics are the same for both molecules suggests that the
optical response for these monolayers is mediated by structure on a length scale

significantly greater than the individual molecules. The origin of the apparently anomalous

-l30-

OH
0'ng! S / \ ...rthH

HO \ / S 11 a
u s / \ S / \ 2:014
(115? \ / S \ / S H b
0
H OH
0'?\/\/\/\/\/\/\p’""”OH C
HO A

Figure 6.1 Structures of oligobisphosphonates used in this work. (a) The donor 2,2'-
bithiophene-S,5'-diylbis(phosphonic acid) (BDP); (b) the acceptor 2,2':5',2":5",2"'-
quaterthiophene-S,5"'-diylbis(phosphonic acid) (QDP); (c) the diluent 1,12-
dodecanediylbis(phosphonic acid) (DDBPA).

-l3l-

relaxation behavior is the structural heterogeneity of the monolayer. Our lifetime data are
consistent with the formation of aggregates. Based on an excitation transport model used
previously for stilbazolium dyes in Langmuir-Blodgett films,“I we predict aggregate
domain sizes on the order of 50 A to 100 A. Atomic force microscopy (AFM) images
reveal that the diameter of aggregates typically ranges from 40 A to 100 A. The nearest
neighbor spacing of these islands varies from 80 A to 200 A, which is sufficient separation
to preclude efficient inter-island excitation transport. The agreement between the
spectroscopic and physical (AFM) measurements of the surface morphology shows that
the domain size sensed by spectroscopic experiments is the same as that measured by
AFM, implying a significant degree of intermolecular alignment within the individual
aggregate domains. These findings indicate that the surface silanol group distribution
determines the initial priming layer heterogeneity, and chromophore aggregation plays an

additional, key role in the development of the observed layer morphology.

6.2 Experimental

Metal-phosphonate monolayer synthesis. Polished silicon (100) substrates
(MultiCrystal Optics) were cleaned by piranha solution etch for 10 minutes and rinsed
with flowing distilled water. Samples were dried with a N2 stream and were primed
immediately. The substrates were primed with 3-(aminopr0pyl) triethoxysilane (APTES)
or 3-(aminopropyl) dimethylethoxysilane (APDMES). Substrates were refluxed in a 1%
v/v solution of APTES in anhydrous octane for 10 minutes, followed by rinses in hexane
and water. Alternatively, they were immersed in a 0.5% v/v solution of APDMES in

anhydrous toluene overnight and rinsed with methanol and water. After drying with N2,

-l32-

the aminated surface was derivatized to the phosphonate using a solution of 0.2 M POCl;
and 0.2 M collidine in anhydrous CH3CN for 20 minutes, followed by rinses with CH3CN
and water. The resulting surface was exposed to Zr" by immersion in a 5 mM solution of
ZrOClz in 60% aqueous ethanol for 10 minutes. The syntheses of 2,2'-bithiophene-5,5'-
diylbis(phosphonic acid) (BDP) and 2,2': 5,2": 5";2'"-quaterthiophene-5 , 5'"-
diylbis(phosphonic acid) (QDP) and 1,lZ-dodecanebisphosphonic acid (DDBPA), were
reported in Chapter 2, and we show their structures in Figure 6.1. Monolayers were
deposited from solutions having a total 1 mM concentration of organobisphosphonate
([BDP] + [QDP] + [DDBPA] = 1 mM) in 80% DMSO, 20% (90°/o EtOH (aq)). The
primed and zirconated substrates were immersed in the bisphosphonate solution overnight
to allow formation of the layer of interest. The relative amount of each constituent in
solution is related directly to the fraction of each in the formed monolayer for this solvent
system.

Lifetime Measurements. Time correlated single photon counting spectroscopy
(TCSPC) was used to measure the fluorescence lifetimes of the chromophores within the
monolayers. This system has been described in Chapter 3, and we recap its salient features
here. Short light pulses (~5 ps FWHM) are generated with a cavity-dumped,
synchronously pumped dye laser (Coherent 702-2) pumped by the second harmonic of the
output of a mode-locked CW Nd:YAG laser (Quantronix 416). The dye laser output is
fi'equency doubled (LiIO3, Type I SHG) to excite the samples at 315 nm (630 nm
fundamental, Kiton Red, Exciton). Monolayer samples were held approximately
horizontally, with 5° tilts away from the horizontal in two directions, toward the excitation

beam and toward the detector. Fluorescence from the sample was imaged through a

-l33-

reflecting microscope objective and collected across the emission bands of the
chromophores with 30 nm FWHM detection bandwidth. The chromophore lifetimes we
report here were fit to the sum of two exponential decays using commercial software
(Microcal Origin v. 5.0) and are reported as the average of the fits of three decays and the
associated standard deviations.

Steady-state optical spectroscopy. The absorbance spectra of oligothiophene
chromophores in solution were measured using a Hitachi 'U-4001 UV-visible
spectrophotometer. Spectra were collected with 5 nm resolution. Fluorescence spectra
for solutions were measured on a Hitachi F-4500 fluorescence spectrometer. Excitation
and emission slits were adjusted according to spectral intensity. Previous Chapters (4,5)
that reported on the linear response of oligothiophene monolayers have shown that their
steady state spectra are very similar to solution phase spectra.

Optical null ellipsometry. The thicknesses of primed substrates and monolayer
samples were determined using a commercial optical null ellipsometer (Rudolph Auto-EL
11) operating at 632.8 nm. Data were acquired and reduced using Rudolph software. For
both primer and ZP layers, we used a value of the refractive index of n = 1.462 + or for

the determination of film thickness.

6.3 Results and Discussion

The focus of this work is on understanding the morphology and structural
heterogeneity within a ZP monolayer made with varying concentrations of oligothiophene
chromophores. The unusual population decay dynamics of the chromophores in these

environments allow for modeling that can shed light on the physical and structural

434-

properties of the monolayers. Because the surface properties extracted from the
spectroscopic data depend on the assumptions and approximations made in the modeling,
it is important to compare these findings to a direct, physical determination of surface
structure and morphology. AF M provides a real space image of surface topography with
high resolution. The surface structure and morphology seen with the AFM is in excellent
agreement with the information derived spectroscopically. Our findings are significant in
and of themselves because it is not clear from first principles that the spectroscopic data
are sensitive to the same island properties as the AFM probes. We consider the two
bodies of information individually.

Optical Spectroscopy. To appreciate the structural information content of the
lifetime data, it is important to consider the result expected for a uniform monolayer first.
The systems we report on here, specifically, mixed BDP and DDBPA monolayers, mixed
QDP and DDBPA monolayers and mixed BDP, QDP and DDBPA monolayers on
oxidized Si(100), all yielded the same non-exponential excited state population relaxation
decays (Figure 6.2). We also note that the regressed time constants for these data are
independent of chromophore concentration. These two findings place substantial limits on
the structural possibilities for these monolayers.

The expected case for a monolayer of chromophore is that Forster energy transfer
will dominate the excited state relaxation dynamics. The expected result for a simple
monolayer is that the fluorescence lifetime of the donor will decrease monotonically as the
concentration of acceptor is increased, due to dipolar donor-acceptor coupling. We
synthesized monolayers from lmM bisphosphonate solutions where the donor BDP is 5%

of the total bisphosphonate concentration (0.05 mM) and varying amounts of acceptor

~135-

1000

 

TIIIIII

8363'“

.0

O.

0"

.0.
O

O

T

 

 

 

0
8° '1
’33 * 0° %
g . O O O ‘t
8 ° ‘
v 100 _— ,3 935g .5 . .
b - 0 £3 0‘ 9
.5.) .i o ‘ .
C1 '- 06% 9 o
8 "o o o; ‘ ’6’ o
E '0 6% . “P g .
t: -. o . 0 $0.03 . o .0.
.9. n8 0 o ouoouwo..oo..
a I 0mm 0 a... -0...
.... in coma) o co 0 0.. 001
E .0 080°C? 0 0...... o
0 lo (D o o o o. -0000
. (000 o o .00. .0
10 ~01) o co
50 ooomo ° . 0...
m 0000mm 0 . .
o o more
'0 o
l 1 l moodocrrmocn b o 1 o
0.0 0.5 [.0 1.5 2.0
time (ns)

Figure 6.2 Representative instrument response function (0) and excited state population
decay for QDP in a monolayer (5%) (O).

