 

 

35%.... ..2$u...._.... .
31mm. 5m

#1... gﬁmr
.2....3....¢. 22......» ..
”mug“... 5n ...r

1......2

g1! 33

2." v 1:...

2... . .39....33:
2......nh29ﬁnmmwumprﬂ
. 2.

1

\ :59:
.tsli!
23.52231!
1".
In...“
25-18:! I.
{giviuit rl

 

 

LIBRARY
Michigan State
University

 

 

 

This is to certify that the
dissertation entitled

INTERROGATING MOLECULAR SCALE ORGANIZATION IN
PLANAR LIPID BILAYERS AND LIPOSOMES

presented by

MONIKA J DOMINSKA

has been accepted towards fulﬁllment
of the requirements for the

PhD. degree in Chemistry

 

 

/g’7 Wes/7s

Major Professor’s Signature

?/z?/o€

Date

 

MSU is an Afﬁrmative Action/Equal Opportunity Employer

—~—.-—.-u—-n-n-o-a-I-n-o--.—.-.-.3

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

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5/08 KzlProlechres/CIRC/DateDue Indd

 

 

INTERROGATING MOLECULAR SCALE ORGANIZATION IN PLANAR LIPID
BILAYERS AND LIPOSOMES

By

Monika J. Dominska

A DISSERTATION

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

DOCTOR OF PHILOSOPHY

Chemistry

2009

ABSTRACT

INTERROGATING MOLECULAR SCALE ORGANIZATION IN PLANAR LIPID
BILAYERS AND LIPOSOMES

By

Monika J. Domiriska

Biomolecules such as membrane proteins and enzymes are attractive candidates
for use in biosensing devices because of their high selectivity, but to maintain the
biological activity of these biomolecules, the interface in which they reside must mimic
their natural environment. The activities of biomolecules associated with biological
membranes are best maintained in lipid bilayers. However, to make a sensing device,
artiﬁcial lipid bilayers must be attached to a transducer surface. Because the organization
of lipids and biomolecules in a biological membrane is a complex mosaic that is
characterized by its ﬂuidity, the biomimetic lipid bilayer structures on transducer surfaces

must have ﬂuid properties similar to those of biological membranes.

We have imbedded selected molecular probes into our biomimetic interfaces to
study their properties as candidates for supported lipid membranes capable of
maintaining the activity of biomolecules such as transmembrane proteins. This research
focuses on the organization of biomimetic interfaces absent protein incorporation. Our
model systems were monolayers and bilayers doped with a reporter molecules. We have
used self-assembly, Langmuir-Blodgett (LB), and Langmuir-Schaefer (LS) methods to
construct planar lipid membranes on hydrophilic substrates (gold and indium-doped tin

oxide (ITO) for electrochemistry, silica for spectroscopy).

Pyrene tethered to the interface acted as our reporter molecule. This probe has
shown to experience signiﬁcant molecular freedom as the only constituent of the
interface. The addition of aliphatic coadsorbates to the interface improves the
organization of the molecules by making the interface more rigid. Capping the
monolayer with top lipid leaﬂet helps the organization of the molecules in the bottom

leaﬂet even further.

We attempted liposome fusion on gold surfaces where the liposomes were
composed of mixed lipids including thio-lipids modiﬁed with polyethylene glycol (PEG)
in the headgroup. The failure to produce surface-attached lipid bilayer structures led to
the need to understand the solution phase lipid assemblies that were presented to the
interface. The addition of the thio-PEG lipid to the lipid assembly produces changes in
organization that are most pronounced in the lipid headgroup region whereas the acyl
chain region is minimally affected. To probe the dynamics of the acyl chain region we

have used a free chromophore (perylene).

ACKNOWLEDGEMENTS

I would like to express my gratitude for the help in completion of this dissertation
to Professor Gary Blanchard. He has been a great advisor to me and he has offered me
all the possibilities that every young researcher needs at the beginning of their career. He
taught me a lot of things and he always had his conﬁdence in me. I truly appreciate it. I

have been very fortunate to join his group.

I would like to thank second advisor Professor Pawel Krysir'iski from University
of Warsaw for introducing me to Professor Blanchard as well as for maintaining the
collaboration with me. I am grateful to him for hosting me in Warsaw and allowing me
to perform some of my experiments in his lab.

As every researcher at the beginning of their career I came across many
challenges while working on my research project. Every time desperate looking for help
I could always count on my Guidance Committee. I would like to thank them all:
Professor Daniel Jones, Professor Greg Swain and Professor Merlin Bruening for always
being approachable and helpful.

I owe special thanks to my parents Krystyna and Henryk Dominski for their
encouragement, constant support, love and prayers. I want to dedicate this dissertation to

them.

I am also very thankful to my great friends and family for all the conﬁdence in
me. I very much appreciate their help in my preparations for the seminars and all the

great discussions that we always have.

iv

TABLE OF CONTENTS

List of Tables ..................................................................................................................... vii
List of Figures ................................................................................................................... viii
Chapter 1 ............................................................................................................................. 1
Introduction ..................................................................................................................... 1
Literature Cited ......................................................................................................... 17
Chapter 2 ........................................................................................................................... 20
Probing Interfacial Organization in Surface Monolayers using Tethered Pyrene. 1.
Structural Mediation of Electron and Proton Access to Adsorbates ............................. 20
Introduction ............................................................................................................... 20
Experimental ............................................................................................................. 23
Results and Discussion ............................................................................................. 27
Conclusions ............................................................................................................... 59
Literature Cited ......................................................................................................... 60
Chapter 3 ........................................................................................................................... 63
Probing Interfacial Organization in Surface Monolayers using Tethered Pyrene. 2.
Spectroscopy and Motional Freedom of the Adsorbates .............................................. 63
Introduction ............................................................................................................... 63
Experimental ............................................................................................................. 65
Results and Discussion ............................................................................................. 69
Conclusions ............................................................................................................... 85
Literature Cited ......................................................................................................... 87
Chapter 4 ........................................................................................................................... 89
Interrogating Interfacial Organization in Planar Bilayer Structures ............................. 89
Introduction ............................................................................................................... 89
Experimental ............................................................................................................. 93
Results and Discussion ............................................................................................. 98
Conclusions ............................................................................................................. 126
Literature Cited ....................................................................................................... 128
Chapter 5 ......................................................................................................................... 131
Constituent-Dependent Liposome Structure and Organization .................................. 131
Introduction ............................................................................................................. l3 1
Experimental Section .............................................................................................. 136
Results and Discussion ........................................................................................... 139
Conclusions ............................................................................................................. 1 6 1
Literature Cited ....................................................................................................... 163
Chapter 6 ......................................................................................................................... 165

V

Conclusions ................................................................................................................. I 65

vi

LIST OF TABLES

Table 3.1 Steady-state band intensity ratios for surface-bound pyrene derivatives as a
function of the solvent in which these interfaces are immersed. .......................... 73

Table 3.2 Summary of time-resolved data for P8 and P19 on silica ................................ 77

Table 3.3 Results calculated for hindered rotor model based on experimental r(0), r(oo)
and THR data. ......................................................................................................... 81

Table 4.1 Redox properties from cyclic (CV) and AC voltammetry of tethered pyrene in
mono- and bilayers. ............................................................................................. 103

Table 4.2 Peak positions for CH2 stretching modes in crystalline and liquid statesa, and
in monolayers and bilayers adsorbed on gold. aFrom reference 26 ................... 108

Table 4.3 Summary of time-resolved data for pyrenedecanoic and pyrenehexadecanoic
acid in monolayers and bilayers .......................................................................... 120

Table 4.4 Results obtained for hindered rotor model based on experimental r(oo), r(0),
and THR data. ....................................................................................................... 121

Table 5.1 Summary of time-resolved ﬂuorescence data for pyrene labeled DPPC in all
the systems reported in this paper. The percentage corresponds to mol % of thio—
PEG-DPPE in the particular system. .................................................................. 147

Vii

LIST OF FIGURES

Figure 1.1 Schematic of the redox reactions characteristic of the pyrene moiety. ............ 6

Figure 1.2 (a) Equivalent circuit used to model the AC impedance data. RSOL =
solution resistance, RCT = charge transfer resistance, CAD = adsorption

pseudocapacitance and CD1. = double layer capacitance. (b) Plot of peak
current/background current ratio vs. log (AC frequency) ....................................... 9

Figure 1.3 (a) Time-resolved emission intensities for polarization parallel (1“(t)) and

perpendicular (Ii (1)) to the excitation polarization (b) Orientational anisotropy
function obtained from data shown in (a). The ﬁt to the data is shown as a solid
line and residuals are distributed around zero ....................................................... 13

Figure 2.1 Cyclic voltammograms of: (a) P7 on gold in 0.1 M HClO4, sweep rate
2mV/s; (b) P13 on gold in 0.1 M HClO4, sweep rate 5 mV/s; (c) P8 on ITO in 1

M HClO4, sweep rate 20 mV/s; ((1) P19 on ITO in 1 M HClO4, sweep rate 20
mV/s. ..................................................................................................................... 28

Figure 2.2 Cyclic voltammograms of: (a) P7 and (b) P13 on gold in 1M HClO4. Sweep
rate 20 mV/s. Charge under the anodic peak is 2.105 uC for (a) and 3.106 uC for
(b). ......................................................................................................................... 30

Figure 2.3 Cyclic voltammograms of: (a) P8 and (b) P19 on ITO in 1M HClO4. Sweep
rate 20 mV/s. Charge under the anodic peak is 1.048 uC for (a) and 1.345 uC for
(b). ......................................................................................................................... 31

Figure 2.4 Pressure-area isoterrns for pyrenebutyric acid and octadecylamine system.
Inset: Area per molecule as a function of mole fraction of pyrenebutyric acid in
mixture of octadecylamine with different concentrations of pyrenebutyric acid. 32

Figure 2.5 Schematic of the redox reactions characteristic of the pyrene moiety (for
clarity this drawing shows only the P7 molecule). The structures shown are not
an exhaustive list of the reaction products, but are representative of the dominant
species. .................................................................................................................. 34

Figure 2.6 Cyclic voltammograms for the pyrene-containing adlayers: (a) P7 on gold;
(b) P13 on gold; (c) P8 on ITO; (d) P19 on ITO, all in 1M, 0.5M, 0.1M, 0.05M

and 0.01M aqueous HClO4. The ionic strength is maintained at 1 M for all
solutions by aqueous LiClO4. The sweep rate was 20 mV/s for all CVs. ........... 36

viii

Figure 2.7 Dependence of the midpoint potential, Em, on the pH of perchloric acid
aqueous solutions for: (a) P7 on gold, slope = -39.5 mV/pH, (b) P13 on gold,
slope = -36.3 mV/pH, (c) P8 on ITO, slope = -43.8 mV/pH, and (d) P19 on ITO,
slope = -37.9 mV/pH. For all scans the sweep rate was 20 mV/s ........................ 37

Figure 2.8 AC voltammograms for P7 (left column) and P13 (right column) on gold in
0.1M HClO4. AC frequencies are indicated in each pane. .................................. 40

Figure 2.9 ipeak/ ibackground vs. log(AC frequency) plots of monolayers of P7 (top
graph) and P13 (bottom graph) on gold. The solid lines represent ﬁts to the data

(black squares) using a Randles equivalent circuit. ’ ...................................... 41

Figure 2.10 AC voltammograms for P8 (leﬁ column) and P19 (right column) on ITO in 1
M HCIO4 ............................................................................................................... 45

Figure 2.11 ipeak / ibackground vs. log(AC frequency) plots of monolayers of P8 (top

graph, kc, = 0.2 s") and P19 (bottom graph, kc, = 0.1 s") on ITO. The lines
represent ﬁts to the data (black squares) using a Randles equivalent circuit ........ 46

Figure 2.12 CVs for mixed monolayers: (a) P7 + diluent on gold; (b) P13 + diluent on
gold; (0) P8 + diluent on ITO; (d) P19 + diluent on ITO in 1, 0.5, 0.1, 0.05 and

0.01 M aqueous HClO4. The ionic strength for all measurements was maintained
at l M by aqueous LiClO4. Sweep rate 20 mV/s. ................................................. 49

Figure 2.13 AC voltammograms for P7 + diluent (left column) and for P13 + diluent
(right column) on gold in 0.1M HClO4. ............................................................... 50

Figure 2.14 Dependence of the midpoint potential, Em, on pH in 1 M HCIO4 aqueous
solutions for (a) P7, slope = -43.2 i 3.9 mV/pH, and (b) P13, slope = -48.3 i 5.9
mV/pH, in two-component monolayer structures, on gold and for (c) P8, .......... 52

Figure 2.15 ipeak/ ibackground vs. log(AC frequency) plots of monolayers of P7 + diluent
(top graph) P13 + diluent (bottom graph) on gold. The solid lines represent ﬁts to
the data (black squares) using a Randles equivalent circuit. ................................ 54

Figure 2.16 AC voltammograms for P8 (left column) and P19 (right column) on ITO in 1
M HClO4 ............................................................................................................... 56

Figure 2.17 ipeak/ ibackground vs. log(AC frequency) plots of monolayers of P8 (top
graph) and P19 (bottom graph) on ITO. The solid lines represent ﬁts to the data
(squares) using a Randles equivalent circuit. The size of symbol used for
experimental data is larger then S.D ..................................................................... 57

ix

Figure 3.1 Structures of pyrene-terminated probes (P8, P19) and diluents used in the
construction of monolayer structures reported in this work. ................................ 68

Figure 3.2 Steady state emission spectra of pyrene derivatives P8 and P19, with and
without monolayer diluent, bound to silica surfaces. P8 (solid line), P8 with
diluent (dashed line), P19 (dotted line), and P19 with diluent (dash and dotted
line), on silica. Panel (a) is for interface immersed in cyclohexane, (b) immersed
in l-pentanol, (c) immersed in water, and (d) exposed to nitrogen gas. The I and
III emission bands are indicated in (a). ................................................................. 71

Figure 3.3 Fluorescence lifetime data for P8 bound to silica and immersed in l-pentanol.
These data are representative of the lifetime data for P8 and P19, with results of
ﬁts to the data shown in Table 3.2. Residuals show agreement of the ﬁt to a two-
component exponential decay. .............................................................................. 76

Figure 3.4 (a) Time resolved emission intensities for emission polarizations parallel and
perpendicular to the (vertical) excitation polarization for P8 + diluent bound to
silica and immersed in 1-pentanol. (b) Induced orientational anisotropy function
constructed from data shown in (a). The ﬁt to the data and residuals are also
shown. ................................................................................................................... 80

Figure 4.1 (a) Scheme of preparation of the Langmuir-Blodgett monolayer on
hydrophilic substrate by withdrawing the substrate from the subphase. (b) Scheme
of preparation of the bilayer using Langmuir-Schaefer method (horizontal touch).
............................................................................................................................... 95

Figure 4.2 Scheme of bilayers containing tethered pyrene ((a) pyrenedecanoic acid, and
(b) pyrenehexadecanoic acid) reported in this work. The structures identical are
not intended to reﬂect the actual conformation of the interfacial constituents. 99

Figure 4.3 Cyclic voltammograms of (a) pyrenedecanoic acid and (b)

pyrenehexadecanoic acid in monolayers and bilayers on gold in 0.1 M HClO4.
The scan rate is 20 mV/s. .................................................................................... 101

Figure 4.4 Cyclic voltammetry of pyrenehexadecanoic acid in a monolayer on gold in

aqueous 1 mM K4[Fe(CN)6] and in 1 mM [Ru(NH3)6]Cl3 in 0.1 M LiClO4.
Sweep rate 100 mV/s. ......................................................................................... 105

Figure 4.5 FTIR spectra of the monolayers and bilayers reported here: (a)
pyrenedecanoic acid and (b) pyrenehexadecanoic acid in monolayers, (c)
pyrenedecanoic acid and (d) pyrenehexadecanoic acid in bilayers. Spectral range

presented here is 3100 - 2700 cm-1 ..................................................................... 107

DPPC vesicles (0 mol % thio-PEG-lipid) in water. (b) Orientational anisotropy
function obtained from data shown in (a). The ﬁt to the data is shown as a solid
line and residuals are distributed around zero ..................................................... 149

Figure 5.6 Angular conﬁnement of pyrene labeled DPPC represented as 60 versus the
concentration of thio-PEG-DPPE in the studied systems. .................................. 151

Figure 5.7 “Wobbling” diffusion coefﬁcient (Dw) plotted versus the concentration of
thio-PEG-lipid in the lipid aggregates. ............................................................... 153

Figure 5.8 Fluorescence lifetime of perylene as a function of thio-PEG-DPPE
concentration in the lipid aggregates. ................................................................. 154

Figure 5.9 (a) Time-resolved emission intensities for polarization parallel (1“(t)) and

perpendicular (IJ_(t)) to the excitation polarization for perylene in DPPC vesicles
(0 mol % thio-PEG-lipid) in water. (b) Orientational anisotropy function obtained
from data shown in (a). The ﬁt to the data is shown as a solid line and residuals
are distributed around zero .................................................................................. 156

Figure 5.10 Reorientation time constants TORJ obtained for perylene as a function of
thio-PEG-DPPE concentration in the lipid aggregates. ...................................... 158

Figure 5.11 Cartesian components of diffusion coefﬁcient (a), Dx and Dz, and their ratio

(b), Dz / Dx, obtained for perylene in lipid aggregates plotted as a function of thio-
PEG-DPPE concentration in the lipid structures. ............................................... 159

xii

CHAPTER 1

INTRODUCTION

The combination of biological recognition with physical transduction to produce a
functional biosensing device holds great appeal because of the potential application of
these devices in ﬁelds such as medical diagnostics, environmental monitoring, and drug
or toxin detection. Depending on the design, the transducer surface can serve either to
monitor the biological reaction or to activate the biomolecules to function in the desired
manner. Biomolecules such as enzymes, antibodies, protein receptors, channels, pumps,

and nucleic acids constitute a broad palette of biologically selective components that can

1-4 . . . . .
be attached to a transducer surface. In developing a biosensor, it IS 1mportant to

enable communication between the biologically selective component and the transducer

surface. Several methods were developed to immobilize biomolecules on the transducer

surface while maintaining their biological activity}6 Very sophisticated methods of

. . . . . . . . 7-10
structurally allgnlng and orienting blomolecules on SOlld supports were Investigated.

Because the native environment for the biomolecules in a living cell is the cytoplasm and
an interfacial region of the cell membrane, changes caused by direct immobilization of
these molecules onto a solid support may affect their function. For example, membrane
proteins functioning as ion channels, pumps, and receptors lose their biological activity
when removed from their natural plasma membrane environment. To better
accommodate and maintain the biological activity of these molecules, the transducer

interface has to be modiﬁed in some manner.

Lipid bilayers provide a suitable environment for membrane proteins. However,
to maintain the protein native activity in the supported lipid bilayer, several requirements
must be met. One of the requirements is that the lipid bilayer must be in a ﬂuid state to
allow diffusion of the incorporated proteins. Another is that both sides of the lipid
membrane must be in contact with aqueous environments in order to accommodate
hydrophilic portions of the inserted proteins. Numerous artiﬁcial lipid bilayer systems on

solid supports have been designed in the last few decades and different bilayer deposition

5,11-27

procedures have been developed depending on the solid substrate properties. One

of the ways to deposit a lipid bilayer on a planar substrate is to add one lipid leaﬂet after
another. The ﬁrst leaﬂet can be prepared using Langmuir-Blodgett (LB) or self-
assembling techniques. To complete the lipid bilayer, the second lipid leaﬂet can be
added by Langmuir-Schafer (LS) deposition, self-assembly from organic solvents, or
vesicle fusion. Another elegant method to generate a complete lipid bilayer is liposome
fusion on a hydrophilic solid substrate. The advantage of this method is that the lipid
bilayers can be prepared easily in a fast one-step procedure. The disadvantage of using
physical adsorption methods is poor stability and the possibility of forming defects in a
created interface, which can affect the quantitative analytical information. Covalent
attachment of the lipid bilayer to the solid support, with a controlled density of covalent
attachment points, seems to be an attractive approach to membrane immobilization.
Electrochemical methods of analysis require the use of conductive substrates as

transducers. Metals such as gold, mercury, and silver have been used as substrates for

ll-l4,l6,l7

bilayer lipid membranes. Gold has been shown to be useful in the

development of well organized and stable self-assembled monolayers (SAMs).28'30 A

2

metal surface modiﬁed with an alkanethiol monolayer can serve as a platform for the
construction of supported membranes. Supported lipid membranes on gold electrodes
generally exhibit good stability which can be estimated by measurement of the membrane

capacitance. Typical capacitance for a stable and isolating lipid membrane is on the

211.,13,15,l6

order of 0.5-1 uF/cm . Many techniques for constructing lipid bilayers on

1 l-l3,15-l9

gold have been described. Most methods take advantage of thiol-gold

chemistry to self-assemble molecules on the gold surface. These monolayers are usually
well organized and relatively defect free, making them suitable for electrochemical
analysis. SAMs of alkanethiols or thiolipids have been commonly used as a hydrophobic
base for the deposition of a second lipid leaﬂet. The second leaﬂet can be deposited
using methods such as Langmuir-Schaefer, physisorption of lipids from organic solvents,

or vesicle fusion. Such interfaces have been shown to be useful for the immobilization of

. 12 . .
protein receptors. However, for transmembrane proteins such as ron channels or

pumps, the interface lacks the lateral ﬂuidity as well as a hydrophilic cushion between the

. . . . . 13
bilayer and the supporting substrate, and these factors preclude protein 1ncorporatron.

To provide a hydrophilic spacer between the lipid bilayer and the gold surface, the latter

can be functionalized with charged SAMs to allow the adsorption of a counter-charged

lipid membrane.15 One disadvantage of this method is the decoupling of the membrane

from the SAM support when high ionic strength solutions are used as electrolytes. To
maximize the hydrophilicity of the bilayer support, and simultaneously minimize the

interaction of the support with the membrane, self—assembled hydrophilic polymers were

employed.” ’22 The disadvantage of using physisorption on a polymer surface is that it

3

typically leads to incomplete vesicle fusion and low surface coverage, precluding

. . . . 21
electrochemrcal examrnatron of surface-attached proteins. Better surface coverage has

been achieved with SAMs of lipids modiﬁed at their headgroup with a thiol-terminated

1 1,17,23,24

hydrophilic polymer. SAMs formed from these modiﬁed lipids can serve as a

platform for the deposition of a second lipid leaﬂet. The second lipid leaﬂet can then be
attached to the modiﬁed surface by self-assembly from organic solvents, detergent

dilution or vesicle fusion. The membrane capacitance for this type of interface has been

2 ll,l7,23,24

reported to be on the order of 035-1 uF/cm . Supported lipid bilayers

constructed this way have been shown to successfully form ﬂuid membranes with an

aqueous region between a lipid membrane and a gold electrode, which allows

. . . . l 1,17,24
transmembrane protern incorporation and proper function.

To monitor biological events at an interface using ﬂuorescence spectroscopy, the
lipid bilayers have to be prepared on transparent substrates such as indium-doped tin

oxide (ITO) or silica. Because of the hydrophilicity of the substrates, vesicle fusion can

25-27,3

be used to deposit a planar lipid bilayer. 1 Another approach to constructing a lipid

bilayer with a hydrophilic region between the bilayer and the surface is to transfer lipids

modiﬁed with hydrophilic polymers in the headgroup and terminated with a silane group,

using the LB method.27 To complete the bilayer formation, the second lipid leaﬂet is

then added using vesicle fusion. Cellulose modiﬁed ITO substrates onto which positively

charged lipid vesicles were fused provided insulating lipid membranes.32 The impedance

analysis showed the capacitance of this kind of membranes was ca. 0.5 uF/cmz.

In this project we used various methods to build interfaces on solid substrates and
investigated the organization of these ﬁlms at the molecular level. Our long term goal is
to construct and examine the properties of supported lipid bilayers that are capable of
accommodating transmembrane proteins and maintaining the protein biological activity.
We have started with simple model systems to get a better understanding of the
complexity of the supported bilayer lipid membrane as a whole. The work accomplished

to date focused on the formation and characterization of mono- and bilayer ﬁlms that are

components of the supported lipid bilayer on transducer surfaces.”-35

Information about the bilayer structure and dynamics can be obtained by
controlled “doping” of the bilayer with probe molecules. These reporter molecules are
capable of interrogating the bilayer local organization. However, for information from
these probes to be useful, they must be localized at predetermined locations within the
interface and this can be accomplished by binding the reporter molecule to the transducer
surface. Using reporter molecules tethered at various distances from the transducer
surface, we have investigated the structure and dynamics of monomolecular ﬁlms as well
as the lipid bilayers. Localization of the probe at various points of the lipid bilayer
provides direct information about the organization and dynamics of molecules at certain
planes of the interface.

A molecular probe used in both spectroscopic and electrochemical studies of
planar lipid bilayers was a family of pyrene derivatives. The electrochemical response of

pyrene derivatives attached covalently to the solid substrates has been described in the

33,35,36

literature (Figure 1.1). In the ﬁrst anodic scan pyrene is oxidized to form radical

cation at potentials higher than 0.8 V vs. Ag I AgCl reference electrode. In the

5

 

 

 

 

 

 

 

 

 

 

OH
.. “0 133+ 0
H20 T
O NH O NH
9 8
Ho I
1.
H20
0
NH NH
3 8
f f Ho 0
' _e HO 3:, 0
H20 :
o o
H $NH
i

 

s"
t

Figure 1.1 Schematic of the redox reactions characteristic of the pyrene moiety.

subsequent cathodic scan, the radical cation reacts rapidly with water to form a
monohydroxypyrene. The monohydroxypyrene can also react with water to form one of

several isomeric forms of dihydroxypyrene. Because the'redox peaks associated with the

formation of monohydroxypyrene (0.6 V vs. Ag | AgCl) disappear in subsequent scans,36

the pyrenedione form must be the most stable product. The two dominant isomeric forms
are 1,6- dihydroxypyrene/ 1,6-pyrenedione and l,8-dihydroxypyrene/1,8-pyrenedione

couples. The redox potential of the 1,8 isomeric form is shifted toward negative

potentials by about 0.1 V relative to the 1,6 isomer.36’37 In the cyclic voltammogram of

pyrene derivatives covalently attached to the electrode surface the redox peaks from these

two couples overlap to form one broad peak at ca. 0.2-0.3 V vs. Ag I AgCl.33’35’36

To interrogate the organization of the interfacial structures doped with tethered
pyrene, we use electrochemical methods such as cyclic voltammetry and AC
voltammetry. Cyclic voltammetry provides information regarding the surface coverage
with redox active probes attached to the monolayer. Knowing the number of electrons
involved in the redox reaction (n) and the electrode surface area (A), the amount of the
redox-active probe tethered to the electrode surface (F) can be estimated from cyclic

voltammograms based on the charge under the peak (Q):

(1.1)

where F is the Faraday constant. In the absence of the Faradaic current, cyclic
voltammograms also allow the estimation of the interfacial ﬁlm thickness by measuring

the charging current and relating it to the capacitance of the interface:

C=— (1.2)

where C is the speciﬁc capacitance, d is the distance between the capacitor plates, which
is the measure of the insulating ﬁlm thickness, 8 is the dielectric constant, and so is the

permittivity of free space.

