unzmndmmﬁ human.

.Nmrwﬁ
kumiazmm was .w.....=.
Whammgih.

‘31..“ we. :

.40
.
A

.. ‘1. Awe-R5
‘44». :1. r a.”
.4153. native. . a

GNU;

 

 

FEE {all}

 

 

gf,
290‘? LIBRARY
Michigan State
University

 

 

This is to certify that the
dissertation entitled

STRUCTURE-FUNCTION RELATIONSHIP OF BORON—DOPED
DIMAOND THIN-FILM ELECTRODES AND APPLICATION FOR IN
VITRO AMPEROMETRIC MEASUREMENTS

presented by

Shihua Wang

has been accepted towards fulﬁllment
of the requirements for the

Ph. D degree in Chemistry

 

 

- Jojo/{A we

Méjor Professor’s Signature

5’5“»960‘8'

 

 

Date

MSU is an affinnative-action. equal-opportunity employer

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

 

Cb
my. '31 ‘1

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5108 KzlProlecc8-Pres/CIRCIDaIeDue.indd

 

STRUCTURE-FUNCTION RELATIONSHIP OF BORON-DOPED DIMAOND
THIN-FILM ELECTRODES AND APPLICATION FOR IN VITRO
AMPEROMETRIC MEASUREMENTS
By

Shihua Wang

A DISSERTATION
Submitted to
Michigan State University
In partial fulﬁllment of the requirements
For the degree of
DOCTOR OF PHILOSOPHY
Department of Chemistry

2008

A bstract

STRUCTURE-FUNCTION RELATIONSHIP OF BORON-DOPED DIMAOND
THIN-FILM ELECTRODES AND APPLICATION FOR IN VITRO
AMPEROMETRIC MEASUREMENTS
By
Shihua Wang

In this dissertation, the structure-function relationship of boron-doped
diamond thin-ﬁlm electrodes was studied. This includes assessing the factors that
inﬂuence the properties of boron-doped diamond thin-ﬁlm electrodes as well as
investigating the electrical and electrochemical mechanism of these electrodes. In
addition, the application of boron-doped diamond microelectrode for in vitro
amperometric measurements was also studied.

The electrical and electrochemical properties of boron-doped diamond thin-
ﬁlm are inﬂuenced by many factors such as growth conditions, grain size, doping
level, impurities (e. g., grain defects, sp2 carbon and adsorbed hydrogen) and surface
chemistry. Growth condition is the primary determinant of the surface morphology of
BDD thin-ﬁlm, although both microcrystalline and nanocrystalline diamond thin-ﬁlm
can be deposited in similar hydrogen-rich gas mixtures. Microcrystalline BDD thin-
ﬁlm has different Raman spectrum and carrier mobility from nanocrystalline ﬁlm but
their electrochemical properties are very similar in macroscopic level. Using CP-AF M
and SECM, we found that the electrical conductivity and electrochemical activity for
microcrystalline BDD thin-ﬁlm are different from those for nanocrystalline ﬁlm.

Microcrystalline BDD thin-film possesses lower fraction of active regions but more

active individual sites compared with nanocrystalline BDD thin-ﬁlm.

Doping level inﬂuences the surface morphology of BDD thin-ﬁlm. More
importantly, CP-AFM and SECM measurements illustrated that both the fraction of
electrically conductive regions and the fraction of electrochemically active regions
increase as the doping level increases and this is true for both microcrystalline and
nanocrystalline BDD ﬁlms.

Surface termination has a signiﬁcant inﬂuence on the properties of CVD
diamond. For example, the fastest electron-transfer kinetics for Fe(CN)63'/4' has been
recorded at hydrogen-terminated BDD electrode but very sluggish reaction has been
recorded at oxygen-terminated BDD electrode. Electrochemical measurements show
that O-terminated BDD electrode has a little larger background current that H—
terrninated BDD electrode. The reason for this might be the O-terminated surface is
more hydrophilic and this may increase the dielectric constant of H20 at the surface.
CP-AF M and SECM studies Show that both the fraction of electrically-active regions
and the fraction of electrochemically-active regions are greater for H-terminated than
for O-terminated BDD ﬁlms.

AS a preclusive work, the BDD microelectrode has been used to study the
denervation of capsaicin on mesenteric blood vessels by monitoring the release of
norepinephrine. The distribution of sensory and sympathetic nerves was visualized
using glyoxylic acid ﬂuorescence histochemistry and anti-TH, anti-substance P and
anti-CGRP immunohistochemistry. The effects of 7 neuro-phamarceutical drugs on
the determination of norepinephrine were also studied. We failed to Show any
evidence for the local communication between the sensory and sympathetic nerves by

using exogenous substance P.

Com/right by
SHIHUA WANG

2008

ACKNOWLEDGEMENTS

The author wishes to express his sincere appreciation to Professor Greg M.
Swain for his encouragement and support for this thesis research. Thanks are also due
to the other members of the author’s guidance committee: Professor Gary J.
Blanchard, Professor Merlin L. Bruening and Professor Tim Hogan. I appreciate your
willingness to serve on my guidance committee for 5 years and thank you for all the
comments and help with my research and study. The author also thanks Professor
James J. Galligan for his guidance on the blood vessel measurements and aspects of
neuropharmacology.

The author is grateful to Dr. Anne E. Fischer and Dr. Jason A. Bennett for
their introducing me to polycrystalline boron-doped diamond. A special note of
thanks is extended to Dr. Jason Stotter and Doug Knigge for the training on AF M and
electrochemisu'y. Thanks also goes to Dr. Prema V. Sonthalia and Dr. Jesse
Seegmiller for assistance with the SECM instrumentation. The author owes a debt of
gratitude to Dr. Jinwoo Park for his training and knowledge sharing on tissue
dissection and pharmacological measurements. Thanks to all Swain group members:
Show-san, Veronika, Grace, Yang, Elizabeth, Audrey, Luther, Martin, Dasha, Iso,
Aihua, Hua, Yingrui, the Matts, Ayten, Jia, and the Galligan group members: Hui,
Xiaochun, Xiaoling, Sandra and Stacie. Thank you all for your help and support!

Special thanks goes to Hua Dong for cooperation in “drug effect” study and

Yingrui Dai for the XPS data.

The author acknowledges the National Science Foundation for support under
the Nanoscale Interdisciplinary Research Team (NIRT, CTS-0210366) program
(Michigan State, Vanderbilt, and Purdue Universities) for the ﬁnancial assistance

which made this thesis research possible.

vi

TABLE OF CONTENTS

LIST OF TABLES .......... . .................................................................. X
LIST OF FIGURES .......................................................................... XII
Chapter 1
Introduction ...................................................................................... 1
1.1 Diamond and the conducting diamond electrode ................................... 1
1.1.1 Diamond .......................................................................... 1
1.1.2 Conducting diamond (Synthetic diamond) .................................. 3
1.2 Electrochemical properties and application of boron-doped diamond . 5
1.2.1 Electrolytic waste water treatment ............................................ 6
1.2.2 Electroanalysis .................................................................. 7
1.3 Motivation and research questions ................................................ 23
1.4 Research goals and Objectives ....................................................... 27
1.5 Dissertation outline .. ..................................................................................... 27
1.6 Reference ............................................................................... 29
Chapter 2
Experimental Section ......................................................................... 40
2.1 Preparation and characterization of boron-doped diamond thin-ﬁlm
electrodes ............................................................................ 40
2.1.1 Chemical vapor deposition conditions ...................................... 40
2.1.2 Characterization of Boron-doped diamond thin ﬁlms ..................... 43
2.1.3 Surface termination ............................................................ 45
2.1.4 Boron-doped diamond microelectrodes ..................................... 50
2.2 Scanning Electron Microscopy (SEM) ............................................ 54
2.3 Raman Spectra ........................................................................ 55
2.4 Electrochemical Characterization ................................................... 56
2.4.1 Planar boron-doped diamond electrode ..................................... 56
2.4.2 Microelectrode .................................................................. 57
2.5 Conductive-Probe Atomic Force Microscopy (CP-AF M) ....................... 59
2.6 Scanning Electrochemical Microscopy (SECM) ................................. 61
2.7 Reagents for electrochemical measurements ..................................... 62
2.8 Experimental methods for Chapter 7 .............................................. 63
2.8.1 Animals ......................................................................... 63
2.8.2 In vitro electrochemical and diameter measurement system ............ 64
2.8.3 Fluorescence histochemistry ................................................. 68

vii

2.8.4 Chemicals and drugs .......................................................... 70

2.8.5 Data analysis ................................................................... 71
2.9 Reference .............................................................................. 72
Chapter 3
Basic Properties of Boron-Doped Microcrystalline & Nanocrystalline
Diamond Thin- Film Electrodes ........................................................... 76
3.1 Introduction ........................................................................... 76
3.2 Scanning Electron Microscopy (SEM) Images ................................... 77
3.2.1 SEM images for microcrystalline boron-doped diamond thin
ﬁlms ............................................................................ 77
3.2.2 SEM images for nanocrystalline boron-doped diamond thin ﬁlms ...... 80
3.2.3 SEM images for boron-doped diamond microelectrodes ................ 83
3.3 Raman Spectra ........................................................................ 84
3.3.1 Raman spectra for microcrystalline boron-doped diamond ﬁlms
on Si substrates ................................................................. 84
3.3.2 Raman spectra for nanocrystalline boron-doped diamond ﬁlms
on Si substrates ................................................................ 89
3.3.3 Raman spectra for boron-doped diamond ﬁlms on Pt substrates ....... 92
3.4 Electrochemical Characterization .................................................. 93
3.4.1 Background CV for boron-doped diamond thin-ﬁlm on Si .............. 93
3.4.2 Electrochemical behavior of boron—doped diamond thin-ﬁlms
with different doping levels ................................................... 97
3.4.3 Electrochemical activities for planar boromdoped diamond thin-
ﬁlm on Si substrate with different terminations ......................... 107
3.5 The Effect of surface hydrogen on the properties of planar boron-
doped diamond thin-ﬁlm on Si substrate ...................................... 115
3.5.1 Contact angle measurement ................................................. 115
3.5.2 Four-point resistivity measurements ....................................... 117
3.5.3 Cyclic voltammetric characterization ...................................... 118
3.6 Conclusions .......................................................................... 121
3.7 Reference ............................................................................ 123
Chapter 4
Heterogeneous Electrical Conductivity of Polycrystalline Boron-Doped
Diamond Films ............................................................................... 128
4.1 Introduction ......................................................................... 128
4.2 CP-AF M images of nanocrystalline ﬁlms ....................................... 131
4.3 CP-AFM images of microcrystalline ﬁlms .................................... 135
4.4 CP—AFM images and topography ............................................... 139
4.5 Effect of bias voltage polarity ..................................................... 142
4.6 CP-AF M images and surface termination ....................................... 146
4.7 Special issues with CP-AFM measurements .................................... 150
4.8 Continuous imaging tunneling spectroscopy (CITS) images .................. 153
4.8.1 CITS measurements for microcrystalline boron-doped diamond
thin ﬁlms ....................................................................... 154
4.8.2 CITS measurements for nanocrystalline boron-doped diamond thin
ﬁlms ............................................................................ 158
4.9 Conclusions .......................................................................... 163

viii

4.10 Reference ............................................................................ 165

Chapter 5
Heterogeneous Electrochemical Activities of Polycrystalline Boron-Doped
diamond ﬁlms ................................................................................ 169
5.1 Introduction ........................................................................ 169
5.2 Preparation of scanning electrochemical microscopy measurements 172
5.2.1 Scanning electrochemical microscopy (SECM) ........................... 172
5.2.2 Tip cleaning and activation .................................................. 174
5.2.3 Calibration of the microelectrode ........................................... 176
5.2.4 Tip approach curves .......................................................... 180
5.3 SECM images for boron-doped diamond thin ﬁlms ........................... 183
5.3.1 Reproducibility of SECM imaging ........................................ 183
5.3.2 Effect of morphology of diamond thin-ﬁlm ............................... 185
5.3.3 Effect of boron doping level .............................................. 187
5.3.4 Effect of surface termination ............................................. 190
5.3.5 Effect of redox system ....................................................... 193
5.3.6 Comparison with other electrode materials ............................... 195
5.4 Mechanism of electron-transfer on polycrystalline boron-doped
diamond .............................................................................. 203
5.5 Conclusions .......................................................................... 212
5.6 Reference ............................................................................. 214
Chapter 6
The Physical and Electrochemical Properties of Two Types of Boron-Doped
Nanocrystalline Diamond Thin-Film Electrodes ........................................ 217
6.1 Introduction .......................................................................... 217
6.2 Results and discussion ............................................................... 218
6.2.1 Morphology .................................................................. 218
6.2.2 Raman Spectra ............................................................... 220
6.2.3 Electrochemical Characterization ........................................... 224
6.2.4 CP-AFM measurements ..................................................... 231
6.3 Conclusions ......................................................................... 233
6.4 Reference ........................................................................... 236
Chapter 7
Diamond Microelectrodes for Continuous Amperometric Recording at Blood
Vessel Surfaces ............................................................................... 239
7.1 Introduction .......................................................................... 239
7.2 Effects of drugs on the electrochemical responses of carbon ﬁber and
diamond microelectrodes .......................................................... 242
7.2.1 Norepinephrine and drug effect ............................................. 242
7.1.2 Results ......................................................................... 246
7.1.3 Discussion ................................................................... 261
7.2 Distribution of sympathetic and sensory nerves .............................. 268
7.2.1 Glyoxylic acid ﬂuorescence histochemistry .............................. 268
7.2.2 Immunohistochemistry. ..................................................... 270
7.3 Contraction and norepinephrine oxidation current at MA and MV ......... 277
7.3.1 Determination of norepinephrine .......................................... 277

ix

7.3.2 Sensory nerve denervation by capsaicin ................................. 278

7.3.3 Discussion .................................................................. 286

7.4 Reference ............................................................................. 288
Chapter 8

Conclusions ................................................................................... 294

LIST OF TABLES

Table 1.1 Properties of type Ila natural and CVD diamond ................ 2
Table 2.1 Growth conditions for microcrystalline and nanocrystalline
boron-doped diamond thin ﬁlms deposited on Si ....................... 43
Table 2.2 Film thickness, resistivity, carrier concentration, and carrier
mobility for typical microcrystalline and nanocrystalline boron-
doped diamond thin ﬁlms .................................................. 43
Table 2.3 Diamond ﬁlms with different doping levels used in this work ........ 45
Table 2.4 Growth conditions for microcrystalline boron-doped diamond
microelectrode on Pt wires .................................................. 52
Table 3.1 One-phonon diamond line position and width observed in
Raman spectra from microcrystalline diamond ﬁlms with
different doping levels ...................................................... 88
Table 3.2 One-phonon diamond line position and width in Raman
spectra ........................................................................ 92
Table 3.3 Cyclic voltammetric ABp values for different redox couples

(0.1 mmol/L analyte in 1 mol/L KCl) at microcrystalline
boron-doped diamond ﬁlms grown with different diborane
Concentrations. Measurements were made with one ﬁlm at
each doping level ........................................................ 100

Table 3.4 Cyclic voltammetric AEp’s for different redox couples (0.1
mmol/L analyte in 1 mol/L KCl) at nanocrystalline boron-
doped diamond films grown with different diborane
concentrations in the source gas. Measurements were made
with one ﬁlm at each doping level ...................................... 106

Table 3.5 XPS results for microcrystalline and nanocrystalline BDD thin
ﬁlms with different surface treatments ................................. 113

Table 3.6 Contact angles of microcrystalline boron-doped diamond thin
films with different doping levels and annealing
treatment .................................................................... 117

xi

Table 3.7 Surface resistivity of boron-doped diamond thin ﬁlms (11 = 6) ...... 118

Table 3.8 Cyclic voltammetric peak separation (AEP, mV) and peak
current (ILA) data for 0.1 mmol/L analytes in 1 mol KCl at 0.5
and 5 ppm boron-doped microcrystalline diamond thin-ﬁlm

electrodes before and after annealing in Ar ﬂow at 400°C ......... 121
Table 5.] Limiting current and calculated microelectrode radius

determined by different redox systems ................................. 180
Table 6.1 Roughness analysis of the MSU and NRL ﬁlms ..................... 219
Table 6.2 Peak separations of the MSU and NRL ﬁlms in different redox

systems ..................................................................... 227
Table 7.1 The structures and ﬁmctions for NE and seven studied drugs ....... 246

Table 7.2 Em values of seven drugs, Fe(CN)63"4' and Ru(NH3)63+’2+ on
three types of microelectrodes .......................................... 247

Table 7.3 The statistic data (n=3 for Em values and current changes of
two probes, Fe(CN)6' ['4 and Ru(NH3)(,+3/+2 before and after
capsaicin and prazosin at P5 5s, T650 carbon ﬁber and diamond
microelectrode .............................................................. 253

Table7.4 Effects of Idazoxan, PPADS and Cocaine on the

electrochemical properties of carbon fiber and BDD
microelectrodes probed by Fe(CN)63"4’ and Ru(NH3)63+’2+ .......... 257

xii

Figure 1.1

Figure 2.1

Figure 2.2

Figure 2.3

Figure 2.4
Figure 2.5

Figure 2.6

Figure 2.7

Figure 2.8

Figure 2.9

Figure 2.10

Figure 3.1

Figure 3.2

LIST OF FIGURES

Diamond cubic crystal structure .........................................

Schematic commercial MPCVD system (ASTex Inc., Lowell,
MA) used for growing microcrystalline and nanocrystalline

l

boron—doped diamond thin ﬁlms .......................................... 41
The tube ﬁimace heating system used for heat treating diamond
thin ﬁlms in Ar ............................................................... 49
Schematic of the assembly used to electrically sharpen the Pt
wire substrates ............................................................... 51
Diagram of the insulated boron-doped diamond microelectrode 54
Design of single compartment, 3 electrode electrochemical cell 57
Schematic of the electrochemical measurement with the
diamond microelectrodes ................................................. 58
Schematic of the CP-AFM system adapted to a NanoScope
IIIa scanning probe microscope ........................................... 60
Schematic of the home-made current-to-voltage converter ............ 60
Schematic of electrical connection between sample and tip for
CP-AF M experimental setup .. ........................................................... 61
Block diagram of the experimental set-up ............................... 65
SEM images for microcrystalline diamond thin ﬁlms grown by
microwave-assisted chemical vapor deposition with different
diborane levels in the source gas: (A) 10, (B) 5, (C) 1, (D) 0.5,
and (E) 0 ppm. The source gas composition was 0.5% CH4 +
99.5% H2, the plasma power was 1000 W, and the pressure was
45 torr. Growth time was 10 hours ....................................... 78

SEM images of boron-doped nanocrystalline diamond thin
ﬁlms grown by microwave-assisted chemical vapor deposition
with different diborane concentrations in the source gas: (A) 50,
(B) 10, (C) 5, (D) 1, and (E) 0 ppm. The source gas

xiii

Figure 3.3

Figure 3.4

Figure 3.5

Figure 3.6

Figure 3.7

Figure 3.8

Figure 3.9

Figure 3.10

Figure 3.11

Figure 3.12

composition was 1% CH4 and 5 % H2 in 94% Ar with diborane,
the plasma power was 800 W, and the pressure was 140 torr.
Growth time was 2 hours ................................................... 81

Bright-ﬁeld TEM image revealing the plan-view morphology
of the nanodiamond thin ﬁlm. The grain boundary contrast
reveals that the ﬁlm possesses a granular structure ............. 82

SAED pattern from the nanodiamond thin ﬁlm. Sharp Bragg
reﬂections are visible indicating good crystallinity ..................... 83

SEM images of a boron-doped diamond microelectrode grown
on a Pt wire and insulated using polypropylene. The white line
indicates the bounarybetween polypropylene coating and
exposed diamond ﬁlm ...................................................... 84

Raman spectrum for a highly boron-doped microcrystalline
diamond ﬁlm on Si, deposited from a 10 ppm diborane source
gas mixture .................................................................. 85

Raman spectra for microcrystalline diamond ﬁlms grown with
different source gas diborane levels. The greater the source gas
diborane concentration, the higher the doping level .................... 88

Relationship between peak-to-baseline (PfB) ratio of Raman
spectrum and the diborane concentration in the source gas
mixture for microcrystalline diamond ﬁlms. The P/B ratio is
deﬁned as the ratio of the intensity of the peak at 1332 cm"1
and the intensity of the baseline at the peak position ................... 89

Raman spectra for nanocrystalline diamond ﬁlms grown with
different source gas diborane levels. The greater the source gas
diborane concentration, the higher the doping

level ........................................................................... 90

Raman spectra for four boron-doped microcrystalline diamond
ﬁlms deposited on Pt wires. The diborane concentration in the
source gas was 4 ppm ....................................................... 93

Background cyclic voltammetric i-E curves for planar diamond
films. Supporting electrolyte: 1 mol/L KC]. Scan rate: 0.1
Background cyclic voltammetric i-E curves for H- and O-
terminated (a) nanocrystalline and (b) microcrystalline
diamond ﬁlms. The curves were recorded in 1 mol/L KCl at a
scan rate of 0.1 V/s starting from -0.2 V (vs.
Ag/AgCl) 97

xiv

Figure 3.13

Figure 3.14

Figure 3.15

Figure 3.16

Figure 3.17

Figure 3.18

Figure 3.19

Figure 3.20

Figure 3.21

Figure 3.22

Figure 3.23

Cyclic voltammetric i-E curves for 0.1 mmol/L Fe(CN)63'/4' in
1 mol/L KC] for microcrystalline boron-doped diamond ﬁlms

grown with different diborane concentrations .......................

Cyclic voltammetric i—E curves for 0.1 mmol/L Ru(NH3)63+’2+
in 1 mol/L KC] at microcrystalline boron-doped diamond ﬁlms
grown with different diborane concentrations in the source

gas ..........................................................................

Cyclic voltammetric i-E curves for 0.1 mmol/L IIC162'/3' in 1
mollL KC] at microcrystalline boron-doped diamond ﬁlms
grown with different diborane concentrations in the source

gas ...........................................................................

Cyclic voltammetric i-E curves for 0.1 mmol/L MV in 1 mol/L
KC] at microcrystalline boron-doped diamond ﬁlms grown
with different diborane concentrations in the source

gas ...........................................................................

Cyclic voltammetric i-E curves for (a) 0.1 mmol/L Fe(CN)63'/4',
(b) 0.1 mmol/L Ru(NH3).3*’2+, (c) 0.1 mmol/L 110.243; and (d)
0.1 mmol/L methyl viologen with 1 mol/L KC] at
nanocrystalline diamond thin ﬁlm electrodes grown with
different diborane concentrations in the source

gas ..........................................................................

Cyclic voltammetric i-E curves for 1 mmol/L Fe(CN)63'/4' in 1
mol/L KC] at 0.1 V/s for 10 ppm micro- and nanocrystalline

BDD thin-ﬁlm electrodes. Both electrodes were H-terminated

Cyclic voltammetric i-E curves for 1 nnnol/L Ru(NH3)63+/2+ in
1 mol/L KC] at 0.1 V/s for 10 ppm micro- and nanocrystalline

BDD thin-ﬁlm electrodes. Both electrodes were H-terminated

Cyclic voltammetric i-E curves for 1 mmol/L Fe(CN)63'/4' in 1

mol/L KC] at 0.1 V/s for 10 ppm H- and O-terminated

nanocrystalline BDD thin-ﬁlm electrodes ..........................

Cyclic voltammetric i-E curves for 1 mmol/L Ru(NH3)63+’2+ in
1 mol/L KC] at 0.1 V/s for 10 ppm H- and O-terminated

nanocrystalline BDD thin-ﬁlm electrodes ...........................

Cyclic voltammetric i-E curves for 1 mmol/L Fe(CN)63'/4' in 1

mol/L KC] at 0.1 V/s for H- and O-terminated microcrystalline

BDD thin-ﬁlm electrodes .................................................

Cyclic voltammetric i—E curves for 1 mmol/L Ru(NH3)¢53+’2+ in
1 mol/L KC] at 0.1 V/s for H- and O-terminated

microcrystalline BDD thin-ﬁlm electrodes ............................

XV

. 100

. 102

. 103

. 104

. 105

110

110

.111

.112

.114

.114

Figure 3.24 Electrochemical response of 0.5 and 5 ppm microcrystalline
BDD thin ﬁlm electrodes before and after annealing in Ar ﬂow
at 400°C for 30 min. a) 0.1 mmol/L Fe(CN)63'/4', b) 0.1
mmol/L Ru(NH3).,3"’2+ and c) 0.1 mmol/L Iron”: The
supporting electrolyte solution was 1 mol/L KC] and the scan
rate was 0.1 V/s ............................................................ 120

Figure 4.1 CP-AFM images (10 x 10 umz) of nanocrystalline BDD thin
ﬁlms grown with different source gas B2H6 concentrations (i.e.,
different doing levels): (A) 50, (B) 10, (C) l and (D) 0 ppm
B2H6. The bias voltage used for all images was +0.2 V vs. the
tip. The current ranged from 0 to 6.5 ILA. (Reprinted with
permission from J. Phys. Chem. 111, 3986 (2007). Copyright
(2007) American Chemical Society.) (From Ref.
49) ........................................................................... 132

Figure 4.2 CP-AFM images (10 x 10 umz) of microcrystalline BDD
thin films grown with different source gas B2H6
concentrations (i.e., different doing levels): (A) 10, (B) 5, (C) 1,
(D) 0.5 and (E) 0 ppm B2H6. The bias voltage used for all
images was +2.0 V vs. the tip. The current ranges from 0 to
0.65 uA. (Reprinted with permission from J. Phys. Chem. 111,
3986 (2007). Copyright (2007) American Chemical Society.)
(From Ref. 49) ............................................................. 136

Figure 4.3 SEM image of a gold-coated, tetrahedral-shaped silicon nitride
tip (Veeco NPG-20) with an apex radius of curvature less than
50 nm and a tip height of ~15-20 pm .................................. 137

Figure 4.4 Schematic representation of how the tip interacts with the (A)
rougher and faceted microcrystalline diamond ﬁlm and (B)
smoother nanocrystalline diamond ﬁlm. (Reprinted with
permission from J. Phys. Chem. 111, 3986 (2007). Copyright
(2007) American Chemical Society.) (From Ref.

49) ............................................................................ 138

Figure 4.5 Topographical (A) and conductivity maps (C) for a 0.5 X 0.5
pm2 region of a nanocrystalline BDD thin ﬁlm grown with 10
ppm B2H6 in the source gas. Corresponding line proﬁles
through the image (B and D) are also presented. The bias
voltage used to obtain the conductivity map was +0.2 V vs. the
tip. The Z-axis scales are 0 to 50 nm for height image and 0 to
0.65 uA for the current image. (Reprinted with permission
from J. Phys. Chem. 111, 3986 (2007). Copyright (2007)
American Chemical Society.) (From Ref.
49) ........................................................................... 140

Figure 4.6 Conductivity (A) and topographical maps (B) for a 2 X 2 umz

region of a microcrystalline BDD thin ﬁlm grown with 10 ppm
B2H6 in the source gas. The bias voltage was +0.2 V vs. the tip.

xvi

The current scale was 0 to 0.65 ILA .................................... 142

Figure 4.7 Conductivity maps (2 X 2 umz) obtained from CP-AF M images
of the same area of a nanocrystalline BDD thin ﬁlm grown
with 10 ppm B2H6 in the source gas at bias voltages of (A)
+0.2 V and (C) -O.2 V vs. the tip. Corresponding line proﬁles
of the same region are presented in (B) and (D). The Z-axis
scale was 0 to 0.65 LIA for both A and C. (Reprinted with
permission from J. Phys. Chem. 111, 3986 (2007). Copyright
(2007) American Chemical Society.) (From Ref. 49) ................ 143

Figure 4.8 CP-AF M section analysis of (l X 1 umz) of a nanocrystalline
BDD thin ﬁlm grown with 10 ppm B2H6 in the source gas at
bias voltages of (A) +0.2 , (B) +1.0 and (C) +2.0 V vs. the tip.
The measurements were made on the same area of the ﬁlm.
(Reprinted with permission from J. Phys. Chem. 111, 3986
(2007). Copyright (2007) American Chemical Society.) (From
Ref. 49) .................................................................... 145

Figure 4.9 CP-AFM images (1 X 1 umz) for “as grown” (a)
microcrystalline and (b) nanocrystalline BDD thin ﬁlms. A
bias voltage of 0.2 V vs. the tip was used for all images. The
current ranged ﬁ'om 0 to 0.65 uA ........................................ 147

Figure 4.10 CP-AF M images (l X 1 umz) for (a) hydrogen-terminated and
(b) oxygen-terminated nanocrystalline BDD thin ﬁlms. A bias
voltage of 0.2 V vs. the tip was used for all images. The current
ranged from 0 to 0.65 ILA for H-terminated ﬁlm and 0 to 0.16
pA for O-terminated ﬁlm ................................................ 148

Figure 4.11 CP-AF M images (1 X 1 umz) for (a) hydrogen-terminated and
(b) oxygen-terminated microcrystalline BDD thin ﬁlms. A bias
voltage of 2.0 V vs. the tip was applied for hydrogen-
tenninated ﬁlm and the currents at some sites were out of range
(> 3.25 uA). For the oxygen-terminated ﬁlm, a bias voltage of
0.2 V vs. the tip was applied and the current ranged from 0 to
0.65 uA ..................................................................... 149

Figure 4.12 CP-AFM images during 2 consecutive scans over the same
area of a 10 ppm boron-doped nanocrystalline diamond thin
ﬁlm (rehydrogenated). The bias voltage applied was +0.2 V vs.
the tip ........................................................................ 150

Figure 4.13 CP-AF M images as a function of the scan number for a 10 ppm
boron-doped nanocrystalline diamond thin ﬁlm. (A) 18t scan,
(B) 5th scan and (C) 10th scan. It shows that the CP-AFM
images changed after 5th scan. The bias voltage applied was
+0.2 V vs. the tip ........................................................... 151

xvii

Figure 4.14 SEM images for (A) an original CP-AFM tip and (B) a
damaged tip after one scan with an overloaded current. The
image was taken using a J SM-6300F (ﬁeld emission) scanning
electron microscope in the Center of Advanced Microscopy at
Michigan State University .............................................. 152

Figure 4.15 (A) CITS and (B) topographic images for microcrystalline
boron-doped diamond thin ﬁlm grown by CVD on highly
doped Si substrate. (C) is the superimposing image of (A) and
(B) ........................................................................... 155

Figure 4.16 Differential electrical properties of two regions of a
microcrystalline BDD thin ﬁlm. (A) a selected region on the
topographic map (white square) and its I-V characteristics
showing ohmic behavior, and (B) a selected region on
topographic map (white square) and its I-V characteristics
showing schottky behavior ............................................. 156

Figure 4.17 Differential electrical properties of two regions of a
microcrystalline BDD thin ﬁlm. (A) a selected region with high
electron emission on the topographic map and its I-V
characteristics, and (B) a selected region with low electron
emission on the topographic map and its I-V characteristics.
Scanned Size: 5 x 5 pm2 .................................................. 157

Figure 4.18 Different electrical properties of two representative regions on a
nanocrystalline boron-doped diamond thin ﬁlm. (A) a selected
region with ohmic behavior on the topographic image and its 1-
V characteristics, and (B) a selected region with Schottky
behavior on the topographic image and its I-V
characteristics .............................................................. 1 59

Figure 4.19 CITS images probed in contact mode with a conductive tip for
a nanocrystalline boron-doped diamond thin ﬁlm. (A) current
image, (B) topographic image, and (C) superimposing of (A)
and (B) ...................................................................... 160

Figure 4.20 Topographic and current images obtain by CTIS in contact
mode for a nanocrystalline highly boron-doped diamond thin
ﬁlm. (A) height image with a marked line selected for section
analysis, (B) height proﬁle for the selected line, (C) current
image with a marked line selected for section analysis, (D)
current proﬁle for the selected line ................................................. 162

Figure5.1 Genera] principle of the positive feedback tip-
generation/substrate-collection mode for SECM. The tip is near
a conductive substrate, feedback diffusion leads to a larger tip
current than when the tip is far away from the
substrate .................................................................. 174

xviii

Figure 5.2 Schematic of the Pt microelectrode probe tip used for imaging.
All Pt microelectrodes were commercial products (CH
Instruments Inc., Austin, TX). The nominal diameter of the
microelectrode (Pt wire) was ca. 2 pm sealed in glass sheath.
The end of the assembly was treated by a special polishing
procedure to make a ﬂat disk shape ................................... 175

Figure 5.3 Typical cyclic voltammetric i-E curve for the Pt microelectrode
after potential cycle cleaning in 0.1 mol/L HClO4 at 0.1 V/s.
All features of Pt are present indicative of a clean surface ........... 176

Figure 5.4 Cyclic voltammetric i-E curves for (a) 1 mmol/L Ru(NH3)63+/2+,
(b) 1 mmol/L Fe(CN)63"“', and (c) 1 mmol/LIrC16243', all in 1
mol/L KC], at a Pt microelectrode. Scan rate = 0.1 V/s ............ 179

Figure 5.5 SECM tip approach curves at the same position on a
nanocrystalline BDD thin film electrode in 1 mM
Ru(NH3)63+/2+ + 1 M KC1. The nominal diameter of the
microelectrode was 2 pm. The microelectrode was held at -0.3
V, and the diamond electrode potential was held at 0 V vs.
Ag/AgCl. i2 is the tip current, in... is the limiting current, and L
= d/a, where d is the distance between the tip and substrate, a is
the radius of the tip. The engaging current ratio for 1St to 4th
approach was 1.5, 1.8, 2.5 and 2.2, respectively ...................... 182