-l36-

QDP (1% - 95%, 0.01 mM - 0.95 mM) are used. The remainder of the monolayer is
comprised of an optically inactive component, DDBPA, which, in its all-trans
conformation, is approximately the same length as the all-anti conformation of the
acceptor QDP.

In a monolayer where each. of the constituents is distributed statistically, the
average distance between a donor and the nearest acceptor depends sensitively on the
concentration of each. The Ferster model predicts that the rate constant for excitation
transfer, and therefore the donor radiative lifetime, depends inversely on the D-A spacing

to the sixth powerml

__ 3x2R§
D“ 210R6 [6.1]

K2 = sin 60 sin 6,, cos¢ — 2cosl90 c0361,

 

where k0... is the rate constant for donor-acceptor transfer, x2 is a geometric factor to
account for the projection of the donor transition moment onto the acceptor transition
moment, To is the donor lifetime in the absence of acceptor, R is the D-A distance and R0
is the critical transfer radius. This latter term contains information on the spectral overlap
between the donor and acceptor and the cross sections of the relevant transitions. 1f the
Ferster model describes excitation transport in the monolayers we report on here, there
should be a marked acceptor concentration (i.e. R) dependence in the donor lifetime
(I‘D-4.1) data. We present the acceptor concentration dependence of the donor lifetimes for
the mixed BDP and QDP system in Figure 6.3. There is no discernible acceptor
concentration dependence on the donor lifetimes. For monolayers composed of only one

chromophore, it may be reasonable to expect a modest concentration dependence due to

-l37-

“a

 

 

 

 

 

 

 

 

 

l --
2000 ~..
1500 - I I
F I
fr]? ..
8: 1-
g 1000 1
.43.: 1-
$—
Q)
g :1 L2:
8
E 250 -
200 - E E E
150 -
l l l l l I I L l 1
0 20 40 60 80 100

% acceptor (5% donor)

Figure 6.3 Dependence of donor lifetimes on acceptor concentration, for a monolayer
containing 5% donor.

-l38-

the spectral overlap dependence of the Forster critical transfer radius, R0, and the
characteristically large Stokes shifts exhibited by the oligothiophene chromophores (Figure
1.1). For the mixed monolayers, however, we expect to see a measurable donor lifetime
dependence on acceptor concentration because of the substantial spectral overlap of the
donor emission and acceptor absorption bands, the high fluorescence quantum yield of the
donor, and presumably favorable geometry (x2 ~ 1) for dipolar coupling between the
donor and acceptor.

There are two possible explanations for our lifetime data. The first is that energy
transfer between donor and acceptor may be too fast to resolve with our instrument. 1f
this is true, it would explain the absence of an exponential growth in the acceptor
fluorescence time-dependence, but not the absence of the donor lifetime dependence on
acceptor concentration. The apparent absence of energy transfer may also be a
consequence of the morphology of the monolayers. An underlying assumption of the
Forster treatment we described above is the uniformity of the monolayers and that the
chromophores behave as individual species. If these assumptions are violated in some
manner, it is possible that excitation transport still proceeds but the signature of the
process is different than that expected from the Forster treatment. We consider the
functional form of the population decays and how these data can be accounted for in a
model of excitation transport that takes into account heterogeneous chromophore
distribution and aggregation of the chromophores.

To understand the interaction between the donor and acceptor chromophores, we
need to understand the functionality of the population decays of the individual

chromophores. In Chapters 4 and 5, we reported the presence of two lifetime components

-l39-

for chromophore monolayers, in contrast to the single exponential decay found for dilute
solutions. We recover the same time constants within the uncertainty of our
measurements for both chromophores (Table 6.1), and these lifetimes are invariant with
respect to concentration of chromophore in the monolayer or the emission wavelength

collected.

Table 6.1 Fits of excited state lifetime data and simulated lifetime data to the firnction f(t)
= Alexp(-t/tl) + Azexp(-t/ 12).

 

 

 

 

 

 

t.(ps) 12(ps)
BDP on silicon 187 i 11 1140 i 377
QDP on silicon 228 i 45 1254 i 152
simulated data 197 1230

 

These data, taken together, suggest that it is not the individual chromophores, but rather
the physical or environmental properties of the monolayers that mediate the optical
response of the collective system. These results are reminiscent of those seen for the
population relaxation of a stilbazolium dye in a Langmuir-Blodgett film'” In that work it
was found that the chromophores aggregated significantly and the majority of the radiative
decay originated from monomer species that were present in more than one environment.
Aggregation is common in conjugated chromophore systems and has been
demonstrated in a variety of structural motifs, including monolayers. In a chromophore
aggregate, the molecules are spaced closely, (we take this spacing to be ~5 A for the ZP
lattice) so excitation transport can compete effectively with radiative decay. As a

consequence, aggregates are typically characterized by a fluorescence quantum yield

-140-

significantly lower than that of the corresponding monomer. The dominant optical
response of such a system is considered to be from the remaining non-aggregated free
monomers. The picture of such a surface is that of aggregate islands of chromophore
separated by regions of predominantly non-chromophore (either DDBPA or unreacted
surface). The proximity of radiative monomer chromophores to the aggregates can, of
course, vary widely, but we consider the two most likely cases for our conditions. Either
the monomer is present as a structural defect within an aggregate or it exists near the
perimeter of an aggregate. The dielectric response of these two environments will differ
significantly and it is therefore reasonable to expect the two types of monomer to be

”‘5' It is important to determine whether or not the

characterized by different lifetimes.
model can account for the population decay behavior we measure experimentally.

Song et al. have developed a model to describe the excitation dynamics of H-
aggregates of stilbazolium dyes in Langmuir-Blodgett layers.“l In this model, depicted in
Figure 6.4, an aggregated island of chromophore molecules, A, is excited by a short pulse
of light. The excitation can decay radiatively (a low probability event), or it can execute
diffusive excitation transport (“hopping”) about the island until it reaches a (radiative)
monomer. The radiative monomer could be located within the island (M.) or in the inter-
island interstices (M2).

A series of coupled differential equations describes the temporal evolution of

monomer and aggregate populations as given in Figure 6.4. These equations can be

solved numerically to yield the time course of the populations of M. and M2.

441-

uni—1

A+M,*+M2

m1

ks

A+M1+M2 —h‘i> A“+M,+M2 111+ A+M,+M2

1%..

A+M1+M2*

m2

Figure 6.4 Kinetic model proposed for population decay dynamics in donor and acceptor
monolayers. A refers to the aggregate species and M1 and M; are monomers in different
chemical environments.

442-

d[A‘]
dt
dlM.’ 1
dt
d[M§]
dr
dlM.]
dr
dle]
dt

 

= -k..14'1(1M. 1+1M.1)—k..14'1

= -k...tM.’1+k..tA‘11M.1

= _kM.tM;1+k..tA‘itM.1 [6°]
=kM.tM:1-k.,t4'1tM.1

:kM2[M;]-kx,[A.][M2]

 

In the numerical simulations, we assume that the aggregates are excited, with the
populations of excited monomers being established by excitation migration. We estimate
k,“ = 1013 s", a value consistent with excitation hopping in LB films of stilbazolium dyes.
We can place limits on this value, as we discuss below. We estimate the non-radiative
decay rate constant kn, = 106 s'l for the aggregates. While the actual value of k... may
differ somewhat from this estimate, its value does not affect the results of the simulation
significantly because km will be smaller than km for essentially any system where excitation
transport is significant. The lifetimes of M1 and M2 were varied to produce a decay that
closely approximates the experimental population decays. The values used in this
simulation (Figure 6.5) were km = 7.0 x 109 5'1 (m; = 197 ps by fitting) and km = 1.6 x
109 (mm = 1230 ps). The simulated curve can be fit to a double exponential decay even
though four rate constants are used to construct it. While the double exponential fit does
not ostensibly account for the dynamics of the chromophores in the monolayer, it provides
a useful empirical representation of the relaxation phenomena we measure with the
TCSPC experiment. Regression of the experimental data using a double exponential

fimction produces residuals that match those shown for the fit to the model in Figure 6.5.