We use AC voltammetry to study redox reaction kinetics. Changes in the peak
current with the frequency of AC voltage provide information about the rate of the
electron transfer between the immobilized redox species and the electrode surface. To

estimate the electron transfer rate constant from AC voltammograms, we use a model

developed by Creager and Wooster.38 In this model the interfacial redox-active system

on the electrode surface is represented by the impedance of a Randles equivalent circuit

(Figure 1.2a) where CD1. is the double layer capacitance, C AD is the adsorption

pseudocapacitance, RCT is the charge transfer resistance, and RSOL is the solution

resistance. The relationship between the circuit elements and the parameters
characterizing the electrode coated with the redox-active monolayer when the average

potential applied to the electrode is the peak potential is as follows:

CDL =63”)
CAD =(F2AF)/(4RT) (1.3)

RCT = (2RT)/(F2A1“ke,)

where C/A is the double layer capacitance per unit area, A is the electrode area, F is the

 

 

 

R30. ——vwv——IH

4.0- o o

3.5: 0
3.0..
2.53 (b)
. 9- 2.0;

/ib

l

1.5-
1.0- ' '

l I I ‘ l I l

1 2 3 4

 

 

o-l

log (AC frequency)

Figure 1.2 (a) Equivalent circuit used to model the AC impedance data. RSOL =
solution resistance, RCT = charge transfer resistance, CAD = adsorption

pseudocapacitance and CDL = double layer capacitance. (b) Plot of peak
current/background current ratio vs. log (AC frequency)

electrode coverage with redox-active species per unit area. Based on these expressions
the electron transfer rate constant is obtained by:

I
k = —— 1.4
e, 2RCTCAD ( )

The analysis of the AC voltammetry data involves plotting the AC peak current (ip) to

background current (ib) ratio versus the logarithm of AC frequency (f) for a series of

voltammograms recorded over a broad AC frequency range:

2'
£=1+ 1 (1.5)

' C
lb RCTCDLf +62%

 

These plots exhibit three distinct regions (Figure 1.2b). At low frequencies, the current
ratio attains a plateau at the value that is indicative of the amount of the redox-active
species on the electrode surface. There is very little kinetic information in this region
because the AC ﬁ'equency is slower than the relevant redox process. At high frequencies,
the peak to background current ratio approaches unity because these AC frequencies are
much faster than the reaction kinetics. At intermediate frequencies, the peak to

background current ratio depends sensitively on the AC frequency and is related to the

standard electron transfer rate constant for the electrode attached redox-active species.

40

To interrogate the rotational diffusion dynamics of the chromophore in our model
systems, we have used time-resolved ﬂuorescence spectroscopy and measured

ﬂuorescence anisotropy decay. With this measurement, the randomly oriented

10

chromophore molecules in the sample are excited with a short, vertically polarized pulse
of light, producing a polarized emission transient from the sample. The molecules with
their absorption transition dipole moments oriented parallel to the excitation light are
excited preferentially, resulting in an oriented population of excited molecules. The
emission from the sample is polarized according to the orientational distribution of the
chromophore emission transition dipole moments, and this distribution reramdomizes as
a ftmction of time. The polarized emission transient is collected at polarizations parallel

and perpendicular to the excitation pulse (Figure 1.3a). The polarized emission

intensities (1”(t) and I i(t)) are then combined (Equation 1.6) to yield the induced

orientational anisotropy function, r(t) (Figure 1.3b):

 

1 (t)—I (t)

_ II i
r(t) — I”(t)+21_L(t) (1‘6)
r(t) = r(0) exp(—t/r0R) (1.7)

For chromophores in an unrestricted environment such as solution, the initial anisotropic
orientational distribution, r(0), can relax to a random orientational distribution following
excitation (Equation 1.7). As the molecules re-randomize, the anisotropy decays with

time to zero, at which point the ensemble of excited molecules are again oriented

randomly. The reorientation time constant (TOR) describes the motion of the

chromophore in its local environment. According to the modiﬁed Debye-Stokes-Einstein

equation, the reorientation time constant for spherical molecules is related to the

. . . . . . 41-43
rotational difquIOn constant, D, and to the vrscosrty of the local envrronment:

ll

 

1 - "Vf (1.8)

2' :— ._
0R 6D kBTS

where I] is the viscosity of the medium surrounding the chromophore, V is the

hydrodynamic volume of the chromophore, f is the frictional term to account for solvent-

solute interactions, k3 is the Boltzmann constant, T is the absolute temperature, and S is a

factor related to the shape of the chromophore.“-43

Because the volume swept out by the probe molecule is typically nonspherical, it

must be described in the context of an ellipsoidal shape, and the anisotropy decay

functionality is related to the shape of the volume swept out by the rotating molecule.44

Depending on the orientation of the absorption and emission transition dipole moments
relative to the ellipsoidal axes, and the shape of the ellipsoidal rotor, it is possible for the
anisotropy decay ﬁinction, r(t), to exhibit more than one exponential decay. This
situation requires a description of molecular motion that is more rigorous than the Debye-
Stokes-Einstein treatment. Chuang and Eisenthal have provided the theoretical treatment

for an ellipsoidal rotor, and their formulation accounts for the range of functional forms

that can be observed in experimental r(t) data.44 The ellipsoidal volume can be

described as either prolate ellipsoid or oblate ellipsoid. Assuming the absorption and
emission transition moments are parallel and lie in the chromophore n-plane, along the

long molecular axis, a prolate ellipsoid is characterized by rotation primarily about the

chromophore long in plane axis (Dx > D = Dz) (Equation 1.9)while oblate ellipsoid is
y

characterized by rotation about the axis perpendicular to the chromophore zr-system (Dz >

12

 

 

 

(a)

2000- 1 t
E? 1500-
G
D
8 .
“ 1000- Ii(t)
a?
(n
G
g 500-

0‘ ﬁr r Y I V I ' I
10000 20000 30000 40000
tune [ps]
0.15- (b)

0.10115" '

r(t)

0.05 -

 

 

 

10000 20000 30000 40000
tune [ps]

Figure 1.3 (a) Time-resolved emission intensities for polarization parallel (1”(t)) and

perpendicular (I J. (t)) to the excitation polarization (b) Orientational anisotropy function
Obtained from data shown in (a). The ﬁt to the data is shown as a solid line and residuals
are distributed around zero.

Dx = Dy) (Equation 1.10).
r(t) = 0.4 exp(—6th) (1.9)
r(t) = 0.3exp(-(2Dx + 4D,)t)+ 0.1exp(—6Dxt) (1.10)

When the motion of the chromophore is restricted in some manner (e. g. conﬁned
by attachment to a lipid membrane), the photo-selected anisotropic distribution cannot
orientationally re-randomize completely. For chromophores in a restricted environment,
the anisotropy at times long after excitation does not decay to zero. The motion of the

chromophore in a constrained environment can be expressed using a hindered rotor

model developed by Lipari and Szabo.45 In this model the orientational anisotropy

decays with time as the chromophore orientational distribution re-randomizes to the

extent possible:

r(t) = r(oo)+(r(0)—r(oo))exp(—t/rHR) (1.11)

where r(O) and r(oo) are initial and inﬁnite time anisotropies, respectively. The zero-time
anisotropy depends on the angle between the excited and emitting transition dipole
moments of the chromophore. The inﬁnite time anisotropy is associated with the degree
of motional restriction imposed on the chromophore by its immediate environment:

r(00)
r(O)

 

=0.5+(cos€0(1+c0860)) (1.12)

14

The decay time constant THR is related to semi-angle of the cone 60 within which the

chromophore is constrained, and is related to the “wobbling” diffusion coefﬁcient Dw,

which describes the motion of the chromophore about its tethering bond

79%
240W

 

THR = (1.13)

The cone angle value provides information about the organization of the molecules in the
studied system. Large cone values are consistent with large volumes being swept out by
the chromophore and represent signiﬁcant motional freedom of the molecule within the
probed environment. Small cone values are indicative of a motion of the chromophore in
a more constrained environment.

We have used various types of pyrene chromophores as tethered probes, with the
speciﬁc choice of tethered pyrene derivative depending on the system under examination.
The ﬂuorescence and electrochemical behavior of this probe is well established and
understood. In Chapters 2 and 3 we investigated both electrochemical and spectroscopic
behavior, respectively, of covalently bound pyrene on solid substrates to probe the
interfacial organization of the molecules in monomolecular ﬁlms. Pyrene in form of fatty
acid analogues was then used to probe the environment of Langmuir-Blodgett bilayers
and these experiments are presented in Chapter 4. The experiments directed toward
construction of a supported lipid bilayer system with a hydrophilic cushion between the
bilayer and the solid support, led us to study the macroscopic structure of the liposomes

and micelles. Attempting the fusion of vesicles modiﬁed with polyethylene glycol on

15

 

solid substrates, we observed polymer concentration dependent changes in the liposome
structure. In Chapter 5 we examine the organization of the molecules in these liposomes
using pyrene tethered to phospholipid acyl chain. In this study we also used a “free”

chromophore which we chose to be perylene. The reorientation dynamics of this probe

. . 46-48 . . . .
are well understood in a range of envrronments. F inal conclusrons are provrded in

Chapter 6.

l6

Literature Cited

(1) Teles, F. R. R.; Fonseca, L. P. Talanta 2008, 77, 606-623.

(2) Dauphas, S.; Corlu, A.; Guguen—Guillouzo, C.; Ababou-Girard, S.; Lavastre, 0.;
Geneste, F. New Journal of Chemistry 2008, 32, 1228-1234.

(3) Pyun, J. C.; Kim, S. D.; Chung, J. W. Analytical Biochemistry 2005, 347, 227-
233.

(4) Cosnier, S. Biosensors & Bioelectronics 1999, 14, 443-456.

(5) Guidelli, R.; Aloisi, G.; Becucci, L.; Dolﬁ, A.; Moncelli, M. R.; Tadini
Buoninsegni, F. Journal of Electroanalytical Chemistry 2001, 504, 1-28.

(6) Willner, 1.; Katz, E. Angewandte Chemie, International Edition 2000, 39, 1181-
1218.

(7) Zhang, J .; Chi, Q.; Kuznetsov, A. M.; Hansen, A. G.; Wackerbarth, H.;
Christensen, H. E. M.; Andersen, J. E. T.; Ulstrup, J. Journal of Physical Chemistry B
2002, 106,1131-1152.

(8) Kamin, R. A.; Wilson, G. S. Analytical Chemistry 1980, 52, 1198-205.

(9) Leggett, G. J .; Roberts, C. J.; Williams, P. M.; Davies, M. C.; Jackson, D. E.;
Tendler, S. J. B. Langmuir 1993, 9, 2356-62.

(10) Edmiston, P. L.; Saavedra, S. S. Journal of the American Chemical Society 1998,
120, 1665-1671.

(11) Naumann, R.; Schmidt, E. K.; Jonczyk, A.; Fendler, K.; Kadenbach, B.;
Lieberrnann, T.; Offenhausser, A.; Knoll, W. Biosensors & Bioelectronics 1999, 14, 651-
662.

(12) Plant, A. L.; Brigham-Burke, M.; Petrella, E. C.; Oshannessy, D. J. Analytical
Biochemistry 1995, 226, 342-8.

(13) Krysinski, P.; Zebrowska, A.; Michota, A.; Bukowska, J.; Becucci, L.; Moncelli,
M. R. Langmuir 2001, 17, 3852-3857.

(14) Becucci, L.; Guidelli, R.; Peggion, C.; Toniolo, C.; Moncelli, M. R. Journal of
Electroanalytical Chemistry 2005, 5 76, 121-128.

(15) Steinem, C.; Janshoff, A.; Ulrich, W.-P.; Sieber, M.; Galla, H.-J. Biochimica et
Biophysica Acta, Biomembranes 1996, 1279, 169-80.

17

(16) Becucci, L.; Innocenti, M.; Salvietti, E.; Rindi, A.; Pasquini, 1.; Vassalli, M.;
Foresti, M. L.; Guidelli, R. Electrochimica Acta 2008, 53, 6372-6379.

(17) Steinem, C.; Janshoff, A.; von dem Bruch, K.; Reihs, K.; Goossens, J .; Galla, H.-
J. Bioelectrochemistry and Bioenergetics 1998, 45, 17-26.

(18) Twardowski, M.; Nuzzo, R. G. Langmuir 2003, 19, 9781-9791.
(19) Ha, J.; Henry, C. S.; Fritsch, I. Langmuir 1998, 14, 5850-5857.
(20) Plant, A. L. Langmuir 1993, 9, 2764-7.

(21) Majewski, J .; Wong, J. Y.; Park, C. K.; Seitz, M.; Israelachvili, J. N.; Smith, G. S.
Biophysical Journal 1998, 75, 2363-2367.

(22) Cassier, T.; Sinner, A.; Offenhauser, A.; Mohwald, H. Colloids and Surfaces, B:
Biointerfaces 1999, 15, 215-225.

(23) Naumann, R.; Schiller, S. M.; Giess, F .; Grohe, B.; Hartman, K. B.; Kaercher, 1.;
Koeper, 1.; Luebben, J .; Vasilev, K.; Knoll, W. Langmuir 2003, 19, 5435-5443.

(24) Naumann, R.; Walz, D.; Schiller, S. M.; Knoll, W. Journal of Electroanalytical
Chemistry 2003, 550-551, 241-252.

(25) Merzlyakov, M.; Li, B; Casas, R.; Hristova, K. Langmuir 2006, 22, 6986-6992.
(26) Wagner, M. L.; Tamm, L. K. Biophysical Journal 2000, 79, 1400-1414.
(27) Merzlyakov, M.; Li, B; Gitsov, 1.; Hristova, K. Langmuir 2006, 22, 10145-10151.

(28) Porter, M. D.; Bright, T. B.; Allara, D. L.; Chidsey, C. E. D. Journal of the
American Chemical Society 1987, 109, 3559-68.

(29) Karpovich, D. S.; Blanchard, G. J. Langmuir 1994, 10, 3315-22.

(30) Schessler, H. M.; Karpovich, D. S.; Blanchard, G. J. Journal of the American
Chemical Society 1996, 118, 9645-9651.

(31) Brian, A. A.; McConnell, H. M. Proceedings of the National Academy of Sciences
of the United States of America 1984, 81, 6159-63.

(32) Hillebrandt, H.; Wiegand, G.; Tanaka, M.; Sackmann, E. Langmuir 1999, 15,
8451-8459.

(33) Dominska, M.; Jackowska, K.; Krysinski, P.; Blanchard, G. J. Journal of Physical
Chemistry B 2005, 109, 15812-15821.

18

(34) Dominska, M.; Krysinski, P.; Blanchard, G. J. Journal of Physical Chemistry B
2005, 109, 15822-15827.

(35) Dominska, M.; Krysinski, P.; Blanchard, G. J. Langmuir 2008, 24, 8785-8793.

(36) Mazur, M.; Blanchard, G. J. Journal of Physical Chemistry B 2004, 108, 1038-
1045.

(37) Moriconi, E. J.; Rakoczy, B.; O'Connor, W. F. Journal of Organic Chemistry
1962, 27, 2772-6.

(3 8) Creager, S. E.; Wooster, T. T. Analytical Chemistry 1998, 70, 4257-4263.

(39) Creager, S.; Yu, C. J .; Bamdad, C.; O'Connor, S.; MacLean, T.; Lam, E.; Chong,
Y.; Olsen, G. T.; Luo, J .; Gozin, M.; Kayyem, J. F. Journal of the American Chemical
Society 1999, 121, 1059-1064.

(40) Weber, K.; Hockett, L.; Creager, S. Journal of Physical Chemistry B 1997, 101,
8286-8291.

(41) Debye, P. Polar Molecules, 1957.

(42) Hu, C.-M.; Zwanzig, R. Journal of Chemical Physics 1974, 60, 4354-7.

(43) Perrin, F. Journal de Physique et Ie Radium 1936, 7, 1-11.

(44) Chuang, T. J .; Eisenthal, K. B. Journal of Chemical Physics 1972, 5 7, 5094-7.
(45) Lipari, G.; Szabo, A. Biophysical Journal 1980, 30, 489-506.

(46) Stevenson, S. A.; Blanchard, G. J. Journal of Physical Chemistry B 2006, I 10,
13005-13010.

(47) Koan, M. M.; Blanchard, G. J. Journal of Physical Chemistry B 2006, 110,
16584-16590.

(48) Lapinski, M. M.; Blanchard, G. J. Chemistry and Physics of Lipids 2008, 153,
130-137.

19

CHAPTER 2

PROBING INTERFACIAL ORGANIZATION IN SURFACE MONOLAYERS USING
TETHERED PYRENE. l. STRUCTURAL MEDIATION OF ELECTRON AND
PROTON ACCESS TO ADSORBATES

Introduction

The development of self-assembling monolayer chemistry, pioneered by Nuzzo

and Allara for disulﬁdes and thiols on gold,1 and subsequently expanded to a range of

different surfaces and bonding chemistries, has served as a versatile and powerful means
for controlling interfacial properties. Self-assembled monolayers (SAMs) provide ready
access to experimental systems where the physical properties of interfaces can be
modiﬁed, and the chemical identities of the interfacial species can be examined by
spectroscopic and electrochemical techniques. Because of the structural versatility of
SAMs and the relative ease with which they can be synthesized, there is a great deal of

practical interest in these materials as potential chemical and biological sensing systems.

Since the original work on thiol/gold chemistry and carboxylic acid/alumina

9

. . 3 . . . .
bonding strategies, there has been a Signiﬁcant effort aimed at making SAM structures

more robust. The gold-sulfur bond responsible for alkanethiol/gold monolayer formation

is modest in strength, and this fact is ultimately responsible for the labile nature of this

family of interfaces.4’5 Indeed, there is interest not only in the formation of enthalpically

more robust mono- and multilayers, but there has also been a sustained effort to create
interfacial layers on oxide surfaces, for a variety of fundamental and practical reasons.

The growth of interfacial mono- and multilayer structures using metal bisphosphonate

20

and related ionic interlayer linking chemistry was developed by the Mallouk, Thompson

6-11 . . . . .
and Katz groups, and IS in Wide use today. Similarly, there have been several reports

of covalent interlayer linking chemistry to form robust, simple mono- and multilayer

interfaces on oxide surfaces such as indium-doped tin oxide (ITO) and SiOx.]2’l3 Until

recently, the chemical methodologies appropriate for layer formation on relatively well
organized metal surfaces have not been accessible to the chemical approaches used on
oxide surfaces. Recent work by Krysir’rski and Blanchard has shown that the

electrochemical formation of an oxide layer on a gold surface can be reacted with acid

chlorides, which serve to stabilize the gold oxide layer by forming a gold-ester.l4 This

advance has brought the chemistry previously restricted to layer growth on oxide surfaces

to more organized metallic substrates.

Given the broad range of means available for the creation of interfacial structures,
it is important to consider the ways in which these interfaces can be characterized. The
“standard” techniques such as optical ellipsometry, contact angle measurement and
infrared (IR) spectroscopy are not well suited to providing direct information on the
molecular scale organization of these interfacial structures. This spatially averaged

information is very useful, but recently it has been shown that such data can be

. . 15 . . . .
misleading. If these 1nterfac1al structures are to be used for chemical sensrng, the

intrinsic selectivity of the interface, which is determined at the molecular level, must be
probed directly. It is this purpose which we address here. We have designed a series of

SAMs that contain pyrene derivatives bound covalently to either gold or oxide (ITO,

SiOx) surfaces. Pyrene is a well known chromophore used to characterize local

21

“polarity”, and both its time- and frequency-domain spectral properties can be used to
measure short range organization within mono- and multilayer structures. The
electrochemistry of pyrene is also well understood, particularly with respect to its
oxidative instability. Interrogating both the spectroscopic and electrochemical behavior
of covalently bound pyrene on metallic and oxide substrates allows us to gain access to
local organization in these systems as a function of the interface chemical identity. In
this ﬁrst of two chapters, we report on the electrochemical response of tethered pyrene
derivatives on Au and ITO substrates. We have found that the organization of several of
these monolayers depends sensitively on both interface composition and on the existence

and polarity of any liquid overlayers on the interface.

22

Experimental

Chemicals. l-Pyrenebutyric acid (97%), l-pyrenemethylarnine hydrochloride
(95%), octadecylamine (97%), 2-mercaptoethylamine (95%), 11-mercaptoundecanoic
acid (95%), 10-hydroxydecanoic acid, octadecylmercaptan (98%), l-hexadecylamine
(98%), N,N’-dicyclohexylcarbodiimide (DCC) (99%), adipoyl chloride (299%), 4-
methylmorpholine (299.5%), triethylamine (299%), perchloric acid (70%),
hexaammineruthenium (III) chloride (98%), potassium ferrocyanide (99%), lithium
perchlorate (99.99%), cyclohexane (99%), l-pentanol (299%), acetonitrile (anhydrous)
were obtained from Sigma-Aldrich. Dichloromethane and chloroform were obtained
from POCH and Mallinckrodt Chemicals. Aqueous solutions were prepared from Milli-

Q water.

Preparation of pyrene on short tether for gold substrate. Pyrene bound to gold
using a “short tether” (cysteamide of pyrenebutyric acid, designated P7, where the
number 7 indicates the number of atoms between the attaching head group and the pyrene

moiety) was synthesized according to a standard procedure for peptide synthesis in

. l6 . . . . . .
solution. Pyrenebutyric acrd and cysteamine were dissolved in dichloromethane and

cooled to 0°C. After 1 h of stirring, DCC in dichloromethane was added and the reaction
proceeded at room temperature for 12 hours. The ﬁnal product was puriﬁed by silica-gel
column chromatography (Silica gel 60 from Merck, particle size 0.040 — 0.063 mm) and

analyzed by IR spectroscopy.

Gold substrate preparation. Gold ball electrodes were cleaned in piranha solution

(3:1 concentrated H2804: 30% H202). Caution! Piranha solution reacts violently with

23

organic matter and should be handled with extreme care! The cleaned electrodes were

rinsed with water and polarized cyclically (scan rate 100 mV/s) in the -250 to +1650 mV

potential range in 1 M HClO4 until reproducible voltammograms were obtained. Next,

electrodes were removed from the cell, washed with distilled water, and dried with a

stream of nitrogen. The area of the gold ball electrode was 0.135 cmz, which was

determined from the cyclic voltammetry (CV) of a K4[Fe(CN)6] solution. 1 7

Silica substrate preparation. Silica slides were cleaned by immersion in piranha

solution for ca. 20 min, then rinsed with water and dried in a stream of nitrogen.

1T0 substrate preparation. Indium-doped tin oxide (ITO) ﬁlms deposited on

silica substrates were cleaned by rinsing with water, ethanol, and acetone, consecutively.

Monolayer preparation. To create a monolayer of P7 on gold, the substrate was
exposed to a solution of P7 in dichloromethane for 12 h. The SAM-covered gold
electrode was then rinsed with dichloromethane and dried in a stream of nitrogen, then
washed with distilled water and dried in a stream of nitrogen. The monolayer of the
nominally buried P7 was obtained by immersing a gold electrode into a solution of P7
and octadecyl mercaptan (ODM) (1:1) in dichloromethane for 12 h. For deposition of a
monolayer of pyrene tethered a longer distance from the gold substrate, the substrate was
ﬁrst exposed to solution of 1 l-mercaptoundecanoic acid (MUA) in dichloromethane for
12 h. The resulting SAM-covered gold substrate was rinsed with dichloromethane, dried
in a stream of nitrogen and immersed in a solution of pyrenemethylamine and DCC (1:1)

in dichloromethane for 12 h, to form a monolayer of mercaptoundecanoic acid

24

pyrenemethylamide (P13). To synthesize a monolayer of “buried” P13, a substrate
covered with a monolayer of MUA was immersed in a dichloromethane solution

containing pyrenemethylamine, hexadecylamine and DCC (1 :1 :1) for 12 h.

Monolayers on ITO substrates were synthesized by reaction of the substrate with
adipoyl chloride (0.3 mL) in dry acetonitrile (10 mL), using 4-methylmorpholine (0.3
mL) as a Lewis base, under reduced pressure for 1 h. The reacted substrates were
removed ﬁom solution, rinsed with dry acetonitrile and ethyl acetate, and dried under a
stream of dry nitrogen. We synthesized monolayers of tethered pyrene either as the only
constituent or as a co-constituent with a long-chain aliphatic moiety. These monolayers
were obtained by exposing the adipoyl chloride covered substrate either to a 2 mM
solution of pyrenemethylamine in dichloromethane, producing a single-component
monolayer of the pyrenemethylamide of adipic acid, P8. For the corresponding
monolayer where P8 is “buried”, a dichloromethane solution of pyrenemethylamine and
hexadecylamine (1:1) is reacted under reduced pressure for 0.5 h. Following this
reaction, the substrate was removed from solution, rinsed with dichloromethane and ethyl
acetate, and dried under a stream of dry nitrogen. For pyrene tethered further from the

substrate, the adipoyl chloride covered substrate was ﬁrst reacted with a 2 mM solution

of 10-hydroxydecanoic acid in dichloromethane. Following the addition of this C10

chain, the substrate was removed from the reaction vessel, rinsed with dichloromethane
and ethyl acetate, and dried under a stream of dry nitrogen. The substrate thus covered
with adipic acid mono-(10-carboxy)-decyl ester was then exposed either to
pyrenemethylamine and DCC (1:1) (to form pyrenemethylamide-10-decyl ester, P19) in

dichloromethane or to pyrenemethylamine, hexadecylamine, and DCC (1 :1 :1) in

25

dichloromethane under reduced pressure for 0.5 h. The reacted substrate was removed
from solution, rinsed with dichloromethane and ethyl acetate, and dried under a stream of

dry nitrogen.

Electrochemistry. Electrochemical measurements were made with a computer-
controlled electrochemical workstation (CH Instruments Model 604B and Model 650),
using a three-electrode cell with ,a Pt wire counter electrode. All potentials are quoted

versus a Ag 1 AgCl | 3 M KCl (-52 mV vs. SCE) reference electrode. Experiments were

carried out in aqueous 0.01 M, 0.05 M, 0.1 M, 0.5 M and 1 M HClO4 (keeping the ionic
strength of 1 M constant by the addition of the appropriate amount of LiClO4), in
aqueous 1 mM K4[Fe(CN)6] in 0.5 M LiClO4, and in aqueous 1 mM [Ru(NH3)6]Cl3 in

0.1 M LiClO4.
Langmuir trough experiments. Langmuir measurements were made with a
commercial trough (N ima Technology). Pyrenebutyric acid and octadecylamine

solutions (2 mg/mL) were used as surfactants. Several solutions with varying

pyrenebutyric acid to octadecylamine molar ratios were spread on the water surface.

26

Results and Discussion

Pyrene derivatives exposed to electrolyte solution. The purpose of this work is to
demonstrate and characterize a group of SAMs modiﬁed with pyrene derivatives, with an
eye toward the application of these monolayer species as spectroscopic and
electrochemical probes of their immediate environment. We have investigated pyrene
derivatives bound covalently to gold and ITO substrates by CV. We show in Figure 2.1
the cyclic voltammograms of our pyrene derivatives bound covalently to gold and ITO.
In the ﬁrst scan for each voltarnmogram we observe an irreversible peak with a maximum
above 900 mV. A pair of new, broad redox peaks appears in subsequent scans between
200 and 300 mV. We assign these features to two pairs of poorly resolved redox peaks,
centered near 210 mV and 320 mV. Previous work has shown that these peaks are

associated with the several possible diol/dione species that are the stable oxidation

products of the pyrene ring system. 18

On the basis of the analysis of electrochemical reactions of surface-bound pyrene

on gold and ITO surfaces reported in the literature,19 the voltammograms presented in

Figure 2.2 and Figure 2.3 can be assigned to 1,6-pyrenedione/1,6-dihydroxypyrene (a

pair of CV peaks centered around 320 mV), and 1,8-pyrenedione/1,8-dihydroxypyrene,

which is shifted to more negative values by ca. 100 mV.18

The electrode coverage with various pyrene derivatives was evaluated with use of

CV. The amount of pyrene derivatives on the electrode surface was calculated from the

charge under the anodic peak from the voltammograms recorded in aqueous l M HCIO4

27

21

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

F.4— N§4~

g 3: E3;

.9 2: .2»?

§ 1'. E]:

'U “U

:25 0» go_

5'10 700 ‘400 ‘600‘80070‘00‘ 5'10 ‘200 ‘400 7008001000

potential [mV] potential [mV]

 

 

 

 

 

 

 

 

 

 

   

 

 

 

 

 

 

 

 

“E “E

0 12? lst 0 $12: 15‘

E 9. -------------- 2nd 3 9* -------------- 2nd

Z." 6' ,,,,,,,,,,,,,,,,,,,,,, 3rd 2" 6' ....................... 3rd

5 3- a 3-

E l ,3 E O ’ .fwmb‘ﬂ...

8 0 E W

_ - . - . - . - . . . _ _3 - . - . - . - . - . 2
5) 30 200 400 600 800 1000 o 0 200 400 600 800 1000
potential [mV] potential [mV]

Figure 2.1 Cyclic voltammograms of: (a) P7 on gold in 0.1 M HClO4, sweep rate
2mV/s; (b) P13 on gold in 0.1 M HClO4, sweep rate 5 mV/s; (c) P8 on ITO in 1 M
HClO4, sweep rate 20 mV/s; ((1) P19 on ITO in l M HClO4, sweep rate 20 mV/s.