Figure 5.6 SECM images (25 x 25 tunz) of a microcrystalline BDD thin
ﬁlm grown with 10 ppm B2H¢ in the source gas in 5 mmol/L
Ru(NH3)63+’2+ + 1 mol/L KC]. The Pt diameter was 2 pm. The
tip potential was -0.3 V and the diamond electrode potential
was 0 V vs. Ag/AgCl. The two images were recorded
consecutively ................................................................ 1 84

Figure 5.7 SECM images (10 x 10 umz) of a (A) microcrystalline and (B)
nanocrystalline BDD electrode, both grown with 10 ppm B2116
in the source gas, in 5 mM Ru(NH3,)¢,3J"2+ + 1 M KC1. In all
cases, the Pt tip diameter was 2 pm. The tip potential was -0.3
V and the diamond electrode potential was 0 V vs. Ag/AgCl.
(Reprinted with permission from J. Phys. Chem. 111, 3986
(2007). Copyright (2007) American Chemical Society.) (From
Ref. 26) ................................................................... 186

Figure 5.8 SECM images (100 x 100 umz) of nanocrystalline BDD thin
ﬁlms grown with (A) 10 ppm, (B) 5 ppm and (C) 1 ppm B2H6
in the source gas mixture using 1 mmol/L Ru(NH3)(,3+’2+ + 1
mol/L KC] as the test redox system. The Pt tip diameter was 2
pm. The microelectrode potential was -O.3 V and the diamond
electrode potential was 0 V vs. Ag/AgCl. Images were
recorded in the TG-SC mode ............................................. 188

xix

Figure 5.9 SECM images (100 x 100 umz) for nanocrystalline BDD thin
ﬁlms grown with 10 ppm (A and D), 5 ppm (B and E), and 1
ppm (C and F) B2I-16 in the source gas mixture using either 1
mmol/L Fe(CN)63"4' ( A, B and C) or 0.5 mmol/L IrC15243'( D,
E and F) as the test redox system. The supporting solution was
1 mol/L KC]. The Pt diameter was 2 pm. Em, = -0.3 V and
E,,.,,,,,, = o v vs. Ag/AgCl when tested with Fe(CN)63'/4'. 13,, =
0.8 V and Esubsmtc = 1.0 V vs. Ag/AgCl when tested with
IrClsz'B'. Images were recorded in the TG-SC mode. ................. 189

Figure 5.10 SECM images (50 x 50 umz) for 1 mmol/L Ru(NH3)63+/2+ + 1
mol/L KC] at nanocrystalline BDD thin ﬁlms: (a) hydrogen-
terminated and (b) oxygen-terminated (Eﬁp = -0.3 V and
Esubsme = 0 V vs. AglAgCl). SECM images (50 x 50 ml) for
1 mmol/L Fe(CN)63-’4- + 1 mol/L KC] at nanocrystalline BDD
thin ﬁlms: (c) hydrogen—terminated and (d) oxygen-terminated
(Em, = 0.5 V and Esubsmc = 0 V vs. AglAgCl). In all cases, the

Pt tip diameter was 2 pm. The z-axis scale shows the tip
current. Images were recorded in the TG-SC mode ................ 191

Figure 5.11 SECM images (100 x 100 ml) for 1 mmol/L Ru(NH3)63+/2+ +
1 mol/L KC] at microcrystalline BDD thin ﬁlms: (a) hydrogen-
terminated and (b) oxygen-terminated (Etip = -0.3 V and
Esubsme = 0 V vs. AglAgCl). SECM images (100 x 100 umz)
for 1 mmol/L Fe(CN)63-/4- + 1 mol/L KC] at micocrystalline
BDD thin ﬁlms: (c) hydrogen-terminated and ((1) oxygen-
terrninated (Em, = 0.5 V and Esubmm = 0 V vs. AglAgCl). In all

cases, the Pt tip diameter was 2 pm. The z-axis scale shows the
tip current. Images were recorded in the TG-SC mode ............. 192

Figure 5.12 SECM images (10 X 10 umz) of a nanocrystalline BDD thin
ﬁlm grown with 10 ppm B2H6 in the source as in (A) 1
mmol/L Ru(NH3)63*’2*, (B) 1 mmol/L Fe(CN)63'4' and (C) 1
mmol/L IrClsm'. The supporting electrolyte in all cases was 1
mol/L KC]. The Pt tip diameter was 2 run. The z-axis scale in
all three images is the same. Images were recorded in the TG-
SC mode. (Reprinted with permission from J. Phys. Chem. 111,
3986 (2007). Copyright (2007) American Chemical Society.)
(From Ref. 26) .............................................................. 194

Figure 5.13 SECM images of glassy carbon in (A) 1 mmol/L Ru(NH3)63+/2+,
(B) 1 mmol/L Fe(CN).3"4' and (C) 0.5 mmol/L IrC152./3'. The
supporting electrolyte in all cases was 1 mol/L KC]. The Pt tip
diameter was 2 pm. Images were recorded in the TG-SC mode.
The left column Shows the original images recorded by SECM.
These images were replotted in right column with a reduced
current scale by a factor of 1/10. Note that the scan size for (B)
and (E) is 100 x 100 um2 and the other are 10 x 10 um2 ............ 198

Figure 5.14 SECM images of a nanocrystalline BDD thin ﬁlm grown with
10 ppm B2H6 in the source gas mixture in (A) 1 mmol/L
Fe(CN)63"4', (B) 1 mmol/L Ru(NH3).3+’2+ and (C) 1 mmol/L
IrCIGZJ '. The supporting electrolyte in all cases was 1 mol/L
KC]. The Pt tip diameter was 2 pm. Images were recorded in
the TG-SC mode. The left column shows the original images
recorded by SECM. These images were replotted in the right
column with a reduced current scale by a factor of 1/ 10 ............ 200

Figure 5.15 SECM images of a gold ﬁlm deposited on a conducting Si
substrate in (A) 1 mmol/L Fe(CN)63'/4' and (B) 1 mmol/L
Ru(NH3)63+/2+. The supporting electrolyte was 1 mol/L KC].

The Pt tip diameter was 2 pm. Images were recorded in the
TG-SC mode. The left column shows the original images
recorded by SECM. These images were replotted in the right
column with a reducd current scale by a factor of 1/10 ............. 202

Figure 5.16 Model of a boron-doped diamond electrode surface and a
diffusion volume unit for an active site. Black dots represent
active sites .................................................................. 204

Figure 5.17 Tip-approach curves at (A) a conductive site and (B) a non-
conductive site of a nanocrystalline BDD thin ﬁlm electrode in
1 mM Ru(NH3)63+’2++ 1 M KC1. The nominal diameter of the
microelectrode was 2 pm. The microelectrode was held at -0.3
V, and the diamond electrode potential was held at O V vs.
Ag/AgCl. i2 is the tip current, in, is the limiting current, and L
= d/a, where d is the distance between the tip and substrate, a is
the radius of the tip. Theoretical curves are calculated
according to the models given in Ref. 21 .............................. 211

Figure 6.1 AF M mages for the MSU (left) and NRL (right) boron-doped
nanocrystalline diamond ﬁlms .......................................... 219

Figure 6.2 SEM images of the (A) MSU and (B) NRL nanocrystalline
boron-doped diamond ﬁlms .............................................. 220

Figure 6.3 Raman spectra (visible excitation, 514.4 nm) for the (A) MSU
and (B) NRL boron-doped diamond ﬁlms ........................... 222

Figure 6.4 Background cyclic voltammetric i—E curves for the MSU and
NRL nanocrystalline boron-doped diamond ﬁlms in 1 mol/L
KC] at 0.1 V/s .................................................................................. 225

Figure 6.5 Cyclic voltammetric i—E curves for (A) 1 mmol/L Fe(CN)63'/4‘
in 1 mol/L KC], (B) 1 mmol/L Ru(NH3),3*’2* in 1 mol/L KC],
(C) 1 mmol/L IrC162'3'+ 1 mol/L KC], (D) 1 mmol/L methyl
viologen in 1 mol/L KC], (E) 1 mmOI/L dopamine in 0.1 mOl/L
HClO4, and (F) 1 mmol/L Few2+ in 0.1 mollL HClO4. Scan

rate: 0.1 V/s ................................................................. 226

Figure 6.6 Relationship between the cathodic peak current and the square
root of the scan rate for the two BDD electrodes in 0.1 mmol/L
Ru(NH3)63’+/2+ + 1 mol/L KC]. The scan rate was valued from
0.1 V/s to 0.9 V/s ......................................................... 227

Figure 6.7 CP-AFM images for the MSU nanocrystalline boron-doped
diamond thin ﬁlm. (A) Height image and (B) Conductivity
map ........................................................................... 232

Figure 6.8 CP-AFM images of an NRL nanocrystalline boron-doped
diamond thin ﬁlm. (A) Height image and (B) Conductivity
map ........................................................................... 233

Figure 7.1 Schematics for (A) the distribution of sensory and sympathetic
nerves around a blood vessel, (B) an arterial neuroeffector
junction, and (C) a venous neuroeffector junction .................. 240

Figure 7.2 An illustration of a simple reﬂex arc ..................................... 242

Figure 7.3 Raman spectra for (a) boron doped diamond thin ﬁlm
deposited on Pt wire, (b) P553 carbon ﬁber, and (c) T650
carbon ﬁber ................................................................. 248

Figure 7.4 Cyclic voltammetric i-E curves for drugs on carbon ﬁber (P553,
T650) and diamond electrodes. (a) Capsaicin, prazosin, UK
14,304 and yohimbine on P553 carbon ﬁber electrode. (b)
Idazoxin, PPADS and cocaine on P553 carbon ﬁber electrode.
(c) Capsaicin, prazosin, UK 14,304 and yohimbine on T650
carbon ﬁber electrode. ((1) Idazoxin, PPADS and cocaine on
T650 carbon ﬁber electrode. (6) Capsaicin, prazosin, UK
14,304 and yohimbine on diamond microelectrode. (f)
Idazoxin, PPADS and cocaine on diamond microelectrode. All
curves were acquired in Krebs solution at 0.1 V/s. The
concentration of the drug was 0.1 mmol/L ............................ 250

Figure 7.5 Cyclic voltammetric i-E curves for NE on P553 (a), T650 (0)
carbon ﬁber electrode and diamond microelectrode (e) with
and without capsaicin and Cyclic voltammetric i-E curves for
NE on P558 (b), T650 (d) carbon ﬁber electrode and diamond
microelectrode (t) with and without prazosin. All curves were
acquired in Krebs solution at 0.1 V/s. The concentration of NE
was 0.02 mmol/L and the concentration of the drug was 0.1
mmol/L ...................................................................... 251

Figure 7.6 Statistic results (11 = 3) for effects of capsaicin and prazosin. (a)
AE1/2 before and after drug on P553, T650 and diamond
electrode for NE detection. (b) Peak current changes before and
after drug for NE detection on three types of electrode. (c)

xxii

Figure 7.7

Figure 7.8

Figure 7.9

Figure 7.10

Figure 7.1 1

Figure 7.12

Figure 7.13

AEi/2 before and after drug was applied for NE detection. ((1)
Peak current changes before and after drug was applied for NE
detection using three types of electrode .............................. 253

Cyclic voltammetric i-E curves for NE on P553 (a), T650 (c)
carbon ﬁber electrode and diamond microelectrode (c) with
and without UK 14,304 and Cyclic voltammetric i-E curves for
NE on P553 (b), T650 ((1) carbon ﬁber electrode and diamond
microelectrode (i) with and without Yohimbine. All curves
were acquired in Krebs solution at 0.1 V/s. The concentration
of NE was 0.02 mmol/L and the concentration of the drug was
0.1 mmol/L .................................................................. 255

Statistic results (11 = 3) for effects of UK 14,304 and
Yohimbine. (a) AEm before and with drug on P553, T650 and
diamond electrode for NE. (b) Peak current changes before and
after drug on NB detection using three types of microelectrode.
(c) AE1/2 before and after drug was applied for NE. ((1) Peak
current changes before and after drug was applied for NE
detection using three types of microelectrode ......................... 256

Cyclic voltammetric i-E curves for norepinephrine on P558 (a),

T650 (c) carbon ﬁber electrode and diamond microelectrode (c)

with and without idazoxan and Cyclic voltammetric i-E curves

for norepinephrine on P558 (b), T650 ((1) carbon ﬁber
electrode and diamond microelectrode (f) with and without
PPADS. All curves were acquired in Krebs solution at 0.1 V/s.

The concentration of norepinephrine was 0.02 mmol/L and the
concentration of the drug was 0.1 mmol/L ........................... 259

Cyclic voltammetric i-E curves for norepinephrine on P558 (a),
T650 (b) carbon ﬁber electrode and diamond microelectrode (c)
with and without cocaine. All curves were acquired in Krebs
solution at 0.1 V/s. The concentration of norepinephrine was
0.02 mmol/L and the concentration of the drug was 0.1
mmol/L ...................................................................... 260

Reaction mechanism for GA and norepinephrine. The circled
groups represent the ﬂuorophore in the molecule ...................... 268

Glyoxylic acid-induced ﬂuorescence images of catecholamines
in the perivascular sympathetic nerves of Sham MA (A), Sham
MV (B), DOCA-salt MA (C) and DOCA-salt MV (D). The
nerve plexus in arteries has a mesh-like arrangement with
ﬁbers oriented along both the longitudinal and circular axes of
the blood vessel. The plexus in veins has a circular
arrangement with very few ﬁbers oriented longitudinally ......... 270

Schematic of immunohistochemical staining of TH, SP and
CGRP. In the sympathetic nerve varicosity, anti-TH antibody

xxiii

(primary antibody) binds to TH, followed by binding of a
secondary antibody with a ﬂuorescent dye, which generated
ﬂuorophoric complex. In the sensory nerve varicosity, anti-SP
antibody binds to SP, followed by binding of a secondary
antibody with a ﬂuorescent dye, which generated ﬂuorophoric
complex. Also in the sensory nerve varicosity, anti-CGRP
antibody binds to CGRP, followed by binding of a secondary
antibody with a ﬂuorescent dye, which generated ﬂuorophoric
complex. The fluorescence intensity depends on the
concentration of ﬂuorescent complex at a particular excitation
wavelength .................................................................. 272

Figure 7.14 Immunohistochemical localization of TH and CGRP in whole
mount preparations of mesenteric artery and vein from a Sham
rat. (A) Localization of TH-ir in Sham MA. (B) Localization of
CGRP-ir in the same tissue. While there are much higher dense
TH-ir nerve ﬁbers, CGRP-ir are less dense. (C) Overlay of (A)
and (B). (D) Localization of TH-ir in Sham MV. (E)
Localization of CGRP-it in the same tissue. (F) Overlay of (C)
and (D). Scale bars in all images are 50 um ........................... 273

Figure 7.15 Immunohistochemical localization of TH and CGRP in whole
mount preparations of mesenteric artery and vein from a
DOCA-salt rat. (A) Localization of TH—ir in MA. (B)
Localization of CGRP-ir in the same tissue. While there are
much higher dense TH-ir nerve ﬁbers, CGRP-ir are less dense.
(C) Overlay of (A) and (B). (D) Localization of TH-ir in MV.
(E) Localization of CGRP-ir in the same tissue. (F) Overlay of
(C) and (D). Scale bars in all images are 50 um ..................... 274

Figure 7.16 Immunohistochemical localization of TH and SP in whole
mount preparations of mesenteric vein from a DOCA—salt rat.
(A) Localization of TH-ir. (B) Localization of SP-ir in the
same tissue. (C) Overlay of (A) and (B). Scale bars in all
images are 200 um ......................................................... 275

Figure 7.17 Immunohistochemical localization of SP and CGRP in whole
mount preparations of DOCA-salt mesenteric arteries. (A)
Localization of SP-ir. (B) Localization of CGRP-it in the same
tissue. While there are numerous CGRP-it nerve ﬁbers, SP-ir is
rare. (C) Overlay of (A) and (B). Note the yellow lines which
are combinations of the red lines (SP-ir) in (A) and green lines
(CGRP-it) in (B). Scale bars in all images are 50 um ................ 276

Figure 7.18 Hydrodynamic i-E curves for NE, SP and capsaicin at a BDD
microelectrode. The measurements were performed in 0.1
mmol/L NE, SP or capsaicin in Krebs’ buffer with a 50 mV
potential step starting from 0 V vs. Ag/AgCl, respectively .......... 279

xxiv

Figure 7.19 Calibration curve for NE oxidation at a BDD microelectrode.
The oxidation currents were measured amperometrically by
perfusing Krebs’ buffer with different NE concentrations .............. 280

Figure 7.20 Typical vasoconstriction and NE overﬂow results recorded for
a Sham mesenteric artery. The top panels show the blood
vessel constriction and the bottom panels show the recorded
NE oxidation current after electrical stimulation. The dash lines
indicate when the electrical stimulus was applied. Column A
shows the diameter and current traces for the blood vessel
control. Column B shows the diameter and current traces after
1 pmol/L capsaicin was applied (30 min) and washed out
(perfused with Krebs’ solution for l h). Column C shows the
diameter and current traces during perfusion of 0.4 umol/L SP.
Column D shows the diameter and current traces after SP wash
out. The time axis and baseline for each current trace were
adjusted for comparison. .................................................. 281

Figure 7.21 Typical results recorded from a DOCA mesenteric artery. The
top panels show the blood vessel constriction and the bottom
panels show the recorded NE oxidation current after electrical
stimulation. The dash lines indicate when the electrical
stimulus was applied. Column A shows the diameter and
current traces for control. Column B shows the diameter and
current traces after 1 umol/L capsaicin was applied (30 min)
and washed out (perfused with Krebs’ solution for 1 hour).
Column C shows the diameter and current traces during
perfusion of 0.4 umol/L SP. Column D shows the diameter and
current traces after SP wash out. The time axis and baseline for
each current trace were adjusted for comparison ..................... 282

Figure 7.22 Typical results for a Sham mesenteric vein. The top panels
Show the blood vessel constriction and the bottom panels Show
the recorded NE oxidation current after electrical stimulation.
The dash lines indicate when the electrical stimulus was
applied. Column A shows the diameter and current traces for
the control. Column B shows the case after 1 umol/L capsaicin
was applied and washed out. Column C shows the case when
0.4 umol/L Substance P was perfused. Column D shows the
cases after Substance P was washed out. The time axis and
baseline current for each trace were adjusted for comparison ...... 283

Figure 7.23 Typical results for a DOCA-salt mesenteric vein. The top
panels show the blood vessel diameter and the bottom panels
show the recorded NE oxidation current after electrical
stimulation. The dash lines indicate when the electrical
stimulus was applied. Column A shows the diameter and
current traces for the control. Column B shows the case after 1
umol/L capsaicin was applied and washed out. Column C
shows the case when 0.4 umol/L Substance P was perfused.

Column D shows the cases after substance P was washed out.
The time axis and baseline current for each trace were adjusted
for comparison ............................................................. 284

Figure 7.24 Percent change in the NE oxidation current (left) and the blood
vessel constriction (right) for mesenteric arteries from 10 Sham
rats and 11 DOCA-salt rats ............................................. 284

Figure 7.25 Percent change in the NE oxidation current (left) and the blood

vessel constriction (right) for mesenteric veins from 4 Sham
rats and 4 DOCA-salt rats ............................................... 285

xxvi

Chapter 1

Introduction

1.1 Diamond and the conducting diamond electrode

1.1.] Diamond

Diamond is a carbon allotrope with a cubic crystal structure (Fig.1.1)
consisting of two interpenetrating face-centered cubic (FCC) lattices, displaced from
each other by one quarter of a body diagonal. Each carbon atom is tetrahedrally
coordinated (3p3 hybridized), creating strong, directed sigma bonds with its four
neighboring carbon atoms. The bond length and lattice constant are 1.54 and 3.56
angstroms, respectively.I The outstanding mechanical, chemical and electronic
properties of diamond stem from the extremely strong and tight 3p3 bonding. Due to
the very large activation barrier between the 3p2 and 3p3 bonding states, diamond is i
for all practical purposes as a stable microstructure, although the most

thermodynamically stable microstructure for carbon is graphite. Based on these facts,

 

. First lattice

G Displaced lattice

 

 

 

 

Figure 1.1 Diamond cubic crystal structure.

diamond is expected to be a promising wide-band-gap semiconductor for electronic
and optical devices, such as ultraviolet light emitting diodes, cold cathode electron
emitters, and high-power and high-frequency devices. Diamond’s well-known
properties include extreme hardness, very high thermal conductivity and optical
transparency, in pure form. In its natural or undoped form, diamond is an excellent
electrical insulator with a bandgap of 5.5 eV.2‘ 3 When intentionally doped with boron,
semiconducting or even semimetallic electrical properties can be imparted to the
material.2’ 3

in order for diamond to be a viable wide-band-gap semiconductor or
semimetallic conductor, several fundamental issues must be resolved including high-
quality diamond growth, different types of doping, contact fabrication and etching
technology. The discovery of chemically vapor deposited diamond at high rates in
198034’ 5 enabled applications such as heat sinks for electronics, wear-resistant layers
on tools, and anti-corrosion layers on metals. 6'8 Nowadays, scientists are seeking to
exploit some of the attractive properties of diamond (see Table 1.1).”'18 For example,
there is a growing interest in the use of doped diamond as an electrochemical

electrode.

Table 1.1 Properties of type IIa natural and CVD diamond

 

 

 

 

 

 

 

 

 

 

 

 

Properties Type IIa CVD Ref.
Density (g/cm3) 3.52 3.52 9
Hardness (Gpa) 57 - 104 50 - 100 10.11

Thermal Conductivity
(W/cm/K) at 298 K 20 - 23 10 - 21 10
Thermal Expansion
Coefﬁcient ( 10'6/K) 0'8 — 1.2 ~ 2.0 12
Heat Capacity
(J/mol/K) 6.2 2.2 1 3
2

Table 1.1 Properties of type IIa natural and CVD diamond (continued)

 

 

 

 

 

 

 

 

 

 

Properties Type Ila CVD Ref.
Optical transmissivity 225 nm to far IR 225 nm to far IR 14
Electrical Resistivity

1016 10'2 — lol6 15
(00cm)
Band gap (eV) 5.45 5.5 1
Electrical breakdown
ﬁeld 10000 9
(kV/cm)
Electron Afﬁni eV
ty( ) 0.4—-1.3 -1.3
(Depends on surface . 16
_ _ (H-termrnated)
termrnatlon)
Electron mobility
2 2200 4500 17
(cm N/s)
Hole mobility
2 1600 3800 17
(cm N/s)
Electron Saturated
. 7 2.7 2.7 18
Velocrty (10 cm/s)
Hole Saturated Velocity
7 1.0 1.0 1
(10 cm/s)

 

 

 

 

 

 

1.1.2 Conducting diamond (Synthetic diamond)

Natural diamond is an electrical insulator with a band-gap of ~5.5 eV and a
resistivity greater than 1012 Q-cm.9 However, diamond grown via CVD in hydrogen-
rich environments can exhibit the so-called “surface conductivity”.'5 The resistivity of
synthetic diamond can be lowered to 106 Q-cm when terminated with hydrogen (or
having it incorporated into the lattice). Synthetic diamond can be made more
conductive via doping with elements such as boron, nitrogen and phosphorous.

Diamond can be made p-type by doping with boron, E2, = 0.37 eV, and n-type by

doping with nitrogen, Ea = 1.7 eV, or phosphorous, Ea = 0.56 eV. These are large
activation energies compared to the activation energy for boron-doped Si (ca. 0.02
eV).19 The candidates for p-type doping are B, Al and heavier group III elements.
Boron is the most commonly used dopant.2' 2° Nitrogen, phosphorous and heavier
group V elements are n-type dopants but the formation of n-type semiconducting
diamond has proven a difﬁcult task. One reason for this is the alternation of the
electric conductivity due to the incorporation of hydrogen into CVD grown ﬁlms from
conventional C/H source gas mixtures. Hydrogen can act as both a donor and an
acceptor. Nitrogen is among a few suitable dopants for CVD diamond, but
unfortunately it forms a deep donor level ~1.7 eV below the bottom of the conduction
band. This is due to its structural distortion from the substitutional position in the
diamond lattice. Therefore, achieving highly conducting (room temperature) n-type
diamond had been a challenge. A breakthrough in synthesizing N-doped diamond
came several years ago through the addition of N2 to the CH4/Ar source gas mixture
for growing ultrananocrystalline diamond.21 Although only B and N have the required
properties of a suitable dopant for diamond, surprisingly, the only proven n-type
diamond is obtained by the introduction of phosphorus during CVD diamond growth.
Koizumi et al reported the growth of n-type diamond on (111)-oriented diamond
substrates by using a mixture of PH3, CH4 and H2 gas in 1997. 22 Following this
success, several studies have also been reported for phosphorous (P) doping during
the chemical vapor deposition of diamond.23’ 2"

The most extensively employed CVD diamond in electrochemistry is the p-
type, boron-doped material. The resistivity of highly boron-doped diamond thin ﬁlms
can be as low as ~0.005 Q-cm. Both high-quality microcrystalline and nanocrystalline

boron-doped diamond thin ﬁlms can be routinely prepared.

Diamond is, in fact, metastable with respect to graphite at low temperature and
pressure as the free energy difference between the two materials at 298 K and 1 atrn is
only 2.900 kJ/mol (-0.03 eV per atom). However, there is a large activation barrier for
conversion.”’ 25 Thus, in theory, diamond growth can be achieved in the metastable
region of the carbon phase diagram as long as graphite formation is suppressed.

Low pressure diamond synthesis via CVD dates back to the 19503 and 603
when research was conducted in both the former Soviet Union and the United
States.”28 Since then, developments in CVD technology have enabled the production
of diamond for a number of applications, such as optical windows, wear-resistant
coatings for cutting tools, thermistors, and electrodes. 29 Several CVD growth
methods exist for producing diamond, all of which utilize a carbon-containing
reactant source gas (e.g., methane and hydrogen), an energy source to promote
reactant dissociation (e.g., microwave, hot-ﬁlament, or DC discharge), and a substrate
(e. g., Si, W, Mo). High-quality (i.e., low defect density and low nondiamond carbon
impurity content) CVD diamond is typically produced at low pressure (10 to 500
Torr), high gas temperature (>1700 °C), high substrate temperature (700 to 800 °C),
and low carbon concentration in the source gas (e.g., 0.5%) with an energy source for

gas dissociation and radical formation.” 29

1.2 Electrochemical properties and application of boron-

doped diamond (BDD)

The ﬁrst paper on the electrochemistry of polycrystalline diamond was
published by Pleskov et al. in 198720’ 30’ 31. This ﬁlm possessed reasonable electrical
conductivity but was not doped. The conductivity likely resulted from nondiamond

impurity phases. Nevertheless, this demonstration of diamond electrodes spawned

conductive diamond used extensively in electrocherrristry. Diamond exhibits several
remarkable properties. First, it has a large potential window in aqueous solution with
low background curren132'34. Second, the electrode formed from polycrystalline
diamond possesses many of the same physical properties of bulk diamond.“ 35'”
Third, the electrode surface microstructure is stable as there is little evidence of
degradation at high potentials and current densities.3944 Fourth, the electrode exhibits
good activity for various redox systems without conventional pre-treatment and
excellent resistance to deactivation. These characteristics make diamond useful for a

number of applications.

1.2.1 Electrolytic waste water treatment

An important electrochemical application of boron-doped diamond is in the
area of electrolytic water puriﬁcation and decontamination.“'50 Electrolytic treatment
is quite attractive for water puriﬁcation and decontamination but usually is limited by
the stability of the electrode. Boron-doped diamond is a good candidate electrode in
this ﬁeld because of its inherent properties. Our group has shown that boron-doped
diamond thin ﬁlm electrodes exhibit superb morphological and microstructural
stability.5 " 52 The wide potential window of boron-doped diamond has allowed poorly
biodegradable compounds53 and persistent contaminants54 to be decontaminated in
waste water. Ruediger et al. reported on the integrated biological and electrochemical
oxidation of industrial wastewater.” The BDD electrode was used to transfer
persistent organic compounds into a biodegradable fraction, followed by high
efﬁcient biological eliminatiOn using specialised bacteria. The advantages of the
process are high efﬁcient chemical oxidation demand (COD) removal with reduced

energy consumption combined with low total residence time because the organic

compounds can be electrochemically oxidized at BDD electrode. Similarly, various
chemical contaminants such as t-butanol,56 chlorophenol,57 phenolic compounds,58
acetaminophen” and organotin compounds60 can be oxidatively degraded with boron-
doped diamond. Kober et a]. invented the method and device for the oxidative
treatment of wastewater containing pesticides based on a boron-doped diamond

electrode.61

The electrooxidation of some chemicals on boron-doped diamond
electrode has also been used for decolourization in industrial wastewater. For example,
Belaid and coworkers used diamond to remove phenolic compounds and decolourize
olive mill wastewater.62 Canizares et al. reported the treatment of Fenton-refractory
olive oil mill wastes by electrochemical oxidation with boron—doped diamond
anodes.63 More recently, Panizza et a]. developed a theoretical model for organic

64 The model was

pollutants mineralization on B-doped diamond electrodes.
formulated for a perfect mixed electrochemical reactor operated as a batch
recirculation system under multiple current steps, in which the applied current was
adjusted during the electrolysis to be close to the limiting value. The model was
experimentally validated by the anodic oxidation of 3, 4, 5-trihydroxybenzoic acid.

Multiple current steps electrolysis and continuous current control allowed obtaining

high oxidation rate and current efﬁciency.

1.2.2 Electroanalysis

CVD diamond has been extensively used in electroanalysis because it offers
advantages over other electrodes in terms of linear dynamic range, limit of detection,
response time, response precision and response stability. Some of the applications of
diamond in electroanalysis reported since 2004 are highlighted below. The review for

the applications reported before 2004 can be seen in the literature.” (’6

A. Metal ion analysis

The electrochemical properties of CVD conducting diamond are ideally suited
for the detection of trace metal ions via anodic stripping voltammetry (ASV). It has
been shown that mercury ions at both the ppb and ppt level can be determined with
BDD electrodes by differential pulse voltammetry (DPV).67’ 68 A modiﬁed BDD
electrode has also been used for the detection of mercury. For example, Salimi et al.(’9
reported on an iridium oxide modiﬁed BDD electrode to detect Hg(I). They deposited
iridium oxide ﬁlms electrochemically on the electrode surface by potential cycling
between —0.2 to 1.2 V from a saturated alkaline iridium (III) solution based on the
following reaction:

Ir203-xH2O (aq) + 20H. —> 21rO2oxH2O (s) + H2O + 2e-

The modiﬁed BDD electrode exhibited an excellent catalytic activity for
oxidation of Hg(I) over a wide pH range and shows excellent analytical performance
for Hg(I) amperometric detection. They suggested that IrO2 catalyzed mercury(l)
oxidation in the following electrochemical catalytic pathway:

Ir (reduced fonn)—»Ir (oxidized form)
Ir (oxidized form) + Hg22+ —» Ir (reduced form) +2 Hg2+

The detection limit, sensitivity, response time and dynamic concentration
ranges are 3.2 nmol/L, 77 nA/(lunol/L), 100 m3 and 5 nmol/L — 5 limol/L. These
analytical parameters compare favorably with those obtained with modern analytical
techniques and recently published electrochemical methods."9

Lead can be detected with BDD7o or modiﬁed BDD71 electrodes by ASV at

low pH and copper coexisting. The low limit of detection (2 nmol/L at BDD and 0.85

nmol/L at modiﬁed BDD) is an advantage compared to other electrode materials. The
technique has been used for the detection of lead in tap water and blood.

Banks et a]. studied the deposition of cadmium on boron-doped diamond with
square-wave anodic stripping voltammetry in quiescent conditions, in the presence of
an acoustic ﬁeld and then in the presence of the neutral surfactant Triton® X—100.
They found the effect of optimized insonation was to increase the sensitivity from
0.63 (under silent conditions) to 3.78 uA/(lrmol/L) and to reduce the limit of detection
by an order of magnitude from 10'8 to 10‘9 mol/L.72

It has been shown that ppb levels of cadmium and lead both can be detected
using square-wave anodic stripping voltammetry with a boron-doped diamond
electrode.”