443-

 

 

 

 

 

 

 

 

 

 

 

 

0.04~
1.0 . g
0.02» .0
._.. ’ . O . O
’ ‘3 0.00 ” e e
72 . 00 o
m 0
0.8 8 -0.02- o g.
”'7
:1 1.0 1.5
3 .
3‘ time(ns)
,,_, 0.6
(I)
C:
0)
4.0
.E
76'
ED 0.4
"'v'i
0.2
0.0 “0* o
1 . 1 t 1 . 1 . 1
O l 2 3 4

time (ns)

Figure 6.5 Simulated population decay (0) using Equation 6.2 (see text) and fit of data
( ) to a double exponential decay of the form f(t) = Alexp(-t/t.) + Azexp(-t/tz). Inset
is a plot of the residuals fit.

 

-144-

Monolayer morphology. The key to understanding the morphology of these
systems is relating the experimental data to the sizes of the aggregates. The agreement
between model and experiment suggests that the physics of the model are appropriate for
our system. There are two system properties that together provide a sensitive measure of
aggregate size. These are k,“ and the number of excitation transport events, n, required
for the aggregate-supported excitation to find a monomer and decay radiatively. The total
hopping time (kx."*n) can be related to average distance traveled by means of a Monte

“' The experimental manifestation of the

Carlo simulation, as has been reported before.
migration of an excitation to a radiative monomer will be a build-up in our single photon
counting signal after time zero. Despite the fact that our instrument response fianction is
short (~35 ps) for this technique, it is longer than the expected build-up time and we do
not observe a build-up directly. In an attempt to resolve whether or not there is a build-up
in the experimental data at early times, we have attempted deconvolution of the response
function from the data. While there is a build-up seen in the deconvoluted data, the actual
time constant we recover for the build-up depends sensitively on the initial parameters
used in the deconvolution procedure. We therefore do not report a value for this build-up
time. Rather, we can place an upper bound of 10 ps on the excitation migration time.
From this upper bound, we can place a corresponding upper bound on the size of the
aggregate island(s) sensed by the measurement. We consider next the relationship
between total excitation migration time and island size.

We need to consider two limiting cases of aggregate-to-monomer transfer because

this step can, in principle, affect the estimate of island size. In the first case, the radiative

monomers are close in proximity to an island, perhaps even within the island but residing

445-

in a physical position that precludes their participation in the aggregate structure. In this
case, the last transfer event, from the aggregate to the monomer, is fast. As detailed
above, we cannot quantitate the lifetime of the aggregate excitation, so we use the upper-
bound value of I,“ z 10 ps. The time constant per excitation transfer event, or hop, is
taken to be 0.1 ps, so in 10 ps, an average of 100 hops can occur. Song, et al. have
published the results of calculations where the dynamics of exciton motion within an H-
aggregate were modeled using a random walk approach.‘ll This calculation produces a
curvilinear relationship (n") between mean distance traveled for exciton diffusion and the
number of steps taken. For 100 hops, the mean distance traveled is 40 A. With an
approximate 5 A spacing between lattice sites in the zirconium-phosphonate lattice, the
exciton can travel ~8 lattice separations within the aggregate. The actual aggregate shape
will, of course, play an important role in determining the island size in this model and it is
clear that the results will depend on the aspect ratio of the island. The simplest case for
this calculation is a square island, for which we obtain an average aggregate size of 64
molecules. At this point, it is fair to consider the accuracy of our estimation of I,“ = 10 ps.
We know from the raw data that it must be significantly faster than ~ 25 ps, and we
speculate that it is at least 5 ps. We calculate from these limits that the island size lies in

the range of 36 to 144 molecules (Table 6.2).

-146-

 

Table 6.2 Calculated aggregate sizes for selected excitation hopping time constants and
total migration times.

 

 

 

excitation diffusion molecules
In (p9 thop (ps) distance (A) in aggregate
0.1 5 30 36
IO 40 64
25 60 144
0.05 5 4O 64
IO 55 121
25 90 324

 

 

 

 

 

 

In addition to uncertainty in the total hopping time, it is possible that our estimate of the
hopping time constant is incorrect. If we double kxl and assume a 10 ps time constant for
transport, we recover a result that is identical to k,“ = 10'3 s'l and t = 20 ps. Thus the
model still holds. It is unlikely that the hopping rate constant would be slower than
1013 5'1 because of the proximity of the chromophore molecules to one another.

In the second limiting case, where the radiative monomer is not in direct contact
with the aggregate, the last transfer step will not be as efficient, consistent with the Forster
R" dependence (Equation 6.1). To estimate the total time required for excitation
transport, we take the critical transfer radius, R0, to be the lattice spacing, 5 A. We realize
that this is a different estimate than is typical for an organic chromophore. The close
proximity and high degree of alignment of neighboring chromophores in this system
effectively “screens” an excited chromophore from coupling efficiently to more distant
acceptors. Using this critical radius and a radiative lifetime of rm, = 200 ps, for I,“ = 0.1

ps the maximum D-A spacing is calculated to be dmax = 17.75 A. In the simplest model of

-l47-

the surface, where all the aggregates are spaced evenly, dmax ~ 2.15\/N. Thus N = 68
molecules, in essentially exact agreement with the first case. The location of the radiative
monomers relative to the aggregate is thus not a deterministic factor in the estimation of
the aggregate size.

To this point, we have postulated the existence of aggregates in these monolayers.
The basis for this postulation is the concentration independence and the functionality of
the chromophore population relaxation within the monolayer. Within this framework, we
have applied an excitation hopping model known to be appropriate for layered assemblies,
and we recover an aggregate size of ~65 molecules. Given the uncertainties in estimating
the values of some of the model parameters, it is fair to bound our island size estimate to
be between 50 and 100 molecules. For a S A inter-site spacing and a square aggregate,
we estimate the aggregate size to be on the order of 50 A x 50 A. This estimate reflects
the domain size sensed by the spectroscopic measurement and it is not clear from first
principles the aggregate domain size measured in this manner bears any relationship to the
physical size of the aggregate. Indeed, it is reasonable to expect that the density and type
of defects within the aggregates will determine the “spectroscopic size”, and not the
physical size. In addition, the spectroscopic measurements are averaged over all the
aggregate domains excited and offer little insight into the distribution of aggregate size
and order. We have used AFM to image the local structure and morphology of these
monolayers in an effort to compare the spectroscopic data to measurable physical
dimensions.

Atomic Force Microscopy Measurements. We have imaged'b' and compared the

surfaces of oxidized silicon substrates, APTES-primed substrates, QDP monolayers on the

-l48-

primed substrates, and monolayers of 5% BDP, 50% QDP and 45% DDBPA on the
primed substrates. Example topographic images are shown in Figure 6.6. We found in
Chapter 5 that the morphological differences between oxidized silicon and silica substrates
mediate the motional freedom of the chromophores in the monolayer while population
dynamics are not affected measurably by the substrate identity. As shown in Figure 6.6a,
the blank substrate is characterized by grooves in the oxidized surface that are 10 A - 50 A
deep and ~500 A apart. The origin of these features is most likely mechanical polishing.
Priming alters this morphology substantially. The grooves fill in and islands begin to
appear when the substrate is reacted with APTES (Figure 6.6b). The overall surface
roughness, 15 A - 30 A as determined by the FWHM of a histogram of the topographic
images, is very similar for all four types of samples examined. The primer layer has a very
distinct surface morphology, in which three-dimensional clusters with lateral dimension
20 A - 200 A are clearly visible. A typical nearest neighbor distance between clusters is
~ 100 A, although the separation between clusters varies from 50 A to 1000 A. Higher
resolution images (e. g. 100 A x 100 A) taken at various locations on the surface reveal no
long range order or periodicity within an island. Thus the island structure originates from
the nature of the initial priming layer chemistry. In contrast to thiol monolayers on gold,l7|
we observe no crystalline ordering or structural regularity in these monolayers over length
scales of more than a few molecules. We believe that the primer layer island formation is
determined by primer polymerization in concert with a non-uniform silanol site distribution

for these substrates.