28

electrolyte solution (Figure 2.2 and Figure 2.3) at low sweep rate (as speciﬁed in ﬁgure

captions). Integration of this peak yielded a surface coverage (assuming a 2 electron/2

proton process”) of ca. 4.4 x 10H mol/cm2 for P7 and 1.7 x 10‘” mol/cm2 for P13.
These same measurements on an ITO substrate yield a surface coverage of 1.2 x 10‘] l

mol/cm2 for P8 and 8.4 x 10.12 mol/cm2 for P19. To assess the maximum theoretical

coverage for our pyrene derivatives on a given substrate, one has to evaluate the
minimum area that can be occupied by a single pyrene ring system with its hydrocarbon
tail outstretched from the surface in an all-trans conﬁguration. For this purpose we have
used the Langmuir technique. Unfortunately, because of the ﬁnite solubility of
pyrenebutyric acid in water, we could not produce a pressure/area isotherm over the
required phase transition range for this compound. Therefore, we have investigated the
mixed pyrenebutyric acid/ octadecylamine system with various pyrenebutyric acid mole
fractions on a water subphase (Figure 2.4a). It was possible to obtain the pressure-area
isotherms of this binary system for up to a 1:1 ratio of the two constituents. By
assumption of additive behavior for such a binary mixture on the interface, a plot of the
area per molecule as a function of the mole fraction of pyrenebutyric acid in

octadecylamine allows us to estimate the area occupied by a single pyrenebutyric acid

molecule to be 312 A2 (Figure 2,4b). This value is comparable to that of sterols in a

monolayer: 37 Az/molecule for cholesterol.20 The value of 3 1 .2 AZ/molecule is entirely

reasonable on geometric grounds. For a rigid molecule, one can estimate an area of ca.

2 . . . .
26 A /molecule based on a molecular mechanics calculation, not taking into account any

29

2
uA/cm
i—I r—d N N U.)
u: o m o m o
I I I I '

 

current density [
«3;. o

 

 

L.
o

l I l I I I L I I

0 200 400 600 800 1000
potential [mV]

 

2
uA/m
r—si—INNU.)
OMOMO
'I'I'I 1'1
0"

 

current density [
LII

Clio

 

 

L.
o

 

l l l I J I I 1 n

200 400 . 600 800 1000
potential [mV]

C

Figure 2.2 Cyclic voltammograms of: (a) P7 and (b) P13 on gold in 1M HClO4. Sweep
rate 20 mV/s. Charge under the anodic peak is 2.105 uC for (a) and 3.106 uC for (b).

30

 

 

6'

m'—' a
E

o

24"

.3.

Z?

82'

0)

"U .

g A
50"

o

 

I
[\J

0 200 400 . 600 800
potential [mV]

V

l

1000

C\
I

A
7

O

 

current density [uA/cmz]
N

I
N

 

l l I l 1 I I

0 200 400 600 800 1000
potential [mV]

Figure 2.3 Cyclic voltammograms of: (a) P8 and (b) P19 on ITO in 1M HClO4. Sweep
rate 20 mV/s. Charge under the anodic peak is 1.048 uC for (a) and 1.345 uC for (b).

31

21

    

 

 

 

 

50 - ‘E 32
2 30
40 " “1:1. 1‘ 2 28
,_. 1'2. E 26
E ‘Ii. ‘5 0 ‘—
2 30_ . ‘i/SOA’PY 8.24
E ‘ 3 22
as: 20_ {VT/40% Py 0.0 0.2 0.4 0.6 0.8 1.0
w 20% P the mole fraction ofP
g . y 30% Py y
it to - 0% Py "
0 _ "Fiﬁ...“
I A l l L- 41 L I
10 20 30 40 50 60

area per molecule [A2]

Figure 2.4 Pressure-area isoterrns for pyrenebutyric acid and octadecylamine system.
Inset: Area per molecule as a function of mole fraction of pyrenebutyric acid in mixture
of octadecylamine with different concentrations of pyrenebutyric acid.

32

motional freedom or alternative conformers of the pyrenebutyric acid molecule. Taking

2 . . . .
the value of 31.2 A /molecule for our pyrene derivatives, we estimate the maxrmum

-10 2 . . .
surface coverage of pyrene on the electrode surface to be 5.3 x 10 mol/cm , indicating

that the surface coverage we obtain is well below that of a full monolayer.

We used cyclic voltammetry to assess the organization of pyrene-containing
monolayers on Au and ITO substrates. Typically, well organized, hydrophobic layers
block electron transfer from hydrophilic redox probes present in the supporting
electrolyte solution to the conductive substrate. For these interfaces, we have used
hexaammineruthenium (III) chloride (1 mM) as the redox-active probe, which produces a
quasi-reversible CV curve (not shown). Our data point to poorly organized monolayers,
consistent with previous literature reports showing that when cysteamine molecules are

used as anchoring moieties to gold (P7), the resulting monolayer adopts a predominantly

. 21-23 . .
gauche conformation. For such a conformation, one would expect a monolayer wrth

numerous access points to the substrate. For P13, where mercaptoundecanoic acid is the
anchoring species, the amide group, formed by reaction of MUA with
pyrenemethylamine to form P13, should be solvated, also resulting in a poorly organized
monolayer. Thus, the hydrophilic hexaammineruthenium (III) chloride redox probe,
regardless of the length of a tether anchoring the pyrene ring system to the substrate, can

gain access to the electrode surface, resulting in quasi-reversible voltammetric behavior.

By consideration of the sub-monolayer coverage and relative ease of diffusion for
the hexaammineruthenium (III) cation through these layers, as well as the fact that

protons participate in the redox reactions (Figure 2.5), we should observe a pH

33

 

 

 

 

 

 

 

 

 

 

 

OH
-. ”0 33+ 0
H20 T
O NH O NH
i 8
H0 is
;e_.
H20
NH O NH
8 S
f f HO o
‘ -6 H0 :33, 0
H20 :
o o
3NH sNH
i i

 

Figure 2.5 Schematic of the redox reactions characteristic of the pyrene moiety (for

clarity this drawing shows only the P7 molecule). The structures shown are not an
exhaustive list of the reaction products, but are representative of the dominant species

34

dependence in the plots of the potential of the midpoint, Em. Between the maximum of

the oxidation peak, Ep , and the potential of the maximum of the reduction peak, EpC, a

gradual decrease in pH should lead to a positive shift of the midpoint potential by about
60 mV per unit pH if the ratio of protons to electrons is close to one (2H+, 2c.) in the
reaction scheme (Figure 2.5) that leads to either 1,6-pyrenedione/ 1 ,6-dihydroxypyrene or

1 ,8-pyrenedione/ 1 ,8-dihydroxypyrene.

Figure 2.6 shows the shifts in peak potentials to more positive values with
decreasing solution pH. On the basis of CVs for pyrene derivatives obtained in different

perchloric acid concentrations we can determine the dependence of the midpoint potential

between the anodic and cathodic peaks, Em, as a function of pH (Figure 2.7). The
experimental data show that the slope of plots of the midpoint potential, Em, is ca. -45

mV/pH unit for both adlayers on gold, and for the ITO substrate, the slope of Em vs. pH

is ca. -53 mV/pH unit. There are several interesting features contained in these data. The

ﬁrst is that the pyrene derivatives anchored to either Au or ITO appear not to undergo a

+ - . . . .
full 2H /2e redox reaction. Rather, the calculated ratio of protons to electrons involved

in the unit redox reaction is close to 0.8. Even though this ﬁnding is similar to data

reported for the ubiquinone/ubiquinol (UQ/UQHZ) redox couple conﬁned to the mercury

surface by a phospholipid monolayer,24—26 where deviations from the expected 60

mV/pH unit were observed for UQ, we believe that the observed deviation from a 1:1

35

 

 

 

 

 

 

 

 

 

 

   

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

   

 

 

 

 

— 1M —— 1M
---------- 0.5M 0.5M
N2— ................ 0.1M N'— 0.1M
g —--—- 0.05M g —---- 0.05M
2 2 6 - b -—--—~— 0.01 M
.3 .3.
3‘ Z?
.5 .5).
§ §
3 0 200 400 600 800 g o 200 400 600 800
potential [mV] potential [mV]
__ 1M —- 1M
N... ---------- 0.5M _. """""" 05M
_ ................ 0.1M N """""""" 0.1M
E 2 c —-- 0.05M g -——-—-— 0.05M
2 --— 0.01 M < -——- 0.01 M
.3. 3.
.2: — b “‘
m "(7)
5’ i?
E E
Q _2 . . - . - i A . - _ . . . i L
g 0 200 400 600 800 if: 20 200 400 600 800
Q 0

potential [mV] potential [mV]

Figure 2.6 Cyclic voltammograms for the pyrene-containing adlayers: (a) P7 on gold;
(b) P13 on gold; (c) P8 on ITO; (d) P19 on ITO, all in 1M, 0.5M, 0.1M, 0.05M and

0.01M aqueous HClO4. The ionic strength is maintained at 1 M for all solutions by
aqueous LiClO4. The sweep rate was 20 mV/s for all CVs.

36

o.)
O
O

m

N
\l
O
0
W
O
O
0

P7 0“ gold P13 on gold

N
p—t
O
N
.5
O

 

 

(13p A+E pC)/2 [mV]
N
.b
O

(Ep A+13 pC)/2 [mV]
N
\l
O

 

 

010‘ 0:5 A 110‘ 15 A 23' 0:0 ‘ 015 ‘ 1f0‘ 1:5 ‘ 20‘
pH pH

330

U)
D.)
O
o

O

D»)

O

O

0

P8 on ITO I P19 on ITO g
.- slope =-43.8 mV/pH . . slope=-37.9 mV/pH ,

NN
A
O

 

 

(Ep A+1»: pC)/2 [mV]
\1
O

(13p A+13 pcyz [mV]
N
\l
O

N
b
O

 

 

0:0 ‘ 0:5 ‘ 110‘ 1:5 ‘ 2:0 ' 0:0 ‘ 015 ‘ 1:0 ‘ 115 ‘ 20’
PH pH

Figure 2.7 Dependence of the midpoint potential, Em, on the pH of perchloric acid

aqueous solutions for: (a) P7 on gold, slope = -39.5 mV/pH, (b) P13 on gold, slope = -
36.3 mV/pH, (c) P8 on ITO, slope = -43.8 mV/pH, and ((1) P19 on ITO, slope = -37.9
mV/pH. For all scans the sweep rate was 20 mV/s.

37

H+:e- ratio is mainly due to uncertainty in the experimental data (the 1M HClO4 solution,

in particular, is of concern).

In the following section, we consider the rate constants for the redox processes that are

responsible for the observed behavior. Our previous work has shown that pyrene redox

. . . . + - . 18
chemlstry lS explained well 1n the context of a 2H /2e reaction, and we evaluate the

rate constants based on this knowledge. We caution, however, that there are other

. . . . + - .
posmble explanat10ns for our data. It IS also possrble for a 1H /1e reactlon to account

for our ﬁndings, particularly for reactants in a conﬁned space, or for more complex

schemes involving disproportionation and/or further reduction to be operative (Figure

2.7) These possibilities have been pointed out previously for the UQ/UQH2 redox pair.25

One should also consider that the chemical reaction, involving a protonation step during

Py(O)2 reduction and a deprotonation in Py(OH)2 oxidation, is usually the rate

determining step, with electron transfer being in quasi-equilibrium.

We have used AC voltammetry to study the redox kinetics of several pyrene
derivatives bound as monolayers to Au and ITO. This technique has proven to be

particularly useful for the investigation of redox reactions in layered assemblies on

27-30 . . . . .
electrodes, because a varlety of 1nformation can be obtained from such data. First,

for a given reversible reaction, the peak position on the voltammogram gives information
on the formal potential of the surface redox reaction. As shown in Figure 2.8,_the peak

potential values are lower by ca. 58 mV than the formal potential values obtained during

the CV experiments by extrapolating the slope of the midpoint potential, Em vs. pH to

38

pH=O. Since the alternating current (AC) experiments were carried out in 0.1 M HClO4

(pH=1), this smaller value is due to the dependency of the overall process on proton
concentration. It is also important to note that the difference in peak position between

P13 and P7 is slightly higher than that obtained from direct current CV, yielding the

A(AGO) value of ca. 7 kJ/mol.

In these AC voltammograms (Figure 2.8) we observe a decrease in the peak-to-
background current ratio with increase in frequency. Such changes in the magnitude of
the AC peak current with the changing frequency of the AC voltage provide information
about the rate constant of the reaction. These data are a series of AC voltammograms
taken at three selected frequencies to illustrate the behavior of the pyrene redox reaction
as a function of tether length and whether the tether length inﬂuences the redox kinetics.
Creager’s group has shown that it is possible to assess the charge-transfer rate constants

from such voltammograms by using a relatively simple Randles equivalent circuit for the

. . . 27-29
impedance of an electrode coated With a redox-active ﬁlm. We note that, for the

systems we consider here, it can be the chemical step (e.g., protonation/deprotonation)

that is rate determining, so the quantitative evaluation of the charge-transfer rate constant,

k6,, should be viewed with some caution.

. . . 27-29
With these caveats 1n mmd, we have used Creager’s methodology and

present the AC data from Figure 2.8 in form of the peak current to background current

ratio, ipeak/ ibackground vs. log(AC frequency) for a series of voltammograms recorded

over a range of frequencies (Figure 2.9). Plotting the AC voltammetry data in the manner

39

 

 

 

 

 

 

 

 

2“ 2‘0 11—IZ 2'1 0.4r
l
31.5. 3 0.31 1H2
H H i
C C.‘
g 1.0 at; 0.2-
o ()5. o 0.1»
U L)
{10.0“ . . . .1 <0.04--+1-. .
O 100 200 300 400 500 600 0 100 200 300 400 500 600
potential [mV] potential [mV]
.— lOO' .— 20-
< IOOHz <
3 30* 3.15. 100112
E 60» E
5 40’ g 10’ A
o 3 5.
2 20' 2
0 <5 150 200 360 400 500 600 O 0 100 200 300 400 500 600
potential [mV] potential [mV]
2? 400 '3 100'
a ' IOOOHZ a. 80l 1000117.
E 300' g 60- A
EZOO- E, 40-
0
81001 O 20»
< 0 . < 0

 

 

 

 

0 100700 300 400 500 600 0T100T20OT300 400500 600
potential [mV] potential [mV]

Figure 2.8 AC voltammograms for P7 (left column) and P13 (right column) on gold in
0.1M HClO4. AC frequencies are indicated in each pane.

40

:7 2.0'
E 1.5
1.0

peak / background
I" Z"
O N

i

 

—-N
ooo

p—a
ON

I

l
p—t
A

I

 

 

_ -1
kct—IZOO s \\

I 8.1). I i

141 A A l J A I l A A L l A 4*;

0 1 2 3 21”
log (frequency [Hz])

 

log (frequency [Hz])

Figure 2.9 ipeak/ ibackground vs. log(AC frequency) plots of monolayers of P7 (top
graph) and P13 (bottom graph) on gold. The solid lines represent ﬁts to the data (black
squares) using a Randles equivalent circuit.27’28

41

shown in Figure 2.9 reveals three distinct regions of the reaction frequency dependence.
At low frequencies, the current ratio attains a plateau value that is indicative of the
amount of pyrene on the electrode surface; at high frequencies this ratio becomes unity
(the AC peak disappears), because the AC sweep rate becomes signiﬁcantly faster than
the reaction kinetics. The intermediate region, where the current ratio is strongly
frequency dependent and exhibits a breakpoint frequency, is related to the standard
charge transfer rate constant for the redox-active species. By ﬁtting the experimental
data (black squares in Figure 2.9) to the Randles model, we extract the rate constants for
the monolayers comprised of P7 and P13. For the Au electrodes the recovered rate

constants depend strongly on the length of the alkane chain anchoring the pyrene moiety

to the electrode surface (kc, = 1200 s.1 for P7 and 25 s.1 for P13).

It is instructive to consider qualitatively the differences in rate constant values for

P7 and P13 in terms of through-bond tunneling that should depend exponentially on the
distance and/or the number of C-C bonds in the anchoring chain. By assumption of

typical bond lengths (1.79 A for the C-S bond, 1.27 A for each C-C bond and 2.6 A for

. 28,3] . . .
the amnde group), we obtam the tunnelmg constant, 3 from the expenmental kc,

values:

ln [M] = ,B(d"- d") (2.1)
kct H

where kc, and d ’ are rate constant and chain length of P7 and k.“ and d .. are the rate

constant and chain length of P13. The use of a single tunneling constant, ,6, for different

length anchoring chains warrants some explanation. The tether molecules, aside from

42

their length, vary in the position of the amide group and the pyrene ring system relative to

the substrate, but as was suggested in the literature, the contribution of an amide group to

. . . . . . 32,33
the electronic coupling 15 substantially Similar to that of two methylene groups.

Because of the structural similarities of P7 and P13, we assumed the same tunneling
constant for both molecules bound to the electrode surface. The structure dependent
tunneling constant is reﬂective of the decay of electronic coupling between the redox
center and the electrode. Therefore, it should also reﬂect the nature of interactions
between the tethered pyrene moieties and their immediate environment, which will
include neighboring molecules within the monolayer and solvent molecules that have

penetrated the interface. For P7 and P13 bound to gold, Equation 2.1 yields a value of ,6

= 0.44/A, approximately half the literature value of 0.85/A for a polymethylene chain.34

A broad range of ,6 values have been reported in the literature, ranging from the lowest

values of 0.4-0.5 for conjugated bridges forming the tether moleculeszg’35 and several

30,35-37

proteins,36 to 0.8-1.0 for saturated alkanethiols and hydrocarbons. However, it is

important to note here that, for a monolayer such as P13 on gold, the chemical system is

much more complex than the simple Randles approximation and it is not appropriate to

. . . . 2
ﬁt the experimental data usmg a Single rate constant (cf. Figure 2.9 bottom, R =O.95). In
our opinion, our results suggest some disorder and dynamical motions in these
monolayers, leading to the presence of multiple populations of redox sites with a

corresponding distribution of rate constants (and therefore the ﬂ values), likely due to a

non-uniform orientational and spatial distribution of pyrene moieties within the

43

monolayer. On that basis, the above discussion should be considered qualitative, with the

central point being the structural complexity of the interface(s) we have reported here.

We have acquired the analogous AC voltammetry data and performed the same analysis
for P8 and P19 on ITO. The data in Figure 2.10 show that the AC peak current

diminishes more rapidly with increasing frequency and the ﬁtting procedure yields much

smaller kc, values that are barely dependent on the anchoring chain length (kc,=0.2 s—I for

P8 and 0.1 s.1 for P19, Figure 2.11). This result appears to be counterintuituve, in light

of the data for gold substrates. We assume that the ester groups present on the tether

chains reduce the rate of electron transfer in a manner analogous to that seen for ether

38 . . . . . . .
groups, making through-chain electron transfer inefﬁc1ent. If this is in fact the case,

the contribution of intermolecular interactions to the electron transfer process may
become signiﬁcant. Such a contribution could be the result of a highly disorganized
interfacial structure, where the tethers of the pyrene derivatives as well as the pyrene
rings form disorganized, aggregated structures, thereby maintaining the pyrene ring
system at roughly the same average distance from the electrode surface for both P8 and

P19.

As was the case for the data taken on a gold substrate, the AC peak position (in 1

M HClO4) denotes the formal potential of these derivatives attached to ITO. We extract

from the data a value of +328 mV, in very good agreement with the value of +324 mV
obtained for a similar pyrenemethylamine derivative tethered to a gold substrate. This

ﬁnding implies that the bond between the adlayer and the substrate plays a negligible role

44

 

 

 

 

 

 

0 100 2003000400 500 600‘

potential [m V]

 

 

 

 

 

'__' 1.0» 1.0-
:1, 0.8» 1H2 ‘5, 0.8 1H2

~ 0-6' '3' 0-6’ W
E C .

“g 0.4- g 0.4

8 0.2- 8 0.2'

0 mo 200 300 400 500 600 ' O 100 200 300 400 500 600
potential [mV] potential [mV]

r-H 8 1— 8_

3 6i 10Hz 3 6: 10112

E . ‘s':’ _ \
g 4 \\ 452 4

8 2- o 2-
00....... 302..-...2
< 0 100 200 300 400 500 600 0 100 200 300 400 500 600

potential [mV] potential [mV]

_ 40- _ 40-
g 30. 50Hz E. 30- 50112

‘5' 20» \\ ﬁg 20- \
t: t: ‘

3 10- 3 10-
SE 0 3: 0

 

0 100 200300 400 500 600
potential [mV]

Figure 2.10 AC voltammograms for P8 (left column) and P19 (right column) on ITO in l
M HClO4.

45

y—i
\]

1.0-
:15? km: 0.2g"
301.4
g 13 \ 1 SD

 

0.0 0.5 1.0 1.5 2.0
log (frequency [Hz])

 

'2 1-4 ' km: 0.l s'1

I SD.

115
i.—i|
N

 

 

 

0:0 ' 0:5 ' 1:0 ‘ 1:5 ‘ 20
log (frequency [Hz])

Figure 2.11 ipeak / ibackground vs. log(AC frequency) plots of monolayers of P8 (top

graph, kc, = 0.2 5-1) and P19 (bottom graph, kc, = 0.1 s") on ITO. The lines represent ﬁts
to the data (black squares) using a Randles equivalent circuit.

46

in the electrochemical responses of these interfaces. Apparently it is only the substituents
in closest proximity to the ring system that can inﬂuence the redox properties of the
pyrene moiety. The aliphatic chain, which mediates the spacing between the pyrene and

the substrate, does not affect the redox properties of the ring system.

Pyrene derivatives nominally buried within the monolayer. The results we have
considered to this point were obtained for pyrene derivatives, both on gold and ITO,
which formed a sub-monolayer of molecules that face the aqueous solution. This system
was characterized by substantial motional and structural freedom, as is seen in their
electrochemical responses. We consider next our results for a family of interfaces where
the pyrene ring system is buried within a mixed monolayer comprised of pyrene-
containing species and aliphatic species. For all of the systems we consider, the mixed
monolayers contained the same pyrene derivatives as were studied above, but they are
coadsorbed with “diluent” molecules of longer hydrocarbon chains. For such systems we
expect to observe changes in the electrochemical response of the pyrene moiety, related
to the low dielectric environment provided by the diluent molecules, even though the
cyclic voltammetry in the presence of the hydrophilic hexaammineruthenium (III)
chloride redox probe yielded quasi-reversible voltammograms (not shown), similar to
those seen for the corresponding one-component monolayers. Octadecanethiol was the
diluent for monolayers containing P7 and the hexadecylamide of mercaptoundecanoic
acid was the diluent for monolayers containing P13 on gold substrates.
Adipoylhexadecylamide was used as the diluent for monolayers containing P8, and the
mono-(9-carboxy)-nonyl ester of adipic acid hexadecylamide was the diluent for

monolayers containing P19 on ITO. Apparently, the coadsorbed diluent molecules do not

47

improve the organization and passivation properties of the mixed monolayers compared
to the one-component interfaces, and the electrochemical probe can diffuse to the

electrode surface.

Figure 2.12 shows the CVs for P7 and P13 with coadsorbed diluents on gold and
P8 and P19 with coadsorbed diluents on ITO. Two features of these voltammograms
deserve closer inspection. The ﬁrst is that the charge under the CV peaks is smaller than
that for the corresponding one-component monolayers, a ﬁnding that we attribute to the

coadsorption of two species that are competing for the same surface sites. For gold

electrodes we recover a surface coverage of 3 x 10.ll mol/cm2 for P7 diluted with
octadecanethiol and 4 x 10.12 mol/cm2 for P13 diluted with mercaptoundecanoic acid
hexadecylamide. For the ITO substrates, we obtain a coverage of 8.4 x 10.12 mol/cm2

for P8 diluted with adipoylhexadecylamide and 5.5 x 10'12 moI/cmz for P19 diluted with

the mono-(9-carboxy)-nonyl ester of adipic acid hexadecylamide.

The other signiﬁcant feature contained in these data is the splitting of the
reduction peak and a “hump” on the oxidation wave of the CV curve for P13 conﬁned
within the two-component monolayer. As noted above, these two peaks can be assigned
to 1,6-pyrenedione/1,6-dihydroxypyrene (a pair of CV peaks centered around 320 mV),
and a contribution from the 1,8-pyrenedione/l ,8-dihydroxypyrene redox couple at ca.
210mV. This behavior is better resolved by AC voltammetry (Figure 2.13, right column,
1 Hz) and can be related to the coexistence of two derivatives in energetically well
deﬁned populations with relatively little interaction between the redox centers. These

features were not seen for the corresponding one-component monolayers, suggesting that

48

 

 

 

 

 

 

 

 

 

 

   

 

 

 

 

 

 

  
 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

— IM
--------- 0.5M
N; 3- N; 3_ .............. 0.IM
i 2 O 2_ -..._..0.05M
i_.i l E l”
g o- a 0»,
.34 — g -1
E '2’ E '2“
g -3 . g . i 1 t; _3L 1 - i A i A 2
o 0 200 400 600 800 a 0 200 400 600 300
potential [mV] potential [mV]
— lM — IM
--------- 0.5M 0.5M
r—a I.5r ....__.......0.]M C Nv—q I.5 .............. 0.1M d
E I O: ———-0.05M I 0 ----0.05M '
o - __ o . __
0.01M 0.0I
2,, 0.5» i 0.5 _
§‘ 0.0» ;_ 4? 0.0 __
8 -0.5 8 -0.5 2......
3 -10- 3 10
La) 1.5 i . _._ 1 r i g 1.5
g ' 0 200 400 600 800 g - ' 0 200 400 600 800
potential [mV] potential [mV]

Figure 2.12 CVs for mixed monolayers: (a) P7 + diluent on gold; (b) P13 + diluent on
gold; (c) P8 + diluent on ITO; (d) P19 + diluent on ITO in 1, 0.5, 0.1, 0.05 and 0.01 M

aqueous HClO4. The ionic strength for all measurements was maintained at 1 M by

aqueous LiClO4. Sweep rate 20 mV/s.

49

 

 

 

 

 

 

 

 

 

 

 

 

.2. 0.4- 0.12- 1H2
.3. 0.3. 1H2 i 0.09. \/‘/\/
g 0.21 E 0.06.
. O)
o 0.1. S 0.03.
U 4 0 l
<"30.0 -. .+.4+.. 00. WCMC
0 IOO 200 300 400 500 600 < 0 I00 200 300 400 500 600
potential [mV] potential [mV]
'5' 100H2 ‘
'2 . - z 64 IOOHZ
.2: 101 3. N
E 1 .. 4‘
C
Q) d) ‘
5
s 5 2.
O C) .
U . . . .-.ﬁ- .. Uo. .-.-.-.---.-
< 0 IOO 200 300 400 500 600 < 0 100 200 300 400 500 600
potential [mV] potential [mV]
E 803 1000Hz .2 60* ”)0on
. 1 WM
_. 60 k *—' 40.
20+ 5 ‘
O I O
00.....-......U ,,,r-rz,,
< 0 IOO 200 300 400 500 600 < 0 100 200 300 400 500 600
potential [mV] potential [mV]

Figure 2.13 AC voltammograms for P7 + diluent (left column) and for P13 + diluent
(right column) on gold in 0.1M HClO4.

50

for the two-component monolayer the different pyrenedione/diol isomers are stabilized
by the presence of diluent molecules that separate the redox centers. For the two-
component monolayer, the surface concentration of the pyrene derivative is almost 10
times smaller than that of the single-component monolayer, reducing the probability of

interactions between different redox centers.

The results presented below are dominated by the 1,6-pyrenedione/1,6-
dihydroxypyrene redox couple, which is seen as a pair of CV peaks centered around 320
mV. The 1,8-pyrenedione/l ,8-dihydroxypyrene redox couple is seen only on gold for
P13, with peaks centered around 130 mV. Figure 2.14 shows the midpoint potential, Em,

vs. pH for the two-component monolayers on gold and ITO substrates. Similar to the

data for the one-component monolayers, the slope of Em vs. pH is ca -46 mV/pH unit for

gold, regardless of the anchoring chain length and ca. -49 mV/pH unit for ITO. We
consider that these data also indicate a 1:1 proton-to-electron ratio to within the
experimental uncertainty in the data, as discussed above for the case of single-component
monolayers. These results show that, for our mixed two-component monolayers, the
presence of diluent molecules has a negligible inﬂuence on the ratio of protons to
electrons involved in the pyrene ring system redox reaction. The extrapolation of the
slope in Figure 2.14 to pH=0 gives the values of the formal potentials for the different
pyrene derivatives in the mixed two-component monolayers. The values of +30] mV for
the pyrenebutyric acid derivatives and +339 mV for the pyrenemethylamine derivatives
are identical to those obtained for the corresponding one-component monolayers,

underscoring the point that the presence of diluent molecules in the monolayer has no

51

 

 

 

 

.>_ 300— a

g, , .

g 270-

113$ 240- 0

+< . P7 + diluent on gold

an. 210,519!” = -4321 3.9 mV/pH ‘ ’ .