Besides anodic stripping voltammetry, cathodic stripping voltammetry (CSV)
at a boron-doped diamond electrode has been used for the detection of metal ions. For
instance, free Mn (11) in seawater samples can be detected with BDD by CSV
utilizing an insonated accumulation protocol. No sample pre-treatrnent was required
with the application of ultrasound providing a sensitive and selective protocol for the
analysis of Mn (II).74’ 75

CSV has also been used to quantify the concentration of Ni (11) in an
electroless nickel deposition bath with a BDD electrode. The methodology relies on
the electrodeposition of Ni onto a boron-doped diamond electrode and the subsequent
oxidative conversion of the metallic layer to Ni(III). The analytical signal was derived
from a cathodic stripping protocol in which the reduction of the Ni(III) layer to Ni(II)
was monitored by the use of differential pulse voltammetry. The stripping peaks
obtained with both solution stirring and electrode rotation were observed at +0.6 V

(vs. SCE) with the response found to be linear over the range 0.3—5 umol/L (solution

stirring) and 0.1—4 umol/L (electrode rotation). The viability of the technique was
evaluated through the determination of nickel in an electroless nickel deposition bath
solution used practically and validated by an independent comparison with the result
of ICP-AES.76

The above discussion shows that BDD electrode can be used to detect
different metal ions using different detection conditions. In fact, BDD electrodes have
become one of the important tools for heavy metal detection. By studying some
analytical parameters of differential pulse anodic stripping voltammetry (DPASV)
method, Tall et a]. demonstrated that it’s possible for the simultaneous determination
of Zn (11), Cd (11) and Pb (11) at very low levels despite the overlapping of Zn and Cd
stripping peaks. Compared to the glassy carbon electrode (without mercury plating), a
lower base line current, wider potential range, higher sensitivity (3 to 5 times higher
than GC) and longevity of the material were noticed for the BDD.77

Because of the high sensitivity and low limit of detection, Hg has historically
been the electrode of choice for ASV but there is an ongoing search for alternate
electrodes and diamond is one of these. Our group compared the performance of
boron-doped diamond for the anodic stripping voltammetric determination of heavy
metal ions (Zn2+, Cd“, Pb“, Cu”, Ag”) with that of Hg-coated glassy carbon (Hg-
GC). Despite the fundamentally different nature of the metal reduction and oxidation
reactions on these two electrodes, BDD provides as good or superior analytical
detection ﬁgures of merit when compared to Hg. The linear dynamic range for BDD
was three to four orders of magnitude (r2 > 0.995), similar to Hg-GC. Although BDD
is less sensitive than Hg-GC, the lower background current and noise for BDD
enabled detection limits as low as those seen for Hg-GC. The minimum concentration

of each metal ion detectable (SN 2 3) with BDD was in the mid to low ppb range

10

(e.g., 50 ppb Zn2+, 1.0 ppb Cd“, 5.0 ppb Pb‘”, 10 ppb Cu2+ and 1.0 ppb Ag"). BDD
exhibited an electrode-to—electrode and run-to-run variability of less than 5%, which is
comparable to that for Hg-GC. The results demonstrate that BDD is a viable alternate

electrode to Hg for ASV.78

B. Other inorganic species

Recently, the concept of an ion-sensitive ﬁeld effect transistor (ISFET) on
diamond using a thin and highly B-doped channel was described.79 The doped
channel is in direct contact with the electrolyte solution. The diamond surface is O-
terrninated to provide pH sensitivity and chemical stability. The ﬁrst pH-sensitive
ISFET microstructure with the extrinsically doped channel was fabricated on a 100-
oriented single crystal diamond substrate using a solid doping source technique and a
wet chemical oxidation for the O-termination. The fabricated structures showed pH
sensitivity close to the Nemst's limit of 59 mV/pH. The ISFET characteristics were
reproducible and did not degrade after repeated cycling between 0.1 mol/L H2804 and
0.1 mol/L KOH solutions.79

Diamond provides a sensitive, reproducible, and stable response for the
electrooxidation of azide, leading to superior detection performance compared with
glassy carbon. Suzuki et a1. investigated the direct detection of sodium azide using
BDD electrodes in three important physiological saline buffers, ADA buffer (N-(2-
acetamido) iminodiacetic acid), sodium EDTA and imidazole. Sodium azide
concentrations of 502 and 1355 ug/mL were detected in two samples with relative
standard deviations (RSD) of 0.07 and 2.37%, respectively. The results show that the
detection of sodium azide with BDD electrode is simple, selective, sensitive and

reproducible.80

ll

Ivandini et a1. studied Pt-modiﬁed diamond electrodes for the oxidation of
hydrogen peroxide and found that the electrodes exhibited high catalytic activity and
excellent electrochemical stability with a very low background current in comparison
to a Pt electrode. The electrode was fabricated for electroanalytical applications by
ion implantation method at 750 keV Pt2+ with a dose of 5 X 1014 cm_2. Very low limit

of detection, 30 nmol/L (S / N = 3), were achieved with RSD of 2.91% (n = 20).8|

Sulfate, which is usually considered to be electrochemically inactive, is
difﬁcult to be detected. However, this situation has been changed because of BDD
electrode. The extremely wide potential window enables BDD to be used for the
electrochemical determination of sulfate according to the following reaction: 82

2H2804 —r 1128203 + 2H+ + 2c-

Diamond electrodes allow the measurement of sulfate on the anodic side of the
potential window through its oxidation into peroxodisulfate, and detection of
peroxodisulfate on the cathodic side through the reverse reduction reaction. In all
cases, the current density was proportional to the concentrations of sulfate and
peroxodisulfate.

Our group reported an accurate method for total inorganic arsenic
determination in real water samples using DPASV and a Au-coated boron-doped
diamond thin-ﬁlm electrode. Keys to the method are the use of a conducting diamond
platform and solid phase extraction for sample preparation. In the method, the As(III)
present in the sample is ﬁrst detected by DPASV. The As(V) present is then reduced
to As(III) by reaction with Na2803 and this is followed by a second detection of
As(III) by DPASV. The method provided reproducible and accurate results for total

inorganic arsenic in two contaminated water samples provided by the US. Bureau of

Reclamation. The total inorganic As concentration in the two samples, quantiﬁed by

12

the standard addition method, was 23.2 :I: 2.9 ppb for UV plant inﬂuent water and
16.4 :I: 0.9 ppb for Well 119 water (11 = 4). These values differed from the speciﬁed
concentrations by less than 4%.47’ 83

Diamond electrodes are also useful for the detection of other electroactive

anion such as 1111111684, chlorine, 85 iodide,86 sulﬁde,87 throsulfate,88 sulfonamldes,8

and dodecylbenzenesulfonate.90

C. Organic species

Boron-doped diamond electrodes have been used to detect aniline by linear-
sweep cathodic stripping voltammetry. The dimeric species (p-aminodiphenylamine
and benzidine) was ﬁrst formed by anodic oxidation of aniline during the
accumulation period, and then it was reversibly reduced in acidic media. The shape of
the stripping voltammetric response is suitable for aniline detection in the micromolar
concentration range. The low background current of conductive diamond is an
advantage compared to other electrode materials and allows a detection limit of 1
umol/L. Weak molecular adsorption and good electrochemical stability are additional
advantages of BDD and even after long-time measurements, the electrode surface can
regain its initial activity by an anodic polarization in the potential region of water
decomposition.“

A mediator-free glucose biosensor, termed a "third-generation biosensor," has
been fabricated by immobilizing glucose oxidase (GOD) directly onto an oxidized
boron-doped diamond electrode.92 It is believed that the surface of the oxidized BDD
electrode possesses carboxyl groups which covalently bind to GOD through
glutaraldehyde. Glucose was determined in the absence of a mediator used to transfer

electrons between the electrode and enzyme. The resulting biosensor exhibited fast

13

amperometric response (less than 5 s) to glucose. The biosensor provided a linear
response to glucose over the range 6.67 x 10-5 to 2 x 10‘3 mol/L, with a detection limit
of2.31 x 10‘5 mol/L. ”~93

Olivia et a]. developed a sensitive and stable glucose biosensor for in vivo
monitoring using Pt-modiﬁed BDD microﬁber (Pt-BDDMF) electrodes. The
electrodes were modiﬁed with Pt nano-particles to detect H202, which was
enzymically produced by glucose oxidase (G0x) immobilized on the electrode
surface. The electrodes exhibited much higher sensitivity compared to Pt-microﬁber
electrodes, Pt electrodes and Pt-modiﬁed diamond thin ﬁlm electrodes.
G0x/overoxidized polypyrrole (0PPy)/Pt-modiﬁed BDDMF electrodes were applied
for continuous interference-free glucose monitoring. Amperometric measurements of
glucose showed a linear response between 1-70 mmol/L, with a relative standard
deviation of 3.7% for ﬁve injections of 100 urnol/L glucose. The electrodes exhibited
good stability over 3 months with no detected anodic current for ascorbic acid (AA),
which is an interfering compound.94

BDD electrodes have also been used for direct and simultaneous determination
of phenol (Ph), hydroquinone (HQ) and 4-nitropheno] (4-NP).95 Results showed that
the oxidative peaks of these 3 phenolic compounds can be completely separated on a
BDD electrode in acidic conditions using cyclic voltammetry. The study also
demonstrates that the BDD electrode is extremely easy to be refreshed to obtain
current values with good reproducibility, even if it is passivated by phenolic
compounds with different adsorption characteristics. For each tested phenolic
compound, the concentration range with linearity is 2 or 3 orders of magnitude in the
presence of other coexisting phenolic compounds with concentrations >1000 times

higher than that of the tested component.

14

Our group reported on the analysis of chlorinated phenols (2-chlorophenol,3-
chlorophenol, 4-chlorophenol, 2,4-dichlorophenol, 2,4,6-trichlorophenol,
pentachlorophenol) in river water using off-line solid-phase extraction (SPE) and
capillary electrophoresis coupled with electrochemical detection. A key to the
sensitive, reproducible, and stable detection of these pollutants was the use of a
boron-doped diamond microelectrode in the amperometric detection mode. Pollutant
recoveries in the 95-100% range with relative standard deviations of 1-4% were
achieved. The oxidative detection of the pollutants was accomplished at +1.05 V vs.
Ag/AgCl without the need for electrode pretreatment. A reproducible electrode
response was observed during multiple injections of the chlorinated phenol solutions
with a relative standard deviation of 5.4%. Good electrode response stability was
observed over many days of continuous use with no signiﬁcant electrode deactivation
or fouling.”98

Flavonoids are phenol derivatives widely distributed in fruits, nuts, seeds,
ﬂowers, vegetable peels, and beverages; they are known to exhibit high antioxidative
activities. Pedrosa et al. described the use of boron-doped diamond electrode for the
amperometric detection and quantiﬁcation of ﬂavonoids in tea samples using a ﬂow
injection system. The results revealed increasing peak current with rutin concentration
in the interval 0.1-2.5 x 104 mol/L with a detection limit of 7.7 x 10" mol/L. The
method was used for the determination of rutin in three different green tea infusions.
The repeatability of current responses for injections of 150 1.1L rutin was evaluated to
be 2.1% (n = 20).99

The attractive features of a boron-doped diamond (BDD) thin-ﬁlm make it a
good choice as a detector for microchip capillary electrophoretic (CE) separations of

dye-related amino-substituted aromatic compounds. Shin et al. applied a diamond

15

electrode for the end-column amperometric detection of 4-aminophenol (4-AP), 1, 2-
phenylenediamine (1,2-PDA), 2-aminonaphthalene (2-AN), 2-chloroaniline (2-CA),
and o-aminobenzoic acid (o-ABA).loo The diamond-based electrochemical detection
system displayed a favorable analytical performance, including lower noise levels,
higher peak resolution with enhanced sensitivity, and improved resistance against
electrode passivation. The diamond detector displayed detection limits of 2.0 and 1.3
umol/L for 4-AP and 2-AN, respectively, and a wide linear response for these
compounds over the 2-50 umol/L range. The enhanced stability was demonstrated by

relative standard deviation (RSD) values of 1.4% and 4.7% for 100 umol/L 1,2-PDA

and 200 umol/L 2-CA, respectively, for repetitive detections (n = 7).")0

D. Pharmaceuticals

Naproxen is a non-steroidal anti-inﬂammatory drug (N SAID) commonly used
for the reduction of moderate to severe pain, fever, inﬂammation and stiffness and the
treatment of primary dysmenorrhea. The electrooxidation of naproxen was studied,
for the ﬁrst time, by cyclic and differential pulse voltammetry (CV and DPV) in
nonaqueous solvent supporting electrolyte system using boron-doped diamond (BDD)
electrode. Naproxen undergoes one-electron transfer resulting in the formation of
cation radical for the ﬁrst electrooxidation step, which follows other chemical and
electrochemical steps such as deprotonation, removal of another electron and the
attack of nucleophile (ECEC mechanism). BDD electrode provided higher signal to
background ratio, well resolved and highly reproducible cyclic voltammograms than
the GC electrode. With a scan rate of 50 mV/s and pulse height of 50 ms, respectively,
the DPV technique was able to determine the naproxen concentrations in the range of

0.5 to 50 mol/L with a detection limit of 30 nmol/L. 10'

16

The wide potential window of BDD electrode makes it very suitable for the
determination of special drugs, which are usually nonelectroactive on normal carbon
electrodes such as glassy carbon. Some examples are discussed here. Lincomycin is a
lincosamide antibiotic that comes from the actinomyces streptomyces lincolnensis.
The electroanalysis of lincomycin was investigated using BDD electrode. BDD
electrode provided well-resolved oxidation irreversible cyclic voltammogram with the
peak position at 1.2 V (vs. Ag/AgCl). The amperometric detection with BDD
electrode was coupled with the ﬂow injection analysis for the determination of
lincomycin. A linear dynamic range of 0.5-125 panel/L and a detection limit of 0.02
pmol/L were obtained. The recovery was shown to be in the range of 96—103%.102

Lidocaine is a common local anesthetic and antiarrhythmic drug. It is used
topically to relieve itching, burning and pain from skin inﬂammations, injected as a
dental anesthetic, and in minor surgery. Recently, a new electroanalytical procedure
was developed for the determination of the drug in commercial anesthetics. The
procedure is based on the use of electrochemical methods, such as cyclic and square-
wave voltammetry, with boron-doped diamond electrodes. The oxidation of lidocaine
in Britton-Robinson buffer (0.1 mol/L) using this type of electrode gives rise to one
irreversible peak in 1.68 V (vs. AglAgCl). The detection and quantiﬁcation limits
obtained from pure water were 10.0 and 34.4 ug/L, respectively. Lidocaine recoveries
ranged from 97.6% to 99.2%.103

Aspartame (or APM) is the name for an artiﬁcial, non-saccharide sweetener,
aspartyl-phenylalanine-l-methy1 ester; i.e., a methyl ester of the dipeptide of the
amino acids aspartic acid and phenylalanine. This sweetener is an ingredient of
approximately 6,000 consumer foods and beverages sold worldwide. It is also one of

the sugar substitutes used by people with diabetes. Medeiros et al. reported on the

17

determination of aspartame in dietary products using square-wave voltammetry in
conjunction with a cathodically pretreated boron-doped diamond electrode. The
samples were analyzed without previous treatment in a 0.5 mol/L H2804 solution. A
single oxidation peak at a potential of 1.6 V vs. Ag/AgCl with the characteristics of an
irreversible reaction was obtained. The linear range is 9.9 x 104’ to 5.2 x 10‘5 mol/L
of aspartame and the detection limit is 2.3 x 10'7 mol/L. The method was applied with
success to the determination of aspartame in several dietary products producing
results similar to those obtained by HPLC at the 95% conﬁdence level. ‘04

Carbaryl (l-naphthyl methylcarbamate) is a chemical in the carbamate family
used chieﬂy as an insecticide. Originally, Union Carbide discovered carbaryl and
introduced it commercially in 1958. It remains the third-most-utilized insecticide in
the United States for home gardens, commercial agriculture, and forestry and
rangeland protection. The anodic voltammetric behavior of carbaryl on a boron-doped
diamond electrode in aqueous solution has been reported.105 The square-wave
voltammetric data obtained for carbaryl were typical of an irreversible electrode
process with mass transport control. The results, obtained by square-wave
voltammetry at 0.1 mol/L Na2SO4 and pH 6.0, allow the development of a method to
determine carbaryl, without any previous step of extraction, clean-up,
preconcentration., or derivatization, in the range 2.5-30.0 x 10'6 mol/L with a
detection limit of 8.2 jig/L in pure water.105

BDD electrode has also been used in pharmaceutical formulations because of
its rapidity, precise and good selectivity for electrochemical analysis. For example,
methods using ﬂow injection (F1) with amperometric detection at anodized boron-
doped diamond (BDD) thin ﬁlms have been developed and applied for the

determination of tetracycline antibiotics (tetracycline, chlortetracycline,

l8

oxytetracycline and doxycycline).'°("108 The anodized BDD electrode exhibited well-
deﬁned irreversible cyclic voltammograms for the oxidation of tetracycline antibiotics
with the higher currents compared to as-deposited BDD and glassy carbon electrodes.
A low detection limit of 10 nmol/L (S/N = 3) was achieved for each drug when using
ﬂow injection analysis with amperometric detection using an anodized BDD electrode.
Linear calibrations were obtained from 0.1 to 50 nmol/L for tetracycline and 0.5—50
nmol/L for chlortetracycline, oxytetracycline and doxycycline.10°"°8

Modiﬁed BDD electrode has also been used for drug determination. For
example, the Ru(bpy)33+-modiﬁed boron-doped diamond electrode has strong
electrocatalytic effects on the redox behavior of Vitamin Be (VB5).109 The oxidization
and reduction peak potentials of VBs are 167.6 and -113 mV, respectively. In the
presence of a large amount of Vitamins B. (VBl) and B2 (VB2), the redox peak
potentials of VB6 shift slightly while VB] and VB2 are not reduced or oxidized on the
modiﬁed electrode. A sensitive square wave voltammetric response for VB; is
obtained covering a linear range from 0.2775 umol/L to 0.3686 mmol/L with a
detection limit of 6.319 X 10’8 mol/L in the presence of a large amount of VB; and
VB2. The modiﬁed electrode showed good stability and reproducibility with

satisfactory recovery results. '09

E. Biosensor

Recently, diamond has been used extensively in this area due to its excellent
response sensitivity, reproducibility, pH independent current and its minimal surface
fouling. “0' '” Dopamine is a hormone and neurotransmitter occurring in a wide
variety of animals, including both vertebrates and invertebrates. In the brain,

dopamine functions as a neurotransmitter, activating the dopamine receptors and their

19

variants. Dopamine is also a neurohonnone released by the hypothalamus. Its main
function as a hormone is to inhibit the release of prolactin from the anterior lobe of
the pituitary. The electrochemical reaction of dopamine on carbon electrodes has been
extensively studied. 1 12" 18

Norepinephrine or noradrenaline is anther catecholamine with dual roles as a
hormone and a neurotransmitter. It is released from the adrenal medulla into the blood
stream as a hormone, and is also a neurotransmitter in the central nervous system and
sympathetic nervous system where it is released from noradrenergic neurons. The
actions of norepinephrine are canied out via the binding to adrenergic receptors. As a
stress hormone, norepinephrine affects parts of the brain where attention and response
actions are controlled. Along with epinephrine, norepinephrine also underlies the
ﬁght-or-ﬂight response, directly increasing heart rate, triggering the release of glucose
from energy stores, and increasing blood ﬂow to skeletal muscle.119 Our group

published several papers'zo'122

on the in vitro continuous amperometric measurement
of norepinephrine with a boron-doped diamond microelectrode. This new
microelectrode, without any protective polymer coating, exhibited a sensitive and
stable voltammetric response for norepinephrine released from sympathetic nerves
innervating a mesenteric artery from a laboratory test animal. The background
voltammetric current was low and stable over time and independent of the solution
pH. The microelectrode was also used in electrochemical detection, coupled with
capillary zone electrophoresis, to analyze solutions for norepinephrine, other
catecholamines and their major metabolites. In the amperometric detection mode, the
microelectrode exhibited good response sensitivity, low mass limits of detection

(100—400 amol, S/N = 3) and excellent response reproducibility (RSD = 2—4%)

without any conventional pretreatment.121 BDD electrodes have also been used to

20

12

- - 12
4 ascorbic acld, 5

determine other biochemical species such as cysteine,123 tyrosine,
. . . . . 129
serotonrn (5-HT),'25’ ‘26 adenosrne,127 myoglobm,128 haemyoglobrn,128 purine, and

pyrimidine. '29

F. Electrosynthesis

Because of the outstanding inert chemistry of the BDD electrode, it has been
used in the electrosynthesis of some chemicals. For example, Canizares et al. reported
the electrosynthesis of ferrates using boron doped diamond (BDD) anodes. Their
results show that BDD electrode leads to efﬁcient processes only limited by the
availability of oxidizable-iron species. The hydroxyl ion concentration, the current
density and the iron raw material were found to inﬂuence greatly on the results. Thus,
the efficiencies of the process were increased using high current density and high
concentration of hydroxyl anions. '30

Electrolytic synthesis of ammonia with a byproduct oxygen gas from water
vapor and nitrogen gas was demonstrated in LiCl-KCl-CsCl melt containing 0.5
mol/L Li3N at 573 K under atrn. pressure. Water vapor reacted with nitride ions to
form ammonia and oxide ions in the melt. The consumed nitride ions were supplied
by the reduction of nitrogen gas at the cathode and the formed oxide ions were
oxidized to oxygen gas at a boron-doped diamond anode. '3 I

Fardel et a1. studied the anodic methoxylation of formaldehyde dimethylacetal
(F ADMA) to tri-Me orthoforrnate (TMOF) in basic MeOH on B-doped diamond
electrodes. Cyclic voltammetry and preparative electrolysis showed that FADMA is

electrochemically inactive in the solvent stability region; nevertheless F ADMA can be

oxidized in the potential region of MeOH oxidation.132

21

G. Spectroelectrochemistry

Spectroelectrochemistry is a hybrid of two techniques: spectroscopy and
electrochemistry. The technique (transmission mode) can be used to record a change
in the spectroscopic signature of an analyte associated with an alteration in redox state
brought about by electron transfer with an optically transparent electrode (OTE).
Boron-doped diamond (BDD) possesses attractive qualities as an OTE: a wide optical
window from the near-UV extending into the far-IR (0.25-100 I‘m) region of the
electromagnetic spectrum, a wide working potential window (>3 V in aqueous media),
low background current, microstructural stability during cathodic and anodic
polarization, and resistance to molecular adsorption and fouling. Our group has been
developing optically transparent diamond for use in the spectroelectrochemical
measurements for some time now.'33'136 The latest review of this ﬁeld has been

published by Dai et a1, 136. 137

H. Fuel cell and power devices

In the development of fuel cells for transportation, there is a need to develop
advanced catalyst-support materials that can stably function, i.e., without
microstructural or morphological degradation or lost catalytic activity at the elevated
temperatures (150—200 °C) in aggressive chemical environments. spz-bonded
amorphous carbons and graphite are the most commonly used support materials. A
limitation of these supports is their susceptibility to corrosion, in the presence of
oxygen or in aggressive chemical environments, at elevated temperature.138 The high
surface area carbon used as fuel cell electrocatalyst (nanoparticles consisting of Pt, Ru
and other metals) supports are being targeted for replacement with diamond to take

advantage of the superior stability and chemical modiﬁability of diamond.‘39 The

22

modiﬁcation of BDD powder with metallic oxides, using the sol-gel method to
prepare high area and very stable electrodes for the MeOH oxidation reaction in

MeOH fuel cell, has been reported.'40

J. Superconductivity

Another attractive property of boron-doped diamond is its superconductivity.
The recent observation of superconductivity emerging near the insulator-to-metal
transition in heavily B-doped diamond has stimulated a discussion on the fundamental
origin of the metallic states responsible for the superconductivity.”" ”2 The physical
properties of lightly doped semiconductors are well described by electronic band-
structure calculations and impurity energy levels. Such properties form the basis of
present-day semiconductor technology. If the doping concentration (11) exceeds a
critical value (m), the system passes through an insulator-to-metal transition and
exhibits metallic behavior; this is widely accepted to occur as a consequence of the
impurity levels merging to form energy bands. Researches show that the doping-
dependent occupied electronic structures are consistent with the intrinsic diamond
bands, indicating that holes in the diamond bands play an essential part in determining
the metallic nature of the heavily B-doped diamond superconductor. This supports
the diamond band approach and related predictions, including the possibility of

143

achieving dopant-induced superconductivity in Si and Ge. This conclusion is still

controversial. 136’ ”2

1.3 Motivation and research questions

BDD exhibits a low background current, making it useful in analytical

applications where low concentrations of analyte need to be detected. ”4’ 145 It also has

23

a wide potential window in aqueous media of 3 - 4 V, which allows electrochemically
active species to be detected that may be masked by the solvent decomposition and
surface reactions at other electrode materials. Other desirable properties, such as the
chemical inertness, as well as its hardness and robustness, have led to applications
where electrode stability is important, such as in wastewater treatment. However, the
relationship between the physical, chemical and electronic properties of BDD and its
electrochemical performance is still incompletely understood. In recent years, several
groups have made progress learning how different factors control the electrochemical
properties. For example, several authors have investigated the kinetics and
mechanisms of heterogeneous redox reactions at BDD, and two general models for

electron-transfer have emerged:3’ ”6"“

(i) electron-transfer mediated by electronic
states within the band gap of BDD with the rate of the electron-transfer being
relatively uniform across the surface but controlled by the potential-dependent density
of states and (ii) electron-transfer through a partially blocked electrode that consists
of highly conductive or active sites isolated by less conductive or inactive sites.
Evidence supporting the second model comes from investigations of electron
emission from diamond; a process that involves electron-transfer across a solid-
vacuum interface and is analogous to electron-transfer across a solid-liquid interface.
It has been reported that boron-doped diamond ﬁlms with higher electrical
conductivity exhibited a lower turn-on voltage for ﬁeld emissionlsz’153 The electron

emission was observed to occur from sites of high electrical conductivity. The

possibility of some channel and/or grain boundary conduction was proposed.'52’ '53

Developing an understanding of electrochemical reaction rates at diamond and

their heterogeneity (i.e., the spatial nature of the electrochemical activity) across an

24

electrode requires the ability to characterize and spatially distinguish the physical,
chemical and electrical properties. Without question, polycrystalline diamond is one
of the most complex carbon electrodes with factors such as the boron-doping level,
hydrogen content, grain boundary density and electric double layer structure affecting
electron-transfer rates. For boron-doped microcrystalline diamond, several papers
describe the spatially inhomogeneous nature of the physicochemical properties, such
as the variable distribution of boron in doped polycrystalline diamond.'54’ ‘55 Raman
imaging, electrogenerated chemiluminescence imaging and photoluminescence
measurements have revealed variations in the microstructure, electrical properties and
electrochemical activity across a microcrystalline diamond thin-ﬁlm surface.'56"58
Recently reported electron emission studies from hydrogen-terminated and
chemically-modiﬁed boron-doped nanocrystalline diamond ﬁlms were consistent with
a heterogeneous surface having different local work functions.159 This conclusion was
reached based on electron energy distribution measurements that were made as a
function of temperature. If a complete picture of the structure-reactivity relationship at
boron-doped polycrystalline diamond is to be developed, then probes of the ﬁlm
properties at the microscopic level are needed.

Conducting probe-atomic force microscopy (CP-AFM) is a technique that is
useful for investigating the electrical properties of conductors and semiconductors
with high spatial resolution.'6°’ '61 Both nanoscale electrical characterization and
topographic imaging are provided by CP-AF M. The technique usually employs a
metal-coated (Pt or Au) cantilever-tip assembly that functions as both a scanning
electrical contact and force sensor. Metal wire probe tips can also be used.162 Thus,

direct measurements of I-V curves at speciﬁc topographical sites on an electrode are

possible. CP-AFM has been used to probe the conductivity across hydrogen-

25

terminated BDD thin-ﬁhn electrodes with different boron-doping levels,163 oxygen-
tenninated BDD164 and undoped diamond ﬁlms.165 The general observation, at least in
some of the work, is that the electrical conductivity is variable across polycrystalline
diamond with a strong dependence on the grain orientation, fraction of grain boundary
and the boron-doping level. 163' 165 On the other hand, Macpherson and coworkers
observed much less spatial variability in the electrical and electrochemical properties
across smooth polished, oxygen-terminated diamond.‘02

Our group has reported on the heterogeneous nature of the microstructure and
electrochemical activity of polycrystalline BDD using Raman microprobe
spectroscopy and electrogenerated chemiluminescence imaging.l5 6’ 15 7 Additionally, a
pattern of variable conductivity and electrochemical activity at hydrogen-terminated
boron-doped microcrystalline diamond electrodes, with different boron doping levels,
were measured using CP-AF M and SECM.'63’ "’6

Although work has been performed to study the electrical and electrochemical
properties of boron-doped diamond thin-ﬁlm electrodes, they still possess some
characteristics that are not fully understood from a fundamental view point. Principle
among these is the mechanism of electron exchange between the diamond surface and
solution redox species, especially the effect of the electrode surface chemistry. To
gain better insight on heterogeneous electron transfer process, research was conducted
to answer the following questions:

1. What is the spatial uniformity of the electrical conductivity of boron-doped
microcrystalline and nanocrystalline diamond thin ﬁlms?

2. What factors contribute to the electrical conductivity (e. g., boron-doping

level, surface hydrogen, surface oxygen and nitrogen compensators)?

26

3. How does the uniformity of the electrical properties correlate with the
spatial uniformity of the electrochemical reaction rates across the electrode surface for

various outer-sphere redox systems?

1.4 Research goals and objectives

The research described herein investigated (i) the electrochemical behavior of
BDD thin-ﬁlm electrodes toward different redox systems using cyclic voltammetry,
(ii) the surface chemistry of BDD electrodes and the effect on electron-transfer
kinetics for various redox systems, (iii) the uniformity of the electrical conductivity
and electrochemical activity across the diamond electrode and gaining a better

understanding of structure-function relationship of these interesting features.

1.5 Dissertation outline

This dissertation is divided into eight chapters. Chapter 1 provides background
information about boron-doped diamond: its properties and applications. In Chapter 2,
experimental details are given about the grth of doped diamond and the
electrochemical measurements performed. Included in this chapter are details of the
surface modiﬁcations used, experimental details for the characterizations of boron-
doped diamond thin ﬁlms including scanning electron microscopy, Raman
spectroscopy and electrochemical measurements. The experimental details about the
in vitro determination of norepinephrine at the blood vessel surface are also presented.
Chapters 3 - 5 present results for the characterization of boron-doped diamond thin
ﬁlm electrodes. In this part, the physical, chemical, and electrochemical properties of
two types of polycrystalline diamond thin ﬁlms are studied. The surface

microstructure and electrochemical characterization of microcrystalline and

27

nanocrystalline boron-doped diamond electrodes are described in Chapter 3. Chapter
4 presents work on applying conductive-probe atomic force microscopy to two types
of polycrystalline BDD thin ﬁlms to investigate the relationship between electrical
conductivity and doping level, morphology, bias level, bias polarity and surface
termination. The relationship between electrochemical activities of BDD thin ﬁlms
and grain size, doping level, surface terminations and redox systems studied via
scanning electrochemical microscopy is presented in Chapter 5. Chapter 6 compares
the physical, chemical and electrochemical properties of two nanocrystalline BDD
ﬁlms from different sources. Chapter 7 describes preliminary work on the use of
boron-doped diamond microelectrodes to measure norepinephrine overﬂow from
sympathetic neuroeffector functions at mesenteric arteries and veins in the presence
and absence of sensory nerve input. Chapter 8 summarizes the conclusions reached

from this work and identiﬁes future direction of the research.

28

Ilti lieyﬁerenrcne

(1)

(2)

(3)

(4)

(5)

(6)

(7)
(8)
(9)

(10)
(11)
(12)
(13)
(14)
(15)
(16)

Fong, C. Y.; Klein, B. M. In Diamond: Electronic properties and applications;
Pan, L. 8., Kania, D. R., Eds.; Kluwer Academic Publishers: Norwell, MA,
1995, pp 3.

Xu, J.; Granger, M. C.; Chen, 0.; Strojek, J. W.; Lister, T. E.; Swain, G. M.
Anal. Chem. 1997, 69, 59lA-597A.

Granger, M. C.; Witek, M.; Xu, J .; Wang, J .; Hupert, M.; Hanks, A.; Koppang,
M. D.; Butler, J. E.; Lucazeau, G.; Mermoux, M.; Strojek, J. W.; Swain, G. M.
Anal. Chem. 2000, 72, 3793-3804.

Spitsyn, B. V.; Bouilovand, L. L.; Derjaguin, B. V. J. Cryst. Growth. 1981, 52,
219-226.

Matsumoto, 8.; Sato, Y.; Kamo, M.; Setaka, N. Jpn. J. Appl. Phys. Part 2
1982, 21, L183-L185.

Buijnsters, J. G.; Rao, R. V. S.; Shankar, P.; Enckevort, W. J. P. V.; Schermer,
J. J.; Gebert, A.; Meulerr, J. J. t. Surface & Coatings Technology 2005, 191,
119-126.

Natishan, P. M.; Morrish, A. Mater. Lett. 1989, 8, 269.
Swain, G. M. J. Electrochem. Soc. 1994, 141, 3382.

Field, J. E. The Properties of Natural and Synthetic Diamond; Academic Press:
New York, 1992.

Sawabe, A.; Inuzuka, T. Thin Solid Films 1986, 13 7, 89.

Ferry, D. K. Phys. Rev. B 1975, 12, 2361.

Angus, J. C.; Hayman, C. C. Science 1988, 241, 913-921.

Tsarkova, 0. G. Physics of Wave Phenomena 2007, 15.

Coe, S. E.; Sussmann, R. 8. Diamond Relat. Mater. 2000, 9, 1726-1729.
Landstrass, M. 1.; Ravi, K. V. Applied Physics Letters 1989, 55, 975-977.

Cui, J. B.; Ristein, J.; Ley, L. Phys. Rev. Lett. 1998, 81, 429-432.

29

(17)

(18)

(19)

(20)

(21)

(22)

(23)

(24)
(25)
(26)
(27)

(28)

(29)

(30)

(31)

(32)

Isberg, J .; Hammersberg, J .; Johnansson, E.; WiksrtrOm, T.; Twitchen, D. J .;

Whitehead, A. J .; Coe, S. E.; Scarsbook, G. A. Science 2002, 297, 1670-
1672.

Sze, S. M. Physics of Semiconductor Devices, 2 ed.; Wiley-Interscience: New
York, 1981.

Miyajirna, 8.; Kim, M.; Ide, Y.; Yamada, A.; Konagai, M. Jpn. J. Appl. Phys.
2003, 42, 3328-3332.

Pleskov, Y. V.; Sakharova, A. Y.; Krotova, M. D.; Bouilov, L. L.; Spitsyn, B.
P. J. Electroanal. Chem. 1987, 228, 19.

Birrell, J .; Carlisle, J. A.; Auciello, 0.; Gruen, D. M.; Gibson, J. M. Appl.
Phys. Lett. 2002, 81, 2235.

Koizumi, 8.; Kamo, M.; Sato, Y.; Ozaki, H.; Inuzuka, T. Appl. Phys. Lett.
1997, 71, 1065-1067.

Okano, K.; Kiyota, H.; Iwasaki, T.; Nakamura, Y.; Akiba, Y.; Kurosu, T.; Iida,

M.; Nakamura, T. Appl. Phys. A 1990, 51, 334.