-149-

3.338%“:

.233 Em: 98 23:8 2: E 950% 2a mEEmeE van 8an 8&3 wEE—oamobou .Z: Nd 1 mm Bo: vow: 8.8 wEwsEm 66
was .52 _.o a E828 028 ESE—.3 of. .< ooom x < ooOm 2.: 533 AS com £688 < 83 x < coo. 3 88 95:38 .88 of.
.Bgmnzm BEE a :0 $235 $9. can x50 Ram dam .x-m mo 3.8—98.: REE a AB 95 defiance x50 53> 88325 BEE“
.2859; uoEta-mm 15¢. BL .08mequ 19$ 80:? cofiExo As Co $8.35 05 .8 33:3 aoc mow—25 oEnEwomoH 96 9...»:

 

 

 

 

 

 

 

 

 

 

3N 3x
oe . am a see sen .ocm :3. 9
Se
U o.
. a
:5.” D
. m can
each w.
xaow
( 2
. ed
fia .<ex
so « mmm 9 seem . sea. a
on c_
m 9
n 6
mean w N
.1”: am (We
~_=m
3
_d

 

 

 

 

 

 

 

 

-150-

 

 

 

 

 

 

 

 

 

 

 

 

A<CN A Ox
8 on o oe- 8n we” 2: c
NQO
2
493m.
n 2
So u 8 w.
moon (
m on
o~o
. A«CN A<Vx
co om o soy cam .oe~ co. c
_ _ _ _
ago
1 c_
egom
. m z
eoom ONMT
”gem
0 am
o~o

 

 

 

 

 

 

 

9:53 5 2:2...

-151-

The adsorption of QDP onto a primed and zirconated substrate results in a surface
morphology similar to that seen for the primer layer (Figure 6.6c). Despite their outward
similarity, there are some subtle differences between the primed and QDP-coated surfaces.
Specifically, the QDP-terminated surface has a broader distribution in terms of aggregate
lateral dimension, and a narrower distribution in terms of roughness. The typical cluster
size is ~ 130 A, with an average nearest neighbor separation of ~ 200 A. A large number
of small clusters 40 A to 100 A in diameter are present in the mixed layer of 5% BDP,
50% QDP and 45% DDBPA, and for this monolayer the typical inter-cluster separation is
~ 130 A. Because of the change in surface morphology associated with the formation of
chromophore layers, we conclude that aggregates form coincident with the growth of
these layers. The growth of chromophore layer is thus not simply a decoration of the
primed substrate, but aggregation among chromophore molecules also plays an important
role. As pointed out earlier, the sizes of most of the aggregates are on the order of
hundreds of angstroms. For our sample preparation conditions, we do not see formation
of bulkier aggregates or thicker films. The thickness of each film was measured using
optical null ellipsometry and AFM. As shown in Figure 6.7, a typical thickness for the
three component mixed layer is 40 A i 10 A, with the uncertainty mainly attributed to the
roughness of the substrate and the aggregates. QDP films exhibit the same thickness as
the mixed layer, as measured by AFM. Considering the firndamentally different nature of
the spectroscopic and AFM measurements, the agreement between the two results is both
surprising and revealing. For a typical organic chromophore, the Forster critical transfer
radius is on the order of 20 A - 50 A, so the separation of the aggregate islands on the

surface serves to limit excitation migration during the excited state lifetime of the

~152-

 

 

 

 

 

 

 

 

 

 

60“
A 40
vs.
DJ
20 '
l
0 1000 o 2000
X (A)
(BDP+QDP+DDBPA)/APTES/Si0x Cursor Plot

Figure 6.7 Topographic image of a mixed monolayer of 5% BDP, 50% QDP, and 45%
DDBPA on a primed substrate. The molecules in the central 2000 A x 2000 A area was
displaced prior to this image using an AFM tip under a force of 1 nN. The overall
scanning area here is 4000 A x 4000 A acquired at a reduced imaging force of 0.05 nN.
The cursor profile reveals the thickness of the layer to be 40 A j: 10 A.

-153-

chromophore.

lhe roles of priming chemistry and substrate heterogeneity. Because the
morphology of these layered interfaces depends on both the priming and adsorption
chemistries, it is important to understand the reason(s) for the formation of the three-
dimensional islands in the first place. In Chapter 5 we reported on the surface relief
characteristics of the substrates. That roughness, while important to the motional
properties of the chromophores, does not address the basis for the island formation seen in
this work.

There are at least three factors contributing to the formation of the islands. The
dominant ones are the intrinsic surface roughness and the apparently heterogeneous
distribution of active silanol groups on the substrate,'8'lol the nature of the priming
chemistry used to form aggregated structures, and the ability of the chromophores to
aggregate either before or during the layer formation process.

In an ideal experiment, we could control selected surface properties independent of
one another by using different materials. Unfortunately, this level of control is not
achievable in materials as complex as silicon oxides. Comparison of mica and SiO.
substrates involves changes in both surface roughness and silanol group density and it is
not possible to fully decouple these surface properties. Nonetheless, it is instructive to
examine the consequences of the priming chemistry on these two surfaces. We have
imaged APTES on mica(001), an atomically flat surface known to possess relatively few
hydroxyl groups. The surface of APTES/mica, with a thickness of 7 i 3 A as measured by
AFM, is much more homogeneous than an APTES/SiOx surface. In addition, the force

required to displace APTES on mica is not as high as it is for APTES on $0,, indicating

454-

the importance of surface silanol group density. On the mica surface, scattered three
dimensional clusters are seen due to the lower density of silanol anchor sites, where on the
APTES/SiOx surface, relatively dense three-dimensional aggregate structure is
characteristic (Figure 6.6).

The primer layer chemistry determines both the stability and the morphology of the
primer layer. These properties, in turn, determine the growth behavior of subsequent
layers. Priming a SiOx or oxidized Si surface with APTES is known to produce a cross-
linked structure because of the three reactive functionalities present on each primer
molecule.“” In an effort to address the role of the Si-O cross-linking, we have also made
monolayers using APDMES. For this primer there is only one ethoxy group per molecule
to react with the substrate, eliminating the possibility of cross-linking. Using this
chemistry, we observe more homogeneous surface. The non-cross-linked APDMES
printing layer is not as stable as the polymerized APTES layer. APDMES can be
displaced with 0.1 nN of imaging force, and greater than 1 nN is required to displace
APTES. Of key importance is that when a chromophore monolayer is formed on the
APDMES-primed substrate, we observe spectroscopic population relaxation dynamics
identical to those seen for an APTES-primed monolayer. Aggregates are seen
spectroscopically even for monolayers where the potentially complicating role of the
priming chemistry has been precluded. This finding provides direct evidence for the

importance of chromophore aggregation during the layer formation process.

~155-

 

6.4 Conclusions

The combination of excited state population decay and AFM measurements has
afforded substantial insight into the morphology of zirconium-phosphonate monolayers
used in this and previous Chapters (4 and 5). The presence of aggregate islands in the
layer accounts for the approximately double exponential lifetime behavior we observe.
The size and separation of the islands, as resolved by AFM, demonstrates the
correspondence between physical and spectroscopic measurements of the interface and
explains the apparent absence of Forster energy transfer between donor and acceptor
chromophores within the layer. The origin of the island structures in the primer layer of
these assemblies appears to lie with the heterogeneous distribution of the active silanol
sites on the substrate. The aggregates formed by the chromophores derive their
characteristic sizes from both the features of the primer layer and their own propensity for
aggregation, either before or during the adsorption process. These findings will be
important in evaluating interlayer excitation transport effects in more complex multilayer

Stl'UCtLlI'CS.