0.0 T05 1.0 1.5 2.0

pH

E. 330’ g C

Q 300» °

1:3. 270:1)3 +diluent on ITO ..

m3. 240. slope = -46.8 1 6.7 mV/pH

v 0.0 To} 4 1:0 ‘ 1:5 A 20 0
pH

(1:p A+13pC)/2 [mV]

(13p A+1~:pC)/2 [mV]

300 ~

270 ~

240

 

 

 

 

I_ P13 + diluent on gold _
slope = -48.3 :t 5.9 mV/pH 0-.
0.0 0.5 1.0 1.5 2.0

pH

.. P19 + diluent on ITO

- slope = -50.9 at 15.8 mV/pH ’
0.0 0.5 1.0 1.5 2.0

pH

Figure 2.14 Dependence of the midpoint potential, Em, on pH in 1 M HCIO4 aqueous

solutions for (a) P7, slope = -43.2 i 3.9 mV/pH, and (b) P13, slope = -48.3 i 5.9 mV/pH,
in two-component monolayer structures, on gold and for (c) P8,

slope = -46.8 i: 6.7 mV/pH and ((1) P19, slope = -50.9 i 15.8 mV/pH, on ITO. Sweep

rate 20 mV/s for all scans.

52

effect on the energetics of redox process. These two-component monolayers are not

sufﬁciently compact to mediate the energetics of the pyrene redox chemistry.

To follow the kinetics of charge transfer in our mixed monolayer systems, we
have performed AC voltammetry experiments analogous to those described above for
one-component systems (Figure 2.13). The AC voltammograms for these two-
component monolayers illustrate that, when the pyrene ring system is conﬁned within a
mixed monolayer interior, the AC peak current disappears at much lower frequencies

than for the corresponding one-component monolayer. This ﬁnding suggests that the

kinetics of charge transfer are slower for the two-component system, and we evaluate kc,
from plots of ipeak/ ibackground vs. log (frequency) (Figure 2.15). These ﬁts yield kc, =

1100 s'1 for P7 + diluent and 21 s.1 for P13 + diluent on gold. These results show that

the presence of diluent molecules slightly improves the spatial disorder of pyrene

moieties within the monolayer.

The data in Figure 2.15 also show that P7 in a two-component monolayer yields a
much more complex signal than the Randles approximation can account for, and the
experimental data are not ﬁt well using a single rate constant (Figure 2.15 top). Similar
behavior has been reported for ferrocene conﬁned within a two-component monolayer

and such results have been explained in terms of multiple populations of redox sites, each

9

. . 2 . . . .
With a different rate constant. The data 1n Figure 2.15 pomt to the mlxed monolayer

structure containing multiple microenvironments for the pyrene rings, resulting in a
distribution of charge-transfer rate constants. Our report of a single rate constant should

thus be viewed as an average of the several rate constants that contribute to the

53

E"
u:

lpeak / lbackground
N
O

.._:
O

1.20

1.15

i—u
I_.
O

1.05

lpeak / lbackground

WU)
GUI

.._.
Ll!

I

I

I

I

 

k =1100s" ° ,
ct

1 SD.

 

1.00

 

.0...1.2 3 4..

log (frequency [Hz])

 

- O
P -1
_ 0 k =21 s
Ct
1
C
C

- 1 SD. M

O 1 2 3

log (frequency [Hz])

Figure 2.15 ipeak/ ibackground vs. log(AC frequency) plots of monolayers of P7 + diluent

(top graph) P13 + diluent (bottom graph) on gold. The solid lines represent ﬁts to the
data (black squares) using a Randles equivalent circuit.

54

experimental data.

Analogous experiments for the case of ITO electrodes covered with mixed
adlayers containing P8 + diluent or P19 + diluent molecules (Figure 2.16, left and right
columns, respectively) show that for these cases the charge-transfer kinetics of pyrene
derivatives follow the same pattern as for P7 + diluent and P13 + diluent on gold. The
presence of diluent molecules causes the AC voltammetry peak to disappear at lower
frequencies as compared to a single-component layer with no diluent. Again, similar to
the data for the gold surface, the pyrene ring system attached via a short tether (P8),
seems to be less affected than P19 by the presence of diluent molecules. The AC
voltammetry data obtained for a range of frequencies are ﬁtted to the Randles model
(Figure 2.17 top for P8 + diluent and Figure 2.17 bottom for P19 + diluent). It appears

from these plots that, due to the available frequency range of the electrochemical
instrumentation, we could get only a threshold of the ipeak/ ibackground plot, as the
charge-transfer kinetics on ITO are very slow. The resulting larger decrease in the

charge-transfer rate constant for the case of P19 + diluent (from 0.1 5.1 without diluent to
0.008 5-1 with diluent), compared to the kc, decrease for P8 (from 0.2 3" without diluent

-l . . . . . . .
to 0.08 5 With diluent), lS 1nd1cat1ve of structural conﬁnement and Improved

organization imposed on the electroactive chromophore by the presence of the diluent
molecules. Relatively small differences in the rate constants, compared to the differences
in chain lengths of the P8 and P19 molecules (18.8 A for P8, 32.3 A for P19), are not
consistent with a tunneling mechanism for these data. An implication of these results, as

for the single-component monolayers on ITO discussed above, is that these

55

.O
O\

 

 

 

 

 

 

 

 

 

 

 

 

'2 06' 1H2 2; 1H2
3 l T 04 W
0.4- ._. ° '
E 1 s
t: ()2. g 0.2-
3 . o .
00.0. u-s-u..s.-. 00014. . . . . .
< 0 100 200 300 400 500 600 < O 100 200 300 400 500 600
potential [mV] potential [mV]
6' 10Hz 6'
i ’ a 10Hz
t: 2- ‘6. 2~
3 1 3 .
00. ......... .- 00.4. .-.1.2.-.2
‘13 0 100 200 300 400 500 600 < 0 100 200 300 400 500 600
potential [mV] potential [mV]
30- 30.
"‘ 1 50Hz ”
Z5. 20. \ 3 20 50Hz
*a ‘a \
{-3 10- “g 10»
8 3
U 0 . . . . . . . U 0 . - . s . . . . .g .
< O 100 200 300 400 500 600 < 0 100 200 300 400 500 600

potential [mV] potential [mV]

Figure 2.16 AC voltammograms for P8 (left column) and P19 (right column) on ITO in 1
M HClO4.

56

1.7
1.6r
1.5
1.4
1.3
1.2
1.1

L0" \F\*-+~«

I I I

_ -1
QFODSS

l I '

lpeak / lbackground

 

l

 

oogdsllk>L13' 20'
log (frequency [Hz])

r=-0I—0r—0
MALI!
r....,.

lpeak / 1background
N

 

p—i
O

O

' '

dol dsl fol ﬁs‘ i0
log (frequency [Hz])

 

Figure 2.17 ipeak / ibackground vs. log(AC frequency) plots of monolayers of P8 (top

graph) and P19 (bottom graph) on ITO. The solid lines represent ﬁts to the data (squares)
using a Randles equivalent circuit. The size of symbol used for experimental data is
larger then S.D..

57

monomolecular layers are substantially disordered, with the pyrene moiety position
within the layer not being dictated primarily by its tethering chain length. We note,

however, that the presence of diluent molecules increases by a factor of 10 the difference

in kc, between the short (P8, kct=0.08 s") and long (P19, kc,=0.008 3") molecules, in
contrast to the data for the corresponding single-component layers (P8, kc,=0.2 3"; P19,

kc,=0.1 s").

58

Conclusions

We have synthesized and characterized a family of SAMs containing pyrene
derivatives on gold and ITO substrates. When the covalently bound pyrene ring systems

are the only constituents bound to the interfaces, their sub-monolayer coverage ranges

from 4.4 x 10-“ mol/cm2 for P7 to 1.7 x 10-” mol/cm2 for P13 on gold. Similar
electrochemical evaluation for the modiﬁed ITO substrates yields values of 1.2 x 10-ll

mol/cm2 for P8 and 8.4 x 10.'2 mol/cm2 for P19. Electrochemical investigations show

that these molecules do not form well-ordered, densely packed monolayers, but enjoy
signiﬁcant structural freedom. The use of the Randles model for the evaluation of charge
transfer kinetics, followed by the application of the tunneling model to qualitatively
evaluate the ﬁlm thickness dependence of the rate constant, reveal that the monolayers on
gold are better organized than those formed on ITO. When the covalently bound pyrene
ring systems are codeposited at the interfaces with longer diluent molecules, the presence
of these diluent molecules imposes some structural conﬁnement on the tethered pyrene
moieties and improved organization can be seen from the charge-transfer kinetics,

' particularly on gold substrates. Yet these two-component monolayers are not sufﬁciently
compact to mediate the energetics of pyrene redox chemistry. The issues of structural
conﬁnement, organization and motional freedom within the monolayer are addressed in

details by means of steady-state and time-resolved emission spectroscopy in Chapter 2.

59

Literature Cited

(1) Nuzzo, R. G.; Allara, D. L. Journal of the American Chemical Society 1983, 105,
4481-3.

(2) Allara, D. L.; Nuzzo, R. G. Langmuir 1985, I, 45-52.
(3) Allara, D. L.; Nuzzo, R. G. Langmuir 1985, I, 52-66.
(4) Karpovich, D. S.; Blanchard, G. J. Langmuir 1994, 10, 3315-22.

(5) Schessler, H. M.; Karpovich, D. S.; Blanchard, G. J. Journal of the American
Chemical Society 1996, 118, 9645-9651.

(6) Lee, H.; Kepley, L. J .; Hong, H. G.; Akhter, S.; Mallouk, T. E. Journal of
Physical Chemistry 1988, 92, 2597-601.

(7) Lee, H.; Kepley, L. J .; Hong, H. G.; Mallouk, T. E. Journal of the American
Chemical Society 1988, 110, 618-20.

(8) Burwell, D. A.; Thompson, M. E. Chemistry of Materials 1991, 3, 730-7.

(9) Vermeulen, L'. A.; Snover, J. L.; Sapochak, L. S.; Thompson, M. E. Journal of the
American Chemical Society 1993, 115, 11767-74.

(10) Putvinski, T. M.; Schilling, M. L.; Katz, H. E.; Chidsey, C. E. D.; Mujsce, A. M.;
Emerson, A. B. Langmuir 1990, 6, 1567-71.

(11) Katz, H. E.; Schilling, M. L.; Chidsey, C. E. D.; Putvinski, T. M.; Hutton, R. S.
Chemistry of Materials 1991, 3, 699-703.

(12) Kohli, P.; Blanchard, G. J. Langmuir 2000, 16, 4655-4661.
(13) Major, J. S.; Blanchard, G. J. Chemistry of Materials 2002, 14, 2574-2581.
(14) Krysinski, P.; Blanchard, G. J. Bioelectrochemistry 2005, 66, 71-77.

(15) Gupta, P.; Ulman, A.; Fanfan, S.; Komiakov, A.; Loos, K. Journal of the
American Chemical Society 2005, 127, 4-5.

(16) Wrébel, J. T.; Kabzinska, K. Preparatyka i elementy syntezy organicznej; PWN:
Warszawa, 1983.

(17) Krysinski, P.; Brzostowska-Smolska, M. Journal of Electroanalytical Chemistry
1997, 424, 61-67.

60

(18) Mazur, M.; Blanchard, G. J. Journal of Physical Chemistry B 2004, 108, 1038-
1045.

(19) Moriconi, E. J .; Rakoczy, B.; O'Connor, W. F. Journal of Organic Chemistry
1962, 2 7, 2772-6.

(20) Serﬁs, A. B.; Brancato, S.; F liesler, S. J. Biochimica et Biophysica Acta,
Biomembranes 2001, 1511, 341-348.

(21) Kudelski, A. Journal of Raman Spectroscopy 2003, 34, 853-862.
(22) Wirde, M.; Gelius, U.; Nyholm, L. Langmuir 1999, 15, 6370-6378.

(23) Michota, A.; Kudelski, A.; Bukowska, J. Surface Science 2002, 5 02-503, 214-
218.

(24) Gordillo, G. J .; Schiffrin, D. J. Faraday Discussions 2000, 116, 89-107.

(25) Moncelli, M. R.; Herrero, R.; Becucci, L.; Guidelli, R. Biochimica et Biophysica
Acta, Bioenergetics 1998, 1364, 373-384.

(26) Moncelli, M. R.; Becucci, L.; Nelson, A.; Guidelli, R. Biophysical Journal 1996,
70, 2716-2726.

(27) Creager, S. E.; Wooster, T. T. Analytical Chemistry 1998, 70, 4257-4263.

(28) Creager, 8.; Yu, C. J .; Bamdad, C.; O'Connor, 8.; MacLean, T.; Lam, E.; Chong,
Y.; Olsen, G. T.; Luo, J .; Gozin, M.; Kayyem, J. F. Journal of the American Chemical
Society 1999, 121, 1059-1064.

(29) Sumner, J. J .; Creager, S. E. Journal of Physical Chemistry B 2001, 105, 8739-
8745.

(30) Sek, S.; Sepiol, A.; Tolak, A.; Misicka, A.; Bilewicz, R. Journal of Physical
Chemistry B 2004, 108, 8102-8105.

(31) Bain, C. D.; Troughton, E. B.; Tao, Y. T.; Evall, J.; Whitesides, G. M.; Nuzzo, R.
G. Journal of the American Chemical Society 1989, 111, 321-35.

(32) Sumner, J. J .; Weber, K. S.; Hockett, L. A.; Creager, S. E. Journal of Physical
Chemistry B 2000, 104, 7449-7454.

(33) Sek, S.; Palys, B.; Bilewicz, R. Journal of Physical Chemistry B 2002, 106, 5907-
5914.

61

(34) Weber, K.; Hockett, L.; Creager, S. Journal of Physical Chemistry B 1997, 101,
8286-8291.

(35) Sachs, S. B.; Dudek, S. P.; Hsung, R. P.; Sita, L. R.; Smalley, J. F .; Newton, M.
D.; Feldberg, S. W.; Chidsey, C. E. D. Journal of the American Chemical Society 1997,
119, 10563-10564.

(36) Kasmi, A. E.; Wallace, J. M.; Bowden, E. F.; Binet, S. M.; Lindennan, R. J.
Journal of the American Chemical Society 1998, 120, 225-226.

(37) Slowinski, K, Chamberlain, R. V. ,Miller, C. J.; Majda, M. Journal of the
American Chemical Society 1997, 119 11910-11919.

(38) Napper, A. M.; Liu, H.; Waldeck, D. H. Journal of Physical Chemistry B 2001,
105, 7699-7707.

(39) Li, J .; Schuler, K.; Creager, S. E. Journal of the Electrochemical Society 2000,
147, 4584-4588.

62

CHAPTER 3

PROBING INTERFACIAL ORGANIZATION IN SURFACE MONOLAYERS USING
TETHERED PYRENE. 2. SPECTROSCOPY AND MOTIONAL FREEDOM OF THE
ADSORBATES

Introduction

The characterization of interfaces is an area of critical importance to materials
science and biology because interfaces mediate processes ranging from chemical sensing
to cell wall permeability. Our group has been actively involved in the spectroscopic and
electrochemical characterization of metallic and oxide interfaces because of their broad
use in chemical sensing. The interfaces that have received the bulk of the attention have
been self-assembling monolayer and multilayer structures because of their comparative

ease of preparation and their ability to mediate a host of properties, such as electron-

. . l . .. 2,3
transfer kinetics and chemical selectlvuy.

The structure and properties of mono- and multilayer interfaces are determined to
a signiﬁcant extent by the substrate. Metallic interfaces such as gold and other coinage
metals produce relatively well organized monolayers using thiols, and such interfaces are

amenable to electrochemical characterization. One problem with conductive substrates is

the quenching of chromophores located within x1/2 of the substrate,4 precluding their use

in optical spectroscopic studies. It is thus difﬁcult to make direct comparisons between
optical and electrochemical experiments, even for partially transmissive interfaces such
as indium-doped tin oxide (ITO) and boron-doped diamond (BDD), owing to their broad

background absorption. We have studied the optical and electrochemical responses of

63

several polycyclic aromatic hydrocarbons (PAHs) and we understand the oxidative

. . . . 5,6 . .
degradatlon reactions characteristlc for these compounds. The kinetlcs of these

reactions depend on the environment of the PAHs, as we have reported in the previous
chapter. Electrochemical data indicate that there is only a fraction of a monolayer of
pyrene-containing adsorbates on gold and ITO substrates and the monolayers formed
using that chemistry exhibit substantial disorder. The purpose of this work is to

understand these monolayers from a spectroscopic perspective.

To perform the spectroscopic measurements, we use silica as the substrate
material. The surface of silica is different from that of ITO, but our previous work has

indicated that the density and distribution of surface —OH groups on these two substrates

are similar, and we have found no evidence to date indicating that SiOx and ITO

substrates give rise to different monolayer structures. The data we report here indicate
that under all circumstances, the formation of a monolayer structure on silica is
accompanied by signiﬁcant disorder, some of which is related to the solvation of the
monolayer and some of which is intrinsic to the monolayer, owing to the spatial

distribution of surface silanol sites from which the monolayers are grown.

64

Experimental

Chemicals. l-pyrenebutyric acid (97%), l-pyrenemethylamine hydrochloride
(95%), octadecylamine (97%), IO-hydroxydecanoic acid, l-hexadecylamine (98%),
N,N’-dicyclohexylcarbodiimide (DCC, 99%), adipoyl chloride (299%), 4-
methylmorpholine (299.5%), triethylamine (299%), cyclohexane (99%), l-pentanol
(299%), acetonitrile (anhydrous) were obtained from Sigma-Aldrich. Dichloromethane

and chloroform were obtained from POCH and from Mallinckrodt Chemicals.

Surface adlayer constituents. Monolayers on silica substrates were synthesized
by reaction of the substrate with adipoyl chloride (0.3 mL) in dry acetonitrile (10 mL),
using 4-methylmorpholine (0.3 mL) as a Lewis base, under reduced pressure for 1 h. The
reacted substrates were removed from solution, rinsed with dry acetonitrile and ethyl
acetate, and dried under a stream of dry nitrogen. We synthesized monolayers of tethered
pyrene either as the only constituent or as a coconstituent with a long-chain aliphatic
moiety. These monolayers were obtained by exposing the adipoyl chloride covered
substrate either to a 2 mM solution of pyrenemethylamine in dichloromethane (producing
a layer we designate as P8, where the number refers to the number of atoms in the tether
between the pyrene moiety and the substrate) or to a dichloromethane solution of
pyrenemethylamine and hexadecylamine (1:1) for 0.5 h. Following this reaction, the
substrate was removed from solution, rinsed with dichloromethane and ethyl acetate, and
dried under a stream of dry nitrogen. For pyrene tethered further from the substrate, the

adipoyl chloride covered substrate was ﬁrst reacted with a 2 mM solution of 10-

hydroxydecanoic acid in dichloromethane. Following the addition of this CIO chain, the

65

substrate was removed from the reaction vessel, rinsed with dichloromethane and ethyl
acetate, and dried under a stream of dry nitrogen. The substrate covered with adipic acid
mono-(10-carboxy)-decyl ester was then exposed either to pyrenemethylamine and DCC
(1:1) (to form pyrenemethylarnide-lO-decyl ester, P19) in dichloromethane or to
pyrenemethylamine, hexadecylamine, and DCC (1 :1 :1) in dichloromethane for 0.5 h.
The reacted substrate was removed from solution, rinsed with dichloromethane and ethyl
acetate, and dried under a stream of nitrogen. The structures of these probes and diluents

are shown in Figure 3.1.

Steady-State Emission Spectroscopy. Excitation and emission spectra were
acquired with a Spex Fluorolog 3 spectrometer. The excitation wavelength was 320 nm
with both the excitation and'emission slits set to a 5 nm band-pass. Experiments were
carried out in a nitrogen environment, and with monolayers immersed in cyclohexane, 1-

pentanol, and water.

Time-Resolved Emission Measurements. Time-correlated single-photon counting
(TCSPC) was used to study the lifetime and motional relaxation properties of the pyrene

derivatives tethered to silica substrates. The TCSPC system used to acquire time-domain

data has been described elsewhere7 and is reviewed brieﬂy here. The second harmonic

of the output of a mode-locked CW Nd:YAG laser (Coherent Antares 76-S) is used to
excite a cavity-dumped dye laser (Coherent 702-2) operated at 640 nm using Kiton Red
laser dye (Exciton Chemical Co.). The output of the dye laser was typically 100 mW

average power at 4 MHz repetition rate with ~5 ps pulses. The 320 nm light incident on

the sample is generated by frequency doubling the dye laser output using a Type I LiIO3

66

SHG crystal. The average optical power at the sample is less than 1 mW. The detection
electronics are characterized by an instrument response function of 35 ps fwhm. The

emission collection wavelength and polarization are computer controlled using National

Instruments LabVIEW® v. 7.0 soﬁware.

67

is"
:21.

P8 diluent P19 diluent

“030‘

NH NH

O
::M

Figure 3.1 Structures of pyrene-terminated probes (P8, P19) and diluents used in the
construction of monolayer structures reported in this work.

68

Results and Discussion

The spectroscopic characterization of monomolecular interfaces can pose a
challenge because of the small number of molecules present. One way to interrogate
interfaces with high sensitivity is to incorporate ﬂuorescent chromophores into the
interface and probe their optical response. We have investigated several tethered PAHs

for this purpose and have found that there is a common pathway for oxidative

. . . . . 5,6 .
degradatlon, resulting 1n the formation of qumones. Pyrene 15 one PAH that we have

made use of, not only to understand its photodegradation behavior, but also because the
emission spectrum of pyrene is known to be sensitive to the “polarity” of its local

. 8-1 7
envrronment.

In this work, as in previous studies, we have used the same reaction chemistry to

modify silica and ITO surfaces. While these surfaces are certainly not identical, their

reactivities toward acid chlorides and other compounds such as POCI3 are very similar.

It is difﬁcult to make direct comparisons between the resulting interfaces because they
are not amenable to being probed by the same means. ITO can be interrogated
electrochemically, but is characterized by a broad absorption band in the spectral range
where PAHs absorb and emit. Thus the spectroscopic characterization of adlayers bound
to ITO yields results that are contaminated by contributions from the conductive layer
itself. Silica has ideal optical properties, but it is not addressable electrochemically,
making direct comparison of these interfaces difﬁcult. We can estimate from absorbance

data that the surface PAH concentrations on these two materials are similar, and neither

69

material exhibits mesoscopic crystallinity. These pieces of information suggest that,
whatever the organization of the chromophores on these interfaces, it is likely similar,
and it is on this basis that we detail the spectroscopic characterization of a series of

pyrene derivatives tethered to silica.

The steady-state absorption and emission spectroscopy of pyrene is well
established and understood. While the polarity sensitivity of pyrene is related closely to
the symmetry of the chromophore, several previous studies have shown that substituents
on the pyrene chromophore do not necessarily eliminate the polarity-sensitivity as long as
the substituent is not conjugated to the ring system. We have found that tethered pyrene

compounds can be used to gauge the polarity of their local environment when

. . 6
1ncorporated into monomolecular layers.

The emission data for tethered pyrene derivatives provide information about the
organization and polarity of these interfaces. We consider that there are two distinct
bodies of information, time domain and frequency domain measurements, and we treat

these datasets separately.

One of the more important properties of the pyrene chromophore is its ability to
sense the “polarity” of its local environment through the ratio of two vibronic emission

bands, designated I and III (Figure 3.2). This sensitivity is due to the relatively close

energetic proximity of the pyrene S] and 82 electronic states, and the fact that the S] <—

80 and $2 +— 80 transitions are polarized nominally orthogonal to one another. The

extent to which these two transitions are coupled is mediated by the dipolar interactions

70

III

 

 

  
 
  

 
 
  
 

 

 

 

 

 

 

 

 

 

 

 

 

   
   

 

 

 

 

 

 

     

 

 

 

 

I __ __
12‘ a P8 d'l mm b E: d'l
. 4 + luent < + luent
i? 1 ' ‘ rm 3 1.01 P19
8 1.01 , 2 . ,
8 . —--——--~—- P19 + diluent 2 0,31 ----~----—~- P19 + diluent
.E 0.3, .E':
B . '3 0.6-'
.s 0.6- .s .
'5 .~ 7:3 0.4-
g 0.4.. E 02<
a 0.2- c 0°01 ““ -
'360 460 4io 440 460 480 560 ' '330 460 4i0 440 460 430 560
wavelength [nm] wavelength [nm]
— P8 — P8
12; C P8+diluent 1 441 d P8+diluent
.é‘ ' « P19 .é‘ ' . ~Pl9
g 1-01 ~-----~——--Pl9+diluent g ., ff---P19+diluent
.5 08- .‘§
“O * "O
5’ 0.6- g
g 0.44, g
8 0.2-' 2
360'460'420'440'46of4§ojsoo '3§0f46614io'440'46094s0'560
wavelength [nm] wavelength [nm]

Figure 3.2 Steady state emission spectra of pyrene derivatives P8 and P19, with and
without monolayer diluent, bound to silica surfaces. P8 (solid line), P8 with diluent
(dashed line), P19 (dotted line), and P19 with diluent (dash and dotted line), on silica.
Panel (a) is for interface immersed in cyclohexane, (b) immersed in l-pentanol, (c)
immersed in water, and (d) exposed to nitrogen gas. The I and III emission bands are
indicated in (a).

71

between the chromophore to its immediate environment. For this transition coupling to

be optimally sensitive to environmental polarity, the symmetry of the chromophore

should be sufﬁciently high that the 82 <— 80 and S] <— 80 electronic transition dipole

moments are close to orthogonal. It has been shown in a variety of previous studies that

the polarity sensitivity of pyrene is not compromised signiﬁcantly upon substitution or

96,

. 8 . . . .
tethering of the chromophore, prov1ded the substitution at the chromophore ring

structure does not perturb the electronic state wavefunctions substantially. For the
substituted pyrene derivatives we report here, we observe experimentally that these
tethered chromophores do indeed retain their polarity-sensitivity, and we use this

information to interrogate the chromophore local environment.

we show in Table 3.1 the pyrene I/III emission band ratios, a quantity related to
environmental polarity, for pyrene in the solvents used here. We also show the I/III
emission band ratios for the surface-bound derivatives in a nitrogen atmosphere and
immersed in the solvents cyclohexane, l-pentanol, and water. For solution phase pyrene,
we obtain the expected polarity dependence, with the I/Ill band ratio being largest for
water (most polar) and smallest for cyclohexane (least polar). For the tethered
derivatives we observe the maximum I/III band ratios for immersion in l-pentanol, and a
decrease in I/III ratio for immersion in water. We note that P19 does not quite follow this
trend, although it is apparently in a comparatively nonpolar environment. This outwardly
surprising result can be understood in the context of the water overlayer causing the
interfacial adlayer to “fold over” and become disorganized, with the pyrene chromophore

seeking an environment that is relatively non-polar. For the arnphiphilic solvent

72

Table 3.1 Steady-state band intensity ratios for surface-bound pyrene derivatives as a
function of the solvent in which these interfaces are immersed.

 

I/III band intensity ratios

 

 

 

 

 

 

dipole dielectric
medium moment constant a p8 + p19 +
(D) so pyrene P8 diluent P19 diluent
nitrogen - 0.68 0.78 0.74 0.79
cyclohexane 0.00 2.02 0.58 0.71 0.76 0.81 0.75
l-pentanol 1.65 13.9 1.02 1.01 1.27 1.06 1.25
water 1.85 78.5 1.87 0.95 1.03 1.09 0.94

 

 

 

 

 

 

 

 

a Experimental values for pyrene from Dong and Winnik treatment.8

73

 

l-pentanol, the interfacial species can be solvated effectively and the pyrene is thus
exposed to the solvent to a greater extent than it is for the water overlayer. The
cyclohexane data show similar I/III band ratios for the solution phase and surface-bound
species, suggesting relatively effective solvation of the interface by this nonpolar solvent.
Taken collectively, these data suggest that the presence of water above the interface gives
rise to signiﬁcant disorganization of the adlayer, and this ﬁnding is consistent with the

time-domain emission data that we consider next.