Alexenko, A. E.; Spitsyn, B. V. Diamond Relat. Mater. 1992, I, 705-709.
Frenklach, M.; Speak, K. E. J. Mater. Res. 1988, 3, 133-140.

Eversole, W. G.; US. Patent No. 3030188, 23.07.1958.

Angus, J. Q; Will, H. A.; Stanko, W. S. J. Appl. Phys. 1968, 39, 2915-2922.

Derjaguin, B. V.; Fedosseev, D. V.; Lukyanoich, V. M.; Spitsyn, B. V.;
Ryabov, V. A.; Lavrentyev, A. V. J. Cryst. Growth 1968, 2, 380.

Asmussen, J.; Reinha, D. K. Diamond Films Handbook; Marcel Dekker: New
York, 2002.

Spataru, N. ; Rao, T. N.; Tryk, D. A.; Fujishima, A. J. Electrochem. Soc. 2001,
148, E112.

Spataru, N.; Sarada, B. V.; Popa, E.; Tryk, D. A.; Fujishima, A. Anal. Chem.
2001, 73, 514.

Prado, C.; Flechsig, G. U.; Grundler, P.; Foord, J. 8.; Marken, F.; Compton, R.
G. Analyst 2002, 127, 329.

30

(33)

(34)
(35)
(36)

(37)

(38)

(39)

(40)

(41)

(42)

(43)

(44)

(45)

(46)

(47)

(48)

(49)

Ohnishi, K.; Einaga, Y.; Notsu, H.; Terashima, C.; Rao, T. N.; Park, S. G.;
F ujishima, A. Electrochem. Solid State Lett. 2002, 5, D1.

Uchikado, R.; Rao, T. N.; Tryk, D. A.; Fujishima, A. Chem. Lett. 2001, 144.
Davis, R. F. J. Cryst. Growth 1994, I3 7, 161.
Bachmann, P. K.; Messier, R. Chem. Eng. News 1989, 67, 24.

Ramesham, R.; Hill, D. 0; Best, 8. R.; Rose, M. F .; Askew, R. F. Thin Solid
Films 1995, 25 7, 68.

Saterlay, A. J.; Wilkins, S. J.; Holt, K. B.; Foord, J. 8.; Compton, R. G.;
Marken, F. J. Electrochem. Soc. 2001, 149, E66.

Holt, K. B.; Campo, J. D.; Foord, J. 8.; Compton, R. G.; Marken, F. J.
Electroanal. Chem. 2001, 513, 94-99.

Granger, M. C.; Swain, G. M. J. Electrochem. Soc. 1999, 146, 4551-4558.

Terashima, C.; Rao, T. N.; Sarada, B. V.; Tryk, D. A.; Fujishima, A. Anal.
Chem. 2002, 74, 895.

Rodrigo, M. A.; Michaud, P. A.; Duo, 1.; Panizza, M.; Cerisola, G.;
Comninellis, C. J. Electrochem. Soc. 2001, 148, D60.

Ramesham, R. Thin Solid Films 1999, 339, 82.

Hagans, P. L.; Natishan, P. M.; Stoner, B. R.; 0'Grady, W. E. J. Electrochem.
Soc. 2001, I48, E298-E301.

Alfraro, M. A. Q.; Ferro, 8.; Martinez-Huitle, C. A.; Vong, Y. M. J. Braz.
Chem. Soc. 2006, I 7, 227-236.

Canizares, P.; Saez, C.; Lobato, J.; Rodrigo, M. A. J. Chem. T echn. Biotechn.
2006, 81, 352-358.

Song, Y.; Swain, G. M. Anal. Chim. Acta 2007, 593, 7-12.

Fang, N.; Jia, J .; Zhong, D.; Wang, Y. Huangjing Wuran Yu Fangzhi 2007, 29,
708-712.

Debus, O. Chemie T echnik (Germany) 2004, 33, 40—42.

31

(50)

(51)

(52)

(53)

(54)

(55)

(56)

(57)

(58)

(59)

(60)

(61)

(62)

(63)

(64)

Comninellis, 0; Duo, 1.; Michaud, P.-A.; Marselli, B.; Park, S.-M. In
Diamond Electrochemistry; Fujishima, A., Ed.; Elsevier: Amsterdam, 2005,
pp 449-476.

Chen, 0.; Granger, M. C.; Lister, T. E.; Swain, G. M. J. Electrochem. Soc.
1997, 144, 3806-3812.

Jolley, 8.; Koppang, M.; Jackson, T.; Swain, G. M. Anal. Chem. 1997, 69,
4099.

Stolte, 8.; Abdulkarim, 8.; Aming, J .; Blomeyer-Nienstedt, A.-K.; Bottin-
Weber, U.; Matzke, M.; Ranke, J.; Jastorff, B.; Thoeming, J. Green Chemistry
2008, 10, 214-224.

Linke-Wienemann, V.; Baune, M.; Thoeming, J. Vom Wasser 2006, 104, 7-11.
Ruediger, A.; Ruediger, 1.; Jurisevic, L. Water Sci. T echn. 2007, 55, 195-200.

Ouattara, L.; Chowdhry, M.; Comninellis, C. New Diamond and Frontier
Carbon Technology 2004, 14, 239-247.

Li, X.; Wang, J.; Chen, D.; Liu, F.; Zhang, Z. Qinghua Daxue Xuebao 2006,
46, 641-644.

Canizares, P.; Lobato, J .; Paz, R.; Rodrigo, M. A.; Saez, C. Water Research
2005, 39, 2687-2703.

Waterson, K.; Wang, J. W.; Bejan, D.; Bunce, N. J. J. Appl. Electrochem.
2006, 36, 227-232.

Arevalo, E.; Calmano, W. J. Hazard. Mater. 2007, 146, 540-545.

Kober, R.; Stadler, R.; Griesbach, U.; Hoppe, C.; Lorenz, K-H.; Huber, R.;

Rycherr, P.; Proven, C.; Pupunat, L.; Daum, A.; Matthies, A.; Doetzer, R.

Method and device for the treatment of wastewater containing pesticides; W0
2006061 192: Germany, 2006061 5.

Belaid, C.; Kalle], M.; Khadlrraou, M.; Lalleve, G.; Elleuch, B.; Fauvarque, J .-
F. J. Appl. Electrochem. 2006, 36, 1175-1182.

Canizares, P.; Martinez, L.; Paz, R.; Saez, C.; Lobato, J .; Rodrigo, M. A. J.
Chem. T echn. Biotechn. 2006, 81 , 1331-1337.

Panizza, M.; Kapalka, A.; Comninellis, C. Electrochimica Acta 2008, 53,
2289-2295.

32

(65) Swain, G. M. In Electroanalytical Chemistry; Bard, A. J ., Rubinstein, I., Eds.;
Marcel Dekker, 2004; Vol. 22, pp 181-277.

(66) Ivandini, T. A. 8., Rika; Makide, Yoshihiro; F ujishima, Akira; Einaga,
Yasuaki Diamond Relat. Mater. 2004, 13, 2003-2008.

(67) Manivannan, A.; Ramakrishnan, L.; Seehra, M. 8.; Granite, E.; Butler, J. E.;
Tryk, D. A.; Fujishima, A. J. Electroanal. Chem. 2005, 577, 287-293.

(68) Manivannan, A.; Seehra, M. 8.; Fujishima, A. Fuel Proc. T echn. 2004, 85,
513-519.

(69) Salimi, A.; Alizadeh, V.; Hallaj, R. T alanta 2006, 68, 1610-1616.

(70) Dragoe, D. 8., Nicolae; Kawasaki, Ryuji; Manivannan, Ayyakkannu; Spataru,
Tanta; Tryk, Donald A.; Fujishima, Akira. Electrochimica Acta 2006, 51 ,
243 7-2441.

(71) Kruusma, J.; Banks, C. E.; Compton, R. G. Anal. Bioanal. Chem. 2004, 379,
700-706.

(72) Banks, C. E.; Hyde, M. E.; Tomlmageik, P.; Jacobs, R.; Compton, R. G.
T alanta 2004, 62, 279-286.

(73) Babyak, C.; Smart, R. B. Electroanalysis 2004, 16, 175-182.

(74) Goodwin, A.; Lawrence, A. L.; Banks, C. E.; Wantz, F.; Omanovié, D.;
Komorsky-Lovric, 3.; Compton, R. G. Anal. Chim. Acta 2004, 533, 141-145.

(75) Banks, C. E; Kruusma, J.; Moore, R. R.; Tombr’k, P.; Peters, J.; Davis, J.;
Komorsky-Lovrié, 8.; Compton, R. G. T alanta 2005, 65, 423-429.

(76) Zhang, Y.; Yoshihara, S. J. Electroanal. Chem. 2004, 573, 327-331.

(77) Tall, 0. E.; Jaffrezic-Renault, N.; Sigaud, M.; Vittori, 0. Electroanalysis 2007,
19,1152-1159.

(78) McGaw, E. A.; Swain, G. M. Anal. Chim. Acta 2006, 575, 180-189.

(79) Denisenko, A.; Jamommam, G.; El-Hajj, 11.; Kohn, E. Diamond Relat. Mater.
2007, 16, 905-910.

(80) Suzuki, A. I., Tribidasari A.; Kamiya, Akihide; Nomura, Satoshi; Yamanuki,
Mikito; Matsumoto, Koichi; Fujishima, Akira; Einaga, Yasuaki Sensors and
Actuators B: Chemical 2007, 120, 500-507.

33

(81) Ivandini, T. A.; Sato, R.; Makide, Y.; Fujishima, A.; Einaga, Y. Diamond
Relat. Mater. 2005, 14, 2133-2138.

(82) Provent, C. H., Werner; Santoli, Eduardo; Rychen, Philippe. Electrochimica
Acta 2004, 49, 3737-3744.

(83) Song, Y.; Swain, G. M. Anal. Chem. 2007, 79, 2412-2420.

(84) Oliveira, R. T. S.; Garbellini, G. 8.; Salazar-Banda, G. R.; Avaca, L. A. Anal.
Lett. 2007, 40, 2673-2682.

(85) Murata, M. I., Tribidasari A.; Shibata, Mamoru; Nomura, Satoshi; Fujishima,
Akira; Einaga, Yasuaki J. Electroanal. Chem. 2008, 612, 29-36.

(86) Chailapakul, 0.; Amatatongchai, M.; Wilairat, P.; Grudpan, K; Nacapricha, D.
Talanta 2004, 64, 1253-1258.

(87) Lawrence, J .; Robinson, K. L.; Lawrence, N. 8.; Schlumberger, K. K. Anal.
Sci. 2007, 23, 673-676.

(88) Suryanarayanan, V. Z., Y.; Yoshihara, 8.; Shirakashi, T. Sensors and
Actuators B: Chemical 2004, 102, 169-173.

(89) Preechaworapun, A.; Chuanuwatanakul, S.; Einaga, Y.; Grudpan, K.;
Motomizu, 8.; Chailapakul, 0. T alanta 2006, 68, 1726-1731.

(90) Weiss, E.; Groenen-Serrano, K.; Savall, A. J. Appl. Electrochem. 2007, 37,
1337-1344.

(91) Spataru, T.; Spataru, N.; Fujishima, A. T alanta 2007, 73, 404-406.
(92) Wu, J.; Qu, Y. Anal. Bioanal. Chem. 2006, 385, 1330-1335.

(93) Zhao, W.; Xu, J .-J .; Qiu, Q.-Q.; Chen, H.-Y. Biosensors &Bioelectronics 2006,
22, 649-655.

(94) Olivia, H.; Sarada, B. V.; Honda, K.; Fujishima, A. Electrochimica Acta 2004,
49, 2069-2076.

(95) Lei, Y.; Zhao, G.; Liu, M.; Xiao, X.; Tang, Y.; Li, D. Electroanalysis 2007, 19,
1933-1938.

(96) Muna, G. W.; Tasheva, N.; Swain, G. M. Environ. Sci. T echnol. 2004, 38,
3674-3682.

34

(97) Muna, G. W.; Quaiserova-Mocko, V.; Swain, G. M. Anal. Chem. 2005, 77,
6542-6548.

(98) Muna, G. W.; Quaiserova-Mocko, V.; Swain, G. M. Electroanalysis 2005, I 7,
1 160-1 170.

(99) Pedrosa, V. A.; Malagutti, A. R.; Mazo, L. H.; Avaca, L. A. Anal. Lett. 2006,
39, 2737-2748.

(100) Shin, D.; Tryk, D. A.; Fujishima, A.; Muck, A. J.; Chen, G.; Wang, J.
Electrophoresis 2004, 25, 3017-3023.

(101) Suryanarayanan, V.; Zhang, Y.; Yoshihara, 8.; Shirakashi, T. Electroanalysis
2005, I 7, 925-932.

(102) Boonsong, K.; Chuanuwatanakul, 8.; Wangfuengkanagul, N.; Chailapakul, 0.
Sensors and Actuators B: Chemical 2005, 108, 627-632.

(103) Oliveira, R. T. 8.; Salazar-Banda, G. R.; Ferreira, V. 8.; Oliveira, S. C.; Avaca,
L. A. Electroanalysis 2007, 19, 1189-1194.

(104) Medeiros, R. A.; Evaristo de Carvalho, A. R.-F., Romeu C.; Fatibello-Filho,
Orlando. Anal. Lett. 2007, 40, 3195-3207.

(105) Codognoto, L.; Tanirnoto, S. T.; Pedrosa, V. A.; Suffredini, H. B.; Machado, S.
A. 8.; Avaca, L. A. Electroanalysis 2006, 18, 253-258.

(106) Wangfuengkanagul, N.; Siangproh, W.; Chailapakul, 0. T alanta 2004, 64,
1183-1188.

(107) Charoenraks, T. Anal. Sci. 2005, 21, 241-245.

(108) Treetepvij it, 8.; Chuanuwatanakul, 8.; Einaga, Y.; Sato, R.; Chailapakul, 0.
Anal. Sci. 2005, 21, 531-535.

(109) Wu, J .; Lei, C.; Yang, H.; Wu, X.; Shell, G.; Yu, R. Sensors and Actuators B:
Chemical 2005, I 07, 509-515.

(110) Chailapakul, 0.; Siangproh, W.; Tryk, D. A. Sens. Lett. 2006, 4, 99-119.

(111) Nebel, C. E.; Rezek, B.; Shin, D.; Uetsuka, H.; Yang, N. J. Phys. D: Appl.
Phys. 2007, 40, 6443-6466.

(112) Poon, M.; McCreery, R. L.; Engstrom, R. Anal. Chem. 1988, 60, 1725-1730.

35

 

(113)

(114)

(115)

(116)

(117)

(118)

(119)

(120)

(121)

(122)

(123)

(124)
(125)

(126)

(127)

(128)

Alehashem, 8.; Chambers, F.; Strojek, J. W.; Swain, G. M.; Ramesham, R.
Anal. Chem. 1995, 67, 2812-2821.

Hochstetler, S. E.; Puopolo, M.; Gustincich, S.; Raviola, E.; Wightrnan, R. M.
Anal. Chem. 2000, 72, 489-496.

Pessiglione, M.; Seymour, B.; Flandin, G.; Dolan, R. J .; F rith, C. D. Nature
2006, 442, 1042-1045.

Suzuki, A.; Ivandini, T. A.; Yoshimi, K.; Fujishima, A.; 0yama, G.; Nakazato,
T.; Hattori, N.; Kitazawa, 8.; Einaga, Y. Anal. Chem. 2007, 79, 8608-8615.

Weng, J.; Xue, J.; Wang, J.; Ye, J.-S.; Cui, H.; Sheu, F.-S.; Zhang, Q. Adv.
F unc. Mater. 2005, 15, 639-647.

Roy, P. R.; Saha, M. 8.; Okajima, T.; Park, S.-G.; Fujishima, A.; Ohsaka, T.
Electroanalysis 2004, 16, 1777-1784.

Rang, H. P. Pharmacology; Edinburgh: Churchill Livingstone, 2003.

Park, J.; Galligan, J. J .; Fink, G. D.; Swain, G. M. Anal. Chem. 2006, 78,
6756-6764.

Park, J .; Quaiserovaia-Mocko, V.; Peckovaiéi, K.; Galligan, J. J .; Fink, G. D.;
Swain, G. M. Diamond Relat. Mater. 2006, 15, 761-772.

Park, J.; Galligan, J. J.; Fink, G. D.; Swain, G. M. J. Physiol. 2007, 584, 819-
834.

Nekrassova, 0.; Lawrence, N. 8.; Compton, R. G. Electroanalysis 2004, 16,
1285-1291.

Zhao, G.; Qi, Y.; Tian, Y. Electroanalysis 2006, 18, 830-834.
Salimi, A.; Mamkhezri, H.; Hallaj, R. T alanta 2006, 70, 823-832.

Patel, B. A.; Bian, X.; Quaiserovaa-Mocko, V.; Galligan, J. J .; Swain, G. M.
Analyst 2007, 132, 41-47.

Xie, 8.; Shafer, G.; Wilson, C. 6.; Martin, H. B. Diamond Relat. Mater. 2006,
15, 225-228.

Zhang, J.; 0yama, M. Microchem. J. 2004, 78, 217-222.

36

(129)

(130)

(131)

(132)

(133)

(134)

(135)

(136)

(137)

(138)

(139)

(140)

(141)

(142)

(143)

(144)

Ivandini, T. A. H., Kensuke; Rao, Tata. N.; Fujishima, Akira; Einaga, Yasuaki
T alanta 2007, 71, 648-655.

Canizares, P.; Arcis, M.; Saez, C.; Rodrigo, M. A. Electrochem. Comm. 2007,
9, 2286-2290.

Murakami, T.; Nohira, T.; Araki, Y.; Goto, T.; Hagiwara, R.; Ogata, Y. H.
electrochem. Solid St. Lett. 2007, 10, E4-E6.

F ardel, R.; Griesbach, U.; Puetter, H.; Comninellis, C. J. Appl. Electrochem.
2006, 36, 249-253.

Zak, J. K.; Butler, J. E.; Swain, G. M. Anal. Chem. 2001, 73, 908.

Stotter, J.; Zak, J.; Behler, Z.; Show, Y.; Swain, G. M. Anal. Chem. 2002, 74,
5924.

Stotter, J .; Show, Y.; Wang, 8.; Swain, G. Chem. Mater. 2005, I 7, 4880-4888.

Dai, Y.; Proshlyakov, D. A.; Zak, J. K.; Swain, G. M. Anal. Chem. 2007, 79,
7526-7533.

Dai, Y.; Swain, G.; Porter, M. D.; Zak, J. Anal. Chem. 2008, 80, 14-22.

Hupert, M.; Muck, A.; Wang, J .; Stotter, J .; Cvackova, Z.; Haymond, 8.;
Show, Y.; Swain, G. M. Diamond Relat. Mater. 2003, 12, 1940-1949.

Gonzalez-Gonzalez, 1.; Tryk, D. A.; Cabrera, C. R. Diamond Relat. Mater.
2006, 15, 275-278.

Salazar-Banda, G. R.; Eguiluz, K. I. B.; Avaca, L. A. Electrochem. Comm.
2006, 9, 59-64.

Sidorov, V. A.; Ekimov, E. A.; Bauer, E. D.; Mel'nik, N. N.; Curro, N. J.;
Fritsch, V.; Thompson, J. D.; Stishov, S. M.; Alexenko, A. E.; Spitsyn, B. V.
Diamond Relat. Mater. 2005, 14, 335-339.

Ohta, Y. New Diamond and Frontier Carbon technology 2007, I 7, 33-44.

Yokoya, T.; Nakamura, T.; Matsushita, T.; Muro, T.; Takano, Y.; Nagao, M.;
Takenouchi, T.; Kawarada, H.; Oguchi, T. Nature 2005, 438, 647-650.

Swain, G. M.; Ramesham, R. Anal. Chem. 1993, 65, 345-351.

37

(145)

Granger, M. C.; Xu, J .; Strojek, J. W.; Swain, G. M. Anal. Chim. Acta 1999,
397, 145-161.

(146) Compton, R. G.; Foord, J. 8.; Marken, F. Electroanalysis 2003, 15, 1349-1363.

(147)

(148)
(149)

(150)

(151)

(152)

(153)

(154)

(155)

(156)

(157)

(158)

(159)

i (160)

(161)

Ferreira, N. G.; Silva, L. L. G.; Corat, E. J .; Trava-Airoldi, V. J. Diamond and
Relat. Mater. 2002, 11, 1523-1531.

Becker, D.; Jﬁttner, K. J. Appl. Electrochem. 2003, 33, 959-967.
Ferro, 8.; De Battisti, A. Electrochimica Acta 2002, 47, 1641-1649.

Lagemaat, J. v. d.; Vanmaekelbergh, D.; Kelly, J. J. J. Electroanal. Chem.
1999, 475, 139-151.

Fischer, A. E.; Show, Y.; Swain, G. M. Anal. Chem. 2004, 76, 2553-2560.

Kim, Y. D.; Choi, W.; Wakimoto, H.; Usami, 8.; Tomokage, H.; Ando, T.
Diamond Relat. Mater. 2000, 9, 1096-1099.

Iseri, Y.; Honda, M.; Kim, Y.-D.; Ando, T.; Choi, W.; Tomokage, H. J. Phys. :
Condens. Matter 2004, 16, 8171-8179.

Ushizawa, K.; Watanabe, K.; Ando, T.; Sakaguchi, 1.; Nishitani-Gamo, M.;
Sato, Y.; Kanda, H. Diamond Relat. Mater. 1998, 7, 1719-1722.

Ushizawa, K.; Nishitani-Gamo, M.; Watanabe, K; Sakagushi, I.; Sato, Y.;
Ando, T. J. Roman Spectrosc. 1999, 30, 957-961.

Wang, J.; Swain, G. M.; Mermoux, M.; Lucazeau, G.; Zak, J .; Strojek, J. W.
New Diamond Front. Carbon T echnol. 1999, 9, 317-343.

Mermoux, M.; Marcus, B.; Swain, G. M.; Butler, J. E. J. Phys. Chem. B 2002,
106, 10816-10827.

Honda, K.; Noda, T.; Yoshimura, M.; Nakagawa, K.; Fujishima, A. J. Phys.
Chem. B 2004, 108, 16117-16127.

Robinson, V. 8.; Show, Y.; Swain, G. M.; Reifenberger, R. G.; Fisher, T. 8.
Diamond Relat. Mater. 2006, 15, 1601-1608.

Loiacono, M. J .; Granstrom, E. L.; Frisbie, C. D. J. Phys. Chem. 1998, 102,
1679-1688.

Kelley, T. W.; Granstrom, E. L.; Frisbie, C. D. Adv. Mater. 2005, 11, 261-264.

38

(162) Macpherson, J. V.; Mussy, J. P. G. d.; Delplancke, J. L. Electrochem. Solid St.
Lett. 2001, 4, E33-E36.

(163) Holt, K. B.; Bard, A. J.; Show, Y.; Swain, G. M. J. Phys. Chem. B 2004, 108,
15117-15127.

(164) Wilson, N. R.; Clewes, S. L.; Newton, M. E.; Unwin, P. R.; Macpherson, J. V.
J. Phys. Chem. B 2006, 110, 5639-5646.

(165) Temyak, 0.; Cimmino, A. A.; Prawer, 8.; Hoffman, A. Diamond Relat. Mater.
2005, 14, 272-278.

(166) Wang, S.; Swain, G. M. J. Phys. Chem. C2007, 111, 3986-3995.

39

 

Chapter 2

Experimental Section

2.1 Preparation and characterization of boron-doped

diamond thin-ﬁlm electrodes

2. 1. 1 Chemical vapor deposition conditions

Microwave plasma-assisted CVD was used to grow both microcrystalline and
nanocrystalline diamond thin ﬁlms. 1'3 For all growths, boron was introduced by
adding diborane to the source gas mixture. A block diagram of a typical reactor is
shown in Figure 2.1. Brieﬂy, the source gases (e.g., Ar, H2, CH4, and B2H6) enter the
chamber at controlled rates regulated by mass ﬂow controllers. Microwave energy is
focused into the quartz cavity and ignites a plasma. The plasma, which is in close
proximity to the substrate, contains the reactive species necessary for growth. The
chamber pressure is maintained by a throttle exhaust valve and rotary pump that are
operated continuously during ﬁlm growth.

The speciﬁc deposition conditions used for the growth of each type of
diamond thin ﬁlm on Si are summarized in Table 2.1. All source gases were ultrahigh
purity (99.999%). The substrate was highly conducting p-Si (111) (0.05 cm thick X 1
cm2 in area, ~ 10'3 ohm-cm, Virginia Semiconductor Inc., Fredricksburg, VA) and
was prepared for ﬁlm grth by mechanically scratching the surface with 0.1 llm
diameter diamond powder (GE Superabrasives, Worthington, OH) slurried in water

on a felt polishing pad. The scratched substrate was then copiously rinsed with ultra-

40

 

 

Microwave l
Generator .s

 

 

 

 

   
  

 

 

Mass Flow
Controllers

 

 

Quartz
Chamber

 

 

Substrate
Holder

    
 
  
  

 

   

 

 

 

 

 

 

 

 

 

 

  

Gas Inlet .;.:=::. 5'5 ,
Pressure
Gauge

 

l 1 Valve

 

H2 cylinder

 

 

R0131")! Pump :To vent

 

 

CH4 cylindeD-IX}

B2H6 cylinder N

 

 

 

 

 

 

 

Figure 2.1 Schematic commercial MPCVD system (ASTex Inc., Lowell, MA)
used for growing microcrystalline and nanocrystalline boron-doped diamond thin

ﬁlms.

pure water (~18 MQ-cm, Bamstead E-Pure), isopropyl alcohol (IPA, Fisher
Scientiﬁc), acetone (Mallinckrodt), IPA, and ultrapure water to remove all polishing

debris from the surface. Embedded diamond powder particles, as well as the polishing
striations, serve as the initial nucleation sites for diamond growth. A high density of
these sites leads to a high instantaneous nucleation density, which is necessary for
forming a continuous ﬁlm at a low nominal thickness. Microcrystalline boron—doped
diamond was deposited at 1000 W using a CH4/H2/B2H6 source gas mixture
consisting of 0.5% CH4 in 99.5% H2 with 10, 5, 1, 0.5 or 0 ppm diborane (0.1% B2Hr,
diluted in H2) added for boron doping. The system pressure was 45 torr, the substrate

temperature was ca. 750 °C, as estimated via an optical pyrometer, and the growth

41

 

time was 10 h. At the end of the deposition period, the CH4 and B2H6 gas ﬂows were
stopped and the ﬁlms remained exposed to an H2 plasma at 1000 W and 45 torr for an
additional 10 min. The plasma power and pressure were then slowly reduced over 5
min to cool the sample in the presence of atomic hydrogen to a temperature estimated
to be less than 400 °C. The plasma power was then turned off and the ﬁlms were
further cooled to room temperature under a ﬂow of H2. Post-grth annealing in
atomic hydrogen served to gasify any adventious nondiamond sp2 carbon impurity, to
minimize dangling bonds, and to ﬁrlly hydrogenate the surface.

Nanocrystalline diamond was deposited at 800 W using a CH4/H2/Ar/B2H6
source gas mixture consisting of 1% CH4 and 5 % H2 in 94% Ar, with 50, 10, 1 or 0
ppm diborane (0.1% B2H6 diluted in H2) added for boron doping. The system pressure
was 140 torr, the substrate temperature was ca. 750 °C, as estimated via an optical
pyrometer, and the grth time was 2 h. At the end of the deposition period, the CH4
and B2f16 gas ﬂows were stopped and the ﬁlm remained exposed to an Ar/H2 plasma
at 800 W and 140 torr for an additional 10 min. The Ar ﬂow was then stopped, and
the plasma power and pressure were then slowly reduced over 5 min to cool the
sample in the presence of atomic hydrogen to a temperature below 400 °C. The
plasma power was then turned off and the ﬁlms were further cooled to room
temperature under a ﬂow of H2.

Boron-doped nanocrystalline diamond thin ﬁlms, deposited on conducting Si,
were also provided from Dr. James Butler (Naval Research Laboratory, Washington,
DC, USA). The exact deposition conditions were not disclosed. These ﬁlms were

used for the comparison with our ﬁlms, as discussed in Chapter 6.

42

Table 2.1 Growth conditions for microcrystalline and nanocrystalline boron-

doped diamond thin ﬁlms deposited on Si.

 

 

 

 

 

 

 

 

 

Microcrystalline Nanocrystalline
Microwave Power 1000 W 800 W
Chamber Pressure 45 Torr 140 Torr
Gas Mixture CH4/H2/B2H6 Ar/I-I2/CH4/B2H6
Methane Concentration 0.5% CH4 1% CH4
Total Gas Flow 200 sccm 100 sccm
Substrate Temperature ~750 °C ~750 °C
Growth Time 10 h 2 h

 

 

 

2.1.2 Characterization of Boron-doped diamond thin ﬁlms

A. Morphology

In an attempt to better understand the inﬂuence of the diamond particle size on
the electrical conductivity and electrochemical activity, both microcrystalline and
nanocrystalline of boron-doped diamond thin ﬁlms were studied. Typical diamond

ﬁlm thicknesses, resistivities, carrier concentrations, and carrier mobilities are listed

in Table 2.2.

Table 2.2 Film thickness, resistivity, carrier concentration, and carrier mobility

for typical microcrystalline and nanocrystalline boron-doped diamond thin ﬁlms.“

 

 

 

 

 

 

 

Carrier Carrier
Type Film Thickness Resistivity Concentration Mobility
(”l“) ( -cm) (cm’3) (cm2/V-s)
Microcrystalline ~ 57 0010.1 ~10” 1040
Nanocrystalline ~ 5 0.2 ~1020 0.1-1

 

 

43

 

 

B. Doping level

Boron is the most efﬁcient dopant for diamond because of the high
concentration that can be achieved without signiﬁcant alteration in the diamond
microstructure, and the low activation energy for carrier production (0.34 eV). It is
not possible to uniformly dope diamond as various grth sectors uptake different
concentrations of the dopant.7'9 Ushizawa et a]. studied how the addition of boron can
affect various factors including crystal morphology and quality.10 They found that a
small amount of diborane added to the source gas improves the diamond crystal
quality. Characteristic changes occur in the spectra for heavily doped compared to
lightly doped ﬁlms.1| More recently, Tokuda et a]. reported surface roughening of
heavily boron-doped diamond ﬁlms caused by the boron concentrations of 10'7 to 5 x
1020 atoms/cm3 .12’ 13 Obviously, the most important role of boron in p-type diamond is
as a dopant. The boron concentration in a ﬁlm determines the electrical conductivity.
Boron substitutes for carbon atoms in the growing lattice and also aggregates in grain
boundariess’ 14’ '5 The former boron is the electrically active type. Therefore, in this
work, both microcrystalline and nanocrystalline diamond ﬁlms with different doping
levels were studied to learn how the addition of boron affects the ﬁlm morphology,
electrical conductivity and electrochemical activity. The properties of the different
microcrystalline and nanocrystalline diamond ﬁlms used in this work are listed in

Table 2.3.

The dOping level (B/C ratio in the ﬁlm) for both ﬁlms increases proportionally
with the concentration of diborane in the source gas mixture. Therefore, in this
dissertation, we refer to the ﬁlms in terms of the diborane concentration used during

ﬁlm deposition.

 

 

 

 

 

 

 

 

 

 

Table 2.3 Diamond ﬁhns with different doping levels used in this work.
Film [B2116] in B/C ratio .
Type thickness source gases in ﬁlm [Bzcigglm 1:3isgg?
(um) (ppm) (ppm)*
0 48 8.5 x 10'8 5180
0.5 230 4.0 x 1019 37.93
Microcrystalline ~ 57 1 325 5.7 x 10'9 0.591
5 2280 4.0 x 1020 0.185
10 3285 5.8 x 1020 0.016
0 58 1.0 x 1019 14638
1 247 4.3 x 10‘9 41.81
Nanocrystalline ~ 5 2 0
10 1401 2.5 x 10 0.264
50 2144 3.8 x 1020 0.0186

 

 

 

 

 

 

 

 

*: Measured by boron nuclear reaction analysis at Case Western Reserve University.
Courtesy of Dr. Alan Mcllwain.

**: Calculated using p = 4.532dr, where d is the ﬁlm thickness (d = 6 pm for
microcrystalline ﬁlms and d = 5 pm for nanocrystalline ﬁlms), r is the resistance

measured in a 4-point conﬁguration using spring-loaded tips.

2.1.3 Surface termination

Since the discovery of “surface conductivity (SC)” by Landstrass and Ravi,l6’
‘7 many studies have shown that the surface termination can inﬂuence the electrical
properties of diamond's'Z'. For boron-doped diamond grown via MPCVD, the “as
grown” diamond surface is hydrogen-terminated because the deposition occurs in a
hydrogen-rich plasma. However, this surface is not immune to oxidization. Therefore,
knowledge of the interaction of hydrogen and oxygen on diamond surface is crucial to

the application of BDD thin ﬁlms in electrochemistry. It is well known that BDD thin

45

ﬁlms with different surface terminations exhibit different performance in
electrochemistry and electron emission6’ '4’ 22’ 23, but the relationship between the
electrical properties and the microstructure and the surface chemistry of
polycrystalline diamond, especially for nanocrystalline diamond, remains unclear. To
address the issue of surface heterogeneity, hydrogen and oxygen terminated boron-

doped diamond ﬁlms were prepared.