-l56-

6.5 Literature Cited

H

2.

Song, Q.; Bohn, P. W. Blanchard, G. J. J. Phys. Chem. B 1997, 101, 8865.
Forster, T. Z. Naturforsch 1949, A4, 321.
Forster, T. Faraday Disc. Chem. Soc. 1959, 27, 7.

Girard, C.; Martin, O. J. F.; Dereux, A. Phys. Rev. Lett. 1995, 75, 3098.

. Drexhage, K. H. J. Lumin. 1970, 1,2, 693.

We thank Prof. G.-Y. Liu and Y. Huang (Wayne State University) for imaging the
samples used in this study.

(a) Chidsey, C. E. D.; Liu, G. Y.; Rowntree, P.; Scoles, G. .1. Chem. Phys. 1989,
91(7), 4421. (b) Poirier, G. E. Langmuir 1994, 13, 2019. (c) McDermott, C . A.,
McDermott, M. T.; Green, J. 8.; Porter, M. D. J. Phys. Chem. 1995, 99, 13257. (C!)
Liu, G. Y., Xu, 8., Cruchon-Dupeyrat, S. In Thin Film: Self-Assembled Monolayers of
Yhiols; Ed. A. Ulman, Academic Press, 1998, 24, 81.

. Wang, H.; Harris, J. M. J. Phys. Chem. 1995, 99, 16999.

Lochmuller, C. H.; Colborn, A. S.; Hunnicutt, M. L.; Harris, J. M. J. Am. ('hem. Soc.
1984, 106, 4077.

10. Lochmuller, C. H.; Colborn, A. S.; Hunnicutt, M. L.; Harris, J. M. Anal. ( 'hem. 1983,

55, 1344.

ll. Haller, I. J. Am. Chem. Soc. 1978, 100, 8050.

457-

Chapter 7

Structural Mediation of lnterlayer Excitation Transport in Zirconium
Phosphonate Multilayers

Summary

We report on our investigation of interlayer excitation transport in zirconium-
phosphonate multilayer structures. We find that dipolar (Forster) excitation transfer
between energetically overlapped oligothiophene chromophores in adjacent layers does
not play a significant role in determining the relaxation dynamics of these systems. This
finding is not consistent with the predictions of the Forster model and we account for this
discrepancy by considering the spatial modulation of the dielectric response of the ZP
multilayer assemblies. The presence of polarizable Zr-bisphosphonate layers between
layers of donor and acceptor chromophores serves to screen dipolar D-A coupling.
Intralayer excitation transfer within individual chromophore aggregates dominates
population relaxation dynamics in these systems. We also find that spacing the optically
active layers away from the primed substrate surface does not eliminate chromophore
aggregation. We discuss these data in terms of the dominant factors in the formation of

ZP layers.

-158-

 

7.1 Introduction

Both interlayer and intralayer processes must be evaluated to understand the
physical limits on the localization of excitations within multilayer assemblies. In Chapter
6 we considered intralayer excitation transport in detail and have found that aggregation
and island formation dominates the structural and optical properties of individual layers.
With this understanding in place, we next consider the efficiency of excitation transport
between layers. To understand and control excitation transport in MP films, we have
constructed a series of optically active MP multilayer films where the spatial relationship
between donors and acceptors is well controlled.

To study excitation transport within and between molecular layers, chromophores
with specific optical properties must be used. In order to facilitate excitation transport
and thus place limits on the possible structural motifs appropriate for optical information
storage in multilayer systems, the emission band of the donor should overlap the
absorption band of the acceptor. The thiophene oligomers 2,2'-bithiophene
diylbisphosphonate (BDP) and 2,2':5‘,5":2",2"'-quaterthiophene diylbisphosphonate
(QDP) are the donor and acceptor chromophores, respectively, and their structures are
shown in Figure 7.1. The emission band of BDP overlaps the absorption band of QDP
(Figure 1.1). The thiophene oligomers are not candidates for optical information storage
applications because of the facile rotation of individual thiophene rings within the
chromophores. This property is beneficial to this work because it precludes significant
steric contributions to excitation transport.

In Chapters 4, 5, and 6, we focused on achieving a fundamental understanding of

the chromophores both individually and in mixed monolayers. We now know that the

-159-

 

 

P
HO’: P 3
H5 \ / S 11
0
on
H5 \ / S \ / S 11
o
0
II OH
HOV£P\/\/\/\/\/\/\P’...010H C
HO II
o
0
II OH
HO"__.P\/\/\/\p'..-HIOH . d
Ho II
0

Figure 7.1 Structures of organobisphosphonates used in this work. (a) 2,2'-bithiophene-
S,5'-diylbis(phosphonic acid) (BDP); (b) 2,2':S',2":5",2"'-quaterthiophene-5,5'"-
diylbis(phosphonic acid) (QDP); (c) 1,l2-dodecanediylbis(phosphonic acid) (DDBPA);
(d) l,6-hexanediylbis(phosphonic acid) (HBPA).

-l60-

properties of the layers are mediated to a substantial extent by the silanol density and
distribution on SiOx and oxidized Si surfaces. Silane-primed SiO, surfaces are
characterized by islands, and the growth of at least the first several organophosphonate
layers is templated by the chemically reactive islands. Atomic force microscopy (AFM)
images of the primed substrates and of monolayers have confirmed the presence of
islands. In addition to the primer island formation, which is mediated by heterogeneity in
the distribution of surface silanol sites, we have found that the oligothiophene
chromophores aggregate during the deposition process, reinforcing the substrate-
mediated island structure. Because of the dominant role of the aggregate islands within
these monolayers and the inter-island spacing, intralayer excitation transport over
mesosc0pic length scales is not efficient.

We are concerned with understanding excitation transport between optically
active layers in systems where there is not a statistical distribution of donors and/or
acceptors within individual layers. We have synthesized MP multilayer structures with
the donor and acceptor chromophores present in varying amounts in different layers. We
expect and observe the same island morphology that we have elucidated previously for
individual layers. We have found that excitation transport between the donor and
acceptor chromophores across different layers is inefficient. We consider the possible
reasons for this experimental finding in the context of the orientational and structural

properties of these systems.

~161-

 

7.2 Experimental

Multilayer Synthesis. Metal-phosphonate multilayers were synthesized on
polished Si(100) substrates (Boston Piezo-Optics) having a native oxide layer of 5-20 A,
as determined by optical null ellipsometry. The substrates were cleaned immediately
before use by immersion in piranha solution for 10 minutes (3 H2504 : 1 H202) rinsed
with flowing distilled water, and dried with a N2 stream. The surface was primed with a
layer of 3-(aminopropyl) dimethylethoxysilane, by immersing overnight in a 0.5% v/v
solution in toluene, followed by rinses with toluene, methanol, and water, and drying
with N2. The resulting aminated surface was derivatized to the phosphonate in a solution
that is 200 mM in both POC13 and collidine in anhydrous acetonitrile for 10 minutes, then
rinsed with acetonitrile and water. A layer of Zr4’ was deposited onto the phosphorylated
surface using a 5 mM ZrOClz solution in 60% EtOH(aq) for 10 minutes. Multiple layers
are synthesized through the alternate deposition of organobisphosphonates (at least four
hour deposition time) and Zr4+ (ten minute deposition time). The organobisphosphonates
were dissolved in 80% DMSO, 20% EtOH solutions with 1 mM total concentration
([BDP] + [QDP] + [alkanebisphosphonate] = 1 mM). The syntheses of BDP, QDP, 1,6-
hexane bisphosphonate (HBPA), and 1,12-dodecane bisphosphonate (DDBPA) have been
reported in Chapter 2 and their structures are shown in Figure 7.1.

Steady-State Optical Spectroscopy. Spectra of 10’5 M solutions of each
chromophore are shown in Figure 1.1. Absorbance spectra were recorded on a Hitachi
U-4001 spectrophotometer with spectral resolution of 5 nm. Emission spectra were
measured on a Hitachi F-4500 spectrophotometer. Slits were adjusted according to the

emission intensity of the samples.