The time-resolved emission response of pyrene and its derivatives has been
studied extensively, and some preparatory discussion of these data is in order before we

consider the information they contain in detail. For the native chromophore, the

dominant excitation is for the 82 0— So transition and the primary emission is from the S]

state. These states are polarized orthogonal to one another, leading to negative
anisotropy values. For substituted pyrene derivatives, the addition of some chemical
functionality to the pyrene ring structure perturbs the electronic states to an extent that
depends on the identity of the substitution. In some cases this perturbation leads to the
two electronic states being polarized at an angle less than 54.7° with respect to one
another, and it other cases, the polarization difference between the two states remains at a

value greater than 54.7°. Accordingly, some tethered pyrene moieties yield positive

. . . . . . 19-22 . .
anisotropy data and other derivatives yield negative anisotropy data. The prediction

of whether or not a negative anisotropy will be observed for a given pyrene derivative is

not feasible at the present time, but computational studies are underway to address this

74

issue. For the substituted pyrene derivatives we use in this work, we observe a positive

anisotropy signal.

The second point that needs to be considered is the ﬂuorescence lifetime of
pyrene derivatives. Pyrene is quenched efﬁciently by oxygen, and thus it is not readily
possible to infer environmental information from lifetime measurements unless the
samples used are rigorously deoxygenated. We have not performed this step in the work
we report here and thus we do not try to interpret the lifetime data in terms of speciﬁc
molecular-scale interactions. Rather, we consider that the number of recovered lifetimes
is indicative of a variety of environments in which the pyrene chromophore resides. In
most cases, we recover two lifetime components for the tethered pyrene derivatives
(Table 3.2), with the characteristic short lifetime being on the order of a nanosecond or
less, and the long lifetime being on the order of several nanoseconds, with considerable
uncertainty attending the determination of these longer lifetimes. In contrast, for solution
phase pyrene, previous work has shown that lifetimes on the order of 300 — 400 ns are
obtained under oxygen-free conditions and 6 — 8 ns lifetimes are seen for ambient oxygen
concentrations in different solvents. All of our lifetimes are short by comparison and the
primary conclusion we can draw from these data is that the chromophore is distributed
among at least two local environments in these interfaces. For the systems we report on
here, we recover predominantly two-component lifetimes, suggesting the existence of
more than one environment for the pyrene chromophores within a given monolayer
(Figure 3.3). For the few cases exhibiting apparently single-exponential decays, we

believe these results to be due to the limited signal/noise ratio of the data. The primary

75

 
 
   

 

 

1000 - f(t) = Alexp(-t/r]) + Azexp(-t/1:2)
, AI = 479 i 14
300 - 53 11:1013i40ps
A2 = 442 i 17
g 600 _ I, = 7998 :1: 520 ps
0
.2.
g 400 L-
C‘.
0)
.§
200 -
0 WWW-WNW»-
: l A i n l 1 J n I
0 1000 2000 3000 4000 5000
time [ps]

Figure 3.3 Fluorescence lifetime data for P8 bound to silica and immersed in l-pentanol.
These data are representative of the lifetime data for P8 and P19, with results of ﬁts to the
data shown in Table 3.2. Residuals show agreement of the ﬁt to a two—component
exponential decay.

76

Table 3.2 Summary of time-resolved data for P8 and P19 on silica.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

medium monolayer composition
P8 P8 + diluent P19 P19 + diluent
air Tﬂ [ps] 152i4, 1028i13 441i36, 338i54,
1127i17 1883i177 2022i111
THR [ps] 224i49 l74:t30 -- --
r(0) 0.12zt0.02 03010.04 0.03i0.01 0
r(oo) 0.02:1:0.0l 0.212t0.01 0.03i0.01 0
water Tﬂ [ps] 785130, 8823:18, 972i88 786i122,
4842i650 9028i4043 3401i] 370
THR [ps] 813:25 00 (3141-326) -- --
r(O) 0.1 li0.02 0.03zt0.02 0 0
r(oo) 0.08zt0.01 0.02i0.01 0 0
pentanol rﬂ [ps] 1013i40, 8163:60, 1244i93, 997i287,
7998i520 5086i359 16500i5000 6264i3286
IHR [ps] 453il3 537i30 397124 --
r(0) 0.20i0.01 0.222t0.01 0.24i0.01 0.05i0.01
r(oo) 0.03i0.01 0.04:1:0.01 0.04i0.01 0.05zt0.01
cyclohexane rﬂ [ps] 449i23, 392:1:28, 200i7, 786i142,
2611i126 2817i130 2418i62 4834i1605
THR [ps] 712t42 319i67 -- --
r(O) 0.10i0.04 0.21i0.03 0.09i0.01 0.10i0.01
r(oo) 0.03:1:0.01 0.1 1:1:0.01 0.09i0.01 0.10:t0.01

 

77

 

value of the lifetime data lies in its ability to reveal the existence of a structurally

heterogeneous environment for the pyrene chromophores.

We turn at this point to a discussion of the anisotropy data for tethered pyrene,
because these data contain information on the local environment that is interpretable in a

straightforward manner. We show in Table 3.2 the quantities r, r(0) and r(oo). These

quantities are related to the experimental time-resolved intensity data l”(t) and Ii(t)

through the following series of equations:

 

 

r(t): [um-110)
l||(l)+21i(t)
r(t) = r(oo) + (r(0) — r(oo)) exp(—t/IHR) (3.1)
T _ 793
HR - 240w

The quantities r(O) and r(oo) are the zero-time and inﬁnite-time anisotropies, Dw is the

“wobbling” diffusion coefﬁcient, representing motion of the chromophore about its tether

. . . 23 . . . . .
Within a cone of semi-angle 60. The inﬁnite time anisotropy is related to the average

organization or extent of motional restriction experienced by the chromophore within its
environment, and r(0) is determined by the angle between the excited and emitting

chromophore transition dipole moments. The quantities of most interest to the

organization of the interfacial structures are r(oo) and THR- We consider the time-domain

dynamical data for P8 and P19 separately.

78

The r(oo) results for the interfaces containing P8 show that, in all cases, the
interface is substantially disorganized, with r(oo) values ranging from 0.02 to 0.08. For
P8 coadsorbed with longer organic adlayer species, we ﬁnd that there is a signiﬁcantly
smaller orientational distribution in air and cyclohexane, with r(oo) values for pentanol
and water being consistent with a relatively disorganized interface (Table 3.2). For all of

these measurements, save for P8 buried in an aliphatic environment, we see a decay of

the induced anisotropy and thus are able to report a value of THR- While it may be

tempting to evaluate this quantity as we would for free rotor reorientation times, we

cannot do so because of the dependence of THR on both Dw and 60. Combining these
data with the r(oo) results, which address the limits that can be placed on 60, we see that

the local chromophore environment can indeed place some motional restrictions on Dw.

We know from the hindered rotor model that

2
r(oo) = r(0)<P2(cost9)> (3.2)
and we can extract the quantities r(oo) and r(0) from the experimental data (Figure 3.4).
It is thus possible to extract information on 6. If we assume that 65:61) at long times, we

can use these data to provide information on both Dw and the intrinsic disorder in the

ﬁlms (Table 3.3).

Before considering the interesting trends shown in Table 3.3, we need to be

speciﬁc about the meanings of the quantities Band Dw. The entries marked ” - “ in

79

N

U]

C
l

 

 

 

 

 

_ a
E 200
C
8
‘3‘ 150
E. 100
C
8
.E 50
O
0 1000 2000 3000 4000 5000
time [ps]
0.4 - b
0.3 -
0.2 " . :‘l:
g 0.1 ’ ’2‘ -:::::"°‘"’ - in; . W.;" 5}. . ”3': 'r.' . ‘53:. 5.0;"; .5455}:
I ”I‘ll: 3 “71' I 'l‘l‘i [3"l'1iv31' l. 7 ml '
00 I ll 1% {kiln-IT “ 'l'r‘l‘ 51.10:”
-0.1 -
-O. L . l . I . l . l . I
0 1000 2000 3 000 4000 5000
time [ps]

Figure 3.4 (a) Time resolved emission intensities for emission polarizations parallel and
perpendicular to the (vertical) excitation polarization for P8 + diluent bound to silica and
immersed in l-pentanol. (b) Induced orientational anisotropy function constructed from
data shown in (a). The ﬁt to the data and residuals are also shown.

80

Table 3.3 Results calculated for hindered rotor model based on experimental r(0), r(oo)

 

 

 

 

 

 

 

 

 

 

 

 

and 111R data.
medium monolayer composition
P8 P8 + diluent P19 P19 +
diluent
9 [deg] 39 i 9 19 :h 4 -- --
air Dw [MHZ] 600 (+1160, - 180 (+150, - -- --
310) 100)
6 [deg] 18 i 7 20 :t 17 -- --
water DW [MHZ] 360 (+630, - -- -- --
260)
9 [deg.] 40 i 3 38 i 3 39 i 3 --
Fenian“ Dw [MHZ] 310 (+60, - 240 e 50 340 (+80, - --
50) 70)
61 [deg.] 33 :t 21 25 i 4 -- --
CYC'Ohexane Dw [MHZ] 1360 (+7500, 170 (+120, - -- --
-1250) 70)

 

 

 

 

 

81

 

Table 3.3 correspond to the absence of detectable decay dynamics.22 This limit

represents the absence of free volume in which the chromophore can move (Dw —> 0),

but does not provide direct information on the tilt angle of the pyrene. It is the value of

r(oo) that is related to average tilt angle. The 6 values in Table 3.3 are related to the free
volume which the chromophore can access, and Dw is a measure of how rapidly the

pyrene moiety is moving (rotating, wobbling) within that free volume.

We note from Table 3.3 that, in general, the P8 chromophores are characterized
by much more motional freedom than the P19 chromophores. We attribute this freedom
of motion for the shorter tethered pyrene moiety to the lower propensity of this shorter

molecule for associating with its neighbors by means of H-bonding or van der Waals

interactions. For P19, the greater length of the tethering chain could give rise to either
stronger intermolecular van der Waals interactions if the molecules are in sufﬁciently
close proximity, or the longer chains could allow for enough degrees of freedom to make
chromophore 7t—71.’ interactions more efﬁcient. We believe this latter explanation is not
operative because of the absence of signiﬁcant excimer signal in the steady state emission

data.
For the P8 chromophore, we observe that, while the cone angle 6 is qualitatively

similar for all interfaces immersed in solvent, the Dw terms are signiﬁcantly slower in

the monolayer containing aliphatic diluent molecules. This ﬁnding can be understood in
the context of the diluent providing a viscous environment in which the chromophore

groups move, and the relationship between local viscosity and motional relaxation time is

82

well established.24 For P8 monolayers without diluent, there are presumably fewer

neighbor molecules to impede the motional relaxation of the pyrene group.

The P8 chromophore also exhibits a signiﬁcant solvent dependence to its
relaxation dynamics. These dynamics do not scale simply with solvent viscosity, and this

ﬁnding provides insight into the nature of solvent-monolayer interactions. For P8 in air,

we recover a value of Dw that is intermediate between water and cyclohexane. For P8

without diluent, the monolayer structure appears to be solvated efﬁciently by
cyclohexane, presumably because of that solvent’s ability to provide a nonpolar
environment in which the pyrene can rotate. In air, the pyrene chromophore layer will
either interact with its neighbors or the substrate, and such interactions would likely
provide an environment which will impede the motion of the pyrene ring system more

than the presence of a low viscosity nonpolar solvent. Water and pentanol solvent

overlayers give rise to the lowest Dw values for the P8 monolayer. We speculate that the
relatively polar solvents cause the nonpolar pyrene to interact most strongly with its
neighbors or the substrate, in effect creating a reasonably viscous environment.

For the monolayers containing P8 + diluent, we observe some variation in Bbut

the values of Dw are solvent independent to within the experimental uncertainty. The

implication of these data is clear that the diluent is providing an aliphatic environment in
which the pyrene can reside and that in this environment the chromophore is substantially
isolated from the inﬂuence of the solvent overlayers. We note also that the only P19

monolayer to exhibit measurable dynamics is the monolayer/pentanol system, and the

83

results we extract are indistinguishable from the P8 data in pentanol. Based on this
ﬁnding, it is also possible that the amphiphilic pentanol solvent provides essentially the
same solvation environment for all of these monolayers, and the resulting dynamics turn
out to be similar to the P8 + diluent dynamics because the effective viscosity of the
diluent environment is similar to that of pentanol. This is admittedly a speculative
explanation, but it is consistent with the experimental data. The absence of dynamics in
the P19 + diluent monolayer immersed in pentanol is likely the result of the length of the

aliphatic diluent precluding solvent access to the pyrene chromophore.

84

Conclusions

The broad picture that emerges from our emission data is that, for chromophores
bound very close to the substrate surface, their environment could be expected to be
relatively rigid because of covalent bonding and limited orientational degrees of freedom.
This effect is compensated by the relatively smaller number of attractive interactions with
neighboring molecules, and the net result is that we observe somewhat constrained
molecular motion for P8, both by itself and with the addition of a diluent. The longer
P19 molecule is characterized by less motion, consistent with the dominant inﬂuence of
neighbor-neighbor intermolecular interactions. The exposure of these monolayers to
solvents has a signiﬁcant effect for P8 by itself, demonstrating the interplay between
solvent and monolayer constituents, sometimes acting to solvate the monolayer
constituents and at other times serving to force a molecular-scale phase separation at the
interface. When P8 is surrounded by aliphatic diluent, we observe that the behavior of
this chromophore is dominated by the diluent and inﬂuenced to a much lower extent by
the presence of the solvent. We have speculated that the amphiphilic solvent pentanol
can in fact solvate the P8 + diluent and P19 monolayers, and that the viscosity of the
resulting monolayer environment is characterized by a viscosity that is on the same order
as that of pentanol (> 3 CF). When these data are viewed in the context of the
electrochemical information presented in the preceding paper, it is clear that any
permeability of the monolayer is characterized by mesoscopic features and not by
molecular-scale penetration of solution phase molecules. The comparative

disorganization of the monolayers found with the electrochemical experiments is not

inconsistent with the spectroscopic data. Our Band Dw ﬁndings suggest that the amount

85

of motional freedom available to the pyrene moiety depends on a number of structural
factors but in all cases it is relatively limited. The basis for this limited freedom is likely
the solvent-monolayer interactions forcing the monolayer into a comparatively tightly
packed conformation. We are presently investigating the properties of these systems in
greater detail to provide insight into what structural controls can be exerted to mediate

interactions between monolayers and solvent overlayers.

86

Literature Cited

(1) Dominska, M.; Jackowska, K.; Krysinski, P.; Blanchard, G. J. Journal of Physical
Chemistry B 2005, 109, 15812-15821.

(2) Major, J. S.; Blanchard, G. J. Langmuir 2002, 18, 6548-6553.
(3) Major, J. S.; Blanchard, G. J. Langmuir 2003, 19, 2267-2274.

(4) Chance, R. R.; Prock, A.; Silbey, R. Advances in Chemical Physics 1978, 3 7, l-
65.

(5) Mazur, M.; Blanchard, G. J. Journal of Physical Chemistry B 2004, 108, 1038-
1045.

(6) Mazur, M.; Blanchard, G. J. Journal of Physical Chemistry B 2005, 109, 4076-
4083.

(7) DeWitt, L.; Blanchard, G. J .; LeGoff, E.; Benz, M. E.; Liao, J. H.; Kanatzidis, M.
G. Journal of the American Chemical Society 1993, 115, 12158-64.

(8) Dong, D. C.; Winnik, M. A. Photochemistry and Photobiology 1982, 35, 17-21.
(9) Dong, D. C.; Winnik, M. A. Canadian Journal of Chemistry 1984, 62, 2560-5.
(10) Carr, J. W.; Harris, J. M. Analytical Chemistry 1986, 58, 626-31.

(11) Carr, J. W.; Harris, J. M. Journal of Chromatography 1989, 481, 135-46.

(12) Bogar, R. G.; Thomas, J. C.; Callis, J. B. Analytical Chemistry 1984, 56, 1080-4.

(13) Wistus, E.; Mukhtar, E.; Almgren, M.; Lindquist, S. E. Langmuir 1992, 8, 1366-
71.

(14) Liu, Y. S.; Ware, W. R. Journal of Physical Chemistry 1993, 97, 5980-6.

(15) Liu, Y. S.; de Mayo, P.; Ware, W. R. Journal of Physical Chemistry 1993, 97,
5987-94.

(16) Liu, Y. S.; de Mayo, P.; Ware, W. R. Journal of Physical Chemistry 1993, 97,
5995-6001.

(17) Karpovich, D. S.; Blanchard, G. J. Journal of Physical Chemistry 1995, 99, 3951-
8.

87

(18) Tulock, J. J .; Blanchard, G. J. Journal of Physical Chemistry A 2000, 104, 8341-
8346.

(19) Tulock, J. J .; Blanchard, G. J. Journal of Physical Chemistry B 2002, 106, 3568-
3575.

(20) Krysinski, P.; Blanchard, G. J. Langmuir 2003, 19, 3875-3882.

(21) Kelepouris, L.; Krysinski, P.; Blanchard, G. J. Journal of Physical Chemistry B
2003, 107, 4100-4106.

(22) Karpovich, D. S.; Blanchard, G. J. Langmuir 1996, 12, 5522-5524.
(23) Lipari, G.; Szabo, A. Biophysical Journal 1980, 30, 489-506.

(24) Debye, P. Polar Molecules.

88

CHAPTER 4

INTERROGATING INTERFACIAL ORGANIZATION IN PLANAR BILAYER
STRUCTURES

Introduction

The development of selective chemical and biological sensing devices has been a
topic of extensive research interest because of their potential importance. Biologically
aCtive components such as transmembrane proteins, nucleic acids, and antibodies are
used in biosensing devices due to their intrinsic ability to perform molecular or biological
recognition functions. For a biosensing device, there must be a biologically active
component that is connected in some manner to a transducer such that the chemical
change associated with molecular recognition can be converted into a quantiﬁable signal.
The sensing components must maintain their functional activity while attached to the
transducer, a condition that can be difﬁcult to achieve. Most biomolecules lose their
biological activity when not in their native environment, and one widely attempted means

of maintaining bioactivity is to create an environment on a transducer surface that mimics

the native environment of the biomolecule.“7 Supported lipid bilayers are a commonly

. . 8-10 . .
used chemical structure for this purpose, but there remain many open questions

relating to the design and fabrication of biomimetic bilayers owing to their ﬂuidity and
compositional complexity. We are interested in understanding the organization of model
lipid and hybrid bilayer systems, with the goal of using this knowledge to make
comparatively simple bilayer structures that are capable of supporting biomolecules in

their active form(s).

89

The organization of lipids and proteins in biological lipid membranes was

. . . l l .
described as a ﬂuid mosaic, but more recent work pomt to a complex, heterogeneous

structure. Any interfacial system we design to host transmembrane proteins or other
biomolecules must be characterized by ﬂuid and compositional properties similar to those
of a cell wall, and the central issue is determining the critical bilayer properties required
for this purpose. The most pressing issues in the investigation of supported bilayers are
their ﬂuidity (top vs. bottom leaﬂet) and how composition is related to both ﬂuidity and
structural heterogeneity. At heart is the construction of the compositionally simplest
biomimetic bilayer capable of supporting biomolecules in their active forms, and
understanding the relationship between bilayer composition and dynamics is our starting
point. We have focused on interactions between the bilayer constituent molecules over
nm length scales. The studies we report on here focus on the bilayers themselves,
without the incorporation of proteins. Before we introduce biomolecules into the
artiﬁcial lipid membranes for sensing purposes, we need to better understand the

properties of the bilayer structure.

To aid in this examination, we want to start with a relatively well organized and
compositionally simple system. It is important to have a membrane free of defects for the
eventual electrochemical characterization of incorporated biomolecules. For chemical
sensing applications, the matrix that serves as the host has to be sufﬁciently ﬂuid to
accommodate the presence of a biomolecule, and from a compositional standpoint, the
bilayer has to have the requisite constituents to support the biomolecule(s) in their active
form(s). These requirements place multiple constraints on the bilayer, and it is important

to examine the organization of the bilayer bottom leaﬂet as well as the full bilayer. To

90

examine these monolayers and bilayers, we have utilized several techniques to obtain
information on the organization of the interface over a variety of length scales.
Information that is averaged over macroscopic length scales, such as ellipsometric
thickness, is very useful in terms of determining the extent of bilayer formation and
organization of the bilayer as a whole. Contact angle measurements likewise provide
information on the presence and, to some extent organization, of the bilayers, and
infrared spectroscopic data provide useﬁrl information on the extent of order and
crystallinity within the nonpolar regions of the bilayer. Probing the system
electrochemically provides additional detail about the kinetics of electron transfer
processes across the bilayer interface, and how it changes with the structure and
organization of the interface. Probing this interface with redox active species present in
the electrolyte solution yields information on the permeability of this interface to small
ions. This information is especially important if we want to study selectively the signal
from redox probes of interest or biomolecules located within the interior of an interface.
Time-resolved ﬂuorescence spectroscopy provides information on local organization of
bilayer molecules surrounding the chromophores. Information from these techniques,
taken collectively, provides a picture of the molecular scale organization of the bilayer

constituents.

We have used tethered pyrene as a spectroscopic and electrochemical probe to
examine the organization of selected biomimetic bilayer interfacial structures. Pyrene
has been shown to be useful both for ﬂuorescence dynamical experiments and for studies
of electron transfer in interfacial systems. The steady state and time-resolved

ﬂuorescence behavior of pyrene is well understood, and the sensitivity of pyrene to the

91

polarity of its local environment has been used widely to interrogate interfacial and bulk

12-16 . . .
systems. The electrochemical properties of tethered pyrene have been characterized

. 17,18 . . .
in recent years as well. Tethering this probe to the transducer surface at different

distances can be used to study the redox reaction kinetics and thereby deduce the
properties of the probe’s local environment. These kinetic data point to limited access of
the tethered probe to the overlayer and/or very slow electron transfer between the probe
and the electrode. Fluorescence lifetime data are consistent with the kinetic results and
indicate that the immediate environment of the tethered probe mediates the access of

oxygen to the probe.

92

Experimental

Chemicals. l-Pyrenedecanoic acid (2 98%), and arachidic acid (299%) were
obtained from Fluka. l-Pyrenehexadecanoic acid was obtained from Molecular Probes,
Inc. 1,2-Dipalmitoyl-sn-glycero-3-phosphocholine (DPPC) (299%), l-octadecanethiol
(98%), perchloric acid (70%), sulfuric acid (99.999%), lithium perchlorate (99.99%),
potassium ferrocyanide (II) trihydrate (99%), hexaammineruthenium (III) chloride (98%),
and gold foil (99.99%) were obtained from Sigma-Aldrich. Chloroforrn (HPLC grade)
was obtained from OmniSolv. Aqueous solutions and the aqueous subphase in

Langmuir-Blodgett trough were prepared from Milli-Q water.

Gold and quartz substrate preparation. Gold coated silica substrates (200 nm of
Au sputtered on 20 nm of Cr on Si wafers) were cleaned in a UV cleaner for 10 min and

stored in Milli-Q water prior to modiﬁcation. Gold foil and quartz substrates were

cleaned in piranha solution for 20 min (3:1 concentrated HZSO4 : 30% H202). Caution!

Piranha solutions react violently with organic matter and should be handled with
extreme care! The clean substrates were then rinsed with Milli-Q water and dried under

a stream of dry nitrogen. The gold foil was then cleaned electrochemically in 0.5 M

H2S04 (scan rate 100 mV/s, potential range from -300 to 1700 mV vs. AgIAgCl

reference electrode) until reproducible cyclic voltammograms were obtained. The gold

foil was then rinsed with Milli-Q water and dried under nitrogen. The working area of

the gold electrode was estimated from cyclic voltammetry using [Ru(NH3)6]C13 as a

redox probe.

93

Langmuir ﬁlms. Langmuir ﬁlms were obtained using a computer-controlled
Langmuir-Blodgett trough (model 612D, NIMA Technology Ltd). The Langmuir ﬁlms

were obtained at room temperature (21 °C) on a water subphase. The ﬁrst layer of the

. . . . 19,20
interface was transferred to the solid substrate usrng the Langmuir-Blodgett method,

where the hydrophilic substrate was initially immersed in the water subphase (Figure
4.1a) prior to the spreading of the Langmuir layer. A volume of 40 uL of 2 mg/mL
solution of arachidic acid, octadecanethiol, and pyrenedecanoic or pyrenehexadecanoic
acid (3:3:1 molar ratio) in chloroform was then spread on the water subphase and
evaporation of the chloroform was complete after 10 min. The molecules on the
subphase surface were compressed to the solid phase state (surface pressure 20 mN/m)
and the substrate was then drawn vertically at constant pressure of 20 mN/m. The
resulting monolayer was dried under nitrogen for at least 3 h. To transfer the second lipid
leaﬂet onto the substrate, a volume of 40 uL of 2mg/mL DPPC solution in chloroform
was spread onto the water subphase and the chloroform was allowed to evaporate for 10
min. The surface layer was next compressed to a surface pressure of 45 mN/m and the

monolayer was transferred onto the substrate horizontally according to the method

developed by Langmuir and Schaefer2] (Figure 4. 1 b). The resulting bilayer was allowed

to dry under nitrogen for at least 3 h.

Ellipsometry. Film thickness was measured using a multiwavelength ellipsometer
(J. A. Woollam Co., Inc.) with a 75 W xenon lamp source. The electric ﬁeld incidence
angle is 75° with respect to the sample surface normal, and the refractive index of the
ﬁlm assumed to be 1.5. The measured thickness of the monolayer was 20 i 4 A

(pyrenedecanoic acid), and 24 i 5 A (pyrenehexadecanoic acid). The thickness of the

94

Weiss

4444111113 11111143144

Figure 4.1 (a) Scheme of preparation of the Langmuir-Blodgett monolayer on
hydrophilic substrate by withdrawing the substrate from the subphase. (b) Scheme of
preparation of the bilayer using Langmuir-Schaefer method (horizontal touch).

95

bilayer was measured to be 57 i 7 A (pyrenedecanoic acid), and 68 :1: 3 A
(pyrenehexadecanoic acid). A monolayer of DPPC on gold was measured to be 35 i 2 A.

The uncertainties in the data presented are for at least three individual determinations.

F T-IR Spectroscopy. Reﬂectance FT-IR spectra of modiﬁed gold substrates were

collected using an FT-IR spectrometer (N icolet, Magna—IR 560) equipped with a Pike

grazing angle (80°) attachment. The spectral resolution was set to 2 cm-1 for these

measurements.

Contact Angle. Water contact angles were measured using a Video Contact
Angle System 2000 (AST Products, Inc.). We obtained a contact angle of 94° 3: 2° for a
monolayer containing pyrenedecanoic acid, and 97° :t 1° for a monolayer containing
pyrenehexadecanoic acid. For bilayer structures we recovered a contact angle of 48° d:

6°. We consider the chemical signiﬁcance of these data below.

Electrochemistry. Electrochemical experiments were performed using a
computer-controlled electrochemical analyzer (CH Instruments Model 650) operating
with a three electrode cell. All potentials are referenced to Ag 1 AgCl I 3M KCI (-52 mV

vs. SCE), used as the reference electrode. Platinum wire served as a counter electrode.

Experiments were carried out in aqueous 0.1 M HClO4. 1 mM K4[Fe(CN)6], and
[Ru(NH3)6]Cl3 in 0.1 M LiClO4 were used in studying the integrity of LB ﬁlms.22 A

solution of 1 mM [Ru(NH3)6]CI3 in 0.1 M LiClO4 was also used to estimate the working

electrode surface area.