A. Oxidation of the diamond ﬁlm

Surface oxidation was accomplished by exposing the ﬁlm to warm (~ 60 °C)
aqua regia (3 parts hydrochloric acid/1 part nitric acid, V/V) for 30 min to oxidatively
remove metallic impurities. The electrode was then copiously rinsed with ultrapure
water, isopropyl alcohol, acetone, IPA, and ultrapure water after cooling to room
temperature. The electrode was then immersed in a warm 30% peroxide hydrogen
solution (Columbus Chemical Industries, Inc.) for another 30 min for further
oxidative removal of nondiamond Sp2 carbon impurity. This was followed by rinsing
with pure water, IPA, acetone, IPA, and ultrapure water. These chemical treatments
not only clean the surface but also introduce surface carbon-oxygen functional groups.
For example, the XPS atomic oxygen to carbon ratio increased from <0.02 to ca. 0.15
after this wet chemical oxidation. Consistent with a hydrophilic, oxygenated surface,

the contact angle for water decreased from >70° to <5°. 6

B. Hydrogenation of the diamond film
Surface hydrogenation was achieved by ﬁrst cleaning the diamond surface in
acid, as described above. This was followed by exposing the oxidized ﬁlm to a

hydrogen plasma. The plasma was generated by MACVD using a commercial reactor

46

(1.5 kW, 2.54 GHz, ASTex Inc., Lowell, MA) at 1000 W using pure H2 gas. The
system pressure was 45 torr, the substrate temperature was ca. 750 °C, as estimated
via an optical pyrometer, and the treatment time was 30 min. At the end of the
hydrogenation, the plasma power and pressure were slowly reduced over 5 min to
cool the sample in the presence of atomic hydrogen to a temperature estimated to be
less than 400 °C. Surface hydrogen, which is the hydrogen bonded to the terminal
carbon atoms, does not desorb at these temperatures. The plasma power was then

turned off and the ﬁlms were further cooled to room temperature under a ﬂow of H2.

C. Heat treatment of the diamond ﬁlm

In 1989, when Ravi and Landstrass ﬁrst reported SC on single crystal and
CVD diamond, the phenomenon was exclusively attributed to the presence of
hydrogen.”’ '7 However, in recent years the role of atmospheric species in the
formation of a hole accumulation layer near the ﬁlm surface has even established. It is
well established that a p-type surface conductive layer (SCL) exists at the surface of
as-grown CVD or hydrogen-terminated poly- and single-crystalline diamond,
independent of crystallographic orientation, although the removal of SCL on different
crystallographic orientations requires different conditions?" 24’ 25 The SCL has been
attributed to p-type carriers with a lateral concentration of up 10l3 cm'2 and a mobility
between 30 and 100 cmZ/Vs showing only weak temperature dependence in the range
of 120 — 400 K.'8' '9 The mobility is similar to the hole mobility measured for boron-
doped polycrystalline diamond.26 Thus, there is general agreement that the carriers are
holes residing in an accumulation layer close to the surface. Furthermore, it has been
shown that the conductivity of the SCL is closely linked to an adsorbate layer, which

covers the diamond surface when exposed to air.27’ 28 However, the mechanism by

47

 

which the hole accumulation layer forms is still controversial. Meier et 01.29 have
proposed an electrochemical model in which electrons are transferred from the
diamond valence band to solvated H30+ ions in a water layer that spontaneously
forms in air. This model is supported by recent theoretical30 and experimental
studieszo' 2'

0n the other hand, experiments have shown that exposure to different gases
can affect the conductivity by many orders of magnitude,31 and that the SCL can be
eliminated by oxidation treatments."" '7 In these cases, the SCL on polycrystalline or
homoepitaxial (001) diamond surfaces was removed and high resistive diamond was
generated. Therefore, many researchers have pointed out that hydrogen atoms play an
impOltant role in the formation of SCL. From these results, it is generally accepted
that no SCL exists on oxidized diamond surfaces. This may not be true. Recent

research by Riedel et a]. conﬁrmed that the SCL can not be eliminated by oxidation

treatment. Their results showed that SCL on the (001) surface was maintained in an

ozone-rich atmosphere up to a partial oxygen coverage of -~40%.32

To test the role of subsurface hydrogen, we investigated the electrochemical
behavior of hydrogen terminated boron-doped diamond thin-ﬁlm electrodes with
different doping levels before and after annealing in Ar at 400 °C. At this temperature
the chemisorbed surface hydrogen remain in tact but B-H couples are broken. Cyclic
voltammograms for Fe(CN)63‘/4‘, Ru(NH3)(53+f2+ and Irle”' were recorded using
hrdogen-terminated diamond electrodes. The concentration for all redox species was
0.1 mmol/L, unless otherwise stated, and the supporting electrolyte was 1 mol/L KCl.
After electrochemical measurements, all electrodes were acid-washed, as described in
2.1.3A and rehydrogenated as described in section 2.1.3B. The electrodes were then

heat treated in a tube furnace to desorb the hydrogen. The system is shown in Figure

48

2.2. Ar was passed continuously through the tube for 30 min before the fumace was
switched on and adjusted to temperature. The temperature was ramped at a rate of 20
°C/min until 400 °C was reached. Once reached, the annealing was done for 30 min
at 400 °C. The power of the furnace was adjusted to decrease the temperature of the
chamber back to room temperature at -20 °C/min. One of the electrodes was then
taken out of the chamber alter cool-down and mounted in an electrochemical cell for

immediate electrochemical studies. The others were kept in the furnace under an Ar

purge until use.

Vent

521%

Electrical Furnace Diamond Samples /

\ /

 

 

 

 

 

 

 

 

 

 

 

c I 02:54:» ....... :3:
Power Temperature \ ArCylinda'
® ® QuaﬂZTube

 

 

 

 

Figure 2.2 The tube furnace heating system used for heat treating diamond thin

ﬁlms in Ar.
2.1.4 Boron-doped diamond microelectrodes

49

 

A. Growth of boron-doped diamond thin films on Pt wire

Boron-doped diamond microelectrodes were prepared by coating sharpened Pt
wire substrates by microwave-assisted chemical vapor deposition (CVD).33’ 34 There
are only a few reports describing the preparation and basic electrochemical
characterization of diamond microelectrodes35'39. Diamond microelectrodes have been
found to provide superior detection ﬁgures of merit, as compared to carbon ﬁbers in
terms of linear dynamic range, limit of detection, response variability and stability“).
The Pt wire was ﬁrst etched electrochemically in 1 M KOH to form a conical shape.
An etching solution containing 7.0 g of CaCI2°H20 in a mixture of 20 mL of ultrapure
water and 20 mL of acetone also functioned well.”’“ A 1.4 cm long piece of Pt wire
(99.99%, Aldrich Chemical, 76 um diam) was used. Both ends of the wire were
sharpened so that two electrodes could be made from one wire. The wire end was
immersed to a depth of l to 2 mm and positioned in the center of four connected
carbon rod counter electrodes during the etching, as shown in Figure 2.3. An AC
polarization of 12 V (60 Hz) was applied between the wire and the counter electrodes
using a variable autotransformer (Staco Energy Products, Dayton, OH). The etching
procedure lasted for approximately 5 3 until gas evolution visibly ceased at the tip.

The etched wire was conically-shaped near the end and cylindrically-shaped
above. Prior to diamond deposition, the sharpened wire was ultrasonically cleaned in
acetone (10 min) and ultrasonically seeded (30 min) in a diamond powder suspension
(5 nm particles, ca. 20 mg in 100 mL of ethanol, Tomai Diamond Co., Tokyo, Japan).
In order to prevent damage to the sharpened end, the wire was cleaned and seeded
while vertically suspended in the agitated solution. During the seeding process, the
surface was scratched by the diamond particles with some particles getting embedded.

Both the scratches and the embedded particles likely serve as the initial nucleation

50

sites for ﬁlm growth. A high instantaneous nucleation density is desired because this
leads to the formation of a continuous ﬁlm in the shortest time with a low nominal
thickness. A thin ﬁlm of boron-doped diamond was then deposited on the sharpened,
cleaned and seeded wire using the same commercial CVD system we used for the
growth of the planar diamond thin ﬁlm. The pretreated Pt wires (~ 5-6 in parallel per
deposition run) were mounted horizontally on the top of a quartz plate (10 x 10 x 1
mm) in the reactor. The quartz plate was placed in the center of the reactor's
molybdenum substrate stage and served to thermally isolate the wires. The thin ﬁlm
of boron-doped diamond was deposited from a 0.5% CH4/H2 (V /V) source gas

mixture with 4 ~ 5 ppm of diborane (0.1 % B2H6 diluted in H2) added for doping.

variable autotransformer

\

 

AC Power supply

 

 

 

 

 

 

 

 

 

 

   
  

Pt wire bolder\

 

Carbon rods (x4)

 

Figure 2.3 Schematic of the assembly used to electrically sharpen the Pt wire

substrates.

51

Table 2.4 summarizes the growth conditions used for depositing boron-doped
diamond on the Pt wires. All source gases were ultrahigh purity grade (99.999%). The
system pressure was 45 Torr, the substrate temperature was ca. 300 °C, or less, (as
estimated with an optical pyrometer), the microwave power was 400 ~ 600 W and the
total gas ﬂow was 200 standard cubic centimeters per min (sccm) during growth. The
deposition time was 10 h. After deposition, the CH4 and B2H6 gas ﬂows were shut-off
and the ﬁlms cooled in the presence of atomic hydrogen (H2 plasma) by slowly
reducing the plasma power and system pressure over a 30 min period. This post-
growth treatment was necessary for removing adventitious sp2 carbon impurity,
minimizing dangling bonds and ensuring ﬁll] hydrogen termination. Under these

conditions, the nominal growth rate was estimated to be about 0.3 rim/h based on the

ﬁnal ﬁlm thickness of 3-5 pm. Once coated, the diameter of the cylindrical portion

was approximately 80 1111293136241

Table 2.4 Growth conditions for microcrystalline boron-doped diamond

microelectrode on Pt wires.

 

 

 

 

 

 

 

 

 

 

 

 

Microcrystalline
Microwave Power 400 ~ 600 W
Chamber Pressure 40 ~ 45 Torr
Gas Mixture CH4/H2/B2H6
Methane Concentration 0.5% CH4
B2H4 Concentration in gas mixture 4 ~ 5 ppm
Total Gas Flow 200 sccm
Substrate Temperature ~300 °C
Growth Time 10 h

 

52

B. Insulation of boron-doped diamond microelectrode

For the electroanalytical measurements, the diamond microelectrode
was attached to a Cu wire with conductive epoxy. Each electrode was then
sealed in polypropylene.“ ‘13 This was accomplished by inserting the
microelectrode into a pipette tip and carefully heating the tapered end using the
heating coil of a P-30 micropipette puller (Sutter Instrument Co. Novato, CA).

Details of the procedure are given below:

1. Cut all diamond-coated Pt wires into two pieces.

2. Cut copper wire into pieces of appropriate length (several cm’s).

3. Prepare conductive epoxy mixture.

4. Dip one end of the copper wire into the epoxy.

5. Contact the copper and Pt wire and allow epoxy to cure. Make sure
the conical end of the Pt wire (etched end) is not contacted by the
epoxy.

6. Leave the epoxy overnight in air.

7. Cut and insert the diamond/copper wire into a pipette tip.

8. Mount the pipette tip and wire in the micropipette puller, make sure
the diamond cone protrudes from the end of the tip.

9. Switch on the heating coil of the micropipette puller to heat up the tip.

10. Gently pull up the copper wire to keep same length of coated Pt wire

exposed when the micropipette tip starts to melt and shrink.

53

11. Turn off the heater when the hole in the tip is totally sealed and wait
for 2 ~ 3 s, then gently push down the copper wire to make the end
of the tip a conical shape.

12. Fix the copper wire inside the tip with normal epoxy and wait
another overnight before the microelectrode can be tested.

The ﬁnal product is schematized in Figure 2.4.

Conductive Epoxy Diamond-Coated Pt

 

Copper Wire

Polwrom’lene PM Tip

Figure 2.4 Diagram of the insulated boron-doped diamond microelectrode.

2.2 Scanning Electron Microscopy (SEM)

The SEM was performed using a JSM-6300F (ﬁeld emission) scanning
electron microscope (JEOL Ltd., Tokyo, Japan, Center for Advanced Microscopy at
Michigan State University). The images were constructed from both secondary and
backseattered electrons with an accelerating voltage of 10 kV.

It is difﬁcult to ﬁgure out the crystalline nature of nanodiamond ﬁlms from the

plan-view SEM images. Therefore, transmission electron microscopy (TEM) and

54

selected-area electron diffraction (SAED) were used to examine the phase
composition of the nanodiamond ﬁlms. All TEM and SAED images were performed
using a JEOL2200FS transmission electron microscope (JEOL Ltd., Tokyo, Japan,
Center for Advanced Microscopy at Michigan State University). The samples used for
TEM were prepared involving chemically etching and ion milling, as shown below.

1. Diamond thin ﬁlms grown on Si were immersed in a mixed acidic solution
of hydrochloride acid (HCl), hydroﬂuoride acid (HF) and acetic acid (HAc)
to get free-standing diamond ﬁlms.

2. Carefully cut the free-standing ﬁlms into pieces of ca. 3 x 3 mmz.

3. Stick a small piece of diamond ﬁlm onto a copper grid slot used for TEM.

4. Mount the grid slot to the sample holder of the ion milling instrument with
melted wax.

5. Set the milling conditions as follow:

High voltage: 4 kV, Incident angle: 4°, Milling mode: one side.
6. Carry out ion milling till the proper thickness of the diamond ﬁlm reaches

(~ 100 nm, the ﬁlm will be totally transparent).

2.3 Roman Spectra

Raman spectra were acquired at room temperature using a RAMAN 2000
spectrograph (Chromex, Inc., Albuquerque, NM) that consisted of a diode-pumped,
frequency-doubled continuous wave (CW) Nd:YAG laser (500 mW at 532 nm,
COHERENT), a Chromex 500 spectrometer (f/4, 600 grooves/mm holographic
grating), and a thermoelectrically cooled 1024 x 256 element charge-coupled device
(CCD) detector (ANDOR Tech, Ltd., South Windsor, CT). Spectra were collected

with an incident power density of ca. 500 kW/cm2 (100 mW at the sample and 5 pm

55

diameter spot size) and a 10 3 integration time. A white-light spectrum was collected
under the same conditions and served as the background for spectral correction. The
spectrometer was calibrated (wavelength position) with 4-acetamidophenol
(CH3C0NHC6H40H) and a high pressure, high temperature (HPHT), single crystal

diamond sample (ﬁrst-order phonon position = 1332 cm").

2.4 Electrochemical Characterization

2. 4. I Planar boron-doped diamond electrode

Electrochemical measurements were at the core of the material
characterization performed on the electrodes. Cyclic voltammetry was used to 1)
assess the electrochemical reaction kinetics for several redox systems at diamond thin
ﬁlms deposited on Si, 2) probe the electrochemical properties of diamond
microelectrodes. Basic theory and operating principles of the electrochemical
techniques used in this work can be found elsewhere." 44

All of the electrochemical measurements, were performed using a CHI650a or
CHI900 computerized potentiostat (CH Instruments Inc., Austin, TX) with a graphite
rod counter electrode and a Ag/AgCl reference electrode (3 M KC1, E °' = -56 mV vs.
SCE). All measurements were made at room temperature (25 :I: 1 °C), unless
otherwise noted, and all solutions were deoxygenated with N2 for at least 10 min prior

to a measurement and blanketed with the gas during the measurement. Details of the

electrochemical cells used for each type of electrode are provided below.

56

 

    
 

 

VitOIIQO-l'illg
Witt. ; é—=.'-;'.~'..-:‘.-.7:'. '- “'T- I "i-T-‘F-‘ii-V‘ " E .
Copper "“‘i .. -2. Worhng
.... “we" m m...

1333

l‘ . l.
.9 -'.‘;'4- -' ... ‘

 

Figure 2.5 Design of single compartment, 3 electrode electrochemical cell.

The diamond thin ﬁlm electrodes on Si were sealed against the bottom of a
single-compartrnent electrochemical cell with a Viton® o-ring (~ 0.2 cm2 area), as
shown in Figure 2.5.3 The backside of the Si substrate was scratched, cleaned, and
coated with graphite or carbon tape in order to ensure good ohmic contact with a
copper current collector plate. All electrodes were tested after a 20 min soak in

distilled isopropyl alcohol.

2. 4.2 Microelectrode

Both boron-doped diamond coated Pt microelectrodes and carbon ﬁber
microelectrodes were tested in a single-compartment, 3-electrode electrochemical cell

as schematized in Figure 2.6.

57

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

CHI650a
Potentiostat
L 1 L\
Computer
Electrolyte
Solution

 

 

 

Figure 2.6 Schematic of the electrochemical measurement with the diamond

microelectrodes.

Voltammetry were performed on CHI 6508 potentiostat (CH Instruments, Inc.,
Austin, TX) using 2 types of carbon ﬁber electrodes, thomel P-553 (d = 10 pm) and
T650 (d = 8 pm) (Amoco, Greenville, SC) and boron-doped diamond microelectrode
(d = 76 um). All measurements were recorded in a three electrode cell as shown in
Figure 2.6. The scan rate used for all redox systems was 0.1 V/s. The potential was
scanned between -0.2 and 0.8 V vs. AglAgCl on carbon ﬁber microelectrode (CFME)
or -0.2 and 1.0 V vs. Ag/AgCl on diamond microelectrode (DME) when probed with
Fe(CN)6344‘. When probed with Ru(NH3)6+3’+2, the potential scan range was from 0.2
V to -O.4 V (CF MB) or -0.6 V (DME) vs. Ag/AgCl. For the measurements involved

Krebs buffer, the potential scan range was between 0 and 1.2 V vs. Ag/AgCl.

58

Background cyclic voltammetry in 1M KC] was measured ﬁrst, then the cyclic
voltammogram for Fe(CN)r,'3/'4 , Ru(l\lH3)6+3’+2 , Krebs’ buffer solution and NE
solution were recorded, followed by the measurement of a drug. And then, cyclic
voltammogram in 1M KC], Fe(CN)(5'3/'4 , Ru(NH3)r,+3/+2 , Krebs buffer solution, and
NE redox systems were measured again to investigate the inﬂuence of the drug on the
electrochemical behavior of this microelectrode. The procedure described above was
one integrated measurement for one type of microelectrode in the experiment. The
other two types of microelectrode will follow the same procedure as measuring the
same drug. The protocol described above would be applied to other six drug
measurements using three types of microelectrodes. For statistic purpose, at least

three measurements were made for each drug and each type of microelectrode.

2. 5 Conductive-Probe Atomic Force Microscopy (CF-AFM)

To perform CP-AF M measurements, a commercial scanning probe
microscope (NanoScope IIIa scanning probe microscope, Veeco Instruments Inc., CA)
was remodeled as schematized in Figure 2.7.

The conductivity measurement was made using a homemade current-to-
voltage converter circuit (see Figure 2.8). This circuit converted the current ﬂowing
through the probe tip to a proportional output voltage that was then sent to a computer
via a Signal Access Module (V eeco Instruments Inc., CA). The measurable current

range was between i6.5 LIA.

59

SPM Head

 

 

 

 

 

 

 

 

 

 

\I

 

F ”1‘, i . . .
I SPM Controller 4::
Scanner 1 Computer
I

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Home-made current-to-
voltage converter Signal Access Module

Figure 2.7 Schematic of the CP-AFM system adapted to a NanoScope Illa scanning

probe microscope.

From Omni 90

   

To Signal
Access
Module

 

-15 v 4.7 m
Figure 2.8 Schematic of the home-made current-to-voltage converter.

A gold-coated, tetrahedron-shaped silicon nitride tip (V eeco NPG-20), with an

apex radius of curvature less than 50 run, a tip height of ~15-20 um, and a full-tip

60

cone angle less than 35°, was mounted on a rectangular cantilever. The boron-doped
diamond thin-ﬁlm was attached to a metal pod, and electrical contact was made with
carbon tape and Ag paint between the pod and backside of the conducting Si substrate
(Figure 2.9). A bias voltage (sample vs. tip) of i0.2 or i2 V from an Omni 90
potentiostat (Cypress Systems, Chelmsford, MA) was applied between the sample and
tip. Conductivity maps were recorded simultaneously with the topographic images.

All measurements were performed in the open air.

DC Power Supply

 

 

Conductive Tip

I1

Diamond Film

Highly Doped Si Substrate

Carbon Tape

 

AFM Scanner Covered with Teflon Tape

 

Figure 2.9 Schematic of electrical connection between sample and tip for CP-

AFM experimental setup.

2.6 Scanning Electrochemical Microscopy (SE CM)

SECM22'45' ‘6 measurements were made using a commercial microscope (CHI
900, CH Instruments Inc., Austin, TX) in a four-electrode conﬁguration. The

electrochemical cell consisted of a Pt wire auxiliary electrode and a 2 pm diameter Pt

61

ultramicroelectrode probe tip (CH Instruments Inc.). BDD was investigated as the
substrate electrode, and a homemade Ag/AgCl electrode was used as the reference.
The probe tip (2 pm Pt ultramicroelectrode) need to be cleaned prior to any
measurements. This can be achieved by running at least 20 cycles of cyclic
voltammetric scanning from 1.25 V to -0.3 V (vs. Ag/AgCI) in 0.1 mol/L HClO4
solution. Good shape of peak associated with hydrogen redox, adsorption and
desorption are seen on the recorded cyclic voltammogram indicating a well-cleaned
probe tip. Otherwise, the probe tip should be cleaned again or just replaced. Tip
placement for all SECM images presented was performed by approaching the
substrate surface in the positive-feedback mode until the current reached a value 2.3
times that for the tip far from the surface. At this point, the tip-to-substrate distance
was estimated to be about 1 um for a 2 11m diameter tip.46 Once positioned, the
protocol was then switched to map the electrochemical activity across the substrate
electrode using the constant distance mode with Ru(NH3)63+/2+, Fe(CN)63"4', and
IrC162'/3' (1 mmol/L) as the redox test systems. The supporting electrolyte for all three
redox systems was 1 mol/L KC]. All measurements, both tip approach curves and x-y

images, were made in the feedback or collection mode.

2. 7 Reagents for electrochemical measurements

All chemicals were analytical-grade quality, or better, and were used without
additional puriﬁcation. Solutions of l and 0.1 mmoVL potassium ferrocyanide
(Fe(CN),“4', Aldrich), hexaammineruthenium (III) chloride (Ru(NH3),3+4*, Aldrich),
potassium hexachloroiridate (IV) (IrClsz‘B', Aldrich), methyl viologen (1,1 ’-dimethyl-
4,4’-bipyridyl, MVZHHO, Aldrich) in 1 mol/L potassium chloride (Spectrum), and 0.1

mol/L perchloric acid (redistilled 99.999%, Aldrich) were made. All redox analyte

62

solutions were prepared daily using ultrapure water (>18 MQ-cm) from an E-pure
water puriﬁcation system in glassware that was cleaned by a three-step process:

ethanol/KOH bath, alconox/ultrapure water solution, and an ultrapure water rinse.

2. 8 Experimental methods for Chapter 7

2. 8.1 Animals

The University Committee on Animal Use and Care at Michigan State
University approved all procedures for handling and caring for the laboratory test
animals. All rats were allowed 2-3 days of acclimatization prior to entry into any
experimental protocol. Pelleted rat chow (Harlan/Teklad 8640 Rodent Diet;
Harlan/Teklad) and water were provided. Rats were housed in temperature and

humidity controlled rooms using a 12-h on/Off light cycle.

A. DOCA-salt hypertension

This model is chosen because its control state is easy to deﬁne unlike many
genetic states of hypertension.47 The DOCA-salt hypertension model is produced by
surgical uninephrectomization followed by implantation of the mineralocorticoid,
called deoxycorticosterone acetate (DOCA), and the administration of excess salt,
NaCl, in the drinking ﬂuid. Arterial blood pressure will signiﬁcantly rise in a few
weeks after such a treatment. DOCA salt-treatment causes signiﬁcant changes in the
cardiovascular system that contributes to the overall pathology of this model. Other

details about this model can be fund in literature.” 48

B. Preparation of DOCA -salt and sham rats

63

Male Sprague-Dawley rats (175-200 g, Charles River Inc., Portage, M1) were
exteriorized and the left kidney was removed under anesthesia with sodium
pentobarbital (50 mg/kg i.p. (intraperitoneal)) after ligation of the renal artery, vein
and ureter with 4-0 silk sutures. DOCA implantation was made between the back
shoulder blades of the animal using a 1 cm incision. DOCA implants (600 mg/kg)
were prepared by mixing deoxycorticosterone acetate in silicone rubber resulting in a
dose of 200 mgfkg. DOCA-implanted rats received standard pelleted rat chow and salt
water (1% NaCl + 0.2% KCl). Sham, norrnotensive rats also underwent left kidney
removal but no DOCA pellet implantation and received unsalted tap water. The
surgery was performed on a heated pad and rats recovered in a heated box. Antibiotics
(enroﬂoxacin, 5 mg/kg s.c.(subcutaneous)) and an analgesic (butorphanol tartrate, 2
mg/kg, so.) were administered immediately after the surgery. After recovery, the rats
were housed under the above listed conditions for 4 weeks. Systolic blood pressure
was measured using the tail-cuff method four weeks after surgery. Rats with mean

systolic blood pressure of > 150 mmHg were considered hypertensive.” 5°

2. 8.2 In vitro electrochemical and diameter measurement system

The combined use of continuous amperometry with diamond and carbon ﬁber
microelectrodes and video imaging were used for the simultaneous in vitro
measurement of NE release from sympathetic nerves innervating a rat mesenteric
artery (MA) and vein (MV) and the evoked contractile response.

These techniques, along with various drugs, were employed to show the
relationship between the oxidation current associated with endogenous NE overﬂow

and blood vessel constriction. A block diagram of the experimental set-up is shown in

Figure 2.10.

64

 

 

 

 

 

 

 

 

 

 

 

 

 

Microscope

Figure 2.10 Block diagram of the experimental set-up.

A. Tissue preparation

The sham and DOCA-salt rats (300 - 430 g,) were euthanized with a lethal
pentobarbital injection (50 mg, i.p.). The abdomen was surgically opened, and the
small intestine carefully removed and placed in an oxygenated (95% O2, 5% CO2)
Krebs buffer solution (pH 7.4) of the following composition: 117 mM NaCl, 4.7 mM
KCI, 2.5 mM CaCl2, 1.2 mM MgCl2, 25 mM NaHCO;, 1.2 mM NaH2PO4, and 11
mM glucose}0 A section of the mesentery close to the ileal wall was carefully out free
from the intestine and transferred to a small silicone elastomer-lined, Teﬂon ﬂow bath
(5 mL volume). Secondary or tertiary arteries and veins (180-330 um outside
diameter) were isolated for in vitro study by carefully removing the surrounding
adipose and connective tissue under a dissecting microscope using ﬁne scissors and
forceps. The Krebs buffer solution entered the bath at one end (center) and ﬂowed out
the opposite end (center). The bath, containing the ﬁxed tissue preparation, was

attached to the stage of an ACCU 3030 inverted microscope (ACCU-SCOPE Inc., NY)

65

and superfused continuously with 37 °C Krebs buffer at a ﬂow rate of 1.6 mlJmin.
The solution ﬂow was controlled by a MasterFlex US Economy drive (Cole-Pamer

Instrument Company, IL) peristaltic pump.

B. Assessment of electrode performance

The diamond and carbon ﬁber microelectrodes were cleaned by soaking in
distilled isopropyl alcohol (IPA) for at least 15 min prior to a measurements 1 The
microelectrode was afﬁxed to a micromanipulator (Fine Science Tools, CA), which
was used to position the electrode on the surface of the MA and MV. The
microelectrode was positioned in the center of the bath containing the ﬁxed tissue
preparation, equidistant from the inlet and outlet solution ports. A Pt wire counter and
a commercial "no leak" Ag/AgCl (3 M KC1, model EE009, Cypress Systems Inc.)
reference electrode were also mounted in the bath to complete the electrochemical cell.
All electrochemical measurements were made with a Biostat four channel potentiostat
(ESA Biosciences, Inc., MA) and a computer running Biostat software (version 2.2.],
BSA Biosciences, Inc., MA). The software was used to record the continuous
amperometric current-time proﬁles, blood vessel diameter changes and temperature of
the solution in the bath. The analog output current from the potentiostat was low-pass
ﬁltered (ﬁlter parameter = 5 in Biostat). All data were digitized using a sampling rate
of 100 Hz. The Krebs buffer ﬂowed over the electrode and the tissue sample for 30

min prior to the start of a series of measurements.

C. Focal stimulation of perivascular nerves

Short trains of electrical stimulation were used to trigger NE release.

Perivascular nerves were stimulated using a bipolar focal stimulating electrode

66

positioned on the surface of the blood vessel at a distance of ca. 200 pm from the tip
of the carbon ﬁber. This positioning minimized the noise introduced into the current
recordings. The focal stimulating electrode consisted of two AgCl-coated Ag wires
inserted into double barrel glass capillary (tip diameter = 180 um). The wires were
connected to a stimulus isolation unit and a Grass Instruments stimulator (S48K,
Grass Telefactor W. Warwick, RI). Trains of 60 stimuli (0.3 ms duration, 30-90 V) at

frequencies ranging between 2 and 20 Hz were used.

D. In vitro video monitoring of vasoconstriction

The output of a black and white video camera (OLY 105, Olympus America
Inc., USA) attached to the microscope was fed to a PC Vision Plus frame-grabber
board (Picolo Pro 2 board, Euresys Inc., IL) mounted in a personal computer.
Changes in blood vessel diameter of 0.1 pm could be resolved. Video images were
analyzed using Diamtrak software (http://www.diamtrak.com, Adelaide, Australia).
The digitized signal was converted to an analog output (NuDAQ 6208/6216 multi-
channel analog output card, ADLink Technology Inc., Taiwan) sent to Biostat
potentiostat and analyzed in a second computer running Biostat software for a

permanent recording of blood vessel diameter as function of time.

E. Protocol for in vitro electrochemical and diameter measurements

For the in vitro electrochemical and diameter measurements, we followed the
protocol given below:

1. Tissue preparation,

2. Set up the measuring system (Amperometry, V = 650 mV vs. Ag/AgCl),

3. Perfuse Krebs for 30 min,

67

4. Find out proper spots for stimulation and current measuring,

5. Record at least 3 stimulations in a 5 ~ 10 min interval as “control”,

6. Perfuse 1 nmol/L capsaicin in Krebs for 30 min,

7. Perfuse Krebs for 1 hr,

8. Record at least 3 stimulations as “after capsaicin”,

9. Perfuse 0.4 1.1M Substance P in Krebs for 10 min,

10. Record at least 3 stimulations as “with substance P”,

11. Perfuse Krebs for 1 hr,

12. Record at least 3 stimulations as “after substance P”.

All current-time and diameter-time traces were recorded by Biostat software

and stored in the computer for data treatment later.

2. 8.3 Fluorescence histochemistry

A. Tissue preparation

Second or third-order veins and arteries from the ileal mesentery, were
isolated by carefully clearing away the surrounding adipose and connective tissues.
Red blood cells were ﬂushed out of the blood vessel lumen with phosphate buffer (0.1
mol/L, pH 7.2) using a 30 ga. hypodermic needle and syringe, and tissues were cut to
1 cm length.
B. Glyoxylic acid ﬂuorescence histochemistry

Catecholamine ﬂuorescence was revealed after incubating the prepared tissues
in a 2% glyoxylic acid/0.2 mol/L phosphate buffer (pH 7.0) solution for 5 minutes at
room temperature. The blood vessels were stretched onto glass slides and then heated

at 80 °C for 5 min. Preparations were mounted in mineral oil and observed using a

68

ﬂuorescence microscope (Nikon Eclipse TE 2000-U) equipped with a ﬁlter set,

UV2E/C (excitation ﬁlter, 340-380 nm and emission ﬁlter, 435-485 nm).

C. Double immunohistochemistry

Mesenteric arcades were stretched tightly in a Sylgard-lined petri dish using
insect pins. A 30-gauge hypodermic needle was used to cannulate the primary artery
or vein and the arcade was ﬂushed with phosphate-buffered saline (PBS, 0.01 M, pH
7.2). The tissues were then stretched tightly on small pieces of balsa wood using
insect pins and were immersed in Zamboni ﬁxative (4% formaldehyde, 2% picric acid
in 0.1 M phosphate buffer, pH 7.4) overnight at 4 °C. The tissues were then cleared
three times at 10-min intervals with dirnethyl sulfoxide followed by 3 washes at 10-
min intervals with PBS. Arteries and veins were then dissected from mesenteric fat
and incubated overnight with primary antibodies. We used a mouse monoclonal anti-
tyrosine hydroxylase antibody (1 :200 dilution in PBS, Calbiochem, EMD Chemicals,
Inc., NJ), a rat monoclonal anti-substance P antibody (1 :200 dilution in PBS, Serotec
Laboratories, MorphoSys US Inc., NC) and a rabbit polyclonal anti-CGRP antibody
(1:200 in PBS, Chemicon, Temecula, CA) to localize immunoreactivity for SP and
CGRP, respectively. After incubation with primary antibodies, tissues were washed 3
times in PBS and then incubated with goat anti-rat IgG conjugated to Cy3 (1:200
dilution in PBS; Jackson Immunoresearch Laboratories, West Grove, PA), goat anti-
mouse IgG conjugated to Cy3 (1:200 dilution in PBS; Sigma-Aldrich, St. Louis, MO)
or to goat anti-rabbit ﬂuorescein isothiocyanate (1:40 dilution in PBS; Sigma-Aldrich,
St. Louis, M0) for 1 h at room temperature. The tissues were washed 3 times with
PBS and then mounted on microscope slides and coverslipped using buffered glycerol

(pH 8.6). Fluorescent images were acquired using a Leitz Laborlux S upright

69

microscope, a PL Fluotar 40x objective (0.7 NA), 8 SPOT-2 cooled color digital

camera (Diagnostic Instruments, Sterling Heights, MI).