-l62-

 

Fluorescence Lifetime Measurements. Fluorescence lifetimes of chromophore-
containing films were measured using time correlated single photon counting (TCSPC).
This system has been described in detail in Chapter 3, and we present only the essential
details here. The second harmonic of the output of a mode-locked CW thYAG laser
(Quantronix 416) is used to pump a cavity-dumped dye laser (Coherent 702-2), operating
at 630 nm (Kiton Red, Exciton). The output of the dye laser is frequency doubled (LIIO3
Type I SHG) to excite the multilayer samples at 315 nm. This wavelength is a maximum
in BDP absorbance spectrum and, simultaneously a minimum in the QDP absorbance
spectrum. The sample surfaces are nearly parallel to the laser table and are held at ~5°
with respect to the incident laser beam in two directions: toward the excitation beam and
toward the detector. Fluorescence from the sample was imaged through a reflecting
microscope objective, and collected in two wavelength ranges with the detection
bandwidth being set to 30 nm FWHM in each case. BDP emission was collected at 400
nm and 425 nm, and QDP emission was collected at 525nm and 550 nm. Decays were
measured at two wavelengths in each chromophore emission range to test for wavelength
dependent dynamics. As we had found in Chapters 4 and 5, the recovered decay time
constants were independent of collection wavelength, so the time constants reported for
each chromophore are the averages of the lifetimes at both wavelengths in their
respective emission range. The typical response fimction for this spectrometer is ~ 35 ps
FWHM, as shown in Figure 7.2. The lifetimes we report are the results of fits of the data

to the sum of two exponential decays using commercial software (Microcal Origin v. 5.0)

~163-

 

fluorescence intensity (counts)

1000

100

 

h"
00
I O
h.°
O
O
.-
'_- O
20%
" O o
O

 

 

 

' ca 0? o..‘
o 8 0 ~ 3'. ‘ o
. 0 0:0-.. &.
:- 8 :-.00 00.0.0...
1-0 O 0.000...-00‘00 .0.
.. O O O O--- 0 no.--
- CDC 0 . .-.”O-.OO
- CDO --“0-0- 0“
- amaze 000...... C)
r mo 0 o to... ---
l OCDCD I 1 0.1. O..-
0 l 2

time (ns)

Figure 7.2 Typical instrument response function (0) and excited state population decay
(O) (425 nm) for a multilayer with firll donor and acceptor layers. The emission at 425

nm is from BDP.

-l64-

and are reported as the average of the fits of at least 3 sets of experimental data in each
wavelength range. The uncertainties represent the standard deviations of the average of

the different sets of data.

7.3 Results and Discussion

The goal of this work is to understand the effects of chromophore orientation and
layer structure on interlayer excitation transport behavior of oligothiophene
chromophores in ZP multilayer structures. In previous studies, we have investigated the
role of photonic energy transfer within a single mixed chromophore layer (Chapter 4) and
the role of substrate identity in determining the structural freedom available to
chromophores within the monolayers (Chapter 5). For a homogeneous monolayer, where
the donor and acceptor chromophores are distributed statistically, Forster energy transfer
is expected to manifest itself as a decrease in the donor lifetime as the acceptor
concentration is increased. For mixed D-A monolayers, we did not observe changes in
donor lifetime as a function of changes in acceptor concentration. The absence of the
expected response was attributed to structural heterogeneity within the layer. The
experimental data are accounted for quantitatively by the model detailed in Chapter 6 for
excitation transport where ~100 A aggregated islands of chromophores are separated by
distances sufficiently greater than 50 A to preclude efficient inter-island excitation
transport. With this understanding of intralayer energy transfer processes in MP
multilayers, we turn our attention interlayer transport.

The TCSPC measurements yield a nonexponential decay for both donor and

acceptor emission that is approximated well using a double exponential decay function.

-l65-

The short time component is ~ 200 ps, which corresponds to that of radiative
oligothiophene monomers, the same as seen in solution, and the long component has a
time constant of ~ 1 ns, which we consider to originate from radiative monomers in an
aggregated chromophore environment.

The model for radiative dynamics of these layers has been developed by Song, et
al. for excitation transport in aggregated LB films of stilbazolium dyes "I The
chromophores in a monolayer are present in two forms: most of the chromophores are in
aggregates, with a smaller number present as monomers. In this model, the aggregate is
excited, and it can either decay radiatively, a low probability event, or its excitation can
be transferred to individual molecules in the aggregate. The excitation can “hop”
randomly within the aggregate until it reaches a radiative monomer “defect” in the
aggregate or one of the radiative, isolated chromophores nearby the aggregate. The
different environments of the two radiative chromophores result in the two fluorescence
lifetimes, which provide the major contributions to the observed decay dynamics.

We present in Figure 7.3 the kinetic scheme relevant to this system. We reported
in Chapter 6 on this scheme for monolayers, where Layer 1 in Figure 7.3 is the final state.
The difference between the monolayer case and that of multilayers lies in the presence of
additional final states, containing A2, M3 and M4, where these species reside in layers
adjacent to the initially excited layer. The only fundamental difference between the
monolayer and multilayer systems is the number of final states. Any measured difference
between monolayer and multilayer systems will result from changes in kn”, km; and km.
Our signal has significant contributions from each of these rate constants and the Forster

model predicts that kml, km; and k.,, depend sensitively on the presence or absence of

-l66-

 

AI +M,'+M2

 

 

 

 

 

 

 

 

 

 

kml
k," K Layer 1 \
hV l"nr Al ‘1' M1 + M2 +
A1+M1+M2—> A1.+M1+M2 —"
A2+M3+M4+A2.+M3‘+M4.
k
Xt k Layer2 j
ka
A1+ M1 + M2.

Figure 7.3 Kinetic scheme for population dynamics in layered structures. A. is the
donor aggregate and A2 is the acceptor aggregate. M. and M2 refer to radiative donor
monomers and M3 and M4 refer to radiative acceptor monomers. For a monolayer (Layer
1), only the species A., M. and M2 are present in the relaxed system. For a multilayer
(both Layer 1 and Layer 2), A2, M3 and M4 are also present in the relaxed system.

-l67-

 

Layer 2. The first questions must therefore be the applicability of the FOrster model and
the size of the expected changes in these rate constants resulting from the presence of
acceptor layer(s). Our previous work on monolayers has demonstrated that dipolar
energy transferm] within aggregates and between aggregates and monomers is operative.
We therefore expect the same basic mechanism to hold for multilayer systems. The
existence of island structures within a given layer precludes efficient energy transfer over
large distances within that layer. Because the growth of one layer on top of another
requires spatial registration, donor and acceptor chromophores in adjacent layers will be
in close spatial proximity to one another. We describe the multilayer structures we have
synthesized to evaluate interlayer excitation transport in Table 7.1.

For monolayers, the early time emission dynamics of the radiative monomers are
mediated by diffusive excitation transport within the aggregate to the monomer "' This
model provides exact agreement with atomic force microscopy data on the surface
morphology of ZP monolayers. (Figure 6.6) This same physical picture holds for each
layer in a multilayer assembly based on the form of our spectroscopic data (vide infra).
The question at hand is to determine whether or not interlayer excitation transport can
compete effectively with intralayer transport in ZP multilayers.

Dipolar excitation transport among like chromophores (D*—>D) as well as
between donor and acceptor chromophores (D*—>A) both proceed within a given layer,
as demonstrated in Chapter 6. The depopulation dynamics of D“ are related to the
characteristic length scale of structural heterogeneity within the layer. For multilayers,
our data point to the donor lifetime being dominated by intralayer (D*—+D) dipolar

excitation transport. Within the framework of the FOrster model, this is an unexpected

-l68-

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

mom ..