96

Time-Resolved Emission Measurements. A time-correlated single-photon
counting (TCSPC) instrument was used to study the lifetime and anisotropy decay
dynamics of tethered pyrene in monolayers and bilayers on quartz substrates. The

experiments were carried out at room temperature and in the presence of atmospheric

. . . . 23 .
oxygen. This spectrometer has been described in detail elsewhere and we recap its

salient features here. The source laser is a mode-locked continuous wave Nd:YAG laser
(Coherent Antares 76-S) that produces ~30 W average power at 1064 nm with 100 ps
pulses at a repetition rate of ca. 76 MHz. The second harmonic of the output of this laser
(~2W average power, 100 ps pulses, 76 MHz repetition rate) is used to excite a cavity-
dumped dye laser (Coherent 702-2) operating at 640 nm using Rhodamine 6G dye
(Kodak, Eastman Fine Chemicals). The dye laser produced 5 ps pulses with average
power of ~100 mW at a 3.8 MHz repetition rate. The 320 nm light used to excite the
sample is generated by frequency doubling the dye laser output using a Type I KDP SHG
crystal. The average power of UV light on the sample is less than 1 mW. The
ﬂuorescence detector is a Hamarnatsu R3 809U microchannel plate PMT and the

instrument response time is ca. 35 ps. The emission collection wavelength and

polarization are computer controlled using National Instruments LabVIEW® v. 7.0

software.

97

Results and Discussion

The primary purpose of this work is to evaluate the organization of bilayer
structures formed using Langmuir-Blodgett and Langmuir-Schaefer technology and to
understand whether or not the organization of the bilayer itself can mediate the dynamics
and electron transfer properties of species contained within these structures. In addition
to addressing these issues, which are of potential importance to trans-membrane signaling
and leakage processes, we are also concerned with utilizing these and similar structures
as biomimetic systems, and as such it is important to understand the role of a probe

species contained within the bilayer.

In this work we have used Langmuir-Blodgett and Langmuir-Schaefer methods to
deposit planar lipid bilayers on gold and quartz substrates. We have chosen these
substrates for their utility as electrodes for electrochemical experiments and as optically
transparent supports for spectroscopic measurements, respectively. For these bilayers,
the bottom leaﬂet was composed of arachidic acid, octadecanethiol, and either
pyrenedecanoic (Ple) or pyrenehexadecanoic (Py16) acid, in a 3:3:1 molar ratio (Figure
4.2). The thiol was used to provide stable attachment of the layer to the gold electrode
and it was also used for layers formed on quartz for the sake of compositional
consistency. We have used different pyrene tether lengths to evaluate how the
electrochemical and spectroscopic signals from the pyrene moiety vary with distance
from the solid substrate. We expect the functional form of the pyrene tether length
dependence to be different for the electrochemical and spectroscopic measurements, and
we are also aware that the pyrene moiety can itself act as a perturbation to the layer

structure. For all of our measurements, the top leaﬂet of the bilayer we deposit on our

98

\

ONE:
\

 

O O O O 0 U0
/ /OH / /H / /§|OH /

Au/quartz Au/quartz

 

a b

Figure 4.2 Scheme of bilayers containing tethered pyrene ((a) pyrenedecanoic acid, and
(b) pyrenehexadecanoic acid) reported in this work. The structures identical are not
intended to reﬂect the actual conformation of the interfacial constituents.

99

planar substrates was composed of the phospholipid DPPC with no additives.

Cyclic Voltammetry. We consider the electrochemical response of the tethered

pyrene moiety ﬁrst. The electrochemistry of this molecule has been reported elsewhere,

and is well understood.”’18 We show in Figure 4.3 the ﬁrst three cycles in the cyclic

voltammograms of tethered pyrene in monolayers and bilayers on gold. We assume here
that thiols react with gold so that they stabilize the monolayer by virtue of their
attachment to the substrate. For both pyrenedecanoic acid (Ple) and
pyrenehexadecanoic acid (Py16), the ﬁrst anodic scan reveals irreversible oxidation of
the pyrene moiety at ca. 1 V, to form an unstable radical cation which, in a subsequent
cathodic scan forms a more stable pyrenediol. The reversible peaks between 200 and 300
mV, seen in scans two and three, are associated with the pyrenedione/pyrenediol redox
couple. The peaks, seen at 260 mV for both Ple and Pyl6, shift toward more positive
potentials with the addition of the DPPC lipid adlayer, producing maxima at 275 mV for
Ple and 290 mV for Py16. These shifts are analogous to those reported previously in
the literature for different redox probes and have been attributed to a combination of local
solvation effects on the redox species potential and the double layer effect, where the
redox potential was shifted to more positive values due to a net charge separation

between the redox species buried within a ﬁlm and their charge-compensating ions

present in the solution.24’25 The pyrenedione/pyrenediol couple undergoes a 2e-/2HJr

reaction, requiring the presence of protons to proceed. The screening of the
pyrenedione/pyrenediol by the DPPC overlayer limits proton access resulting in a

positive shiﬁ of peak potential for bilayers compared to monolayers. The screening effect

100

 

 

 

 

 

 

 

, monolayer
5 L ------ bilayer 3
NE 4 _
o D
2
3.
A?
U)
C
d)
'O
E
l:
“.3
O

 

 

 

 

 

potential [V]

 

 

 

 

 

 

 

6 - b
monolayer
5 _ ------ bilayer
N'— J
E
Q
<
.2:
.«E‘
(I)
ﬁ
0
'O
is
t:
:3
U

 

potential [V]

Figure 4.3 Cyclic voltammograms of (a) pyrenedecanoic acid and (b)

pyrenehexadecanoic acid in monolayers and bilayers on gold in 0.1 M HClO4. The scan
rate is 20 mV/s.

101

can also be seen in the amount of pyrene moieties undergoing redox reaction. The
loading density of pyrene derivatives is constant for both mono- and bilayers and can be

estimated from the isotherms obtained during the Langmuir-Blodgett deposition. The

. 2
surface area per molecule from the measured isotherms (data not shown) are ca. 20 A

for arachidic acid, 22 A2 for octadecanethiol, 25 A2 for pyrenedecanoic acid, and 39 A2

for pyrene hexadecanoic acid. Based on this information we calculated the number of

moles of pyrene derivatives per electrode area (assuming the transfer ratio is l) to be 150

pmol/cm2 for Ple and 120 pmol/cm2 for Py16. These values are signiﬁcantly higher

than the corresponding information on electrochemically active pyrene moieties extracted
from cyclic voltammograms (Table 4.1) and we believethat this discrepancy can be
interpreted in terms that the electrochemistry data report on the number of
electrochemically active pyrene moieties rather than the total number present on the
surface. The number of electroactive pyrene moieties contributing to the AC
voltammetric signal (calculated from AC voltammograms at low AC frequencies) is three
times that determined from CVs for monolayers, but in bilayers it is similar for both
techniques. AC voltammetry shows that only about 10% of the pyrene moieties
measured for the monolayer contribute to the AC voltammetric signal when the bilayer is
present. Another, even more noticeable feature in the cyclic voltammograms presented in

Figure 4.3 is a two to four-fold decrease in the capacitive current for bilayers relative to

monolayers (20 to 8 uF/cm2 for Ple, and 24 to 5 uF/cm2 for Pyl6). This change in the

capacitance is the result of an increase in ﬁlm thickness and it can be understood in terms

of a capacitor with increased distance between the capacitor plates,

102

Table 4.1 Redox properties from cyclic (CV) and AC voltammetry of tethered pyrene in

mono- and bilayers.

 

 

 

 

 

 

 

 

 

 

 

ch 1" AC CAC 8AC _1
2 2 2 ket [S ]
[pmol/cm ] [pmol/cm ] [uF/cm ]
Ple-monolayer 9 d: 2 32 i 3 2.8 :l: 0.6 6.3 :1: 1.2 283 :1: 27
Py16-monolayer 7 i 2 20 d: 8 2.8 :1: 0.4 7.7 :l: 1.0 262 i 43
Ple-bilayer 6 i 1 6 3c 1 0.3 :t 0.03 6.4 i 0.2 434 :1: 23
Py16-bilayer 8 :t 1 2 :t 1 0.8 i 0.01 6.0 :1: 0.1 142 :l: 32

 

103

 

(3.—.._... (4.1)

where C is the speciﬁc capacitance, dis the distance between the capacitor plates, 8 is the

dielectric constant, and 80 is the permittivity of free space. The capacitive current values

shown in Table 4.1 are extracted from AC voltammograms where the observed
capacitance is the result of charging and discharging of a dielectric membrane only. In
CV experiments there must be additional Faradaic components that perturb the system
because the capacitance values extracted from CVs (data not shown) are higher than
expected for this type of interface. Based on capacitance measured with electrochemical
techniques and the ﬁlm thickness measured with Ellipsometry (vide infra), we can also
estimate the values for dielectric constants for different systems. The estimated values of
dielectric constant for monolayers and bilayers are 6-8. These values are three to four
times higher than dielectric constants of 2-3 commonly accepted for alkanethiol ﬁlms.
The reason for that may be the constrained motional freedom of the mono- and bilayer

structures as well as the presence of small defects.

To obtain a selective voltammetric response from the pyrene moieties present in
the surface-bound ﬁlm, the monolayers and bilayers have to be relatively well organized

and defect free. To evaluate the ﬁlms for the presence of defects, we used two different

redox probes: K4[Fe(CN)6] and [Ru(NH3)6]CI3. We have chosen these two probes

because they differ in the value of heterogeneous electron transfer rate constants (1.5 x

10"2 cm/s for Fe(CN)64- and 1.8 x 10'2 cm/s for RuCNH_~,)63+).22 For both redox probes,

104

 

 

 

 

 

 

 

_ Ru(NH3)6 F C(CN):
150 -
100 -
NH '
g 50 -
E o — -
£7 -50 - ..--;';'—':’.:;’.-'
m _ “52:: """
0:) V \\ 'I
'6 '100 h \‘I”
E _150 _ —— bare Au
3 ------ monolayer
0 -200 - ---------- bilayer
—250
-0.6 -0.4 -O.2 0.0 0.4 0.6

potential [V]

Figure 4.4 Cyclic voltammetry of pyrenehexadecanoic acid in a monolayer on gold in
aqueous 1 mM K4[Fe(CN)6] and in 1 mM [Ru(NH3)6]Cl3 in 0.1 M LiClO4. Sweep rate

100 mV/s.

105

the cyclic voltammograms were recorded at 100 mV/s scan rate. Cyclic voltammograms

recorded for [F e(CN)6]4-/3- were characterized by an exponential shape typical for a

defect-free ﬁlm (Figure 4.4, potential window from -0.2 to 0.6V), while CVs recorded for

[Ru(NH3)6]3+/2+ exhibited well resolved peaks, indicative of the presence of defects

(Figure 4.4, potential window from -0.5 to 0.2V). It has been reported elsewhere that

because of the charge of these redox probes, the transfer of the [F e(CN)6]4- ion from
aqueous solution to a low dielectric medium requires 80% more energy than the transfer

of [Ru(NH3)6]3+.22 Since the transfer of the [Fe(CN)6]4' ion trough the ﬁlm was

completely blocked, but the ﬁlm was permeable to [Ru(NH3)6]3+ ion, our ﬁlms likely
contain some relatively small defect sites.

F TIR Data. We are cognizant of the fact that the introduction of a ﬂuorescence
probe such as pyrene in a monolayer will give rise to disorder within the monolayer.

This structural factor may contribute to the presence and detection of defects in our

bilayer structures. In an attempt to evaluate the extent of organization of our L-B ﬁlms

on reﬂective substrates, we have used FT-IR spectroscopy. We use the CH2 asymmetric

and CH2 symmetric stretches of the aliphatic chains in our bilayers as a gauge of

organization (Figure 4.5). The band position for these vibrations depends on the

organization of the medium. For an adlayer characterized by quasi-crystalline

organization, the CH2 stretches are seen to shift to lower energies than those seen for

disordered or liquid-like ﬁlms.26-29 This effect is seen for a variety of interfacial

106

 

normalized absorbance
o O.

 

L J

3 100 3000 2900 2800 2700

frequency [cm-1]

Figure 4.5 F TIR spectra of the monolayers and bilayers reported here: (a)
pyrenedecanoic acid and (b) pyrenehexadecanoic acid in monolayers, (c) pyrenedecanoic
acid and (d) pyrenehexadecanoic acid in bilayers. Spectral range presented here is 3100 -

2700 cm".

107

Table 4.2 Peak positions for CH2 stretching modes in crystalline and liquid statesa, and

in monolayers and bilayers adsorbed on gold. aFrom reference 26

 

band positions [cm' 1]
CH2

 

Py10- Py16- Ple- Ple-

stretch crystallinea liquida . .
monolayer monolayer bilayer bilayer

 

asymmetric 2918 2924 2921 2918 2918 2919

 

symmetric 2851 2855 2850 2849 2850 2850

 

 

 

 

 

 

 

 

 

108

systems that contain aliphatic chains. For example, well ordered self-assembled

alkanethiol monolayers on gold have CH2 asymmetric and CH2 symmetric stretch peaks
at 2918 and 2851 ch , respectively,26 and the band positions for gel phase DPPC
monolayers are 2920 cm-1 and 2851 cm-1 in the liquid condensed (LC) phase and 2925
cm-1 and 2854 cm.1 in the liquid expanded (LE) phase.30 The CH2 stretching bands shift

to higher energies by ca. 5 cm-I for disorganized monolayers. The F TIR data obtained

for the bilayer systems we report here are given in Table 4.2. For both monolayers and

. . . . . -l
bilayers containing tethered pyrene, the asymmetric CH2 stretch lS centered at 2920 cm

. . -l . .
and CH2 symmetric stretch 1S centered at 2850 cm . These data are conSistent With our

ﬁlms exhibiting signiﬁcant order compared to a liquid-like system, but not quite the same

level of organization seen for long-chain alkanethiol-gold monolayers. Our data are

consistent with a DPPC layer in the LC phase,30 suggesting that the extent of defects in

these ﬁlms is limited. However, pyrene moieties present in the bilayer may disrupt the
crystalline bilayer structure on short length scales, as was seen in the redox probe
diffusion experiments. Such “cracks” resulting from the presence of the probe may not

be obvious on a macroscopic length scale.

Electron transfer across the interface. We expect, as noted above, that the
kinetics of electron and proton transfer for pyrene derivatives in these ﬁlms will depend
on the details of ﬁlm organization. We use AC voltammetry to study the kinetics

associated with oxidation/reduction reaction(s) of tethered pyrene. We have acquired AC

109

voltammograms over the frequency range of 1 Hz to 10 kHz. For these data, the peak

current to background current ratio decreased with increasing AC frequency, in

. 18 .
correspondence With the rate of electron transfer. The data are modeled usmg a

Randles equivalent circuit (Figure 4.6). Shown in Figure 4.7 are a series of AC
voltammograms acquired at three different frequencies to illustrate the behavior of redox
reaction of tethered pyrene with different tether lengths in our mono- and bilayer
structures. It is the frequency-dependence of the data shown in Figure 4.7 that are

modeled by the Randles equivalent circuit. To estimate the electron transfer rate constant

from these data, we have used a model developed by Creager and Wooster.“ In this

model, the interfacial redox-active system on the electrode surface is represented by the

impedance of a Randles equivalent circuit (Figure 4.6), where CDL is the double layer
capacitance, C A D is the adsorption pseudocapacitance, RCT is the charge transfer

resistance, and RSOL is the solution resistance.

The analysis involves plotting the AC peak current (ipeak) to background

(baseline) current (ibackground) ratio versus the logarithm of AC frequency for a series of

voltammograms recorded over a broad AC frequency range. These plots exhibit three

distinct regions. At low frequencies ipeak / [background ratios approach a constant value

that is related to the amount of tethered pyrene on the electrode surface. There is

relatively little kinetic information in this region because the frequency is slower than the

relevant redox process. The second region is at high frequencies where ipeak/ ibackground

110

 

 

~—’V\W —-
Ree. ——~\W—l+-

 

 

Figure 4.6 Schematic of equivalent circuit used to model the AC impedance data. RSOL
= solution resistance, RCT = charge transfer resistance, C AD = adsorption

pseudocapacitance and CDL = double layer capacitance.

111

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

40. 120+
.._. 1.0+ H 30. .._,
r: 0 3. 1 Hz :: c: 804 A a
d) - d) O.)
5 0.6+ 5 20+ 5 60
O 0.4] U '04 O 40.
U 021 U U 204 1000 HZ
< 0'0 0:0 0.1 02 0:3 0:4 0351 < 0 00 oil 02 0:3 0:4 0T5 < 0 070 07.1 0303 04 03
potential [V] potential [V] potential [V]
3 :4 4 :4
H 0.8- 1 H2 H 30‘ 100 HZ H 100- A
a 0.6- t: . c: b
Q) Q) 20‘ d) 75-
. 10+
(in) 0.2 U U 25‘ 1000 Hz
0'0 0:0 oil 0:2 0:3 04 0f5' < 0 do 0Tl 0T2 0T3 0:405 < 0 do oil 0T2 03 04 0f5'
potential [V] potential [V] potential [V]
3. 020+ :i. :i. :3”
E 0.15 1H2 E? 6 IOOHZ E A C
a) a) a) 304
g 0.10« E 15" 20,
o 0.05- 0 2. 0 '0, 1000 Hz
22 . 22 4 22
0'00 0f0'0il'0i2 0f3'014'0f5‘ 0 010'0fl'0f2'0f3'oi4'0f5' 0 0f0'0fl'0i2'0f3'0f4'0f5'
potential [V] potential [V] potential [V]
'2 '2 '2‘
.2. .3 A .=_'-. 25*
e. 0.04 L e. 2.5 ... 20‘
3 0.0+ 6 f-g; g .5. d
0.024 ' .
o 00. 1H7- 5 I-O: ‘00 HZ 6 '0 1000 Hz
< 0’00 0:0 0: 1 0:2 0:3 0:4 0:5‘ 0'0 0:0 oil 02 0:3 0:4 03 0 00 oil 02 0:3 04 0:5
potential [V] potential [V] potential [V]

Figure 4.7 AC voltammograms recorded for pyrenedecanoic acid in a monolayer (a),
pyrenehexadecanoic acid in a monolayer (b), pyrenedecanoic acid in a bilayer (c),

pyrenehexadecanoic acid in a bilayer (d) in 0.1 M HClO4. Frequencies shown: 1, 100,
and 1000 Hz.

112

ratios approach unity because the redox reaction kinetics are much slower than the

oscillation in the AC potential. The intermediate region is a transition between the low

and high frequency limits where the ipeak/ ibackground ratio depends sensitively on the
31-33 . . . ~
AC frequency. It 18 this region, where the AC frequency is of the same order as the

electron-transfer rate constant (ket) for the reaction.

We show in Figure 4.8 plots of ipeak/ ibackground ratios vs. log(AC frequency) for a

series of AC voltammograms recorded in the 1 Hz to 10 kHz range for tethered pyrene in
mono- and bilayers. The average values obtained for at least three different samples are
provided in Table 4.1. From the model, which assumes a single value for the electron
transfer rate constant, we get a sigmoidal frequency dependence, and we see this
functional form in Figure 4.8a-c. For Figure 4.8d, however, the absence of a clear low
frequency plateau means that there are multiple electron transfer rate constants operating,

and the lowest frequency ones are responsible for the plateau’s absence. We interpret the

calculated value of electron transfer rate constant from nonsigmoidal ipeak/ ibackground

curves as an average from multiple electroactive sites and/or rate constants. We note,
additionally, that the recovered surface coverage values are not constant for our ﬁlms
(Table 4.1). We believe that these data report on the number of electrochemically active

pyrene moieties rather than the total number present on the surface.

. . . - + . . .
Because the relevant electrochemical reaction is a 2e /2H process, it is likely that

113

  
  

  

 

 

 

 

  

 

 

 

 

8i b
“g 6- 23 71 °
32‘ e 6-
on on l
6 4- '2‘» 5‘.
,_B .48 41
\ \ 3..
x :4
E- 2. -1 a; 2‘ -1
-— .kqt=272.¢2llsl . . . . -- .. tea-3214.37? . . . a
0 1 2 3 4 0 i 2 5 4
log (frequency [Hz]) log (frequency [Hz])
1.5-
4 , . l -
'0 C 1: 1.41 . d
s s +
O
30"!) 3' «5.0 1.31
‘6 ‘6 '
g g 1.2-
C 2‘ C ‘
‘é 3‘5 1.1*
-l 0 ' ' -
".31, et=429458s .—°- 1.0-ket=156e22s‘
0 ' i ' i ' é ' 4 0 ' i ' i ' é ' 4
log (frequency [Hz]) log (frequency [Hz])

Figure 4.8 Plots of peak current/background current ratio vs. log (AC frequency) for
pyrenedecanoic acid in a monolayer (a), pyrenehexadecanoic acid in a monolayer (b),
pyrenedecanoic acid in a bilayer (c), pyrenehexadecanoic acid in a bilayer (d). The solid
red lines represent the ﬁt of the data (circles) to model developed by Creager's group.

114

the accessibility of the chromophores to protons mediates the recovered F values. For all
of our monolayer and bilayer structures, the values obtained for electron transfer rate
constants are quite small and are essentially the same, suggesting a similar environment
for the electrochemically accessible tethered pyrene in all of these interfaces. One
structural motif consistent with this ﬁnding is that the adlayer(s) sequester the pyrene
moiety so that it has limited access to the aqueous overlayer. This type of organization is
not without precedent. Finklea et al. have reported that burying the redox centers within

the monolayer interior gives rise to small electron transfer rate constants, with values

. . 34
recovered Similar to those we report here. In that work, the slow electron transfer was

from Ru (II/III) redox centers that were attached to a gold electrode. They interpreted the
decreased apparent electron transfer rate constant for redox centers embedded in the outer
portion of the monolayer relative to redox centers located in the electrolyte in terms of
greater heterogeneity of environments around the redox centers. Sumner and Creager

have studied “exposed” and “buried” ferrocene groups self-assembled on a gold

24 . . . . .
electrode. Their work has shown that the kinetics assoc1ated wrth ferrocene redox

reactions were much more sluggish when ferrocene groups were buried within the
monolayer interior than when the ferrocene groups were exposed to electrolyte. The
values of electron transfer rate constants for a monolayer with buried ferrocene moieties
were two orders of magnitude smaller than those obtained for the exposed ferrocene
moieties. The kinetics associated with buried ferrocene oxidation/reduction were also

independent of the tether length.

115

Based on previous studies of various redox probes located at different levels
within aliphatic monolayers, we infer that the slow kinetics seen for the pyrene redox
reaction in the systems we report here are the result of a shielding effect produced by the
aliphatic chains. For monolayer ﬁlms, this shielding occurs within the upper levels of the
monolayer, since the pyrene tether is always shorter than the monolayer constituents and,
for bilayers, additional shielding results from the deposition of the top lipid leaﬂet. The
details of this shielding process are difﬁcult to discern, since the relevant pyrene redox
chemistry involves two protons and two electrons, and this environment could prove to

hinder the transport of both these species. Marcus has presented the idea that ion-pairing

. . . . . 35
interactions could affect the overall redox reaction kinetics. He showed that for

processes where electron transfer is coupled to the ion transport, the apparent electron
transfer rate constant is related to both the electron transfer rate constant and the ion
transfer constant. If electron transfer is fast and ion transport is slow, then ion transport
will dominate the apparent rate constant for the reaction. It appears that, based on our
data (Table 4.1), the comparatively slow rate constants for the pyrene diol/dione redox
chemistry are limited by access of the redox couple to protons, and possibly by the
electron transfer rate constant from the redox-active species to the electrode. The fact
that the frequency-dependence of the AC voltammetry data produce a sigmoidal form
that is modeled well by our equivalent circuit impedance analysis suggests that the
distribution of environments in which the pyrene resides is comparatively narrow. A

broad (energetic and/or geometric) distribution of local environments would give rise to a

. . . 36 . . . . . . .
“linear”, or nonsrgmordal form in the plots shown in Figure 4.8. With this information

116

in mind, we consider next the information we can extract from the ﬂuorescence dynamics

of these pyrene derivatives.

Fluorescence anisotropy. To interrogate the dynamics of tethered pyrene in our
monolayer and bilayer structures, we have used time-resolved ﬂuorescence spectroscopy.
The most effective means for extracting molecular-scale motional information on this
system is through ﬂuorescence anisotropy measurements, where the sample is excited
with a short, vertically polarized pulse of light, and polarized emission from the sample is
collected at polarizations parallel and perpendicular to that of the excitation pulse (Figure
4.9a). The polarized emission transients are combined (Equation 4.2) to produce the
induced orientational anisotropy function, r(t) (Figure 4.9b):

_ III‘IL

r(t) -
I” + 211

(4.2)

For chromophores in an unconstrained environment (e.g. solution), the anisotropic
orientational distribution photoselected by the excitation pulse can relax to a random
orientational distribution. When the motion of the chromophore is restricted in some
manner (e. g. tethering to a surface), the initial anisotropic distribution cannot
orientationally re-randomize completely and r(t) does not decay to zero, as it would for a

free chromophore. For such constrained systems, the hindered rotor model developed by

. . . 37 . . . . .
Lipari and Szabo 1S useful. In this model, the time course of the induced orientational

anisotropy, r(t), is given by Equation 4.3:

r(t) = r(oo) + (r(O) — r(oo))exp(—t/IHR) (4.3)

117

 

 

 

 

 

   
 

 

 

 

400-
H a
'2 300-
:3
O
.2.
200-
.9
.,. l
5
.§ 100-
l
0 . - - J ‘ + ‘ J
0 2000 4000 6000 8000
time [135}
0.6- b
0.4- . -.. "-2:
j: 0-2 ‘ 1' {rear}; ‘1’.“
q - '. El] '
(ill t‘Iilll "
0.0 _. 1 .
ll 1 ‘ll {l
. A 1‘
-0.2+..--1-'
0 2000 4000 6000 8000

time [ps]

Figure 4.9 (a) Time-resolved emission intensities for polarization parallel and
perpendicular to the excitation polarization for pyrenedecanoic acid in a lipid bilayer. (b)
Orientational anisotropy function obtained from data shown in (a). The ﬁt to the data is
shown as the solid line and residuals are distributed around zero.

118

r(0°)

 

 

r(0) = 0.5 + (cos 60 (1 + cos 60)) (4.4)
792
rHR = 24 131W (4.5)

The quantities r(O) and r(00) are initial and inﬁnite-time anisotropies, which are
measured experimentally. The initial anisotropy r(0) is determined by the angle between
the excited and emitting chromophore transition dipole moments. This model predicts
that the anisotropy decays exponentially in time (Equation 4.3) as the chromophore
orientational distribution rerandomizes to the extent possible. The inﬁnite time

anisotropy is related to the extent of motional restriction imposed on the chromophore by

its immediate environment. The decay time constant THR (Equation 4.5) is related to
angular conﬁnement of the chromophore, expressed in terms of a cone with semiangle 60
(Equation 4.4), and the “wobbling” diffusion coefﬁcient Dw, which characterizes the

motion of the chromophore about its tethering bond.

We show in Table 4.3 the quantities THR: r(0), and r(oo) obtained for tethered

pyrene chromophores in our interfaces. From the r(0) and r(oo)data we extract 60 using

Equation 4.4 and from the time dependence of r(t) we determine Dw (Table 4.4). The 60

values shown in the Table 4.4 are related to the free volume accessible to tethered pyrene,

and Dw values are related to the viscosity of the pyrene local environment. We note

from these 6b data that, in general, tethered pyrene molecules reported here are

119

Table 4.3 Summary of time-resolved data for pyrenedecanoic and pyrenehexadecanoic
acid in monolayers and bilayers.