2. 8.4 Chemicals and drugs

All chemicals were reagent-grade quality, or better, and used without
additional puriﬁcation. The chemicals and drugs used for electrode performance
testing were K4Fe(CN)6 (lmmol/L) and K3Ru(NH3)6 (lmmol/L) made with lmol/L
KC] solution, norepinephrine (20 umol/L), pyridoxal-phosphate-6-azophenyl-2',4'-
disulfonic acid (PPADS, 100 umol/L), [4-(4-amino-6,7-dimethoxy-quinazolin-2-yl)
piperazin-l-yl]-(2-furyl)methanone (prazosin, 100 umOI/L), l7a-hydroxy-yohimban-
l6a-carboxy1ic acid methyl ester (yohimbine, 100 umol/L), 5-bromo-N-(4,5-dihydro-
1H-imidazol-Z-yl)-6-quinoxalinamine (UK 14,304, 100 umOl/L), methyl
(1R,2R,3S,SS)-3- (benzoyloxy)-8-methyl-8-azabicyclo[3.2.1] octane-2-carboxylate
(cocaine, 100 umol/L), (3:)-2-[1,4-benzodioxan-2-yl]-2-imidazoline hydrochloride
(idazoxan, 100 umol/L) and 8-methyl-N-vanillyl-trans-6-nonenamide (capsaicin, 100
umol/L) made with Krebs buffer solution. Krebs buffer solution has the following
composition: 117 mmol/L NaCl, 4.7 'mmol/L KC], 2.5 mmol/L CaCl2, 1.2 mmol/L
MgCl2, 25 mmol/L NaHC03, 1.2 mmol/L NaH2P04, and 11 mmol/L glucose.so
Capsaicin (1 pmol/L) and substance P (0.3 umol/L) were used for in vitro study. All
chemicals and drugs except substance P were obtained from Sigma Chemical
Company (Saint Louis, MO). Glyoxylic acid, mouse monoclonal anti-tyrosine
hydroxylase antibody, rat monoclonal anti-Substance P antibody, rabbit polyclonal
anti-CGRP antibody, goat anti-mouse IgG conjugated to Cy3, goat anti-rat IgG

conjugated to Cy3 or to goat anti-rabbit ﬂuorescein isothiocyanate, Zamboni ﬁxative

7O

(4% formaldehyde, 2% picric acid in 0.1 M phosphate buffer, pH 7.4) were used for
ﬂuorescent staining. All these chemicals, antibodies and substance P were from Dr.
James J. Galligan at Department of Pharmacology and Toxicology, Michigan State
University. Ultrapure water (distilled, deionized, and passed over activated carbon,
17-18 MQ, Bamstead E-PureSystem) was used for all solution preparation, and

glassware and electrode cleaning.

2. 8. 5 Data analysis

Data were obtained from > 50 sham rats and >50 DOCA-salt rats for the
research. Mean systolic blood pressure from sham and DOCA-salt rats was 121 mm
Hg and 198 mm Hg, respectively. The mean weight of sham rats was 420 g, and the
mean weight of DOCA-salt rats was 310 g. Data are presented as mean 1 SEM;
where SEM is standard error of the mean. MA and MV constrictions are expressed

as a percentage of the initial resting diameter (in pm) of the blood vessel.

71

2.9

(1)
(2)

(3)

(4)

(5)

(6)

(7)

(8)

(9)

(10)

(11)

(12)

(13)

(14)

(15)

Reference
Angus, J. C.; Hayman, C. C. Science 1988, 241, 913-921.

Asmussen, J .; Reinha, D. K. Diamond Films Handbook; Marcel Dekker: New
York, 2002.

Field, J. E. The Properties of Natural and Synthetic Diamond; Academic Press:
New York, 1992.

Swain, G. M. In Electroanalytical Chemistry; Bard, A. J., Rubinstein, 1., Eds;
Marcel Dekker, 2004; Vol. 22, pp 181-277.

Show, Y.; Witek, M. A.; Sonthalia, P.; Swain, G. M. Chem. Mater. 2003, 15,
879-888.

Granger, M. C.; Witek, M.; Xu, J.; Wang, J .; Hupert, M.; Hanks, A.; Koppang,
M. D.; Butler, J. E.; Lucazeau, G.; Mermoux, M.; Strojek, J. W.; Swain, G. M.
Anal. Chem. 2000, 72, 3793-3804.

Locher, R.; Wagner, J .; Fuchs, F .; Maier, M.; Gonon, P.; Koidl, P. Diamond
Relat. Mater. 1995, 4, 678.

Spitsyn, B. V.; Bouilovand, L. L.; Derjaguin, B. V. J. Cryst. Growth. 1981, 52,
219-226.

Janssen, G.; Enckevort, W. J. P. v.; Vollenberg, W.; Giling, L. J. Diamond
Relat. Mater. 1992, I, 789-800.

Ushizawa, K; Watanabe, K.; Ando, T.; Sakaguchi, I.; Nishitani-Gamo, M.;
Sato, Y.; Kanda, H. Diamond Relat. Mater. 1998, 7, 1719-1722.

Bernard, M.; Deneuville, A.; Muret, P. Diamond Relat. Mater. 2004, 13, 282-
286.

Tokuda, N.; Saito, T.; Umezawa, H.; Okushi, H.; Yamasaki, 8. Diamond Relat.
Mater. 2007, 16, 409-411.

Tokuda, N.; Umezawa, H.; Saito, T.; Yamabe, K.; Okushi, H.; Yamasaki, 8.
Diamond Relat. Mater. 2007, 16, 767-770.

Wilson, N. R.; Clewes, S. L.; Newton, M. E.; Unwin, P. R.; Macpherson, J. V.
J. Phys. Chem. B 2006, 110, 5639-5646.

Wang, 8.; Swain, G. M. J. Phys. Chem. C2007, 111, 3986-3995.

72

(l6) Landstrass, M. 1.; Ravi, K. V. Applied Physics Letters 1989, 55, 975-977.
(17) Landstrass, M. 1.; Ravi, K. V. Appl. Phys. Lett. 1989, 55, 1391-1393.
(18) Jiang, N.; Ito, T. J. Appl. Phys. 1999, 85, 8267.

(19) Hayashi, K.; Yamanaka, S.; Watanabe, H.; Sekiguchi, T.; Okushi, H.;
Kajimura, K. J. Appl. Phys. 1997, 81, 744-753.

(20) Looi, H. J.; Pang, L. Y. S.; Molloy, A. B.; Jones, F.; Foord, J. 8.; Jackman, R.
B. Diamond Relat. Mater. 1998, 7, 550-555.

(21) Riedel, M.; Ristein, J.; Ley, L. Phys. Rev. B 2004,69, 125338.

(22) Holt, K. B.; Bard, A. J .; Show, Y.; Swain, G. M. J. Phys. Chem. B 2004, 108,
15117-15127.

(23) Colley, A. L.; Williams, C. G.; Johansson, U. D. H.; Newton, M. E.; Unwin, P.
R.; Wilson, N. R.; Macpherson, J. V. Anal. Chem. 2006, 78, 2539-2548.

(24) Foord, J. S.; Lau, C. H.; Hiramatsu, M.; Jackman, R. B.; Nebel, C. E.;
Bergonzo, P. Diamond Relat. Mater. 2002, 11, 856-860.

(25) Larsson, K.; Ristein, J. J. Phys. Chem. B 2005, 109, 10304-10311.
(26) LOOi, H. J.; Jackman, R. B.; Foord, J. 8. Appl. Phys. Lett. 1998, 72, 353.

(27) Ri, S.-G.; Nebel, C. E.; Takeuchi, D.; Rezek, B.; Tokuda, N.; Yamasaki, 8.;
Okushi, H. Diamond Relat. Mater. 2006, 15, 692-697.

(28) Ri, S.-G.; Takeuchi, D.; Nebel, C. E.; Tokuda, N.; Yamazaki, Y.; Yamasaki,
8.; Okushi, H. Diamond Relat. Mater. 2007, 16, 831-835.

(29) Maier, F.; Riedel, M.; Mantel, B.; Ristein, J .; Ley, L. Phys. Rev. Lett. 2000, 85,
3472-3475.

(30) Goss, J. P.; Jones, R.; Heggie, M. 1.; Ewels, C. P.; Briddon, P. R.; Oberg, 8.
Phys. Rev. B 2002, 65, 115207.

(31) Ri, 8.; Tashiro, K.; Tanaka, 8.; Fujisawa, T.; Kimura, H.; Kurosu, T.; Iida, M.
Jpn. J. Appl. Phys. Part 1 1999, 38, 3492.

(32) Riedel, M.; Ristein, J .; Ley, L. Diamond Relat. Mater. 2004, 13, 746-750.

73

(33) Park, J .; Show, Y.; Quaiserova, V.; Galligan, J. J .; Fink, G. D.; Swain, G. M. J.
Electroanal. Chem. 2005, 583, 56-68.

(34) Muna, G. W.; Quaiserova-Mocko, V.; Swain, G. M. Electroanalysis 2005, I 7,
1160-1170.

(35) Holt, K. B.; Hu, J.; Foord, J. 8. Anal. Chem. 2007, 79, 2556-2561.

(36) Cvacka, J .; Quaiserova, V.; Park, J.; Show, Y.; Alexander Muck, J.; Swain, G.
M. Anal. Chem. 2003, 75, 2678-2687.

(3 7) Park, J .; Quaiserovaia-Mocko, V.; Peckovaia, K.; Galligan, J. J .; Fink, G. D.;
Swain, G. M. Diamond Relat. Mater. 2006, 15, 761-772.

(38) Shin, D.; Sarada, B. V.; Tryk, D. A.; Fujishima, A.; Wang, J. Anal. Chem.
2003, 75, 530-534.

(39) COOper, J. B.; Pang, 8.; Albin, 8.; Zhong, J.; Johnson, R. M. Anal. Chem. 1998,
70, 464-467.

(40) Muna, G. W.; Quaiserova-Mocko, V.; Swain, G. M. Anal. Chem. 2005, 7 7,
6542-6548.

(41) Muna, G. W.; Tasheva, N.; Swain, G. M. Environ. Sci. T echnol. 2004, 38,
3674-3682.

(42) Zhou, Z.; Misler, S. J. Biol. Chem. 1996, 271, 270-277.
(43) Chow, R. H.; von Ruden, L.; Neher, E. Nature 1992, 356, 60-63.

(44) Bard, A. J .; Faulkner, L. R. Electrochemical Methods F undationals and
Applications, Second ed.; John Wiley & Sons Inc., 2001.

(45) Kwak, J.; Bard, A. J. Anal. Chem. 1989, 61, 1221-1227.
(46) Kwak, J.; Bard, A. J. Anal. Chem. 1989, 61, 1794-1799.

(47) Schenk, J. 0.; McNeil], J. H. J. Pharmacol. T oxicol. Methods 1992, 27, 161-
170.

(48) Park, J. Dissertation, Michigan State University, East Lansing, 2006.

(49) Fink, G. D.; Johnson, R. J .; Galligan, J. J. Hypertension 2000, 35, 464-469.

74

(50) Luo, M.; Hess, M. C.; Fink, G. D.; Olson, L. K.; Rogers, J.; Kreulen, D. L.;
Dai, X.; Galligan, J. J. Auton. Neurosci. 2003, 104, 47-57.

(51) Ranganathan, 8.; Kuo, T.-C.; McCreery, R. L. Anal. Chem. 1999, 71, 3574-
3580.

75

Chapter 3

Basic Properties of Boron-Doped Microcrystalline &
Nanocrystalline Diamond T hin-F ilm Electrodes

3.1 Introduction

Developing an understanding of how the electrochemical reaction rate varies
across a boron-doped diamond electrode requires that one characterize and spatially
distinguish the physical, chemical, and electrical properties. Without question,
polycrystalline diamond is one of the more structurally-complex carbon electrodes
with factors such as the boron-doping level, hydrogen content, grain boundary density,
and electric double layer structure affecting electron-transfer rates. Our group has
reported on the heterogeneous nature of the microstructure and electrochemical
activity of polycrystalline BDD.1"° As a follow-up to this, the present work included
detailed investigation of the electrical and electrochemical properties of boron-doped
diamond thin ﬁlm. In this chapter, we report the effect of boron-doping level and
surface termination on the physical and electrochemical properties of boron-doped
diamond electrodes fabricated by CVD. The diamond thin ﬁlms were
comprehensively characterized by scanning electron microscopy (SEM), transmission
electron microscopy (TEM), Raman spectroscopy, and cyclic voltammetry. Several
speciﬁcally selected, aqueous-based redox couples, including Fe(CN) 6344’,
Ru(NH3)63+/2+, IrC162°’3' and methyl viologen, were used to evaluate the

electrochemical properties of boron-doped diamond thin-ﬁlm electrodes.

76

3.2 Scanning Electron Microscopy (SEM) Images

3. 2. 1 SEM images for microcrystalline boron-doped diamond thin
ﬁlms

SEM images of typical microcrystalline BDD ﬁhns are presented in Figure 3.1.
Triangular crystalline facets, typical of polycrystalline diamond grown from
hydrogen-rich CH4/H2 source gas mixtures, are apparent. Many of the
microcrystallites have a base diameter of ca. 1 pm. The triangular {111}-oriented
crystal facets are the preferred growth phase, which is independent of the doping level.
The grain boundaries and facet surfaces, for the most part, are devoid of smaller
secondary diamond growths at all doping levels. As a general rule, the fraction of
large grains (~ 1 pm) increases but the fraction of small grains (less than 500 nm)
decreases as the doping level increases. The exception is the sample with a doping
level of 5 ppm (see Figure 3.1B). In other words, the surface of boron-doped diamond
became rougher when the doping level increased from 0 to 10 ppm (concentration of
diborane in source gas mixture). This is consistent with the results reported by Kim et
a].ll They observed nanoscale secondary grains to be on top of the isolated crystal,
and surface roughness generally increased with increasing doping concentration.
Twinning, which is evident as re-entrant grooves at the edge of some of the triangular
crystallites, is clearly seen for 10 ppm diamond ﬁlm but is not so apparent for other
ﬁlms.‘2

0n the basis of the SEM images, several conclusions can be drawn:

1. The triangular {111} crystal facets are the preferred growth phase for 0.5 —
10 ppm ﬁlms, which suggests that boron-doping, at least within this range, doesn’t

signiﬁcantly inﬂuence the growth orientation, although the addition of boron does

77

   
 

A. 10 ppm BDD ﬁlm
B. 5 ppm BDD ﬁlm
C. 1 ppm BDD ﬁlm
D. 0.5 ppm BDD ﬁlm
E. 0 ppm BDD ﬁlm

Figure 3.1 SEM images for microcrystalline diamond thin ﬁlms grown by
microwave-assisted chemical vapor deposition with different diborane levels in the
source gas: (A) 10, (B) 5, (C) l, (D) 0.5, and (E) 0 ppm. The source gas composition
was 0.5% CH4 + 99.5% H2, the plasma power was 1000 W, and the pressure was 45

torr. Growth time was 10 hours.

inﬂuence the growth rates of {111} facets (triangular) and {100} facets (square). ”"4

In fact, Wurzinger et al. found out that {111} facets predominate with increasing

78

boron concentration. '3 Ushizawa et a]. reported that {100} facets tend to predominate
when the boron-to-carbon ratio (B/C) in the source gas mixture is less than 80 ppm
but the situation reversed when B/C increased to 200 ppm.14 In other words, {111}
facets are the preferred grth phase. Our results are very similar to this because, in
our system, 0.5 ppm B2H6 concentration corresponds to a B/C of 200 ppm in the gas
phase. Therefore, having {111} facets predominate is not unexpected.

2. The surface of boron-doped diamond became rougher with increasing
diborane in the source gas, which indicates that the incorporation of boron does
inﬂuence the morphology of the diamond surface. This is also consistent with the
literature. Tokuda et 81. studied the surface roughening of heavily boron-doped
diamond ﬁlms using atomic force microscopy (AF M). Their results showed that the
root mean square (RMS) value of the heavily boron-doped diamond surfaces
increased with the increasing B/C ratio in the gas phase.” '6

3. Twinning is clearly evident in Figure 3.1. A twin band is a structure with a
pair of stacking errors in parallel {111} planes containing numbers of layers of correct
stacking between the errors.'2 A single stacking error is a rotation of adjacent layers
by 60° from their correct positions in the diamond-cubic crystal. The intersection of a
pair of parallel {111} stacking errors with the surface of a diamond-cubic crystal
gives rise to a re-entrant comer and these re-entrant corners play a critical role in the
growth of {111} faceted diamond crystals from the vapor because the re-entrant
comer provides a self-regenerating site for the nucleation of new layers and leads to
the rapid growth of the crystal within the plane of the twin band. Therefore, twin
bands can be easily seen in any growth where {111} facet is preferred. This is true for

all growths studied here. From the twining mechanism, one would expect that boron

would not inﬂuence the twin structure too much. As shown in Figure 3.1, all images

79

show {111} facets dominating the surfaces and many re—entrant grooves can be seen
independent of the boron concentration, except for undoped diamond ﬁlm (Figure

3.1E) because {111} facets didn’t dominate the surface.

3. 2.2 SEM images for nanocrystalline boron-doped diamond thin ﬁlms

The nanocrystalline diamond ﬁlm, compared to the microcrystalline ﬁlm, is
much smoother and has feature sizes of around 50 nm, as shown in Figure 3.2. These
nodular features are actually aggregates of diamond grains about 15-20 nm in
diameters’ 7 The smaller feature and grain size is typical of diamond ﬁlms deposited
from hydro gen-poor CH4/H2+Ar source gas mixtures and results from a high
renucleation rate during growth. 17

This set of SEM micrographs demonstrates the transition of the
nanocrystalline diamond microstructure as a function of the dopant concentration in
the source gas mixture. For the undoped nanocrystalline diamond ﬁlm, most of the
individual particles were found to be polycrystals, which had rugged surfaces,
showing little indication of well-developed crystal facets. This is a characteristic of
diamond single crystals (See Figure 3.2E). When 1 ppm diborane was added, particles
with rugged facets largely decreased in number although most of the individual
particles were still polycrystals with a smaller grain size (See Figure. 3.2D). Dramatic
changes happened when more than 5 ppm diborane was added to the source gas
mixture. Particles with rugged facets disappeared and the particle size decreased.
More importantly, a new morphology, a nodular structure emerged. The sizes of these
nodular features are ca. 50 nm which made a cauliﬂower-shape surface of the

nanocrystalline diamond ﬁlm (See Figure. 3.2A, B and C).

80

It is difficult to determine the crystalline nature of these ﬁlms from the plan-
view SEM images shown in Figure 3.2 because of the nodular features. Instead,

transmission electron microscopy (TEM) and selected-area electron diffraction

(SAED) were used to examine the phase composition of the nanodiamond ﬁlms.

 

50 ppm BDD ﬁlm
B. 10 ppm BDD ﬁlm
C. 5 ppm BDD ﬁlm
D. 1 ppm BDD ﬁlm
E. 0 ppm BDD ﬁlm

   

Figure 3.2 SEM images of boron-doped nanocrystalline diamond thin ﬁlms grown
by microwave-assisted chemical vapor deposition with different diborane
concentrations in the source gas: (A) 50, (B) 10, (C) 5, (D) 1, and (E) 0 ppm. The
source gas composition was 1% CH4 and 5 % H2 in 94% Ar with diborane, the plasma

power was 800 W, and the pressure was 140 torr. Growth time was 2 hours.

81

Figure 3.3 is a bright-ﬁeld transmission electron micrograph of a 10 ppm
boron-doped diamond ﬁlm with a fairly uniform distribution of conglomerates made
up of diamond crystallites. The black area in the center of the image was from the
instrument not the sample. The small dots uniformly distributed over the whole image
are diamond particles introduced in seeding procedure before the growth. The bright
traces are grain boundaries or caused by overlapping of different grains along the
incident electron beam direction.17 TEM image showed the structural homogeneity of

the nanocrystalline ﬁlm because of higher spatial resolution.

 

(1!) [mm

Figure 3.3 Bright-ﬁeld TEM image revealing the plan-view morphology of the
nanodiamond thin ﬁlm. The grain boundary contrast reveals that the ﬁlm possesses a

granular structure.

Selected-area electron diffraction (SAED) was used to reveal the phase purity
of the nanodiamond thin ﬁhn. The technique was used because either spz-bonded
glassy carbon or predominantly sp3-bonded diamond-like amorphous carbon (ta-C)
gives rise to recognizable diffraction intensities at positions different from those of the

crystalline carbons.l7 Figure 3.4 is a SAED pattern obtained from a typical

82

nanocrystalline diamond ﬁlm. Sharp Bragg reﬂections are located in concentric
circles, indicative of the ramdomness of the crystallite orientation. No appreciable
scattering intensity from either graphite crystallites or amorphous carbon (glassy

carbon or diamond-like amorphous carbon) was observed, suggesting that the ﬁlm is

phase-pure diamond.

 

Figure 3.4 SAED pattern from the nanodiamond thin ﬁlm. Sharp Bragg

reﬂections are visible indicating good crystallinity.

3.2.3 SEM images for boron-doped diamond microelectrodes

Figure 3.5A shows SEM images of a boron-doped diamond microelectrode
grown on a 76 pm diam. Pt wire and insulated by polypropylene. The conical shape
and the polycrystalline morphology are evident. There are two regions: the exposed
diamond and the isolated diamond. Because the polymer coating was not electrically
conductive, it appears brighter in SEM image. The exposed microelectrode, on the
other hand, is conductive so it appears darker. The polymer layer appears thin and

uniform over the electrode. The microelectrode diameter at the narrowest point is 10

83

um and at the widest point is ca. 80 pm. The length of this particular exposed
electrode is ~50 pm.

A higher magniﬁcation image of the coated wire is presented in Figure 3.5B.
The polycrystalline ﬁlm consists of randomly oriented, well-faceted crystallites
ranging in diameter ﬁom 0.5 to 2 pm. Triangular crystalline facets, typical of
polycrystalline diamond grown from hydrogen-rich CH4/H2 source gas mixtures, are
seen. Twinning is evident as re—entrant grooves at the edge of some of the triangular
crystallites. In contrast to the planar ﬁlms, there is a wide crystallite size distribution

with smaller secondary crystallites existing on the larger primary crystallite facets.

 

—
50 mm

Figure 3.5 SEM images of a boron-doped diamond microelectrode grown on a Pt
wire and insulated using polypropylene. The white line indicates the boundary

between polypropylene coating and exposed diamond ﬁlm.

3.3 Raman Spectra

3.3.1 Raman spectra for microcrystalline boron-doped diamond ﬁlms

on Si substrates

84

Raman spectroscopy is ﬁequently used to characterize CVD diamond ﬁlms,
because of its sensitivity to the carbon microstructure. Raman scattering in the visible
region is ca. 50 times more sensitive to the presence of spz-bonded carbon than sp3-
bonded carbon, which makes the technique a very useful probe of the nondiamond
phase in diamond thin ﬁlms.“"26 The Raman spectrum obtained for a 10 ppm boron-
doped microcrystalline diamond ﬁlm is shown in Figure 3.6.

The sharp peak at 1332 cm"l with a ﬁill width at half maximum of 8 cm“1 is the
characteristic one-phonon diamond line. This line is asymmetric with a dip on the low
wavenumber side of the peak and a shoulder on the high wavenumber side. Similar to
heavily doped silicon,27 this asymmetry is atnibuted to a Fano-type interference
between the scattering by the zone-center optical phonon and scattering by an
electronic continuum (related to the formation of an impurity band above 1019 B cm'3)

whose energy overlaps that of the phonon.28

, 522 cm“

    

1332 cm'1

   

 

  

1232 cm"

 

Intensity (arb. units)
1

 

 

I
I
: I 1450 cm'1
I : | 10 ppm
I . '
I I '
i ' I

a ﬁ I ' T ' l

500 1000 1500 2000
Raman Shift (cm'1)

Figure 3.6 Raman spectrum for a highly boron-doped microcrystalline diamond

ﬁlm on Si, deposited from a 10 ppm diborane source gas mixture.

85

Beside the appearance of a Fano asymmetry, a band near 1200 cm'1 is another
feature present in the spectrum for heavily doped ﬁlms. Along with this band, a very
broad and intense band around 500 cm'l is also seen. Some have attributed the band at
1200 cm"1 to bonds between boron atoms.29 However, a similar band was reported by
Buckley and co-workers for undoped diamond ﬁlms.30 Consequently, these bands do
not appear to be related to boron. The theoretical phonon density of states for
diamond3 I shows a maximum around 1200 cm", which matches closely with the band
observed for this ﬁlm. Gonon et al. proposed that the bands observed at 1200 and 500
cm'I are related to maxima in the phonon density of states of diamond and arise from
scattering by phonons outside the center of the Brillouin zone.28 May et a1. gave
another explanation for these hands. They also noted that the band is near two
maxima in the phonon density of states (PDOS). They attributed these bands to
relaxation of the wavevector selection rules caused by boron incorporation into the
lattice, rather than the hole concentration.32

Recently, Vlasov et a1. studied the vibrational modes of heavily boron-doped
diamond using isotopic substitution of boron and carbon.33 They observed that none
of the bands in the Raman spectra shifted upon 10B-substitution, whereas the bands
shifted to lower frequency upon 13C-substitution. These isotopic substitution
experiments are consistent with the 500 cm’1 and 1200 cm’1 bands being associated
with diamond phonons, revealing the phonon density of states, and not with boron.

From the above discussion, we may interpret these bands in the following way:
In relatively small regions of the crystal, a high boron concentration causes structural
disorder (SD) which breaks the translation symmetry of the crystal lattice. As a
consequence, the wave-vector conservation rule for Raman scattering is broken. As a

result, phonons with a large range of wave vectors in the Brillouin zone are excited

86

and the resultant bands tend to be broad and reﬂect the phonon density of states. In
general, the 500 and 1200 cm'1 bands do not directly arise from but are related to
boron atoms incorporated into the lattice.

The broad peak around 1500 cm'1 is attributed to non-diamond sp2 carbon in
the grain boundaries. It is known that the Raman scattering cross-section for graphite
is about 50 times higher than that for diamond.34 The low intensity of this peak
indicates the dominance of the sp3 bonding in boron-doped microcrystalline diamond
ﬁlms. The sharp peak at 522 cm'1 is from the one-phonon mode of Si 28’ 35 and the
broad peak centered at 950 cm'l is the second-order scattering peak of Si.29

To further reveal the inﬂuence of boron on the Raman spectrum, Figure 3.7
shows Raman spectra for microcrystalline boron-doped diamond ﬁhns grown from
different diborane concentrations in the source gas mixture. The broad band centered
ca. 1500 cm", assigned to the nondiamond sp2 carbon, is strong for the undoped
sample (0 ppm diamond ﬁlm). The inﬂuence of boron is clearly revealed in the
spectral features. The graphite band decreased in intensity with increasing boron
concentration. Above 1020 B cm'3 (5 ppm diamond ﬁlm, see Table 2.3 in section
2.3.2), the graphite band almost diminished, while the one-phonon Raman peak at
1332 cm'1 slightly shifted to lower wavenumbers. This shift is in accordance with the
observation in previous Raman studies of boron-doped polycrystalline diamond ﬁlms
and is typically attributed to internal tensile strain.22 At the same time, the full width
at half maximum increased with increasing boron concentration. The intensity of the
one-phonon diamond line decreased with increasing boron concentration due to the
greater optical density. This makes the penetration depth less with increasing boron

concentration so the number of scatterers is less. Table 3.] summarizes the position

87

and width for the diamond line observed for microcrystalline boron-doped diamond

 

 

 

 

 

 

 

 

thin ﬁlms.
16 q 1 _1
. 522 cm' 1332 cm
14 -
12 - -1 .1
E . 1232 cm . 1450 cm 0 ppm
S 10 - 950 cm'1 I
'-' I
£9,
2 8- I
'2 __ 0.5 ppm
g 6 -1 I m _....._
4 _ ' 1 ppm
a ' | |
’- ‘ 4 I L J 5 ppm
i | IT' '
; I 10 ppm
0 I u I 'fJ ‘1 -1- I v
500 1000 1500 2000

Raman shift (cm'1)

Figure 3.7 Raman spectra for microcrystalline diamond ﬁlms grown with
different source gas diborane levels. The greater the source gas diborane

concentration, the higher the doping level.

As the boron-doping level is increased, a new, broad band emerged around
1200 cm'l appeared and its intensity increased indicating a high concentration of

impurities or lattice defects, as discussed before.

Table 3.1 One-phonon diamond line position and width observed in Raman

spectra from microcrystalline diamond ﬁhns with different doping levels.

 

 

 

[B2115] in the source gas 10 5 1 O. 5 0
(ppm)
Peak position (cm'l) 1330.2 1330.7 1331.8 1331.9 1332.0
FWHM (cm") 12.7 10.8 10.1 10.0 8.9

 

 

 

 

88

The spectra also reveal that the presence of boron enhanced the growth of sp3-
band carbon, as shown in Figure 3.8. Although the intensity of one-phonon diamond
line decreased with increasing diborane concentration, the background
photoluminescence also decreased. The ratio of the peak-to-baseline (P/B) near 1332
cm'l increases with increasing diborane concentration from 0 to 5 ppm. When the
boron-doping concentration increased to 10 ppm, the ratio of the peak-to-baseline

decreased indicates the gradual degradation in diamond microstructure in the ﬁlm.

4.5 —

.1
4.0 - /

3.5 -I

9’
o
I m

Peak/Background (arb. unit)
a
l

N
o
r

 

 

1.5 I I T l ' I ' I ' I

2 4 6 8 10
[82H6] in the source gas (ppm)

O-I

Figure 3.8 Relationship between peak-to-baseline (P/B) ratio of Raman spectrum
and the diborane concentration in the source gas mixture for microcrystalline diamond
ﬁlms. The P/B ratio is deﬁned as the ratio of the intensity of the peak at 1332 cm'1

and the intensity of the baseline at the peak position.

3. 3.2 Raman spectra for nanocrystalline boron-doped diamond ﬁlms

on Si substrates

89

Raman spectra for nanocrystalline boron-doped diamond ﬁlms with different
doping levels are shown in Figure 3.9. Several distinct, broad peaks can be seen: 1145,
1330, 1469, and 1545 cm". The peak at 1545 cm"l is clearly the G-band Raman peak,
which is well known for disordered carbon ﬁlms and arise from the in-plane
stretching modes of the spz-bonded carbon at the grain boundaries of nanocrystalline

diamond. 36

_ 1469 cm‘1

1334cm'

9
(J1
O
3|

      

1145 cm"

Intensity (arb. units)

 

 

I I ' T ' ' .I ' I
500 1000 1500 2000
Raman shift (cm'1)
Figure 3.9 Raman spectra for nanocrystalline diamond ﬁlms grown with different
source gas diborane levels. The greater the source gas diborane concentration, the

higher the doping level.

The broad peak at 1330 cm'1 has been attributed to two different origins: the
one-phonon diamond line that is broadened and down-shifted from the expected 1332
cm'1 position because of the small size of the diamond grains, or the D-band (1350

cm!) scattering by graphitic edge plane phonons. In disordered spz-bonded carbon, D-

90

band stretching arises from the breathing modes in small aromatic clusters. The
argument for the assignment of the broad peak at 1330 cm'1 to the spz-bonded carbon
D-band Raman peak is based on the fact that the peak is absent in the UV Raman
spectra37 because the D-peak is believed to arise from a double resonant Raman
process (when the incident photon wave vector k=0.5q), where the intensity is highest
for low energy excitation and decreases strongly with increasing excitation energy.36

The shoulder at 1145 cm'1 and the smaller peak observed at 1470 cm'l, often
observed for nanocrystalline diamond ﬁlms, have been attributed to trans-(CH)x
(trans-polyacetylene) at grain boundaries.38 The evidence from Ferrari and Robertson,
and Birrell is that these peaks arise from spz-bonded carbon of the grain boundaries
and not sp3-bonded carbon.

Some changes were observed when the concentration of diborane in the source
gas mixture increased. Comparing the 0 to 50 ppm samples in Figure 3.9, we notice
that the shoulder at 1145 cm'1 and the smaller peak observed at 1470 cm'1 decreased
in relative intensity to the D-band peak (at 1330 cm'l) as the doping level increased.
The origins of these peaks has been attributed in the literature to various features,
either due to relaxation of the phonon selection rules for nanocrystalline materials or
due to the presence of speciﬁc carbon—hydrogen bond in the grain boundaries of
nanocrystalline diamond?” We believe that the changes of the peaks at 1145 and
1470 cm'l are due to the presence of greater amounts of boron at the grain boundaries
of nanocrystalline diamond for two reasons. First, the intensities of these peaks in the
visible Raman spectra correlates with the amount of boron used during growth; and
second, there is no evidence for these peaks in the UV Raman spectrum.”’ 32 This

reduction in intensity with excitation wavelength is consistent with the results that

91

assign the peak to spz-bonded carbon, although the exact nature of these bands is as

yet unknown.

3.3.3 Raman spectra for boron-doped diam and ﬁlms on Pt substrates

Figure 3.10 shows Raman spectra for several diamond ﬁlms deposited on the
76 um diam. Pt wires. All spectra possess an intense one-phonon diamond line at
1331 cm". The fact that there is only small downshift of the peak position (1 cm")
means that the diamond ﬁlms are not under signiﬁcant tensile or compressive stress.‘12
There is some scattering intensity near 1580 cm'l, which is attributable to amorphous
spZ-bonded carbon impurity. We suppose that much of this impurity responsible is
located at the interface between the Pt and diamond ﬁlm, and does not reside on the
ﬁlm surface. The linewidth, or the full width at half maximum (FWHM), is inversely
related to the phonon lifetime. The larger linewidth for the ﬁlms on Pt is caused by
increased phonon scattering as a result of the grain boundaries and defects in the
polycrystalline ﬁhn.6’ 43

A broad but weakly intense peak around 586 cm'1 is seen in all the spectra. We
believe this peak is of a similar origin to the peak at 522 cm'1 seen for the
microcrystalline boron-doped diamond ﬁlms. In other words, this peak does not arise
from boron itself, but appears whenever a high concentration of impurities or lattice
defects is present. Table 3.2 summarizes the Raman spectra analysis for
microcrystalline boron-doped diamond thin ﬁlms on Pt substrate.