 

 

 

 

 

 

 

 

 

 

 

“50 I
face D
25:33 BEE I
Emma £52 580 $8_ Emma 03: $58 $02 30 88. gm :8. _
Emma c>8. Emma 39: émoo {52 $58 $8_ Em :52 N
Emma 38. Emma c>8. Emma $02 $58 $8. M
$58 $8. $58 :52 $58 9.? Emma :8. as
$58 $8. E80 ..\.8_ Emma cso? E80 :3 m
Em: XS
mom gem “50:8. 80:3 snags. o
«mam x8
Eo<§_ Smxm $58 $8. 8m ..\..8_ 5
8m e>8. m
o m v m N ~ 92
295m

 

 

.vs 8:3.» E 3288a 83 06:0:— ofi 2 mic—53280 8388 @893. mo cozqtomoo fin 033—.

-l69-

result that reveals the important role of the metal bisphosphonate layers in mediating
interlayer excitation transport.

In a heterogeneous ZP monolayer, the aggregate size determines the observed
early time dynamics. In a multilayer film, the average distance between layers depends
on the thickness of the spacer layers separating the chromophore layers. Both spacing
and donor orientation relative to the acceptor will determine the rate constant for energy

transfer in the F6rster modelml

_ 31(2R6

[hi—2101?: [7.1]

 

The rate constant for energy transfer from donor to acceptor is kD-,., I.) is the excited
donor radiative lifetime, R is the distance between donor and acceptor, and R0 is the
Forster critical radius, a term containing spectroscopic overlap and fluorescence quantum
yield information. The term K2 is a geometrical factor describing the relative orientation
of donor and acceptor transition moments,

K = sin 60 sin 6,, cos¢ - 2cos€D c053,, [7.2]
where the angles 9.) and 6A are the angles of the donor and acceptor with respect to the
vector connecting them, and d) is the azimuthal angle between the transition moments.
Angles of 6.) = 9... = 90° and (b = 0° correspond to a monolayer where the donor and
acceptor are parallel to one another. For a multilayer where the donor and acceptor are
stacked on top of one another, 6.) = 9,. = 0° and the azimuthal angle d) becomes
coincident with the angles 9. The terms K2 and R are the experimental variables pertinent
to the interpretation of our data. R0 and I.) are constants for the purposes of the

experiments we report here. If interlayer transport competes effectively with intralayer

-170-

 

 

transport, we expect the donor lifetime in a multilayer assembly to be sensitive to the
distance between donor and acceptor layers. Experimentally, the lifetimes we measure
for multilayers having adjacent donor-acceptor chromophore layers are the same as those
we measure for individual chromophore layers (Figure 7.4). We observe that, for the
layered oligothiophenes, interlayer F6rster energy transfer does not compete efficiently
with intralayer relaxation.

We attempt to understand these results in terms of the orientation and proximity
of the donor and acceptor chromophore transition moments. The thiophene oligomers are
a good choice for the investigation of energy transfer because of their large absorptivities,
characteristically high fluorescence quantum yields and substantial spectral overlap of the
donor emission and acceptor absorption bands (Figure 1.1). These factors are consistent
with a value of R0 ~ 25 — 50 A for either D*—~>A or D*—>D excitation transport, a range
typical for such chromophores. In either case, the intermolecular distance lies within the
critical radius, R0, suggesting efficient interlayer and intralayer excitation transport.

The orientation of the donor and acceptor transition moments can also play an
important role in determining the form of the observed data, and this factor is accounted
for in the Forster model through the term K2. The term K2 can vary between 0 and 4
(Equation 7.2 and Figure 7.5) depending on the relative orientation of the donor and
acceptor transition moments. For isotropic systems and cases where orientational
averaging occurs on a time scale that is fast relative to excitation transport, <K2> = 2/3.
Within a single ZP layer, x2 ~ 1 because of aggregation and strong interactions that give
rise to parallel molecular alignment. We now consider the value of K2 for interlayer

transport. If chromophores in the donor and acceptor layers are aligned with their

-l71-

 

 

1600 -
b E o _r
1200 - E 0 __

 

 

 

’5? “

Cl. 800 -

V

G)

E I; 1’
1:; a If
9'5

:—

O 250 h-

C

O

“U

200 E

 

 

 

 

 

150 L-
l 1 L 1 l 1 l 1 l 1 l
l 2 3 4 5 6
sample

Figure 7.4 Fluorescence lifetimes measured for various layered systems as described in
Table 7.1. Decays were fit to the firnction f(t) = A.exp(-t/t.) + A2exp(-t/t2).

-172-

 

Figure 7.5 Calculated values of x2 from Equation 7.2 as a function of donor and
acceptor orientation.

-l73-

transition moments both perpendicular to the layer plane, where 9.) and 9,. = 0°, K2 ~ 4,
independent of «1). In the limit of 8.) and 02, = 0°, the angle (1) becomes coincident with the
angles 6 (Equation 7.2 and Figure 7.5). The exact determination of intermolecular
distance is not possible, owing to the delocalization of the molecular transition moments,
but if we assume that Rimmyfl ~ Rgmhya, then we would expect interlayer excitation
transport to be four times more efficient than intralayer transport. For ZP layers, we
found in Chapter 5 that 9 ~ 30° with respect to the surface normal. Using this value for
9.2 and 92., we recover K2 = 1.65 for (b = 0° and K2 = 2.25 for d) = 90° (see Figure 7.5).
The dependence of x2 on (b is small for a tilt angle of 30°. For donor and acceptor tilt
angles of ~30°, interlayer transport is expected to be approximately twice as efficient as
intralayer transport processes, assuming Rum.aycr ~ Rummy“.

The fact that we observe no interlayer excitation transport stands in sharp contrast
to the predictions of the Forster model. This unexpected deviation from Forster behavior
can be explained in the context of the structural properties intrinsic to the multilayer
assembly. Dipolar excitation transport has been shown to explain our monolayer data
quantitatively. We expect the Forster model to hold equally well for multilayers. An
important assumption of the Forster treatment is that the dielectric response of the space
between donor and acceptor is uniform. For the multilayer assemblies we have
constructed, this assumption is not satisfied. Specifically, the layered structure of ZP
films produces a spatial modulation of the dielectric response of the system. The
presence of the Zr-bisphosphonate layer between donor and acceptor can, in principle, act
to screen the oscillating transition dipole moment of the donor from the transition

moment of the acceptor. The polarizable phosphonate anions can act as “polarizability

-174-

screens”, precluding the efficient transfer of excitation between individual layers.
Although we are not aware of any direct measurements of the refractive index of bulk
ZI‘(PO3R)2 compounds where R is a simple substituent such as -H or -CH3, we estimate
this value to be n ~ 1.65 and for organic constituents, we estimate n ~ 1.46. Regardless
of the exact values, the modulation of n is sufficient to interfere with direct dipolar
coupling of chromophore transition moments in adjacent layers.

While the screening of one transition moment from another is consistent with our
experimental data, it is unfortunately not possible to acquire another body of data to
evaluate this model directly. The nature of ZP multilayer chemistry does not allow
variation of the multilayer structure in such a way as to preclude polarizability screening.
Elimination or variation of the MP network will serve to change the chemistry of the
system and thus perturb the structural and spectroscopic properties of interest. Further
experimentation using different interlayer linking chemistry, where the interlayer
connections possess fewer polarizable functionalities, is underway in our laboratory. We
do note that no apparent screening effect is seen in multilayer structures where the
density and organization of polarizable moieties within the multilayer structure is
limited.”