 

 

 

 

 

 

 

 

 

 

 

Quantity Py10- Py16- Ple-bilayer Py16-bilayer
monolayer monolayer
6371 10 44li6 59521223 671 :1: 149
70,1 (08)
,ﬂ 2 (ps) 3550 :h 63 3278 :1: 34 5089 d: 136 4943 i 361
THR (ps) 1498 i 258 852 i 350 271 :t 199 807 i 243
r(0) 0.15 :1: 0.01 0.22 :t 0.02 0.29 i 0.03 0.22 d: 0.03
r(oo) 0.07 d: 0.01 0.17 i 0.01 0.21 i 0.01 0.15 :t 0.03

 

120

 

Table 4.4 Results obtained for hindered rotor model based on experimental r(oo), r(0),

 

 

 

and THR data.
Quantity Py l 0- Py l 6- Pyl O-bilayer Pyl 6-bilayer
monolayer monolayer
61) (deg.) 394:1 2313 274:2 25:1:2
DW (MHZ) 92 21:12 62 :1: l4 91:1:12 74:1: 18

 

 

 

 

 

 

121

 

characterized by a relatively constrained environment, with 60 values ranging from 39°

for Ple in a monolayer to ca. 25° for the other systems studied. We account for these
data as follows. The 39° value for the Ple-monolayer system indicates that there is
signiﬁcant motional restriction imposed on the chromophore even in the monolayer. For
the Ple-bilayer, the presence of the overlayer serves to restrict the motional freedom of
the underlayer, thereby restricting the motion of the chromophore further. In the case of
the Pyl6-monolayer, the presence of the chromophore so close to the “top” of the
interface allows for there to be relatively strong interchain interactions, which further
constrains the available space into which the chromophore can motionally relax. For the
Pyl6-bilayer, the same situation is seen as before, with the interactions between the
adlayers producing a comparatively restricted environment. Owing to the extent of the

orientational conﬁnement already present in the monolayer, the addition of the overlayer

changes this structure comparatively little. The values of 6]) that we recover in all cases

suggest that the pyrene chromophore is conﬁned rather tightly in the surface layer(s), a

result that is entirely consistent with the comparatively small values of ket that we have

- . . +
recovered. We attribute these small values of ket mostly to the limited access of H to

the pyrene dione species in the adlayer. The efﬁciency of the electron transfer between
the pyrene diol/dione and the electrode surface may additionally depend on the distance

of the redox moieties from the electrode surface.

We consider next the values of the wobbling diffusion constant that we recover
from our experimental data. We note that this quantity is very close to the same for all

systems, with there being an apparent slight differentiation between the Ple and Py16

122

chromophores. This slight difference in Dw for these two chromophores does not

correlate with the variations in 6b, and likely represents a difference in motional freedom

for the two chromophores that is somehow related primarily to the tether length. In both

cases, the values of ca. 80 MHz that we recover for our layers suggests a viscosity on the

order of 40 cP, if we assume that the Debye-Stokes-Einstein (DSE) relationship38’39

holds for this system. Regardless of whether or not the DSE relationship is quantitatively
correct, it points to the adlayer in which the chromophore resides as being a
comparatively rigid and conﬁning environment. This qualitative description is consistent
with the nature of the adlayer in which the pyrene chromophore is tethered. Speciﬁcally,
there are a number of Au-thiol attachment points, and while these points may be able to
translate to a limited extent, their motion is impeded by the existence of the arachidic acid

in the layer.

Fluorescence lifetime. We note that the ﬂuorescence lifetime data for tethered

. . . . . . 16,40
pyrene in mono- and bilayers lS consrstent With that which we have seen elsewhere.

For surface-bound pyrene species, a two-component exponential decay (Figure 4.10) is
typically observed. The mechanistic reasons for this decay functionality remain to be
determined fully, but it is possible that these data represent a distribution of lifetimes
rather than simply two discrete components. It is well known that pyrene ﬂuorescence is
quenched efﬁciently by oxygen, and we have not performed any deoxygenation on our

samples. This is the reason for the comparatively short lifetimes we measure for this

chromophore. The most prominent feature in these data is the difference in 72 values

123

between monolayers and bilayers. The longer of the two ﬂuorescence time constants for
the tethered pyrene derivatives in our bilayer structures is consistently longer than it is in
the monolayer structures. We postulate that this difference is related to the accessibility

of oxygen to the chromophore, which should be reduced in the bilayer.

124

800 -

 

 

L :
'5‘ 600 - 2
c o
=3 s
o
o l
“-‘ 400 -
.3:
(D
s: ' ._
8 200 - ° . ".-
.E I V -

O i . _ ‘ _ ‘ ° . ' A : ﬂﬁwor's
0 2000 4000 6000 8000
time [ps]

Figure 4.10 Fluorescence lifetime data for pyrenehexadecanoic acid in a monolayer on
quartz. These are representative for all the lifetime data obtained for the systems reported
in this paper. The ﬁts to all the data are presented in Table 3. Residuals show the
agreement of the ﬁt to a double exponential decay function.

125

Conclusions

There are several important parameters that have to be considered when designing
biosensing devices. Biomolecules such as transmembrane proteins require specially
designed matrices that mimic the natural environment of the transmembrane protein.
Transmembrane proteins span the biological membrane in such a way that the
hydrophobic region(s) reside in the interior of the lipid bilayer, while hydrophilic
portions of the protein are located in aqueous compartments present on both sides of the
lipid membrane. An important aspect of the construction of supported biomimetic
systems is the incorporation of “reporter” molecules that are capable of interrogating the
bilayer local organization. For information from such molecules to be useful, they must
be localized at predetermined locations within the bilayer structure, and one way to
achieve this goal is to tether the probe molecule to the supporting interface. In this paper
we have focused on the organization of a bilayer structure absent protein incorporation.
Our model system was a bilayer comprised of a bottom leaﬂet of fatty acids and fatty
acid analogues and a top leaﬂet of phospholipids. Some of the fatty acid analogues were
pyrene derivatives that we use to probe the organization and electron-transfer
characteristics of the fatty acid monolayer and lipid-capped bilayer. We have shown that
electron transfer from tethered pyrene buried within the mono- and bilayers is relatively
slow and almost independent of the length of the tether connecting pyrene to the
electrode surface. These data suggest that the bilayer is organized sufﬁciently well to
prevent ready access of the tethered chromophore to the electrode surface. Consistent
with this level of organization is the sequestration of the probe away from the aqueous

overlayer and access to protons. Time-resolved ﬂuorescence anisotropy decay

126

measurements indicate that the tethered pyrene resides in a relatively constrained
environment, as detected by the small volume accessible for the chromophore to rotate
within, and the slow rate of motion. These data suggest subtle differences in mono- and
bilayer organization that are expected based on the capping properties of the lipid
overlayer. The information we have gained from this work provides a foundation for
investigating phospholipid bilayer structures with variable amounts of tethering
connections to their supporting substrate. We anticipate these studies to yield a rational
approach to the creation of simple biomimetic structures capable of supporting selected

transmembrane proteins in their active forms.

127

Literature Cited

(1) Ataka, K.; Giess, F.; Knoll, W.; Naumann, R.; Haber-Pohlmeier, S.; Richter, B.;
Heberle, J. Journal of the American Chemical Society 2004, 126, 16199-16206.

(2) Giess, F .; Friedrich, M. G.; Heberle, J .; Naumann, R. L.; Knoll, W. Biophysical
Journal 2004, 87, 3213-3220.

(3) Kiessling, V.; Tamm, L. K. Biophysical Journal 2003, 84, 408-418.

(4) Majewski, J .; Wong, J. Y.; Park, C. K.; Seitz, M.; Israelachvili, J. N.; Smith, G. S.
Biophysical Journal 1998, 75, 2363-2367.

(5) Merzlyakov, M.; Li, B; Gitsov, 1.; Hristova, K. Langmuir 2006, 22, 10145-10151.
(6) Tien, H. T.; Ottova, A. L. Electrochimica Acta 1998, 43, 3587-3610.
(7) Wagner, M. L.; Tamm, L. K. Biophysical Journal 2000, 79, 1400-1414.

(8) Tanaka, M.; Sackmann, E. Nature (London, United Kingdom) 2005, 43 7, 656-
663.

(9) Tanaka, M.; Sackmann, E. Physica Status Solidi A: Applications and Materials
Science 2006, 203, 3452-3462.

(10) Chan, Y.-H. M.; Boxer, S. G. Curr. Current Opinion in Chemical Biology 2007,
11, 581-587.

(11) Singer, S. J.; Nicolson, G. L. Science (Washington, DC, United States) 1972, 175,
720-31.

(12) Dong, D. C.; Winnik, M. A. Photochemistry and Photobiology 1982, 35, 17-21.

(13) Kalyanasundaram, K.; Thomas, J. K. Journal of the American Chemical Society
1977, 99, 2039-44.

(14) Karpovich, D. S.; Blanchard, G. J. Langmuir 1996, 12, 5522-5524.

(15) Mazur, M.; Blanchard, G. J. Journal of Physical Chemistry B 2005, 109, 4076-
4083.

(16) Dominska, M.; Krysinski, P.; Blanchard, G. J. Jaurnal of Physical Chemistry B
2005, 109, 15822-15827.

(17) Mazur, M.; Blanchard, G. J. Journal of Physical Chemistry B 2004, 108, 1038-
1045.

128

(l8) Dominska, M.; Jackowska, K.; Krysinski, P.; Blanchard, G. J. Journal of Physical
Chemistry B 2005, 109, 15812-15821.

(19) Blodgett, K. B. Journal of the American Chemical Society 1934, 56, 495.
(20) Blodgett, K. B.; Langmuir, 1. Physical Review 1937, 51, 964-82.

(21) Langmuir, 1.; Schaefer, V. J. Journal of the American Chemical Society 1938, 60,
1351-60.

(22) Krysinski, P.; Brzostowska-Smolska, M. Journal of Electroanalytical Chemistry
1997, 424, 61-67.

(23) DeWitt, L.; Blanchard, G. J .; LeGoff, E.; Benz, M. E.; Liao, J. H.; Kanatzidis, M.
G. Journal of the American Chemical Society 1993, 115, 12158-64.

(24) Sumner, J. J .; Creager, S. E. Journal of Physical Chemistry B 2001, 105, 8739-
8745.

(25) Twardowski, M.; Nuzzo, R. G. Langmuir 2003, 19, 9781-9791.

(26) Porter, M. D.; Bright, T. B.; Allara, D. L.; Chidsey, C. E. D. Journal of the
American Chemical Society 1987, 109, 3559-68.

(27) Allara, D. L.; Nuzzo, R. G. Langmuir 1985, 1, 45-52.

(28) Laibinis, P. E.; Whitesides, G. M.; Allara, D. L.; Tao, Y. T.; Parikh, A. N.;
Nuzzo, R. G. Journal of the American Chemical Society 1991, 113, 7152-67.

(29) Persson, H. H. J.; Caseri, W. R.; Suter, U. W. Langmuir 2001, 1 7, 3643-3650.

(30) Mitchell, M. L.; Dluhy, R. AJournal of the American Chemical Society 1988,
1 10, 712-18.

(31) Creager, S. E.; Wooster, T. T. Analytical Chemistry 1998, 70, 4257-4263.

(32) Creager, S.; Yu, C. J .; Bamdad, C.; O'Connor, 8.; MacLean, T.; Lam, E.; Chong,
Y.; Olsen, G. T.; Luo, J .; Gozin, M.; Kayyem, J. F. Journal of the American Chemical
Society 1999, 121, 1059-1064.

(33) Weber, K.; Hockett, L.; Creager, S. J. Journal of Physical Chemistry B 1997, 101,
8286-8291.

(34) Finklea, H. 0.; Liu, L.; Ravenscroft, M. S.; Punturi, S. Journal of Physical
Chemistry 1996, 100, 18852-18858.

129

(35) Marcus, R. A. Journal of Physical Chemistry B 1998, 102, 10071-10077.

(36) Li, J .; Schuler, K.; Creager, S. E. Journal of the Electrochemical Society 2000,
147, 4584-4588.

(37) Lipari, G.; Szabo, A. Biophysical Journal 1980, 30, 489-506.
(3 8) Debye, P. Polar Molecules.
(39) Perrin, F. Journal de Physique et le Radium 1934, 5, 497-511.

(40) Mazur, M.; Blanchard, G. J. Bioelectrochemistry 2005, 66, 89-94.

130

CHAPTER 5

CONSTITUENT-DEPENDENT LIPOSOME STRUCTURE AND ORGANIZATION

Introduction

In the bio-sensing community there is a great deal of effort focused on the
development of biomimetic interfaces on planar substrates for the purpose of supporting
transmembrane proteins in their active form(s). There are many issues that need to be
considered in the creation of a functional biomimetic interface, and among them is the
ability to attach the biomimetic interface to a transducer (e. g. electrode, spectroscopic
window). The biomimetic interface is a lipid bilayer-based structure, and a key property
of lipid bilayers is their ﬂuidity. Indeed, the structural heterogeneity and ﬂuidity of lipid
bilayers are prerequisites for the incorporation of transmembrane proteins. Most
biomolecules lose their biological activity when removed from their native environment,
and one approach to maintaining their bioactivity is to create a local environment on a
transducer surface that mimics the native environment of the biomolecule. Because
transmembrane proteins in their native environment experience an aqueous medium on
both sides of the bilayer, it will be important to create analogous organization on the
surface of any viable biosensor. To construct such an interface on a gold (electrode)
surface, we have designed and synthesized a lipid molecule conjugated with a short
polyethylene glycol (PEG) chain terminated with a thiol functionality. Our goal in
creating such a molecule is to form vesicles enriched with this modiﬁed lipid so that
vesicle fusion on a gold substrate would produce a stable supported bilayer structure that

retains some ﬂuidity in the bottom (attached) leaﬂet. Our ﬁndings are that the addition of

131

our modiﬁed thio-PEG containing lipid serves to do more than simply enhance binding to
an Au surface. The presence of the thio-PEG lipid produces changes in the morphology
of the solution phase lipid structures, which serve to alter the ability of the lipid adlayer
to bind to the Au interface. It is these changes in liposome morphology and local

organization that we are concerned with in this work.

The notion that the bilayer constituents can give rise to changes in the
morphology of the liposomes is not new. There is a body of literature that points to the
effect of micelle-forming lipids, such as those containing PEG functionalities, driving the
conversion of spherical lipid bilayers into small micelles. Hristova et al. suggested that
the main reason for liposome transformation into a micelle is the shape of the PEG-lipid
molecule and that the steric barrier to liposome formation resulting from the presence of a

head group-bound polymer chain determines the maximum amount of PEG-lipid that can

be incorporated into a spherical bilayer.l Phase separation and aggregation of PEG-

lipid/phospholipid mixtures has been investigated in several different ways. The
inﬂuence of polymer length on the point of bilayer saturation with PEG-lipids has been
explored, revealing that the gel phase of the lipid bilayer is maintained with up to 8-10%

of PEG-lipids present, where the PEG polymer lengths range from 350 to 5000 Da, and

that above 10%, the lipid bilayer organization is dependent on PEG chain length?-4

Shorter PEG chains and intermediate concentrations of longer PEG chain-containing

lipids have been used to mediate the formation of integrated bilayers, while high

. . . . . . 24
concentrations of longer PEG-containing lipids were shown to form micelles. The

132

most efficient incorporation of PEG-lipids into liposomes has been shown to take place

when the PEG polymer chain matches the length of a lipid acyl chain.4

The conversion of a vesicle into a micelle has been also examined for various

lipid acyl chain lengthsl’s’6 In that work it was found that micelles formed more readily

with lipids possessing long acyl chains. Rovira-Bru et al. used molecular mean ﬁeld

theory to calculate the optimal size and structure of spontaneously formed liposomes

composed of lipid/PEG-lipid mixtures.7 According to their predictions, the

compositional asymmetry between the two lipid leaﬂets in a vesicle is much larger for a
vesicle containing PEG-lipids than for a vesicle composed of unmodiﬁed lipids, and that
the size of the liposome inﬂuences the extent of asymmetry, with smaller liposomes
producing more asymmetric lipid distributions. These authors also indicated that there is
a strong tendency for the PEG head group functionality to locate on the outer surface of
the vesicle, thereby placing limits on the loading density of PEG-modiﬁed lipids for
which vesicles are stable. Also, according to their calculations, a very small polymer
loading excess causes the small micelles to become the most energetically favorable
structure. There is evidence in the literature suggesting PEG-modiﬁed lipids have a
signiﬁcant inﬂuence on the structure of lipid aggregates. PEG-modiﬁed lipids have been
shown to induce “open” bilayer disc or micellar disc structures, with exposed bilayer

edges, when introduced into vesicles in concentrations lower than that at which mixed

spherical micelles start to forms’8 The formation of disc-like and open bilayer fragments

. . . . . 9-12
has been seen during solubilization of liposomes by a number of surfactants.

133

However, disc or bilayer fragments are not stable in solution and tend to close on
themselves to form lipid vesicles. In addition to forming self-closing structures, the
amelioration of exposed edges and consequent stabilization of discs and bilayer

fragments can be achieved by incorporation of molecules into the bilayer that are capable
of shielding the open edges. 13'] 5 PEG-lipids with a bulky, hydrophilic polymer chain

attached to the lipid headgroup can easily accommodate to highly curved lipid monolayer
(or bilayer) regions that are present at the edges of discoidal micelles. The presence of
the bulky polymer at the edge of the disc can prevent closure of bilayer fragments or
fusion of discoids. PEG-lipid concentrations higher than 10 mol% in lipid assemblies
trigger a transition from a lamellar phase to a micellar phase, which proceeds through the

intermediate state of a discoidal bilayer.

To gain insight into the organization and dynamics of molecules comprising
mixed lipid aggregates, we have prepared small liposomes containing controlled amounts
of thio-PEG lipid and a small amount of either pyrene labeled DPPC or perylene as
optical probe molecules. The maximum diameter of the aggregates we produced was
controlled by the pore size of the polycarbonate membrane used in the extrusion process.
We have used time-resolved spectroscopy to measure the ﬂuorescence lifetime and
anisotropy decay behavior of perylene as a function of liposome composition to
understand how the local environment of the probe changes with the addition of thio-

PEG lipids. This chromophore is a hydrophobic polycyclic aromatic hydrocarbon
(PAH), for which there is signiﬁcant knowledge of its spectroscopy and dynamics.‘649

We have also examined the steady state and time-resolved spectroscopic response of

I34

. . . 20-22 . .
pyrene-tethered DPPC incorporated into our liposomes. Fluorescence lifetime data

indicate that the immediate environment of tethered pyrene changes with the addition of

thio-PEG lipids. The ﬂuorescence lifetime of the chromophore is sensitive to the

presence of 02, and the addition of thio-PEG lipid to the lipid assemblies allows 02 in

solution more access to the chromophore, resulting in the quenching of pyrene
ﬂuorescence. The dynamics of tethered chromophore also change with the addition of
PEG-lipid to the liposomes, while the dynamics of perylene remain constant. This
ﬁnding indicates that the presence of the PEG head group has a limited inﬂuence on the

lipid acyl chain region.

135

Experimental Section

Chemicals. 1,2-Dipalmitoyl-sn-glycero-3-phosphocholine (DPPC) (299%), 1-
Palmitoyl-Z-(pyrene- l -yl)decanoyl-sn-glycero-3 -phosphocholine (Py-DPPC) (295%),
1,2-dipalmitoyl-sn-glycero-3-phosphoethanolamine (DPPE), 0-(2-Carboxyethyl)-0'—(2-
mercaptoethyl) heptaethylene glycol (thio-PEG), N-(3-Dimethylaminopropyl)-N’-
ethylcarbodiimide hydrochloride, pyrene (99%), perylene (299%) were purchased from

Sigma-Aldrich. Chloroform (HPLC grade) was purchased from OmniSolv.

Thio-PEG lipid synthesis, puriﬁcation and characterization. Thio-PEG

conjugated DPPE lipid (thio-PEG-DPPE) was synthesized consistent with the procedure

described by Krysinski et al.23’24 A solution of DPPE (100 mg, 0.14 mmol) and thio-

PEG (66 mg, 0.14 mmol) in chloroform was stirred under nitrogen. To this solution a
drop of triethylamine was added to deprotonate the lipid amine group. Next EDC (30
mg, 0.14 mmol) in chloroform was added dropwise to the reaction mixture with
continuous stirring. The reaction solution was stirred at room temperature for 24 h. The
product was puriﬁed using silica gel column chromatography (Silica gel from Spectrum,
6-16 mesh; chlorofomi/methanol/water 20/5/1 v/v/v eluent) and analyzed by mass
spectrometry (Waters QTOF Ultima with electrospray ionization). The yield of this

reaction was 77% after puriﬁcation.

Vesicle preparation. DPPC, thio-PEG-DPPE, and either Py-DPPC (0.5 mol %),
pyrene or perylene (0.1 mol %) were mixed in chloroform in appropriate proportions.
Chloroform was then evaporated and Milli-Q water (Millipore puriﬁcation system,

Bedford, MA) was added to produce a 1.4 mM lipid solution. Each mixture was

136

 

processed three times through a freeze-thaw-vortex cycle to ensure complete lipid and

ﬂuorescent probe mixing. In every cycle the solution was frozen by immersion in N2(l)

for 5 min, thawed in 55 °C water bath for 5 min and vortexed for l min. After the three
freeze-thaw-vortex cycles, the solution was extruded eleven times through a
polycarbonate membrane with 100 nm pore diameter (Avanti Polar Lipids Inc.,
Alabaster, AL). Extrusions were performed above the DPPC transition temperature

(41°C, NIST Lipid Thermotropic Phase Transition Database).

Transmission Electron Microscopy (T EM) Imaging. Liposome samples were
stained with l % uranyl acetate. A 5 uL drop of stained liposome solution was placed on
a Formvar copper grid (200 mesh) and incubated for several minutes. The excess
solution was removed with ﬁlter paper. The microscope used is 100 CX J EOL

instrument (Japan) operating at 100 kV.

Steady-State Emission Spectroscopy. Excitation and emission spectra were
acquired using a Spex Fluorolog 3 spectrometer. The excitation wavelength was 325 nm
for pyrene and Py-DPPC containing samples and 415 nm for samples containing

perylene. Both excitation and emission monochromator slits were set to 1 nm band-pass.

Time-Resolved Emission Measurements. A time-correlated single photon

counting (TCSPC) system was used to acquire time-domain polarized ﬂuorescence

intensity decays. The source laser is mode-locked CW Nd:YVO4 laser (Spectra Physics

Vanguard) that produces 2.5 W average power at 355 and 2.5 W average power at 532
nm with 13 ps pulses at 12.5 ns intervals. The 532 nm output of this laser is used to

excite a cavity-dumped dye laser (Coherent 702-2) operating at 640 nm using Kiton Red

137

620 dye (Kodak Eastman Fine Chemicals). The dye laser produces 5 ps pulses with ~200
mW average power at a 4 MHz repetition rate. A Type I KDP SHG crystal is used to
frequency double the dye laser output and produce 320 nm pulses to excite pyrene-

containing samples. The average power of the 325 nm light at the sample is less than 1

mW. For the samples containing perylene as a probe, the 355 nm output of the Nd:YVO4

laser is used to excite a cavity-dumped dye laser (Coherent 702-2) operating at ca. 430
nm with Stilbene 420 dye (Exciton). This dye laser produces 5 ps pulses with ~100 mW
average power at a 4 MHz repetition rate. Polarized emission from the sample was
collected using a 40x reﬂecting microscope objective, and routed to two detection
channels. The collection wavelength was set using subtractive double monochromators
(Spectral Products CM112) and detected using microchannel plate photomultipliers
(Harnamatsu R3 809U-50). The two signal channels were routed to a Becker & Hickl
SPC-132 dual channel module containing the time-correlation electronics (constant
fraction discriminators, time-to-amplitude converters, multichannel analyzers). The
reference signal was collected using a photodiode (Becker & Hickl PHD-400-N) and
routed through a Becker & Hickl DEL-350 programmable delay module to the reference

inputs of the SPC-132 electronics. The hardware was controlled using a National

® . . . .
Instruments LabVIEW program written in-house. The instrument response function for

this system is typically ca. 30 ps.

138

Results and Discussion

The primary purpose of this work is to better understand the effect of thio-PEG
lipids on the organization and dynamics of molecules comprising liposomes and micelles.
As noted above, the presence of thio-PEG lipid has a substantial effect on the shape of
lipid aggregates, with a progression of structures from liposomes at low thio-PEG lipid
concentrations to discs and/or micelles at high thio-PEG lipid concentrations. In the
work we present here, we interrogate the nonpolar region of vesicles, intermediate
structures in transition from lamellar to micellar phases, and micelles, to elucidate the
molecular-scale consequences of adding PEG-containing lipids to lipid structures. We
have used thiol-terminated PEG-lipids because we are also interested in fusion of
liposomes to planar substrates and the thiol terminal functionality provides the required

afﬁnity for a gold substrate.

Before considering the dynamics of the probe molecules in detail, it is instructive
to examine pictorially the changes in lipid organization with the addition of thio-PEG-
DPPE. Representative TEM images acquired for lipid structures composed of a range of
DPPC/thio-PEG-DPPE molar ratios are shown in Figure 5.1. The images show the
transition from spherical unilamellar vesicles for low thio-PEG-DPPE concentrations ( <
10 mol %) to small spherical micelles at high thio-PEG-DPPE concentrations ( > 50 mol
%), with an intermediate state of discoidal lipid assembly. Similar results were presented
by J ohnsson and Edwards who characterized their PEG-lipid containing aggregates with
cryo-transmission electron microscopy and dynamic light scattering.8 Using‘these

techniques for comparable systems with longer polymer chains, they found out the onset

139

 

Figure 5.1 TEM images of liposomes containing 2 mol % (a), 50 mol % (b) and 100 mol
% of thio-PEG—DPPE.

140

chain used in our study is much shorter than the polymers investigated by Johnsson and
Edwards, and this factor may account for the presence of stable vesicles at thio-PEG-
DPPE levels up to 10 mol %. We observe discoidal lipid assemblies (Figure 5.], center)
for thio-PEG-DPPE concentrations between 20 and 50 mol %. Small spherical micelle
formation becomes more evident at thio-PEG-DPPE concentrations above 50 mol %. It
is clear that the presence of thio-PEG-DPPE has a signiﬁcant inﬂuence of the mesoscopic
morphological properties of the lipid assemblies. The reasons for these structural
changes are likely complex, mediated by both steric and solvation effects, and a key
question is how these changes in nanoscale organization inﬂuence molecular scale

organization. We use spectroscopic means to evaluate this issue.

We have used ﬂuorescent probes to study molecular-scale dynamics in lipid
assemblies with shapes inﬂuenced by thio-PEG-DPPE over a range of modiﬁed lipid
concentrations. For these studies, we have incorporated a small, constant amount pyrene-
modiﬁed DPPC into lipid structures to interrogate local organization. The steady-state
absorption and emission spectroscopy of pyrene is well established and understood, and
can provide some insight. One of the most important properties of this chromophore is
its ability to sense the polarity of its local environment through the ratio of ﬁrst (~3 72
nm) and third (~383 nm) vibronic bands.25'27 The pyrene I/III emission band ratio varies
from 0.6 for nonpolar solvents such as hexane to 1.9 for polar solvents such as water.
This effect is known to be the result of vibronic coupling in the pyrene molecule, where
the local solvent dielectric response serves to mix contributions to vibronic coupling
between two orthogonal electronic states.27 Substituents to the pyrene ring structure

break the symmetry of the chromophore and can inﬂuence the details of vibronic

141

coupling differently than is the case for the symmetric chromophore. For this reason,
tethered pyrene has been reported to exhibit a somewhat decreased sensitivity to
environmental polarity.22’28 Emission from pyrene labeled DPPC in vesicles and micelles
containing controlled amounts of thio-PEG-DPPE does not exhibit substantial changes to
the U11] band ratio (Figure 5.2a). The I/III band ratio recorded for these systems is ~2.5
and is independent of thio-PEG lipid concentration. This value for the I/III band ratio
exceeds the limits seen for a range of solvents using the unsubstituted pyrene
chromophore.26 We attribute the form of these data to the presence of the substituent on
the pyrene ring system, and it appears that there is limited information available from this
tethered chromophore, save from the fact that this band ratio is likely consistent with a
very nonpolar environment, consistent with expectations. To elucidate the actual polarity
of the probed environment, we measured ﬂuorescence emission of (un-tethered) pyrene
in the analogous lipid systems. The pyrene l/IIl emission band ratios for our lipid
structures are ca. l.0 - 1.1 over the range of thio-PEG-lipid concentrations used (Figure
5.2b). This band ratio value corresponds to intermediate polarity on the py scale,
suggesting an environment similar to solvents such as n-butanol or n-propanol. Another
interesting feature seen for both pyrene-labeled DPPC and free pyrene emission is the
appearance of an excimer emission band at low and intermediate thio-PEG-lipid
concentrations (Figure 5.3). More excimers were formed when the free probe was used,
but for both bodies of data, the excimer intensity decreased at high thio-PEG-DPPE
concentrations. This ﬁnding suggests that the chromophores are solvated more
effectively by the lipid acyl chain environment or are partially partitioned into PEG-

containing domains at high thio-PEG-DPPE concentrations. Dilution of the probe,

142

 

 

0%
.......... 1%
............... 2%
~—»~ 5%
“-._..- 10% (a)
—-—— 20%
—— 50%
—— 70%
— 100%

 
 
  
 
 
   

 

 

 

normalized ﬂuorescence emission

 

 

'I'l'l'l‘l'l'l'l
340360380400420440460480500

wavelength [rim]

 

 

 

 

 

0%
l 4 - .......... 1%
a J ............... 2%

.3 1.2 _____ _ 5%

'3 10 ' —-—-— 10% (b)
3 ' _ —— 20%
§ 0.3 - ‘— 50%
in: . —— 70%

s 0.6- —— 100%
3

a u

E 0.4- \

2: ' /_,.//,, ~ ”ﬁr-5‘ \
E 0.2 - - ”'7‘ : Fxf
g u

0.0 -

 

 

' I ' I ' I ' I ' I ' I ' I ' I
340 360 380 400 420 440 460 480 500
wavelength [nm]

Figure 5.2 Steady-state ﬂuorescence of: pyrene-labeled DPPC (a) and pyrene (b) in
liposomes and micelles containing various concentrations of thio-PEG-lipid. Spectra
normalized to emission peak I (472 nm).