Table 3.2 One-phonon diamond line position and width in Raman spectra.

 

 

 

 

 

 

Sample 1 2 3 4
Peak position (cm'l) 1330.8 1330.9 1330.7 1330.6
FWHM (cm") 12.9 13.0 12.7 12.5

 

 

 

 

92

 

 

   

 

 

 

 

 

1 1331cm"
. 586 cm“ 1586 cm"
W.; fr - Electrode 1
1 __
' I
e - ' .
s : '
‘ I
{5 ‘ , | Electrode 2
£3 ' I
E " ' I
'2 ' I
.9 ‘ W I Electrode 3
E - - __ Aw “-
' I
1 I l
' I
‘ M ' Electrode 4
. - _ W _.L
I . '“W ‘
' I I r ' ' I
500 1000 1500 2000

Raman shift (cm'1)

Figure 3.10 Raman spectra for four boron-doped microcrystalline diamond ﬁlms

deposited on Pt wires. The diborane concentration in the source gas was 4 ppm.

3. 4 Electrochemical Characterization

3. 4.1 Background CV for boron-doped diamond thin-film on Si

Figure 3.11 shows the background cyclic voltammetric i-E curves for “as
grown” microcrystalline and nanocrystalline BDD thin-ﬁlms in 1 mol/L KCl at 0.1
V/s. Curves for glassy carbon are shown for comparison. The geometric area of both
electrodes was the same (0.183 cmz). The curves for both electrodes are featureless
between -200 and 800 mV (vs. AglAgCl). There are no redox peaks for diamond as
there are for glassy carbon due to the presence of redox-active surface carbon-oxygen
functionalities that populate the edge plane sites.“ 45 The background current for the
nanocrystalline BDD thin-ﬁlm electrode is slightly larger than that for the

microcrystalline electrode. In the absence of any redox-active solution species, the

93

background voltammetric current (ibkg) is related to the electrode area (A, cmz), the
double-layer capacitance (Cd, F/cmz) and the potential scan rate (0, V/s), and is given

by the following equation:46

ibkg = ACdU (3.1)

There are two reasons for the difference in background current between the two
diamond ﬁlm types. First, as revealed in CP-AF M measurements (see Chapter 4), the
fraction of area that is highly conducting is larger for the nanocrystalline ﬁlm because
of the smaller grain size. Second, the nanocrystalline ﬁlm, because of the 1: states in
the grain boundary, likely has a higher canier concentration than does the
microcrystalline ﬁlm and this leads to a greater excess surface charge density at a
given potential. This results in a greater double layer capacitance and a larger
voltammetric background current. Even though the background current for
nanocrystalline diamond is slightly larger than that for microcrystalline diamond, both
currents are approximately a factor of 5 - 10 less that that for comparably-sized glassy
carbon.4 The larger current for glassy carbon results in large part from the
pseudocapacitance associated with electroactive and/or ionizable surface carbon-
oxygen functionalities.3’ 47 Diamond is devoid of such functionalities. The lower
background current for diamond leads to improved signal-to-background ratios in

electroanalytical measurements.

94

 

 

 

 

 

 

‘ Glassy carbon (GO-30) electrode J
9 ——
3.0x10 3 :
Nanocrystalline BDD electrode 3
at: l
E 0 0 - A. ‘ '
Q) ﬂ—"ﬂ
t
3
O
Microcr tallin B elec rode
-e .
-3.0X10 " :
I ' I W l
0.0 0.4 0.8

Potential (V, vs. Ag/AQCI)

Figure 3.11 Background cyclic voltammetric i—E curves for planar diamond ﬁlms.
Supporting electrolyte: 1 mol/L KC]. Scan rate: 0.1 V/s. The geometric area of all

electrodes was ca. 0.2 cm2.

It is well-known that the surface termination of diamond ﬁlms strongly
inﬂuences their properties. This was conﬁrmed in the present work. Figure 3.12
compares background cyclic voltammetric i—E curves for hydrogen- and oxygen-
terminated nanocrystalline and microcrystalline BDD thin-ﬁlm electrodes in 1 mol/L
KC] at 0.1 V/s. Clearly, the background current for nanocrystalline electrode is
inﬂuenced by the surface termination as the current for the oxygen-terminated
electrode is larger by a factor of 2-3. Hydrogen- and oxygen-terminated
microcrystalline electrodes show similar results. The increased background current
does not appear to be due to increased surface area because of any surface toughening,
as SEM images of ﬁhns before and after oxidation revealed no morphological

changes.4 It seems that the oxygen termination increases the apparent electrode

95

capacitance possibly by causing an increase in the water contact area. The contact
angle decreases signiﬁcantly after oxidation, which is consistent with a more
hydrophilic surface. Yagi et al. reported similar results that showed the background
current for oxidized diamond electrode was 3—5 times larger than that for the as-

grown electrode.48

   

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2.0x10‘« a
O-termlnaled nanocrystalline BDD electrode
l
1.0x10*« *
s”:
g 0.0.
1:
a
0
-1.0x10'6-
-2.0x10*5 . . . . .
0.0 0.4 0.8
Potential (V. vs. AglAgCI)
2.0x10‘°~ b
i O—terminated microcrystalﬂne BDD electrode
1.0x10"-
§ 0.0- ﬂ
1:
8 . t
8
-1.0x10‘61 ,
l
‘ 1 -ermin ted 'cr I'n e tr d
-2.0x10’5~
0.0 0.4 0.8

Potential (V, vs. AglAgCl)

Figure 3.12 Background cyclic voltammetric i-E curves for H- and O-terminated (a)
nanocrystalline and (b) microcrystalline diamond ﬁlms. The curves were recorded in

1 mol/L KCl at a scan rate of 0.1 V/s starting from -0.2 V (vs. Ag/AgCl).

96

Gerischer and coworkers demonstrated that the double layer capacity of
graphite is relatively low in comparison to metal electrodes because of a low electron
density of states (DOS) near the Fermi level.”51

Another possibility is the increase in the dielectric constant of water at the
hydrophilic diamond surface. The organization of water at the hydrophobic hydrogen-
terminated surface is expected to be different from that at the hydrophilic oxygen-
terrninated surface. The ability of water to isolate charge at the hydrophilic surface

would be greater, hence the larger dielectric constant.

3. 4.2 Electrochemical behavior of boron-doped diamond thin-ﬁlms

with different doping levels

Four speciﬁcally chosen redox systems were used to evaluate the
electrochemical behavior of the BDD electrodes: Fe(CN)63'/4', Ru(NH3),3*’2*, Itch”
and methyl viologen.3’ 4’ 7 For carbon electrodes, the heterogeneous electron-transfer
rate constant is known to be strongly inﬂuenced by the amount of exposed edge plane
on sp2 carbon electrodes (i.e., electronic properties), but weakly inﬂuenced by the
oxygen functionalities terminating the edge plane carbon atoms.”54 Surface
cleanliness is also important as are the electrolyte type and concentration.52°54
Fe(CN)(,3"4’ was used as a probe because the electrode reaction kinetics for this couple
are strongly inﬂuenced by the amount of exposed edge plane on spz-bonded carbon,
as well as the surface cleanlinesss4 Therefore, its heterogeneous electron-transfer rate
constant is quite sensitive to the surface cleanness and chemistry of sp2 and sp3 carbon
electrodes.4’ “‘57 This redox system often behaves anomalously on carbon electrodes

and should not be considered typical outer-sphere. The electrochemical kinetics of

Fe(CN)63'/4' has been studied extensively. For example, studies show that there are

97

several factors that can inﬂuence the electrode kinetics at metal and sp2 carbon
electrodes. First, the rate constant is strongly inﬂuenced by the fraction of edge plane
exposed on sp2 carbon electrodes (i.e., electronic properties), but relatively insensitive
to the surface oxygen functionalities terminating the edge plane carbon atoms as long
as a thick, anionically charged oxide ﬁlm is not present.”’ 53’ 58’ 59 Second, surface
cleanliness is important, as are the electrolyte type and concentration.52'54’ 60’ 61 For
example, the involvement of speciﬁcally adsorbed cations (e.g., K‘) through a
possible surface-bridging interaction has been proposed.60 The rate of reaction
increases with electrolyte composition in the order of LiCl < NaCl < KC1.6°’ 6‘ At the
1 mol/L electrolyte concentration, the rate is a factor of ~10 higher in KCl than in
LiCl at both gold and glassy carbon electrodes.60 Third, adsorbed monolayers on 8p2
carbon electrodes can decrease the rate of reaction. The McCreery group observed an
increase in AEp from 5 to 140 mV after modiﬁcation of the polished glassy carbon
surface with adsorbed monolayers.“ 53 The level of increase depends on the type and
coverage of the adsorbate.

Different from spz-bonded carbon electrodes, the physicochemical properties
of boron-doped diamond strongly inﬂuence the electrode kinetics of Fe(CN)63‘/4'. AEp
of Fe(CN)63'/4' is very sensitive to the surface termination with the smallest AEp
observed at the clean, hydro gen-terminated surface. Aﬁer oxygen termination, AEp
increases by over 125 mV but is reversibly reduced to the original value after removal
of the oxygen functionalities by hydrogen plasma treatment.62 The sensitivity of the
electrode kinetics to surface oxygen at diamond is in sharp contrast to the minor
effects these functionalities have on the response at sp2 carbon electrodes.

The Fe(CN)153 '/ 4' rate constant is also sensitive to the electrolyte composition

and ionic strength with the largest values observed in KCl and the smallest in LiCl.

98

 

This dependence is presumed to result from the formation of a dimeric species
consisting of both the oxidized and reduced forms of the anion coupled through a
bridging cation and sabilized by some speciﬁc surface interaction.60 However, the
difference at the 1 mol/L concentration level is only a factor of 2-3 rather than 10, as
is the case for metal and glassy carbon electrodes.63 All the evidence at sp2 carbon and
diamond electrodes suggests the involvement of a clean, non-oxide surface site.

In the present work, we investigated the inﬂuence of the boron-doping level
on the electrochemical activity of diamond thin-ﬁlm electrode using 0.1 mmol/L
Fe(CN)63"4' in 1 mol/L KC]. The results are shown in Figure 3.13 and Table 3.3. Our
results show that the cyclic voltammetric A13p decreases with increasing the diborane
concentration in the source gas (i.e., increased doping level). For a ﬁlm grown with
0.1 ppm diborane, ABp was 420 mV possibly indicating a sluggish electron-transfer
kinetics or an increased iR effect As doping level increased to 0.5 ppm, well-deﬁned
peak shape i-E curve was recorded and ABp was 208 mV implying much faster
electron transfer kinetics. Another explanation for the reduced A13p with increased
doping level is a reduced iR effect When doping level increased over 1 ppm, AEp
decreased to less than 82 mV (Table 3.3). The near-Nemstian AEp indicates that these
boron-doped diamond electrodes possess the requisite surface structure and properties

for a relatively rapid rate of charge transfer.

99

Current (A)

-4.0x10’5 -

 

 

-8.0x10‘6

Figure 3.13 Cyclic voltammetric i—E curves for 0.1 mmol/L Fe(CN)63"4' in 1 mol/L

KC] for microcrystalline boron-doped diamond ﬁhns grown with different diborane

concentrations.

Table 3.3 Cyclic voltammetric ABp values for different redox couples (0.]
mmol/L analyte in 1 mol/L KCl) at microcrystalline boron-doped diamond ﬁlms

grown with different diborane concentrations. Measurements were made with one

-0.4

ﬁlm at each doping level.

0.1 mBMDEe ro e

 

10 ppm MD Electrode

5 ppm BMD Electrode

0.0

0.4
Potential (V, vs. AglAgCl)

1

0.8

 

 

 

 

 

 

 

 

 

 

 

 

AE m
BMD 11.... mm... Ramps} V) MV.... m...
10 ppm 72 56 51 64
5 ppm 78 6] 52 63
1 ppm 82 68 57 63
0.5 ppm 208 76 62 76
0.1 ppm 420 78 59 N/A

 

100

 

Ru(NH3)63+/2+ involves simple electron transfer at most electrodes including
diamond“; 57’ 64 and sp2 carbon.52'54’ 65 The electrode kinetics for Ru(NH3)63+/2+ are
relatively insensitive to the surface microstructure, surface oxides, and adsorbed
monolayers on sp2 carbon electrodes.52'54’ 65 The electron transfer rate constant of this
redox couple is insensitive to surface modiﬁcation, which strongly implies that
electron transfer does not depend on an interaction with a surface site or functional
group.54 The most important factor affecting the rate of reaction is the electronic
properties of the electrode, speciﬁcally the density of electronic states near the formal
potential of the redox system.

For metal and glassy carbon electrodes, a low density of electronic states is
seldom an issue. However, for semiconducting/semimetallic diamond, the potential
dependent electronic density of states is an important factor governing the response.
This is why AEp increases at boron-doped diamond thin ﬁlm electrodes with
decreasing boron doping level.64 AEp for 10 ppm diamond ﬁlm was 56 mV (Table 3.3)
and increased to 78 mV when the doping level decreased to 0.1 ppm. The formal
potential (i.e., cyclic voltammetric Ep/2 value) for this couple is -218 mV vs. SCE.
The valence band position of boron-doped polycrystalline diamond has been
estimated to be ~550 mV vs. SCE.2’ 66 Considering the 5.4 eV band gap, this means
that the formal potential falls within the band gap (i.e., between the valence and
conduction band positions). Therefore, this redox system is not expected to exchange
charge directly with either the valence or the conduction band. The nearly reversible
response indicates that there must be a sufﬁcient density of electronic states present
within the band gap at this potential (Figure 3.14). The midgap density of states in the
polycrystalline ﬁlms can be inﬂuenced in a combined way by the boron—dOping level

and lattice hydrogen content, the inherent grain boundaries and other morphological

101

defects, and the nondiamond carbon impurity phases at the surface. However, in this
case, surface nondiamond carbon phases are clearly not an important factor
inﬂuencing the density of states since Raman spectra showed that the ﬁlms were sp3'-
band carbon dominated especially for the 5 and 10 ppm ﬁlms (Figure 3.7). As will be
discussed below, the evidence suggests that the boron-doping level and the
chemisorbed hydrogen are the most important factors controlling the density of

electronic states in this potential region.

 

 

 

 

 

8.0x10’<3 -
1031121118.121.2519 street;
4.0x10’6 -
S, 0.1 . m BMD Electrod 9“
E 0 0 —
g \\
a ‘ 1 ppm BMD Electrode
.5
-4.0X10 “ \\ 599m EMD Electrode
_6 O. 5 ppm BM I Electrode
“8.0)(1 0 ' I ' I

-O. 4 0.0 0.4
Potential (V, vs. AglAgCl)

Figure 3.14 Cyclic voltammetric i-E curves for 0.1 mmol/L Ru(NH3 )6 3+/2+ ' m1
mol/L KCl at microcrystalline boron-doped diamond ﬁlms grown with different

diborane concentrations in the source gas.

Similar to Ru(NH3)63+/2+, IrCléz‘B' involves simple electron transfer at most
electrodes including diamond3’ 66 and sp2 carbon.54 The koapp is relatively insensitive to

the surface microstructure, surface oxides, and adsorbed monolayers on sp2 carbon

102

electrodes.52'54’ 65 Surface cleanliness is important, although to a lesser degree than for
Fe(CN)63"4’. The most inﬂuential factor is the electronic structure of the electrode. As
shown in Figure 3.15, all electrodes showed near-Nemstian responses to this redox
couple. AEp changed from 63 to 76 mV when the diborane concentration in the source
gas decreased from 10 ppm to 0.5 ppm (Table 3.3). Table 3.3 also shows that AEp
was largely unaffected by changing boron-doping level. The formal potential of this
redox couple is 767 mV vs. Ag/AgCl, which is positive of the proposed valence band

position.66 This means that charge exchange for this couple likely occurs through the

valence band and not through midgap states.

...Q._p_0_r_:l__8__t~l.Qﬁiegtregle

4 0 Us \ 5 m BMD Electrode
. 1 -
x \1/

 

  
  

.0
O
l

 

Current (A)

 

‘ 0.5 ppm BMD Eﬁctrode
0.6 0.8 1.0
Potential (V, vs. AglAgCl)

Figure 3.15 Cyclic voltammetric i-E curves for 0.1 mmol/L IrClbz'B' in 1 mol/L

KC] at microcrystalline boron-doped diamond ﬁlms grown with different diborane

concentrations in the source gas.

103

 

Methyl viologen involves simple electron transfer at diamond and most other
electrodes. The k0,.pp at diamond is relatively insensitive to surface oxides, grain
boundaries, and defect density and the presence of nondiamond carbon impurities.
Like Ru(NH3)63+/2+, the most important factor inﬂuencing the rate is the density of
electronic states at the formal potential. The formal potential of -685 mV vs Ag/AgC]
for MVZW is well into the band gap region, even more negative than the formal

potential for Ru(NH3)153 "2+. Therefore, methyl viologen likely undergoes charge

transfer through midgap surface states.

 

  

 

8.0x10'6 5 0.1 m BMD lectrode
, / ‘1 ppm BMD Electrode
4.0x10’as -
S, 0.0 -
q—o
E
g -l
3 .5
o -4.0x10 3 - 5 ppm BMD Electrode
-6
-8.0x10 - ! .
1 0.5 ppm BMD Electrode J..Q-.L?le_EaI‘-_‘LD-.E_i§,‘g;e.i_13:/§l_g

 

 

I T

-1.2 ' -0'.8 , -0.4
Potential (V, vs. AglAgCl)

Figure 3.16 Cyclic voltammetric i-E curves for 0.1 mmol/L MV in 1 mol/L KC] at

microcrystalline boron-doped diamond ﬁlms grown with different diborane

concentrations in the source gas.

Figure 3.16 shows the cyclic voltammetric i-E curves for methyl viologen on

different boron-doped diamond thin ﬁlm electrodes. Both near-Nerstian redox waves

104

associated with the MVZW and MVJ”o transitions were observed. The near-Nemstian
AEp values indicated that all ﬁlms showed semimetal behavior at the E0. values of ca.
-0.7 and -1.0 V (vs. Ag/AgCl) for methyl viologen. Figure 3.16 and Table 3.3 also
clearly shows that the positions of the redox waves, and therefore the electron transfer
kinetics, were largely unaffected by changing the boron doping level, similar to the

243'. Therefore, we can conclude that the rate constant for this

observation for IrCl6
system is relatively insensitive to surface sites and chemical functionality. Surface

cleanliness and electronic effects appear to be the most inﬂuential factors determining

the rate, and this redox system behaves as outer-sphere.

 

 

 

 

 

 

 

 

 

 

1 m
40- pp
i 4.0 ‘
g 0 + 10 0.0:
,3- m
0 pp 40
40.
5 ppm
0.0 074 0.8
1 m 8.0
40 pp 1
A 4.01
:5: 0.0 0 0
5 l -4.0
0 10 ppm l
-4.0' 5 ppm -8 O
0.8 0:8 10 4:2 ' -0'.8 ' 0.4 T
Potential (V vs. AglAgCl) Potential (V vs. Ag/AgCl)

Figure 3.17 Cyclic voltammetric i—E curves for (a) 1 mmol/L Fe(CN)63'/4', (b) 0.1
mmol/L Ru(NH3)63+’2+, (c) 0.1 mmol/L 1rC162'B', and (d) 0.1 mmol/L methyl viologen
with 1 mol/L KC] at nanocrystalline diamond thin ﬁlm electrodes grown with

different diborane concentrations in the source gas.

105

Similar voltammetric measurements were made to evaluate and compare the
response of nanocrystalline boron-doped diamond thin ﬁlms. The results for
Fe(CN)63’/4‘, Ru(NH3)63+/2+, IrClém', and methyl viologen are presented in Figure
3.17 and Table 3.4. For all the redox systems except F e(CN)63 44', the data shown were
acquired in 0.] mmol/L redox couple with 1 mol/L KC] at a scan rate of 0.1 V/s. The
data for Fe(CN)63'/4‘ were obtained in 1 mmoVL analyte with 1 mol/L KC] at a scan
rate of 0.] V/s.

It can be seen that similar results were obtained with 1, 5 and 10 ppm
nanocrystalline boron-doped diamond ﬁlms, as compared to the microcrystalline
boron-doped diamond ﬁlms with corresponding boron-doping level. AEp’s for
Fe(CN)6344', Ru(NH3)(,3m+ and methyl viologen at both types of ﬁlms are nearly the
same. This means that the electrode resistance is largely unchanged by the change in
doping level. The big difference between the nanocrystalline and microcrystalline

ﬁlms is the AEp’s for IrClom', which are 10 mV larger at nanocrystalline ﬁlms.

Table 3.4 Cyclic voltammetric AEp’s for different redox couples (0.] mmol/L
analyte in 1 mol/L KCl) at nanocrystalline boron-doped diamond ﬁlms grown with
different diborane concentrations in the source gas. Measurements were made with

one ﬁlm at each doping level.

 

 

 

 

 

 

AEp (mV)
BND films * 34¢ 3+/2+ 243-
Fe(CN)6 Ru(NH3)6 MV Ier
10 ppm 68 60 55 74
5 ppm 75 59 56 74
1 ppm 94 62 54 8O

 

 

 

 

 

*: the concentration for Fe(CN)63‘M" was 1 mmol/L.

106

 

The near-Nemstian responses for F e(CN)153 J4“, Ru(NH3)53+/2+, erk2'/3' and
methyl viologen all indicate that resistance or semiconductor effects are minimal even
at very negative potentials when the boron-doping level is over 1 ppm for both
microcrystalline and nanocrystalline diamond ﬁlms. The hydrogen terminated (or new
as grown) diamond surface is chemically inert, hydrophobic, and resists adsorption of
impurities such that a high degree of surface cleanliness is maintained. This avoids the
need for frequent surface pretreatment in order to active the electrode.

The voltammetric data presented above reveal that, for Fe(CN)1.r,3 44',
Ru(NH3)63+/2+, IrCl62'I3' and methyl viologen, the clean, as-grown microcrystalline and
nanocrystalline boron-doped diamond thin ﬁlms possess similar macroscopic

electrochemical propertiess’ 7

3. 4.3 Electrochemical activities for planar boron-doped diamond thin-

ﬁlm on Si substrate with different surface terminations

The surface chemistry of carbon materials and its effect on the electrochemical
behavior is a complicated issue. For example, glassy carbon needs to be activated to
create reproducibly active electrodes due largely to the complex nature of the surface
chemistry. To activate glassy carbon, usually, mechanical polishing with successively
smaller grades of alumina is ﬁrst used, oﬁen followed by additional treatments such
as heat treatment,67 electrochemical polarization,68 plasma treatment69 or laser
activation.70 Although it’s not normally necessary to activate a boron-doped diamond
electrode, the electrode response can be affected by surface contaminants, surface
functional groups, and microstructural defects. Oxygen from air or H20 chemisorbs to
most sp2 carbon surfaces, so it is very difﬁcult to avoid oxygen functional group

formation when using these electrodes in air. The same holds true for diamond but the

107

oxidation reaction rate is slower than that for other spzcarbon materials. In addition,
there is often a low level of sp2 carbon impurity present on CVD diamond electrodes,
which is located mainly in the grain boundary and has been reported to be more active
than sp3 diamond,"6 especially for nanocrystalline diamond. The presence of oxygen
on the diamond surface has a signiﬁcant inﬂuence on the chemical reactivity,44
electrical conductivity,45 and electron afﬁnity.71 On the other hand, oxidation is a
means to clean the BDD surface. At present, several procedures have been used by
different labs to oxidize the boron doped diamond surface including: (i) chemical
treatment, (ii) electrochemical treatment, and (iii) oxygen plasma treatment. These
procedures can remove all detectable impurities except oxygen but form hydroxyl and
ether functional groups, carbonyl and carboxyl groups on the diamond surface. The
distribution of particular oxygen containing functional groups is fairly difficult to be
determined and may itself vary with time and sample history. This makes the
relationship between the electrochemical properties and the surface chemistry of
boron doped diamond more complicated. The “as grown” diamond surface is
hydrogen-terminated because the deposition of diamond ﬁlms is undertaken in
hydrogen plasma. However, this surface is not resistant to oxidization. Therefore,
knowledge of the interaction of hydrogen and oxygen on diamond surface is crucial to
the application of BDD thin-ﬁlms in both electrochemistry and electron emission
ﬁelds. It is well known that BDD thin-ﬁhns with different surface terminations
exhibit different performance in electrochemistry and electron emission,4 but the
relationship between the electrical properties and the microstructure and the surface
chemistry of polycrystalline diamond, especially for nanocrystalline diamond,

remains unclear.

108

In the present work, we characterized highly boron—doped microcrystalline and
nanocrystalline thin-ﬁlm electrodes with H- and O-tenninations using cyclic
voltammetric method. Our results show that O-termination blocked the
electrochemically-active sites on the electrode surface that lowered the electron-
transfer rate and this inhibiting effect was much more pronounced for nanocrystalline
diamond than for microcrystalline diamond.

Figure 3.18 shows the cyclic voltammetric i-E curves for 1 mmol/L Fe(CN)63’
M’ in 1 mol/L KC] at a scan rate of 0.1 V/s at H-terrninated microcrystalline and
nanocrystalline BDD electrodes. Well-deﬁned, peak-shape voltammograms are
observed for both electrode types without any conventional treatment. AEp values of
73 i- 5 mV (n = 3) were seen for the microcrystalline BDD electrode, while the peak
separation for the nanocrystalline BDD electrode was 73 i 4 mV (n = 3). The small
AEp value is indicative of relatively rapid electrode-reaction kineticss Currents
limited by semi-inﬁnite linear diffusion were found as the peak current for the
forward scan varied proportionally with the square root of the potential sweep rate.“
Based on the conductivity maps (see section 4.6), it appears that the current passed
ﬂows through only selected regions of the diamond surface. In other words, the

effective current density is locally high in some regions.

Similar results were obtained for 1 mmol/L Ru(NH3)153+/2+ in l moi/L KCl as
shown in Figure 3.19. ABp value of 63 i- 4 mV and 69 i 6 mV (n = 3) were observed
for microcrystalline and nanocrystalline BDD electrodes, respectively. These results
show that, for “as grown” or H-terminated‘ BDD electrode, the macroscopic

electrochemical properties of both BDD types are similar.5’ 7

109

 

‘ Microcmstalline BDD electrode
-5
4.0x10 -
Et:
‘5' 0.0 -1
t
:1
O
-5
-4.0x10 -
Nanocrystalline BDD electrode

 

 

 

0.0 0.4 0.8
Potential (V, vs. AglAgCl)
Figure 3.18 Cyclic voltammetric i-E curves for 1 mmol/L Fe(CN)?“ in 1 mol/L

KC] at 0.1 V/s for 10 ppm micro- and nanocrystalline BDD thin-ﬁlm electrodes. Both
electrodes were H-terminated.

5 Microc stalline BDD Electrode
4.0x10'

it, on _
5
l:
3 '1
O

-4.0x10 -

Nanocrystalline BDD Electrode

 

 

T

 

-0.4 0.0 0.4
Potential (V, vs. AglAgCl)
Figure 3.19 Cyclic voltammetric i—E curves for 1 mmol/L Ru(NH3)63+/2+ in 1 mol/L

KC] at 0.1 V/s for 10 ppm micro- and nanocrystalline BDD thin-ﬁlm electrodes. Both
electrodes were H-terminated.

110

Figure 3.20 shows the cyclic voltammetric i-E curves for 1 mmol/L Fe(CN)63'
’4’ in 1 mol/L KC] at 0.] V/s for H- and O-terminated nanocrystalline BDD. A nearly
reversible response was seen for the H-terminated electrode. Aﬁer oxidation, a large
increase in A13p was observed indicating that this surface termination causes a
signiﬁcant decrease in rate of electron transfer for this redox couple. The anodic peak
current decreased from 330 i 21 uA/cm2 (n = 3) for the hydrogen termination to 202
i 11 ILA/cm2 (n = 4) for oxygen termination and NSF increased from 73 i 4 mV (n = 3)
to 289 :l: 31 mV (n = 4) after oxidation. The data show that the electrochemical

kinetics for this redox couple strongly depends on the surface termination.62

O-terminated ngnocrvstalline BDD electrode

4.0x10'5- /

 

 

3:
g 0.0 -
t
:3
O
-4.0x10'5 ~
H-terminated nanocmstalline BDD electrode
1 T l 7 f
0.0 0.4 0.8

Potential (V, vs. AglAgCl)
Figure 3.20 Cyclic voltammetric i—E curves for 1 mmol/L Fe(CN)63’/4‘ in 1 mol/L

KC] at 0.1 V/s for 10 ppm H- and O-terminated nanocrystalline BDD thin-ﬁlm

electrodes.

11]

When probed using 1 mmol/L Ru(NH3)63+/2+ in 1 mol/L KC], similar trends
were found (Figure 3.21). The anodic peak current changed from 358 i 25 trA/cm2 (n
= 4) for the H-terminated electrode to 206 i 30 ptA/cm2 (n = 3) for O-terminated
electrode, and A131,, increased from 71 i 6 mV (n = 4) to 180 i 47 mV (n = 3). These
data clearly revealed a comprehensive effect of termination change from H to O and
surface carbon-oxygen functionalities. Oxygen blocks the reactive sites of the

diamond surface which makes more sluggish reaction on O-terminated than that on H-

terminated BDD electrode.

O-terminated Nanocrystalline BDD Electrode

4.0x10'5e

 

i

40:: 0.0 -
t:

3

O
-4.0x10'5*

 

ﬂ-terminated Nanocgstalline BDD Electrode

‘ T———'_——I_——"——'_—I

-0.4 0.0 0.4
Potential (V, vs. AglAgCl)

Figure 3.21 Cyclic voltammetric i-E curves for 1 mmol/L Ru(NH3)63+/2+ in 1 mol/L

KC] at 0.1 V/s for 10 ppm H- and O-terminated nanocrystalline BDD thin-ﬁlm

electrodes.

The cyclic voltammetric i-E curves for 1 mmol/L Fe(CN)?” in 1 mol/L KC]

at H and O-terminated microcrystalline boron-doped diamond thin-ﬁlm electrodes at

112

a scan rate of 0.1 V/s are shown in Figure 3.22. The anodic peak current decreased
from 331 ..+_ 8 ILA/cm2 (n = 3) for the H-terminated electrode to 295 i 28 p.A/cm2 (n =
3) for the O-terminated electrode, and Alip changed from 73 i 5 mV (n = 3) for H-
terminated electrode to 106 i 12 mV (n = 3) for O-terminated electrode. These are
surprising results compared with those for nanocrystalline BDD ﬁlms. It seems that
O-termination of microcrystalline BDD ﬁlm has much less inﬂuence on the
electrochemical kinetics of Fe(CN)63”4'. A possible reason for this is that the
microcrystalline BDD ﬁlm is more resistive to surface oxidation such that there is less
oxide coverage. This is partially conﬁrmed by XPS measurements on a set of similar
BDD ﬁlms, as shown in Table 3.5.

The cyclic voltammetric i-E curves for 1 mmol/L Ru(NH3)63+’2+ in 1 mol/L
KCl at 0.1 V/s on H- and O-terminated microcrystalline BDD conﬁrmed this
hypothesis. AEp changed from 63 i 4 mV (n = 4) for H-terminated electrode to 71 :t 6
mV (n = 3) for O-terminated electrode revealed that only few active sites were

blocked by oxygen (Figure 3.23).

Table 3.5 XPS results for microcrystalline and nanocrystalline BDD thin ﬁlms

with different surface treatments.*

 

 

 

 

 

 

 

Cls (%) Ols (%) O/C

As grown 97.96 2.04 0.02

BMD Rehydrogenated 98.07 1.93 0.02
Acid washed 87.52 10.79 0.12

As grown 95.04 4.96 0.05

BND Rehydrogenated 97.56 2.09 0.02
Acid washed 80.32 13.58 0.17

 

 

 

 

 

 

 

*: The author acknowledges Yingrui Dai, a member of Dr. Swain’s group, who
provided the XPS data.

113

 

   
   

 

O-terminated Microcrystalline BDD electrode
-5
4.0x10 - /
if.
‘5'; 0.0 '1
t:
:3
O
-5
-4.0X10 ‘1 \
H-terminated Microc s lline BDD electrode

 

 

01.0 ' 0:4 ' 0.8
Potential (V, vs. AglAgCl)

Figure 3.22 Cyclic voltammetric i-E curves for 1 mmol/L Fe(CN)63‘/4' in 1 mol/L

KC] at 0.1 V/s for H- and O—terminated microcrystalline BDD thin-ﬁlm electrodes.

 

 

 

- e in t Micr l' c

4.0x10'5 -
i
E 0.0 —
t
3
O l

-5
-4.0x10 -
O-termingted Microcrystalline BDD Electrode

 

-0'.4 ' 0:0 - 0:4
Potential (V, vs. Ag/AQCI)

Figure 3.23 Cyclic voltammetric i-E curves for 1 mmol/L Ru(NH3)(,3+/2+ in 1 mol/L

KC] at 0.1 V/s for H- and O-terminated microcrystalline BDD thin-ﬁlm electrodes.