In addition to understanding interlayer transport, we can also add to our overall
understanding of the morphology of the film as layers are built up. The priming
chemistry and the chromophore aggregation give rise to island formation within a
monolayer as reported previously, but there is some evidence in the literature that this
structure is not preserved with the addition of multiple layers. One possibility for the

growth of multiple layers is that one will grow in registration to another, preserving the

~175-

 

initial island structure. This structural motif has been seen by AFM for two C.6BPA
layers!” It is also possible that the layers could “fan out” and grow together to form a
more uniform surface with the addition of several layers. There is also some microscopic
evidence for this process. To address this question from a spectroscopic point of view,
we synthesized films with chromophore layers that were spaced from the surface by five
DDBPA layers, in order to probe the local structure further from the substrate. The
optical response of the chromophore layers spaced away from the substrate is identical to
that of monolayers formed directly above the substrate. This result points to the
importance of chromophore aggregation coincident with the formation of the layers.
Even if the films do “smooth out” as layers are added, we measure the same time
constants for all cases, suggesting that the chromophores aggregate regardless of what
type of surface they are deposited on. This finding suggests that the formation of
statistically distributed chromophore layers will require a different chemical approach.
Another concern in the synthesis of chromophore-containing films is the effect of
the surrounding dielectric environment on the optical properties and dynamic behavior of
the chromophores. In a film, a chromophore layer can be adjacent to several possible
environments: the primed substrate, another chromophore layer, an alkane layer, or air.
As shown in Table 7.1, we have varied both the donor and acceptor environments in the
series of films we studied. In Chapter 4, we found that surrounding alkane layers had no
effect on the dynamics of a single donor layer. For the multilayer films in this study, we
find similar independence of the dynamics of both chromophores on the dielectric
environment. This further points to the role of the polarizable ZP network between layers

in restricting interlayer interaction in these MP films.

~176-

 

7 .4 Conclusions

We have found that excitation transport between layers of oligothiophene
chromophores in a ZP multilayer structure does not compete efficiently with intralayer
transport, in contrast to the predictions of the Forster model. The absence of interlayer
excitation transfer can be accounted for by screening of adjacent chromophore layers by
the Zr-bisphosphonate portion of layers. This screening is, in effect, the result of the
spatial modulation of the non-resonant dielectric response of the system. These results
are important for understanding the morphological, structural, and optical properties that

will be usefirl for the incorporation of layered organic materials in future technologies.

-177-

 

7.5 Literature Cited

1. Song, Q.; Bohn, P. W.; Blanchard, G. J. J. Phys. Chem. B 1997, [0], 8865.
2. Forster, T. Z. Naturforsch 1949, A4, 321.

3. Forster, T. Faraday Disc. Chem. Soc. 1959, 27, 7.

4. Kaschak, D. M.; Mallouk, T. E. .1. Am. Chem. Soc. 1996, 118, 4222.

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

-l78-

 

Chapter 8

Conclusions

This dissertation has demonstrated the profound influence of the morphological
and dielectric properties of metal-phosphonate films on the energy transfer behavior of
the film. In the current MP literature, most of the focus is on the average properties of
the films, without any detailed study of how the properties are affected by the
morphology of the layers. ZP layers formed on SiOx substrates are inhomogeneous,
characterized by islands, presumably corresponding to high-density areas of silanol active
sites. Our transient and steady-state spectroscopic measurements and modeling in
Chapters 4 — 6 point to the aggregation of chromophores in a ZP layer, which serves to
reinforce the island structure. This aggregation complicates the optical response of the
film, as it creates two distinct environments for the chromophore: aggregated and non-
aggregated, each with a unique optical response. The separation of the chromophore
islands precludes Forster energy transfer from donor to acceptor species, as shown in
Chapter 6, but we find that excitation transport within the chromophore islands is
dominant. The detailed excitation hopping model that corresponds to the observed
behavior also was presented in Chapter 6. With the complementary technique of atomic
force microscopy (Chapters 5 and 6), we conclude that both substrate and layer
morphologies influence the optical properties of the layer.

The segregation of chromophores into microscopic islands appears attractive for
optical information storage applications, but there are several drawbacks to this structural

motif. The most glaring is our inability to control the island dimensions or placement,

-l79-

which bears directly on the ability to control the arrangement of “bits” in a storage
device. Defects will be associated with any substrate chosen, but different types of
defects (e. g. microscopic roughness vs. atomic defects) affect the structural properties of
the films in various ways. Metal substrates, especially those that are not susceptible to
oxidation, are known to be well-ordered on a microscopic scale. For instance, crystalline
Au (111) has dense, uniformly packed active sites, and MP films have been built on gold

1 Films on metal substrates have found

substrates, using a thiol priming layer.”4
application in certain investigations, such as surface plasmon resonancem or QCM“I
studies. Unfortunately, metal substrates are not useful for optical studies of thin films,
because of direct dipolar coupling to the metal, which results in lifetime quenching of the
chromophores“ Future studies of these films should include efforts to identify
compatible substrates that do not interfere with the optical response and that have active
sites that are more statistically distributed.

Another aspect of the MP films that needs improvement is the priming step. We
observed in Chapter 6 that the silane primers smoothed the surface on average, filling in
grooves or depressions in the surface of the substrate. Although they seemed to “build
up” in the recessed areas, the silanes reacted primarily in regions dense with active sites,
forming islands on the surface. The primer we used did not form any extensive network
parallel to the surface, as we might expect for trifunctional silanes. Building several
subsequent silane layers could serve to smooth out the island structure, but silane

chemistry is difficult to control, and unwanted polymerization perpendicular to the

surface is likely to interfere.

-180-

Problems with aggregation in the layer may be diminished with a different set of
chromophores. Lifetimes of the chromophores used in this dissertation averaged 200 ps.
If Forster energy transfer was operative in these systems, the lifetimes would approach
values that would be too fast to measure easily with our instrumentation. Chromophores
that are less rotationally facile would be a benefit to this investigation, because (1) they
have stronger tendencies to order within a layer and (2) the excited state fluorescence
lifetimes are typically longer,[6] allowing us a larger time window to observe the
population dynamics. Identifying chromophores that are less likely to aggregate is also
important, for instance, a conjugated molecule derivatized with constituents that prevent
the close packing required for aggregation. If the chromophores are sterically unable to
aggregate, the constituents of each island could be controlled. Despite island formation,
this average signal would approximate a homogeneouly mixed film better than the signal
we observe for the films with aggregated chromophore islands. We found that spacing
the optically active layers away from the primed substrate surface does not eliminate
chromophore aggregation.

In Chapter 7, we examined energy transport between layers in ZP multilayers,
having gained an understanding of intralayer behavior in earlier Chapters. Because of the
proximity of chromophore layers built directly on top of one another and the favorable
alignment of chromophore transition moments, interlayer Forster energy transfer was
expected to be operative. However, no energy transfer processes except the intra-island
excitation transport were observed. We concluded that the spatial modulation of the

dielectric response between the organic and ZP portions of the film in the multilayer

-181-

structures caused this behavior. The ZP network seems to act as a polarizability screen of
the dipolar oscillations of the chromophores.

It is entirely possible that the screening property of ZP films will find use in
optical storage applications, as it prevents optical cross-talk between neighboring
chromophore layers. The spacing of the optically active layers with optically inactive
layers is another straightforward method of controlling this process, but we were unable
to determine this because of the intervening ZP network. Other multilayer chemistries
may not have the same screening properties, and it will be important to learn how to
control the exchange of photonic energy in these materials. Multilayer chemistries
having a lower density of interlayer polarizable groups or a more continuous dielectric
response, which won’t interfere with the optical response of interest, need to be
developed in order to fully determine the feasibility of using molecular assemblies for

practical devices.

-182-

 

8.1 Literature Cited

1. 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.

2. O'Brien, J. T.; Zeppenfeld, A. G; Richmond, G. L.; Page, C. J. Langmuir 1994, I 0,
4657.

3. Lee, H.; Kepley, L. J.; Hong, H.-G.; Akhter, S.; Mallouk, T. E. J. Phys. Chem. I988,
92, 2597.

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

5. (a) Chance, R. R.; Prock, A., Sibley, R. J. Chem. Phys. 1975, 62, 2245. (b)
Kurczewska, H.; Bassler, H. J. Lumin. 1970, 1, 693. (c) Alivatsos, A.; Waldeck, D.
H.; Harris, C. B. J. Chem. Phys. 1985, 82, 541.

6. Karpovich, D. S.; Blanchard, G. J. J. Phys. Chem. 1995, 99, 3951.

-183-