143

 

O tethered pyrene
0.40 - 0 free pyrene

 

 

 

0.35 -

0.30 -

' O
0.25 - @

0.20 -

‘ o
0.15- '9

0.10-

normalized excimer emission
O

0.05 -
d . .

 

0.00

 

1

' l ' l ' T T ' 1
0 20 40 60 80 l 00
thio-PEG-lipid concentration [mol %]

Figure 5.3 Excimer emission intensities plotted versus the amount of thio-PEG-lipid in
the liposomes. Intensities are normalized to emission peak I (472 nm).

144

whether it resides inside the micelle or in the PEG-containing headgroup region, may be

inﬂuenced by the presence and density of the bulky polymer in the lipid headgroup.

The steady state emission data for pyrene provide some insight into the
chromophore local environment, but do not yield any signiﬁcant information on how that
environment changes with the addition of thio-PEG-DPPE. Time-domain emission
measurements can provide complementary information and it is these data we consider
next. Fluorescence lifetime data obtained for pyrene labeled DPPC point to two-
component exponential decay (Figure 5.4) which has been seen previously for tethered
pyrene derivatives?"29 The two-component lifetime can be understood in the context of
the tethered chromophores residing in a range of environments. The ﬂuorescence
lifetime data for all the systems we have studied here are listed in the Table 5.1. Both
ﬂuorescence lifetime components decrease signiﬁcantly for thio—PEG-DPPE
concentrations above 110 mol %. The shorter ﬂuorescence lifetimes are consistent with
increased access of dissolved 02 to the chromophore when it is incorporated in micellar
structures relative to its incorporation in liposomes. These lifetime data, considered
together with the MU emission band ratio data for tethered and free pyrene imply that the
chromophore resides in a dielectric gradient, and the most likely region for this to occur
for a lipid structure is in a region that spans the acyl chain and headgroup functionalities.
The quenching of ﬂuorescence by 02 in micellar systems suggests a possible dilation of

the lipid headgroup region due to the presence of the thio-PEG terminal functionalities.

To gauge the dynamics of tethered pyrene in lipid vesicles, micelles, and
transitional structures, we measured ﬂuorescence anisotropy decay. For this

measurement, the sample is excited with short, vertically polarized light pulse, producing

145

A'

6000-

 

5000-
3:" 4000-
s:
5
O
8.. 3000-
.E‘
E 2000-
8
.E

1000-l

04:5

 

 

 

‘ I T l ' l ' I ' l
l 0000 20000 30000 40000 50000
time [ps]

Figure 5.4 Fluorescence lifetime data for pyrene labeled DPPC in DPPC vesicles (0 mol
% 0f thio-PEG-lipid) in water. This is representative for all the lifetime data obtained for
the systems reported in this work. The ﬁts to all the data are presented in the Table 1.
Residuals (centered around zero) show the agreement of the ﬁt to a double exponential
decay function.

146

Table 5.1 Summary of time-resolved ﬂuorescence data for pyrene labeled DPPC in all

the systems reported in this paper. The percentage corresponds to mol % of thio-PEG-
DPPE in the particular system.

 

 

 

 

 

 

 

quantity 0% 1% 2% 5% 10% 20% 50% 70% 100%
I [ps] 5789 6482 6678 6787 6811 6182 5419 4716 5310
“v' :t 179 :l: 675 :l: 345 :t 238 :l: 506 :l: 778 :l: 233 i 24 :l: 294
1 [ps] 78026 91:40 76:29 76:23 69299 5731E 88 58:62 53307 67:63
ﬂ 2
’ :l: :t :l:
5 11 2692 1762 2114 2218 3513 2620 594 1325
8415

I [ S] 6781 i 7954 8027 7285 5733 5705 4250 3480
“R p i719 1636 i940 i414 3:789 3:242 :l: 192 :l: 178 :l: 144
r(O) 0.12 i 0.12 :l: 0.11 :l: 0113 0.12 :l: 0.114 0.12 i 0.15 :l: 0.14 :l:

0.02 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01
r(oo) 0.07 i 0.07 i 0.07 i 0.06 :l: 0.06 :l: 0.04 i 0.03 :l: 0.02 i 0.02 :t

0.02 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01

 

 

 

 

 

 

 

 

 

 

147

 

a polarized emission transient from the sample. The polarized emission transient is
collected at polarizations parallel and perpendicular to the excitation pulse (Figure 5.5a).
The polarized emission intensities (In and I 1) are then combined (Equation 5.1) to yield
the induced orientational anisotropy function, r(t) (Figure 5.5b):

1.. (t) - 11(t)
Ill (t) + 21,0)

 

r(t) = (5-1)

For chromophores in an unrestricted environment such as solution, the initial anisotropic
orientational distribution can relax to a random orientational distribution following
excitation. When the motion of the chromophore is restricted, however, (e. g. by
attachment to a lipid membrane), the initial photo-selected anisotropic distribution cannot
relax to a fully random orientational distribution. For chromophores in a restricted
environment, the orientational anisotropy at times long after excitation does not decay to
zero. The motion of the chromophore in a constrained environment can be expressed
using a hindered rotor model developed by Lipari and Szabo.3O In this model the
orientational anisotropy decays with time as the chromophore orientational distribution

re-randomizes to the extent possible (Equation 5.2):

 

 

r(t) = r(oo) + (r(0) — r(oo)) exp(—t/2'HR) (5.2)
r(w) = 0.5x(cose (1 +coso )) (5.3)
r(O) 0 0

79,2
THR ‘ 24Dw (5-4)

148

2000-

1500-

1000-

intensity [counts]

 

 

I“(t)

1,0)

(a)

 

0.15 -'

10000

0.10- - ‘

r(t)

0.05 -

 

If I I
20000 30000 40000

time [ps]

 

 

I
10000

I I
20000 30000
time [ps]

Figure 5.5 (a) Time-resolved emission intensities for polarization parallel (111(1)) and
perpendicular (Ii(t)) to the excitation polarization for pyrene labeled DPPC in DPPC
vesicles (0 mol % thio-PEG-lipid) in water. (b) Orientational anisotropy function
obtained from data shown in (a). The ﬁt to the data is shown as a solid line and residuals

are distributed around zero.

149

 

where r(O) and r(oo) are initial and inﬁnite time anisotropies, respectively. The zero-time
anisotropy depends on the angle between the excited and emitting transition dipole
moments of the chromophore. The inﬁnite time anisotropy is associated with the degree
of motional restriction imposed on the chromophore by its immediate environment
(Equation 5 .3). The decay time constant THR is related to the semi-angle of the cone, 60,
within which the chromophore is constrained, and is related to the “wobbling” diffusion
coefﬁcient, Dw, which describes the motion of the chromophore about its tethering bond

(Equation 5.4).

We show in Table 5.1 the quantities r(0), r(OO), and THR obtained for tethered
pyrene in lipid aggregates as a function of thio-PEG-DPPE concentratiOn. From the
quantities r(0) and r(oo) we can calculate the angular conﬁnement of the chromophore, 60,
using Equation 5.3. Dw values, which are indicative of chromophore local freedom, can
be extracted from Equation 5.4. The 60 values are associated with the organization of the
molecules in the lipid bilayer or in the micelle. These values are shown as a function of
thio-PEG-DPPE concentration in Figure 5.6, revealing a change in the organization of
lipid assembly upon addition of thio-PEG-DPPE. We note from this plot that low
concentrations (up to 10 mol %) of thio-PEG-DPPE in the liposomes do not have a large
effect on the organization of the molecules in the lipid bilayer. At low concentrations of
thio-PEG-DPPE, 00 values range from 30° to 38°, indicating a relatively conﬁning

environment. As the concentration of thio-PEG-DPPE increases, there is progressively

150

 

 

 

30

d

T I T

60

1

r I
80 100

thio-PEG-lipid concentration [mol %]

Figure 5.6 Angular conﬁnement of pyrene labeled DPPC represented as 60 versus the
concentration of thio-PEG-DPPE in the studied systems.

151

 

disorder, as shown by the increase in cone angle accessible to tethered pyrene. The
maximum level of disorder is reached at thio-PEG-DPPE concentrations as high as 70
mol %, where 60 reaches 60° and remains at this level for pure thio-PEG-DPPE micelles.
Examining Dw as a function of thio-PEG-DPPE concentration (Figure 5.7) shows that the
lamellar phase, present at low thio-PEG-DPPE concentrations, does not produce
signiﬁcant changes in Dw, and the transition from lamellar to micellar phase above 10
mol % thio-PEG-DPPE, where Dw increases with thio-PEG-DPPE concentration. This
increase in Dw values with thio-PEG-DPPE concentration corresponds physically to a
decrease in the viscosity of the chromophore local environment. For liposomes with thio-
PEG-DPPE below 10 mol %, the 15 MHz value of Dw suggests a relatively viscous
environment. With increasing thio-PEG-DPPE concentration, the chromophore
environment becomes signiﬁcantly less viscous for tethered pyrene, with Dw increasing
to 90 MHz for micelles composed of 100 mol % thio-PEG-lipid; a factor of six change in
Dw. The variations in local freedom of motion reﬂected by changes in Dw correlate well
with changes in 60, likely corresponding to an organizational change in the headgroup

region as a consequence of the presence of the polymer chain terminal groups.

We have also used the anisotropy decay dynamics of perylene to interrogate the
lipid acyl chain region(s) of our lipid assemblies. We recover single component
ﬂuorescence lifetimes for this chromophore, indicating that perylene resides in a
homogenous environment. In contrast to the pyrene ﬂuorescence lifetime data, the
perylene ﬂuorescence lifetime changes only slightly (~500 ps) with changes in the thio-
PEG-DPPE concentration (Figure 5.8). This ﬁnding could be the result of two factors;

the comparative insensitivity of the perylene ﬂuorescence quantum yield to the presence

152

 

100-

80-

60-

[MHZ]
6.4

W

40-

D
I-Gi

20 -
ii- ‘3
0 f ' I ' I ' r ' I
0 20 40 60 80 100
thio-PEG-lipid concentration [mol %]

 

 

' ‘r

Figure 5.7 “Wobbling” diffusion coefﬁcient (Dw) plotted versus the concentration of
thio-PEG-lipid in the lipid aggregates.

153

 

6600 ~

6500-: .§+ O

8%
.85

lifetime [ps]

6200 3

6100-

6000-

 

 

5900

' I ' I I I I I . I
0 20 40 60 80 100
thio-PEG-lipid concentration [mol %]

d

Figure 5.8 Fluorescence lifetime of perylene as a function of thio-PEG-DPPE
concentration in the lipid aggregates.

154

of oxygen, and the likelihood that, regardless of thio-PEG-DPPE concentration, perylene
resides in an environment that is comparatively well isolated from the terminal PEG

moieties.

The reorientation dynamics of perylene are well understood in a variety of
environments.”19 Perylene can reorient either as a prolate rotor or an oblate rotor. The
rotor shape of this chromophore can be deduced from the functional form of the
orientational anisotropy decay. A single exponential anisotropy decay indicates that
perylene reorients as a prolate rotor (DJ: > Dy = D:) and a double exponential anisotropy
decay points to oblate rotor reorientation (D: > Dx = Dy). We observe two-component
exponential anisotropy decays for all lipid aggregates reported here (Figure 5.9),
indicating perylene reorients as an oblate rotor. For an oblate rotor the induced

orientational anisotropy function, r(t), is given by:
r(t) = 0.1exp(—(2D, + 40, )z)+ 0.3exp(—6D,i) (5.5)

Where the terms D,- are the Cartesian components of the rotational diffusion constant.
Reorientation times obtained from double exponential anisotropy decays are related to

Cartesian components of rotational diffusion coefﬁcients:3 '

 

2' : 5.6
1 6 Di ( )
— 1 (5 7)
2 ZDX +4D: '

155

 

 

 

 

 

 

 

 

2,00- I”(t) (a)
0
_. 20004
33
a 1
a 1500-
.; 11(1)
3 1000-
8
.E
500-
,J
I T T r
2000 4000 6000 8000
time [ps]
1
(b)
0.10-1-
1 o
A 0.05- '3, -:..' ‘
CL 5;
In
I 1151" l 1‘ 1 '
000%.}1" It I1II1II'II'1I'FIIII'IH 1ru’1lll1i... Ii 1
' 1 1. 1
M1. l 1" 5.
. 11
-0.05 . , , . ﬁ f
2000 4000 6000 8000
time [ps]

Figure 5.9 (a) Time-resolved emission intensities for polarization parallel (111(0) and
perpendicular (11(0) to the excitation polarization for perylene in DPPC vesicles (0 mol
% thio-PEG-lipid) in water. (b) Orientational anisotropy function obtained from data
shown in (a). The ﬁt to the data is shown as a solid line and residuals are distributed

around zero.

156

The shorter time constants recovered from the perylene anisotropy decay data are shown
as a function of the concentration of thio-PEG-DPPE (Figure 5.10). These data are
largely independent of the amount of polymer-bound lipids present in the lipid
assemblies. The time constants are comparatively short (ca. 350 ps) relative to perylene
reorientation constants reported for liposomes of similar size but comprised of a shorter
lipid (ca. 1000 ps for perylene in DMPC).19 For DMPC vesicles larger than ca. 1 pm in
diameter, perylene exhibited a 400 ps reorientation time. In another report, reorientation
time constants on the order of 200 and 270 ps were measured for perylene reorienting in
decanoate vesicles and micelles, respectively.l7 Our data indicate that perylene
reorientation times on the order of 350 ps for DPPC vesicles and micelles can be
attributed to the type of the phospholipid head group and to the length of the lipid acyl

chain.

The Cartesian components (Dx, D5) of the rotational diffusion coefﬁcient D
(Figure 5.11a) and their ratios D/Dx (Figure 5.11b) (~15) remain largely unchanged with
variations in the concentration of thio-PEG-DPPE in the lipid assemblies. The lack of
signiﬁcant change in the diffusion coefﬁcient components suggests a constancy of the
environment probed by perylene. Based on calculated Cartesian components of the
rotational diffusion coefﬁcient we can estimate the viscosity of the immediate
environment of perylene. The most appropriate model for this estimation is the modiﬁed
Debye-Stokes-Einstein relation:

_1_ :12: (5 8)
6D kBTS '

157

600 1
550 d

500- §

450 -

400- 1
350 - 1 . §

300-
. . 1
250-

200-

tom [ps]

 

 

150-4

 

 

IOO

l ' l ' l ' I ' I

O ' 20 4O 60 80 100
thio-PEG-lipid concentration [mol %]

Figure 5.10 Reorientation time constants tom obtained for perylene as a function of thio-
PEG-DPPE concentration in the lipid aggregates.

158

1600- O Dx

 

 

 

' 0 D2 (a)
1400-
1200-
; 1000-
\ 800- %
QN -
*5 600- Q
X . O
Q 400-
200-
0. 9.. 'r . Q 6 . .
o 20 40 60 80 100
thio-PEG-lipid [mol %]
25-
(b)
20- 1'
.9
E
>< 15 15%
Q .
SN 1
10— _
5

 

 

I ' r ' I ' l ' I ' I
0 20 4O 60 80 IOO
thio-PEG-lipid concentration [mol %]

Figure 5.11 Cartesian components of diffusion coefﬁcient (a), Dx and D_., and their ratio
(b), D: / Dx, obtained for perylene in lipid aggregates plotted as a function of thio-PEG-
DPPE concentration in the lipid structures.

159

where I] is the viscosity of the medium, V is the hydrodynamic volume of the
chromophore, T is the temperature, k3 is the Boltzmann constant, f is associated with
frictional interactions between chromophore and the medium,32 and S is a factor related to
the shape of the chromophore.33 The average viscosity calculated for all the samples is
on the order of 4.5 cP assuming S = 0.7 and f = 1,'6 a value consistent with results for this
chromophore in large DMPC vesicles.'8"9 We recognize that the model used for the
estimation of viscosity is intended for use in bulk solutions, and its utility for ﬂuid lipid
systems is questionable. While it may be tempting to compare these data directly to
results for perylene in n-alkane solvents, we refrain from making this comparison
because of the intrinsic conﬁnement of the acyl chains of phospholipids. Such a
structural limitation is sufﬁcient to preclude direct comparison to reorientation data for

perylene in bulk solvents.

160

f

Conclusions

The long-term goal of this work is to construct a robust, biomimetic interface.
The potential applications of such an interface are signiﬁcant both in terms of numbers
and speciﬁc type. The work we present here deals with changes in the lipid bilayer
organization when other lipids that are not structurally identical are present. Owing to
the complexity of biological membranes, it is important to evaluate structural and
organizational changes in model systems as a function of composition. We have
synthesized a lipid conjugated with a hydrophilic polymer in its headgroup. The speciﬁc
purpose of this structurally modiﬁed lipid is to provide anchor points to Au surfaces,
thereby producing a chemically robust bilayer. The intended role of the hydrophilic
polyethylene glycol polymer is to accommodate the hydrophilic portion of

transmembrane protein between the lipid bilayer and the transducer surface.

Deposition of our mixed composition lipid structures was to be accomplished by
lipid fusion of mixed lipid vesicles on a gold surface would be the simplest way to build a
supported lipid bilayer on transducer surface. The failure to produce surface-attached
lipid bilayer structures led to the need to understand the solution phase lipid assemblies
that were presented to the interface. TEM images of the liposomes composed of mixed
lipids revealed a dependence of the liposome structure on the concentration of modiﬁed
lipids in the lipid assembly (Figure 5.1). We have interrogated the dynamics of
chromophores contained in these lipid assemblies and our ﬁndings indicate that, in
general, the addition of thio-PEG-DPPE to DPPC lipid assemblies gives rise to changes
in organization that are most prominent in the region of the lipid head group and the

adjacent PEG-containing region. Reorientation dynamics of perylene show that the

161

organization of the lipid bilayer acyl chain region is affected to a limited extent by
changes in the microscopic morphology of the lipid assemblies. Our data point to the
importance of lipid assembly morphology in determining the manner in which the
assemblies interact with interfaces, and at the same time provide some insight into how
changes in molecular scale organization depend on composition for such systems. We
anticipate that the results presented in this paper will ultimately allow for the construction
of stable, supported planar lipid bilayer structures with adjustable molecular-scale

organization and dynamics.

162

 

 

Literature Cited

(1) Hristova, K.; Kenworthy, A.; McIntosh, T. J. Macromolecules 1995, 28, 7693-9.

(2) Belsito, S.; Bartucci, R.; Montesano, G.; Marsh, D.; Sportelli, L. Biophysical
Journal 2000, 78, 1420-1430.

(3) Kenworthy, A. K.; Simon, S. A.; McIntosh, T. J. Biophysical Journal 1995, 68,
1903-20.

(4) Leal, C.; Roegnvaldsson, S.; Fossheim, S.; Nilssen, E. A.; Topgaard, D. Journal
of Colloid and Interface Science 2008, 325, 485-493.

(5) Edwards, K.; Johnsson, M.; Karlsson, G.; Silvander, M. Biophysical Journal
1997, 73, 258-266.

(6) Belsito, S.; Bartucci, R.; Sportelli, L. Biophysical Chemistry 2001, 93, 11-22.

(7) Rovira-Bru, M.; Thompson, D. H.; Szleifer, I. Biophysical Journal 2002, 83,
2419-2439.

(8) Johnsson, M.; Edwards, K. Biophysical Journal 2003, 85, 3839-3847.
(9) Edwards, K.; Almgren, M.; Bellare, J .; Brown, W. Langmuir 1989, 5, 473-8.

(10) Silvander, M.; Karlsson, G.; Edwards, K. Journal of Colloid and Interface
Science 1996, I 79, 104-13.

(11) Vinson, P. K.; Talmon, Y.; Walter, A. Biophysical Journal 1989, 56, 669-81.

(12) Walter, A.; Vinson, P. K.; Kaplun, A.; Talmon, Y. Biophysical Journal 1991, 60,
1315-25.

(13) Edwards, K.; Almgren, M. Langmuir 1992, 8, 824-32.

(14) F romherz, P. Chemical Physics Letters 1983, 94, 259-66.

(15) Lasic, D. D. Biochimica et Biophysica Acta, Biomembranes 1982, 692, 501-2.
(16) Jiang, Y.; Blanchard, G. J. Journal of Physical Chemistry 1994, 98, 6436-40.

(17) Stevenson, S. A.; Blanchard, G. J. Journal of Physical Chemistry B 2006, 110,
13005-13010.

(18) Koan, M. M.; Blanchard, G. J. Journal of Physical Chemistry B 2006, 110,
16584-16590.

163

(19)

Lapinski, M. M.; Blanchard, G. J. Chemistry and Physics of Lipids 2008, 153,

130-137.

(20)
(21)
(22)

(23)

Karpovich, D. S.; Blanchard, G. J. Langmuir 1996, 12, 5522-5524.
Dominska, M.; Krysinski, P.; Blanchard, G. J. Langmuir 2008, 24, 8785-8793.
Mazur, M.; Krysinski, P.; Blanchard, G. J. Langmuir 2005, 21 , 8802-8808.

Krysinski, P.; Zebrowska, A.; Michota, A.; Bukowska, J .; Becucci, L.; Moncelli,

M. R. Langmuir 2001, I 7, 3852-3857.

(24)

Krysinski, P.; Zebrowska, A.; Palys, B.; Lotowski, Z. Journal of the

Electrochemical Society 2002, I 49, E189-E194.

(25)

Kalyanasundaram, K.; Thomas, J. K. Journal of the American Chemical Society

1977, 99, 2039-44.

(26)
(27)

(28)
8346.

(29)
(30)
(31)
(32)

(33)

Dong, D. C.; Winnik, M. A. Canadian Journal of Chemistry 1984, 62, 2560-5.
Karpovich, D. S.; Blanchard, G. J. J. Phys. Chem. 1995, 99, 3951-8.

Tulock, J. J .; Blanchard, G. J. Journal of Physical Chemistry A 2000, 104, 8341-

Ding, L.; Fang, Y.; Blanchard, G. J. Langmuir 2007, 23, 11042-11050.

Lipari, G.; Szabo, A. Biophysical journal 1980, 30, 489-506.

Chuang, T. J .; Eisenthal, K. B. Journal of Chemical Physics 1972, 5 7, 5094-7.
Hu, C.-M.; Zwanzig, R. Journal of Chemical Physics 1974, 60, 4354-7.

Perrin, F. Journal de Physique et le Radium 1936, 7, 1-11.

164

 

 

CHAPTER 6

CONCLUSIONS

The long-term goal of this work is to construct supported lipid bilayers which are
capable of accommodating transmembrane proteins and maintaining their biological
activity and examine their properties. There are several important parameters that have to
be considered when designing such an interface on solid substrates. One of the
requirements is that the lipid bilayer must be in a sufﬁciently ﬂuid state to allow diffusion
of the incorporated proteins. Another requirement is that both sides of the lipid
membrane must be surrounded by aqueous compartments in order to accommodate
hydrophilic portions of the inserted proteins. In our work we have focused on the
organization of a bilayer structure absent protein incorporation. An important aspect of
the construction and interrogation of the supported biomimetic systems is the
incorporation of “reporter” molecules that are capable of probing the local environment
of an interface. For information from these reporter molecules to be useful, they must be
localized at predetermined locations within the biomimetic structure. One way to

accomplish this goal is to tether the probe molecule to the supporting interface.

We have designed a series of self-assembled monolayers (SAMs) containing
pyrene covalently bound to the gold, indium-doped tin oxide (ITO) or silica surfaces.
Electrochemical investigations show that tethered pyrene itself does not form well
organized and insulating SAMs, but rather enjoys signiﬁcant molecular freedom. The
evaluation of electron transfer kinetics in Chapter 2 reveals that the monolayers on gold

are better organized than those formed on ITO. The charge transfer kinetics particularly

165

on gold also show that codeposition of tethered pyrene with aliphatic adsorbates
improves the organization of SAM by imposing some structural conﬁnement on tethered
pyrene. However, even the two component SAMs are not sufﬁciently compact to
mediate the energetics of pyrene redox chemistry. Time-resolved ﬂuorescence
anisotropy decay data on the structural conﬁnement, organization, and motional freedom
of tethered pyrene within these monolayers are presented in Chapter 3. These data show
that chromophores bound very close to the substrate surface are characterized by
molecular motion that is constrained to only a limited extent even when aliphatic
coadsorbates are present. We attribute this effect to a relatively small number of
attractive interactions with neighboring molecules. The chromophores tethered further
away from the substrate exhibit more hindered motion, consistent with inﬂuence from

neighbor-neighbor intramolecular interactions.

We have also used Langmuir-Blodgett (LB) and Langmuir-Schaefer (LS)
methods to construct planar lipid membranes on gold and silica substrates. Our model
system in Chapter 4 was a bilayer comprised of a bottom leaﬂet of fatty acids and fatty
acid analogues and a top leaﬂet of phospholipids. Some of the fatty acid analogues used
in this work were labeled with pyrene. We employed this tethered chromophore to probe
the organization and electron transfer characteristics of the monolayer and bilayer.
Electrochemical data identify a relatively slow and almost distance-independent electron
transfer from tethered pyrene to the electrode surface. These data point to relatively well
organized monolayers and bilayers. Time-resolved ﬂuorescence anisotropy decay
measurements indicate that the tethered pyrene resides in a moderately constrained

environment, as detected by the small volume accessible for the chromophore to rotate

166

 

within, and the slow rate of motion. These experimental results suggest subtle
differences between monolayer and bilayer organization that are expected based on

capping properties of the lipid overlayer.

To provide a hydrophilic compartment for accommodation of the hydrophilic
portion of a transmembrane protein, we have synthesized a lipid conjugated with a short
thiol-terrninated polyethylene glycol (PEG) chain in its headgroup. The speciﬁc purpose
of this thiolated lipid was to provide anchor points to Au surfaces, thereby producing a
chemically robust bilayer. Deposition of a lipid bilayer with a hydrophilic cushion
between the bilayer and solid substrate was to be accomplished by vesicle fusion on a
gold surface. However, the failure to produce surface-attached lipid bilayer structures led
to the need to understand the solution phase lipid assemblies that were presented to the
interface. Chapter 5 deals with a dependence of the liposome structure on the
concentration of modiﬁed lipids in the lipid assembly. We have interrogated the
dynamics of chromophores contained in these lipid assemblies and our ﬁndings indicate
that, in general, the addition of thio-PEG-lipid to phospholipid assemblies gives rise to
changes in organization that are most prominent in the region of the lipid head group and
the adjacent PEG-containing region. The organization of the lipid bilayer acyl chain
region has been shown to be affected to a limited extent by changes in the microscopic

morphology of the lipid assemblies.

In general, the information we gained from this work provides a foundation for
investigating supported lipid bilayer structures. Solution phase lipid aggregate data point
to the importance of lipid assembly morphology in determining the manner in which the

assemblies interact with solid substrates, and at the same time provide some insight into

167

how changes in molecular scale organization depend on composition for such systems.
We anticipate that the results we have presented here will ultimately allow for the
construction of stable, supported planar lipid bilayer structures with adjustable molecular-

scale organization and dynamics.

168

   

mmullInlllllulummImlwuullilfllll“