114

 

 

 

 

 

O-termmgted Microcrystglline BDD electrode
-5
4.0x10 -
3‘.
E 0.0 -
t
3
O
-5
-4.0x10 - \
H-termina_ted MicrocrLs lline BDD electrode

 

ofo ' 0:4 ' 0:8
Potential (V, vs. AglAgCl)

Figure 3.22 Cyclic voltammetric i—E curves for 1 mmol/L Fe(CN)63'/* in 1 mol/L

KCl at 0.1 V/s for H- and O-terminated microcrystalline BDD thin-ﬁlm electrodes.

 

 

- ' t Mi r stall‘n r
4.0x1oid
,
s”:
E 0.0—
t:
3
o l
-5
-4.0x10 -
O-terminated Microcrystalline BDD Electrode

 

 

-0'.4 ' 0:0 f 0:4
Potential (V, vs. AglAgCl)

Figure 3.23 Cyclic voltammetric i-E curves for 1 mmol/L Ru(NH;;)63’+/2+ in 1 mol/L

KC] at 0.1 V/s for H- and O-terminated microcrystalline BDD thin-ﬁlm electrodes.

114

3. 5 The Effect of surface hydrogen on the properties of

planar boron-doped diamond thin-ﬁlm on Si substrate

Surface conductivity (SC) on single crystals and chemical vapor deposition
diamonds was ﬁrst reported in 1989 by Ravi and Landstrass. The phenomenon was
exclusively attributed to the presence of hydrogen at the begining.72’ 73 However, in
recent years the role of atmospheric species on the formation of a hole accumulation
layer has become known. In several experiments it has been demonstrated that
exposure to different gaseous species varies the SC by many orders of magnitude.74
Maier et al.75 have proposed an electrochemical model in which electrons are
transferred from the diamond valence band to solvated H3O+ ions in a solution layer
that forms spontaneously in air (condensation). This model is supported by recent

176 and experimental studies.“ 78 Both physisorbed and chemisorbed

tlreoretica
hydrogen exist in diamond ﬁlms grown by CVD. Although the inﬂuence of adsorbed
hydrogen on the properties of diamond ﬁlms has been studied by several groups,
however, few investigations about the effect of the surface hydrogen on boron-doped
diamond ﬁlms have been done. Here we investigate the properties of boron-doped
diamond thin ﬁlms upon annealing in Ar and exposure to lab atmosphere. Our goal is

to distinguish the different inﬂuences of physisorbed hydrogen and chemisorbed

hydrogen on the physical and chemical properties of boron—doped diamond thin ﬁlms.

3. 5.1 Contact angle measurement

The hydrogen-terminated diamond surface is hydrophobic and resists
adsorption of polar molecules such that a high degree of surface cleanliness is

maintained. On the other hand, the oxygen-terminated diamond surface is hydrophilic

115

and is less resistant to the adsorption of polar impurities because of the terminal
oxygen functional groups. This difference of wettability was conﬁrmed by contact
angle measurements in distilled and deionized H20, as shown in Table 3.6.

The contact angles were measured by commertial instrument, FTA200
(Edmund Scientiﬁc Co., Harrington, NJ), using ultrapure water. The “as grown” ﬁlms
were stored in the laboratory atmosphere for weeks prior to the measurements so the
surface had been partially oxidized. Our results show that the as grown diamond ﬁlms
were highly hydrophobic and the ﬁlm became more hydrophilic as the doping level
increased. A possible explanation for this is that the defects introduced by boron on
the surface, especially in the grain boundaries, increase as the concentration of boron
increases. These defects are more hydrophilic than those terminated hydrogen because
the interstitial boron (aggregated mostly in the grain boundaries) tends to introduce
the sp2 bond79 resulting in more hydrophilic surface. After annealing at 400°C, the
contact angle signiﬁcantly decreased for all tested samples. This can be attributed to
the removal of physisorbed hydrocarbon species and the partial oxidation of the
surface as the gas ambient used for heating was not totally devoid of oxygen. For
example, physisorbed hydrocarbon species desorb from the diamond surface lower

than 400°C80 but oxygen-containing species do not.81

116

Table 3.6 Contact angles of microcrystalline boron-doped diamond thin ﬁlms

with different doping levels and annealing treatment.

 

 

 

 

 

 

 

 

Film Contact angle (degree, aver. :l: std., n=6)
As grown After annealing
0.1 ppm 60.3 d: 2.0 42.4 :1: 4.0
0.5 ppm 60.8 i 2.5 38.5 :1: 3.4
1 ppm 62.3 :1: 1.9 33.2 :t 1.7
5 ppm 48.5 d: 2.4 32.0 :h 3.1
10 ppm 49.7 :1: 1.3 31.0 :1: 3.6

 

 

 

 

3. 5.2 F our-point resistivity measurements

Table 3.7 summarizes the electrical resistivity data for two microcrystalline
boron-doped diamond thin ﬁlms with different doping levels. The resistivity of the
ﬁlm was measured using a home—made 4-point probe, with spring-loaded tips, in
combination with a Hewlett Packard 3478 multimeter (HP, USA). The resistivity was

calculated using the following equation:

p= 4.532dr (3.2)

where p is the resistivity (Q-cm), d is the ﬁhn thickness (for microcrystalline diamond
thin ﬁlm, d = 6 pm), r is the resistance measured in a 4-point probe conﬁguration ((2).

The data show that annealing in Ar at 400°C signiﬁcantly increased the
resistivity of 0.5 ppm boron-doped diamond thin ﬁlm (from 6.4 :l: 1.4 to 11.6 :1: 1.3
Q-cm) but did not change the resistivity of 5 ppm boron-doped diamond thin ﬁlm. A
possible interpretation for this is that the surface conductivity induced by hydrogen

has a greater inﬂuence on the resistivity of 0.5 ppm diamond ﬁlm because of the

117

lower doping level and hence a weaker conductivity. In other words, surface hydrogen
acting as an electron acceptor is the “extra” dopant here. During annealing surface
hydrogen (i.e., electron acceptors) is desorbed and the surface conductivity was
decreased.” 82 For the 5 ppm boron-doped diamond ﬁhn, because of the high doping
level, the contribution to the electronic properties by the boron dopant was great

enough that the desorption of hydrogen had little effect on the resistivity.

Table 3.7 Surface resistivity of boron-doped diamond thin ﬁlms (n = 6).

 

 

 

 

 

Film 0.5 ppm BMD 5 ppm BMD
Resistivity As-grown After annealing As-grown After annealing
Average (SI-cm) 6.4 11.6 0.150 0.148
S.D. (fl-cm) l .4 1 .3 0.006 0.007

 

 

 

 

 

 

3. 5.3 Cyclic voltammetric characterization

The electrochemical response for 0.5 and 5 ppmboron-doped microcrystalline
diamond ﬁlm electrodes towards Fe(CN)63”4', Ru(NH3)63+/2+ and Irle“' was
measured before and after the annealing treatment and shown in Figure 3.24a - c.
Figure 3.24a shows voltammetric i-E curves for 0.1 mol/L Fe(CN)63J4‘ with 1 mol/L

KCI- Fe(CNTo344' is an inner-sphere redox system" 6' 62

with a heterogeneous electron-
transfer rate constant that is sensitive to the electronic properties, surface cleanliness
and the presence of surface oxides on diamond electrodes as discussed before (see
section 3.4.2). At 0.5 ppm diamond electrode, a sluggish curve and a large peak
separation (225 mV) were recorded. While large ABp values can be related to both

sluggish kinetics and to the presence of the ohmic potential loss (iR), the large value

for this ﬁlm is mainly due to surface contamination.

118

Current (A)

Current (A)

Current (A)

8.0x10‘

 

 

 

     
 

 

 

 

   
 

 

 

4.0x10" -

0.0 4
4.03110" 3
-8.0x10" I ' I ' I

0.0 0.4 0.8
Potential (V vs. AglAgCl)
b

4.0x10‘ -

0.0 .l

O. s r w
0.5 32m (after annealing)
4 g 5 22m [as grown)
.Ox10 - 5 l [I I! l
-8.0x10“ . . , . ,
-O.4 0.0 0.4
Potential (V vs. AglAgCl)
_. c

4.0x10 -1

0.0 _
4.0x10" - 0.5 ppm (alter annealing)

5 m as row
« LEW
-a.0x1o" . . , a
0.6 0.9

Potential (V vs. AglAgCl)

Figure 3.24 Electrochemical response of 0.5 and 5 ppm microcrystalline BDD thin-

ﬁlm electrodes before and after annealing in Ar ﬂow at 400°C for 30 min. a) 0.1

mmol/L Fe(CN)63'/4', b) 0.1 mmol/L Ru(NH3)63+’2+ and c) 0.1 mmol/L IrClgz'B'. The

supporting electrolyte solution was 1 mol/L KCl and the scan rate was 0.1 V/s.

119

After annealing at 400°C in Ar, it appeared that physisorbed hydrocarbon
species are removed80 creating clean “sites” for Fe(CN)63‘/4' electron-transfer.
However, annealing at 400°C in Ar may only remove the physisorbed hydrocarbon
species but not oxygen-containing species and chemisorbed hydrogen.” 81’ 82 In other
words, this treatment may just partly change the surface conductivity induced by
hydrogen-termination. It’s possible that this change has signiﬁcant inﬂuence on the
electrochemical response of diamond ﬁlm with a low doping level such as 0.5 ppm
when probed with a inner sphere redox couple like Fe(CN)53"4'. For highly doped
diamond electrode (5 ppm in our case), the effect of annealing is weak enough to be
negligible even when probed with Fe(CN)6344' because of the high density of
electrochemically active sites induced by boron doping. This can also explain why
there was no signiﬁcant difference for the electrochemical response of diamond
electrodes recorded in Figure 4.24 b and c. As discussed in section 4.4.2, Ru(NH3)6
3H2“ and IrClaz‘B' are both out sphere redox couples and involve simple electron
transfer on most electrodes including diamond” 57’ 64 and sp2 carbon?“ 65 The
electrode kinetics for these two redox couples are relatively insensitive to the surface
microstructure, surface oxides, and adsorbed monolayers on sp2 carbon electrodes.52'54’
65 The most important factor affecting the rate of reaction is the electronic properties
of the electrode. We also demonstrated in section 4.4.2 that, for these redox species, a
doping level of 0.5 ppm was good enough for diamond electrodes to obtain near-
Nemstian response. Annealing in Ar at 400°C did not destroy the hydrogen
termination, which meant the electronic properties of the electrodes were largely
unaffected by this treatment. Therefore, the peak separations and peak currents for
Ru(NH3)6 3+/2+ and IrCl62'l3' did not change too much as did Fe(CN)63’/4' before and

after the annealing treatment.

120

Table 3.8 summarizes the changes of peak separation (A5,) and peak current
for 0.5 and 5 ppm boron—doped diamond thin ﬁlm electrodes before and after

annealing at 400 °C in Ar ﬂow.

Table 3.8 Cyclic voltammetric peak separation (AEp, mV) and peak current (ILA)
data for 0.1 mmol/L analytes in 1 mo] KC] at 0.5 and 5 ppm boron-doped

microcrystalline diamond thin-ﬁhn electrodes before and after annealing in Ar ﬂow at

 

 

 

 

 

 

 

 

400°C.
Redox film As grown After Annealing
AEP in. i9,c AEI, i” inc
3% 0.5 ppm 225 3.35 3.31 123 4.53 4.39
Fe(CN)6 5 ppm 82 6.74 6.6 72 6.97 6.92
Ru(NH3)63+’2+ 0.5 ppm 74 5.57 5.88 66 5.86 5.98
5 ppm 62 6.48 6.57 64 6.78 6.80
243- 0.5 ppm 76 5.54 5.95 74 5.54 5.98
"C16 5 ppm 68 5.84 6.71 64 6.01 6.42

 

 

 

 

 

 

 

 

 

 

3.6 Conclusions

Boron-doped microcrystalline and nanocrystalline diamond thin-ﬁlm
electrodes with different doping levels were evaluated using Raman, SEM, TEM and
cyclic voltammetry. The microcrystalline diamond ﬁlms possessed some sp2 carbon
impurity at lower doping levels (< 0.5 ppm diborane in source gas mixture), but this
impurity was reduced at higher doping levels (> 1 ppm diborane in source gas
mixture), based on Raman spectrum. The nanocrystalline diamond ﬁlms have Raman

spectra dominated by the spz-bonded carbon in the grain boundaries. The one-phonon

121

peak at 1332 cm"1 was partially buried by the stronger scattering by grain boundary
carbon in Raman spectra. Although lots of nodular features were seen in SEM for
nanocrystalline diamond ﬁlms, TEM and SEAD images demonstrated that these ﬁhns
consisted of phase-pure cubic diamond crystallites. Additionally, electrochemical
characterization with Fe(CN)(,3"4', Ru(NH3)63+/2+, IrC16243' and MV showed that the
clean, as-grown microcrystalline and nanocrystalline boron-doped diamond thin ﬁlms
possess similar macroscopic electrochemical properties. The present study also
showed that the surface termination strongly affected the electrochemical properties
of boron-doped diamond thin ﬁlm electrodes, especially when probed with Fe(CN)63'
W. The oxygen-terminated diamond electrodes had greater voltammetric background
currents than the hydrogen-terminated counterparts. Annealing at 400 °C in Ar ﬂow
inﬂuenced the physical and electrochemical properties of boron-doped diamond
electrodes with a low doping level (0.5ppm), but did not affect the highly doped

diamond electrodes (5 ppm).

122

3.7

(1)

(2)
(3)

(4)

(5)

(6)

(7)

(3)

(9)
(10)
(11)

(12)

(13)

(14)

(15)

Reference

Xu, J.; Granger, M. C.; Chen, Q.; Strojek, J. W.; Lister, T. E.; Swain, G. M.
Anal. Chem. 1997, 69, 591A-597A.

Swain, G. M.; Anderson, A. B.; Angus, J. C. MRS Bulletin 1998, 56.

Granger, M. C.; Xu, J .; Strojek, J. W.; Swain, G. M. Anal. Chim. Acta 1999,
397, 145-161.

Granger, M. C.; Witek, M.; Xu, J.; Wang, J.; Hupert, M.; Hanks, A.; Koppang,
M. D.; Butler, J. E.; Lucazeau, G.; Mermoux, M.; Strojek, J. W.; Swain, G. M.
Anal. Chem. 2000, 72, 3793-3804.

Fischer, A. E.; Show, Y.; Swain, G. M. Anal. Chem. 2004, 76, 2553-2560.

Bennett, J. A.; Wang, J .; Show, Y.; Swain, G. M. J. Electrochem. Soc. 2004,
151, E306-E313.

Show, Y.; Witek, M. A.; Sonthalia, P.; Swain, G. M. Chem. Mater. 2003, 15,
879-888.

Holt, K. B.; Bard, A. J .; Show, Y.; Swain, G. M. J. Phys. Chem. B 2004, 108,
15117-15127.

Stotter, J .; Show, Y.; Wang, S.; Swain, G. Chem. Mater. 2005, I 7, 4880-4888.
Wang, S.; Swain, G. M. J. Phys. Chem. C 2007, 111, 3986-3995.

Kim, Y. D.; Choi, W.; Wakimoto, H.; Usami, S.; Tomokage, H.; Ando, T.
Diamond Relat. Mater. 2000, 9, 1096-1099.

Angus, J. C.; Sunkara, M.; Sahaida, S. R.; Glass, J. T. J. Mater. Res. 1992, 7,
3001-3009.

Wurzinger, P.; Pongratz, P.; Hartrnann, P.; Haubner, R.; Lux, B. Diamond and
Relat. Mater. 1997, 6, 763.

Ushizawa, K.; Watanabe, K.; Ando, T.; Sakaguchi, I.; Nishitani-Gamo, M.;
Sato, Y.; Kanda, H. Diamond Relat. Mater. 1998, 7, 1719-1722.

Tokuda, N.; Umezawa, H.; Saito, T.; Yamabe, K.; Okushi, H.; Yamasaki, S.
Diamond Relat. Mater. 2007, 16, 767-770.

123

(16)

(17)
(18)
(19)

(20)

(21)

(22)

(23)

(24)

(25)

(26)

(27)

(28)

(29)

(30)

(31)

(32)

Tokuda, N.; Saito, T.; Umezawa, H.; Okushi, H.; Yamasaki, S. Diamond Relat.
Mater. 2007, 16, 409-411.

Gruen, D. M. Annu. Rev. Mater. Sci. 1999, 29, 211-259.

Nie, S.; Emory, S. R. Science 1997, 275, 1102-1106.

Kenneth Ray, 1.; McCreery, R. L. Anal. Chem. 1997, 69, 4680-4687.
F ryling, M. A.; Lhao, J.; McCreery, R. L. Anal. Chem. 1995, 67, 967.

Ashida, K.; Kanamori, K.; Watanabe, K. J. Vac. Sci. T echnol. A 1988, 6,
2232-2236.

Zhang, J.; Zimmer, J. W.; Howe, R. T.; Maboudian, R. Diamond Relat. Mater.
2008, 1 7.

May, P. W.; Ludlow, W. J .; Hannaway, M.; Heard, P. J.; Smith, J. A.; Rosser,
K. N. Chem. Phys. Lett. 2007, 446, 103—108.

Mermoux, M.; Jomard, F.; Tavare‘s, C.; Omne‘s, F.; Bustarret, E. Diamond
Relat. Mater. 2006, 15, 572-576.

Newton, R. L.; Davidson, J. L.; Lance, M. J. diamond Relat. Mater. 2005, 14,
1 73-1 78.

Bernard, M.; Deneuville, A.; Muret, P. Diamond Relat. Mater. 2004, 13, 282-
286.

Cerdeira, F .; Fjeldly, T. A.; Cardona., M. Phys. Rev. B 1973, 8, 4734-4745.

Gonon, P.; Gheeraert, E.; Deneuville, A.; Fontaine, F.; Abello, L.; Lucazeau,
G. J. Appl. Phys. 1995, 78, 7059-7062.

Ager 111, J. W.; Walukiewicz, W.; McCluskey, M.; Plano, M. A.; Landstrass,
M. 1. Appl. Phys. Lett. 1995, 66, 616-618.

Buckley, R. G.; Moustakas, T. D.; Ye, L.; Varon, J. J. Appl. Phys. 1989, 66,
3595-3599.

Tubino, R.; Piseri, L.; Zerbi, G. J. Chem. Phys. 1972, 56, 1022.

May, P. W.; Ashfold, M. N. R.; Mankelevich, Y. A. J. Appl. Phys. 2007, 101,
053115.

124

(33)

(34)

(35)

(36)

(37)

(38)
(39)

(40)

(41)

(42)

(43)

(44)

(45)

(46)

(47)

(48)

(49)

Vlasov, I. 1.; Ekimov, E. A.; Basov, A. A.; Goovaerts, E.; Zoteev, A. V.; Los
Alamos National Laboratory, Condensed Matter, 2008, pp 1-7.

Wang, W. L.; Polo, M. C.; Sanchez, G.; Cifre, J .; Esteve, J. J. Appl. Phys.
1996, 80, 1846.

Ricci, P. C.; Anedda, A.; Carbonaro, C. M.; Clemente, F.; Corpino, R. Thin
Solid Films 2005, 482, 311-317.

Ferrari, A. C. ; Robertson, J. Phys. Rev. B 2000, 61 , 14095.

Birrell, J .; Gerbi, J. E.; Auciello, 0.; Gibson, J. M.; Johnson, J.; Carlisle, J. A.
Diamond Relat. Mater. 2005, 14, 86-92.

Ferrari, A. C.; Robertson, J. Phys. Rev. B 2001, 63, 121405R.
Nemanich, R. J.; Solin, S. A.; Martin, R. M. Phys. Rev. B 1981, 23, 6348.

Nemanich, R. J .; Glass, J. T.; Lucovsky, G.; Schroder, R. E. J. Vac. Sci.
T echnol. A 1988, 6, 1783.

Sowers, A. T.; Ward, B. L.; English, S. L.; Nemanich, R. J. J. Appl. Phys.
1999, 86, 3973.

Dennison, J. R.; Holtz, M.; Swain, G. Spectroscopy 1996, 11, 38-46.

Mermoux, M.; Marcus, B.; Swain, G. M.; Butler, J. E. J. Phys. Chem. B 2002,
106, 10816-10827.

Foord, J. S.; Hian, L. C.; Jackman, R. B. Diamond Relat. Mater. 2001, 10,
710-714.

Shirafuji, J .; Sugino, T. Diamond Relat. Mater. 1996, 5, 706-713.

Bard, A. J .; Faulkner, L. R. Electrochemical Methods F undationals and
Applications, Second ed.; John Wiley & Sons Inc., 2001.

DeClements, R.; Swain, G. M.; Dallas, T.; Holtz, M. W.; I], R. D. H.; Stickney,
J. L. Langmuir 1996, 12.

Yagi, 1.; Notsu, H.; Kondo, T.; Tryk, D. A.; Fujishima, A. J. Electroanal.
Chem. 1999, 473, 173-178.

Gerischer, H. J. Phys. Chem. 1985, 89, 4249—4251.

125

(50) Gerischer, H.; Ekardt, w. Appl. Phys. Lett. 1983,43, 394-396.

(51) Gerischer, H.; McIntyre, R.; Scherson, D.; Storck, W. J. Phys. Chem. 1987, 91,
1930-1935.

(52) Kneten, K. R.; McCreery, R. L. Anal. Chem. 1992, 64, 2518-2524.

(53) Chen, P.; Fryling, M. A.; McCreery, R. L. Anal. Chem. 1995, 6 7, 3115-3122.
(54) Chen, P.; McCreery, R. L. Anal. Chem. 1996, 68, 3958-3965.

(55) Hu, 1. F.; Karweik, D. H.; Kuwana, T. J. Electroanal. Chem. 1985, 188, 59-72.
(56) Kuo, T.-C.; McCreery, R. L. Anal. Chem. 1999, 71, 1553-1560.

(57) Wang, J.; Swain, G. M.; Mermoux, M.; Lucazeau, G.; Zak, J.; Strojek, J. W.
New Diamond Front. Carbon T echnol. 1999, 9, 317-343.

(58) Chen, Q.; Swain, G. M. Langmuir 1998, 14, 7017.

(59) Deakin, M. R.; Stutts, K. J.; Wightrnan, R. M. J. Electroanal. Chem. 1986,
I82, 113.

(60) Peter, L. M.; Dilrr, W.; Bindra, P.; Gerischer, H. J. Electroanal. Chem. 1976,
71 , 31.

(6]) Noel, M.; Anantharaman, P. N. Analyst 1985, 110, 1095.
(62) Granger, M. C.; Swain, G. M. J. Electrochem. Soc. 1999, 146, 4551-4558.
(63) Wang, J .; Swain, G. M. J. Electroanal. Chem.

(64) Vinokur, N.; Miller, B.; Avyigal, Y.; Kalish, R. J. Electrochem. Soc. 1996,
143, L238.

(65) Ranganathan, S.; Kuo, T.-C.; McCreery, R. L. Anal. Chem. 1999, 71 , 3574-
3580.

(66) Alehashem, S.; Chambers, F .; Strojek, J. W.; Swain, G. M.; Ramesham, R.
Anal. Chem. 1995, 67, 2812-2821.

(67) Pagan, D. T.; Kuwana, T. Anal. Chem. 1985, 57, 2759.

(68) Engstrom, R. C. Anal. Chem. 1982, 54, 2310-2314.

126

(69) Evans, J. F .; Kuwana, T. Anal. Chem. 1979, 51 , 358-365.

(70) Rice, R. J.; Pontikos, N. M.; McCreery, R. L. J. Am. Chem. Soc. 1990, 112,
4617-4622.

(7]) Saby, C.; Muret, P. Diamond Relat. Mater. 2002, 11, 851-855.
(72) Landstrass, M. 1.; Ravi, K. V. Applied Physics Letters 1989, 55, 975-977.
(73) Landstrass, M. 1.; Ravi, K. V. Appl. Phys. Lett. 1989, 55, 1391-1393.

(74) Ri, S.; Tashiro, K.; Tanaka, S.; Fujisawa, T.; Kimura, H.; Kurosu, T.; Iida, M.
Jpn. J. Appl. Phys. Part I 1999, 38, 3492.

(75) Maier, F.; Riedel, M.; Mantel, B.; Ristein, J.; Ley, L. Phys. Rev. Lett. 2000, 85,
3472-3475.

(76) Goss, J. P.; Jones, R.; Heggie, M. 1.; Ewels, C. P.; Briddon, P. R.; Oberg, S.
Phys. Rev. B 2002, 65, 115207.

(77) Looi, H. J .; Pang, L. Y. S.; Molloy, A. B.; Jones, F .; Foord, J. S.; Jackman, R.
B. Diamond Relat. Mater. 1998, 7, 550-555.

(78) Snidero, E.; Tromson, D.; Mer, C.; Bergonzo, P.; Foord, J.; Nebel, C.;
Williams, 0.; Jackrnann, R. J. Appl. Phys. 2003, 93, 2700.

(79) Hu, X. J.; Dai, Y. B.; Li, R. B.; Shen, H. 8.; He, X. C. Physica B 2003, 32 7,
39-42.

(80) Mantel, B. F.; Stammler, M.; Ristein, J .; Ley, L. Diamond Relat. Mater. 2001,
10, 429-433.

(8]) Hossain, M. Z.; Kubo, T.; Aruga, T.; Takagi, N.; Tsuno, T.; Fujimori, N.;
Nishijima, M. Surf Sci. 1999, 436, 63-71.

(82) Riedel, M.; Ristein, J .; Ley, L. Phys. Rev. B 2004, 69, 125338.

127

Chapter 4

Heterogeneous Electrical Conductivity of
Polycrystalline Boron-Doped Diamond Films

4.] Introduction

Polycrystalline, boron-doped diamond (BDD) thin ﬁlm, grown by chemical
vapor deposition (CVD), has been used as an electrochemical electrode material for
more than '10 years. 1'7 However, the relationship between the physical, chemical, and
electronic properties of BDD and its electrochemical performance is still incompletely
understood. In recent years, much has been learned about how different factors
control the electrochemical properties. For example, several authors have investigated
the kinetics and mechanisms of heterogeneous redox reactions at BDD, and two
general models have been proposed: 8'” (i) electron transfer via potential-dependent
electronic states within the band gap with a relatively uniform rate constant across the
surface and (ii) electron transfer through a partially blocked electrode consisting of
highly conductive or active sites separated by less conductive or less active regions. In
our early work, we hypothesized that the ﬁrst model explained the electrochemical
behavior of BDD.6 Now, based on recent work and results presented herein, we
suppose a combination of these two models best describes electron transfer at
diamond. Additional evidence in support of the second model comes from
investigations of electron emission from diamond, an analogous process that involves

electron transfer across a solid-vacuum interface.

128

It has been reported that BDD ﬁlms with higher electrical conductivity exhibit
a lower tum-on voltage for ﬁeld emission.”’ '6 The electron emission was observed to
occur from sites of high electrical conductivity. The possibility of some channel
and/or grain boundary conduction was proposed. 15’ '6

Developing an understanding of how the electrochemical reaction rate might
vary across the diamond electrode requires that one be able to characterize and
spatially distinguish the physical, chemical, and electrical properties. Without
question, polycrystalline diamond is one of the more structurally complex carbon
electrodes with factors such as the boron-doping level, hydrogen content, grain
boundary density, and electric double layer structure affecting electron-transfer rates.
For boron-doped microcrystalline diamond, a few papers have described spatially
inhomogeneous physicochemical properties, such as a variable distribution of boron
in doped polycrystalline diamondn‘ ‘8

Raman imaging, electrogenerated chemiluminescence imaging, and
photoluminescence measurements have revealed variations in the microstructure,
electrical properties, and electrochemical activity across a microcrystalline diamond
thin-ﬁlm surface.”21 Recently, electron emission studies from hydrogen-terminated
and chemically-modiﬁed boron-doped nanocrystalline diamond ﬁlms were consistent
with a heterogeneous surface having different local work functions”, This conclusion
was reached based on electron energy distribution measurements that were made as a
function of temperature. If a complete picture of the structure-reactivity relationship at
boron-doped polycrystalline diamond is to be developed, then probes of the ﬁlm
properties at the microscopic level are needed. Conducting probe-atomic force

microscopy (CP-AFM) is one such technique that is useful for investigating the

electrical properties of conductors and semiconductors with high spatial resolution?”

129

24 Both nanometer electrical characterization and topographic imaging are provided
by CP-AF M. The technique usually employs a metal-coated (Pt or Au) cantilever-tip
assembly that functions as both a scanning electrical contact and force sensor. Metal
wire probe tips can also be used.25 Thus, direct measurements of I-V curves at speciﬁc
topographical sites on the sample are possible.

CP-AF M has been used to probe the conductivity across hydrogen-terminated
microcrystalline diamond thin-ﬁlm electrodes with different boron-doping levels,26
oxygen-terminated BDD,27 and undoped diamond ﬁlms.28 The general observation, at
least in some of the work, is that the electrical conductivity is variable across
polycrystalline diamond with a strong dependence on the grain orientation, fraction of
grain boundary, and boron-doping level.“’ 28 These studies highlight the importance
of charge carrier concentration and mobility in BDD in controlling electrochemical
activity.27 Local electrochemical activity is clearly linked to the underlying
conductivity of individual grains, which is governed by the boron incorporation.

Our group has reported on the heterogeneous nature of the microstructure and
electrochemical activity of polycrystalline BDD using Raman microprobe
spectrosc0py and electrogenerated chemiluminescence imaging. 19’ 20 Additionally, a
pattern of variable conductivity and electrochemical activity at hydrogen-terminated
microcrystalline diamond electrodes, doped with boron at different concentrations,
was measured using CP-AF M and SECM.26 As a follow-up to this, we report herein
on a detailed investigation of the electrical properties of boron-doped diamond thin
ﬁlms using CP-AF M.

CP-AF M was used to simultaneously map the topography and electrical

conductivity of the thin ﬁlms as a function of the boron-doping level. The dependence

130

of the electrical conductivity on the boron-doping level, surface topography, and the

bias voltage (sample vs. tip) polarity was investigated.

4.2 CP-AFM images of nanocrystalline ﬁlms

Figure 4.]A-D shows conductivity maps (different current scales) for
nanocrystalline BDD ﬁlms grown with 50, 10, 1 or 0 ppm diborane added to the
source gas mixture. The boron-doping level decreased with decreasing B2H6
concentration in the source gas. The image contrast results from different levels of
current ﬂow between the sample and tip, which is dependent on the contact resistance
between sample and tip (a value that is inversely related to the contact radius of the
tip), the local charge carrier concentration at the electrode surface and the bias
voltage.” 29 In all of these images, the sample was biased +0.2 V relative to the tip.
To ensure that removal of the Au coating from end of the tip was not a complicating
factor, a new tip was used for imaging each ﬁlm and was checked for integrity by
multiple-scan imaging an area of the surface with multiple scans. Importantly, as has
been observed in recent work with hydrogen-terminated microcrystalline ﬁlms,26 the
images reveal variable electrical conductivity across the ﬁlm surface with regions of
high conductivity isolated by regions of low conductivity. The ability of the ﬁlm to
support current ﬂow increased proportionally with the boron-doping level as the
current was highest for the 50 ppm ﬁlm and lowest for 0 ppm ﬁlm at the 0.2 V bias.
Some of the more conductive regions are clustered together, whereas others are more
isolated. Some of more conducting regions have lateral dimensions of 2 to 4 pm,
which means they encompass multiple grains and grain boundaries (refer to 50 ppm
boron doped ﬁlm). The more conducting zones are separated from one another, in

many cases, by a micrometer or less. Only the undoped ﬁlm possesses very few

13]

conductive regions and, since there is little boron-doping (there will be some due to
reactor contamination with the species), we suppose the conductive regions are
associated almost exclusively with the grain boundaries in this ﬁlm. A similar
conclusion was reached by Temyak et al. in their CP-AF M imaging work of undoped
nanocrystalline diamond ﬁlm.28 The resolution of the conductivity maps in the present

work was not high enough to reveal the grain boundaries.

      
 

Figure 4.1 CP-AF M images (10 X 10 umz) of nanocrystalline BDD thin ﬁlms
grown with different source gas BzH6 concentrations (i.e., different doing levels): (A)
50, (B) 10, (C) 1 and (D) 0 ppm B2115. The bias voltage used for all images was +0.2
V vs. the tip. The current ranged from 0 to 6.5 11A. (Reprinted with permission from J.
Phys. Chem. 111, 3986 (2007). Copyright (2007) American Chemical Society.) (From

Ref. 49).

The fact that the contact resistance between the diamond and tip decreased

with increasing boron-doping level indicates that the electrical properties are strongly

132

inﬂuenced by the boron-dopant and not solely by the existing grain boundary or any
adventitious non-diarnond sp2 carbon impurity. Therefore, we conclude, as was the
case in past work, that the conduction pathway is not solely through the grain
boundary but rather through a combination of the doped diamond grains as well as the
grain boundary.30 The same conclusion has been reached by Macpherson and
coworkers in their work with doped diamond electrodes.27 It is expected that the
boron substitutionally inserted into the diamond grains is the primary contributor to
the charge carrier concentration in the ﬁhn, but it is also possible that aggregated
boron atoms in the grain boundary could also contribute to the carrier concentration

by increasing the sp2 carbon bonding (It states).3 ‘

The contact area between the tip and sample is an inﬂuential factor when
measuring polycrystalline diamond because of the tip curvature, which is ca. 50 nm.

This can be seen below in the expression that describes the contact resistance, R,

R=amm2 an

where p and 1, respectively, denote the resistivity and the electron mean free path in
the electrode, and a is the tip radius.29 The radius of curvature of the tips used in our
work is comparable to the roughness of the nanocrystalline ﬁlm. The pattern of
conductivity indicates that the thin ﬁlm consists of regions of high electrical
conductivity isolated from one another in a more insulating matrix, at least within our
detection ability, which is < 0.1 ILA for current or > 10 M9 for resistance. One reason
for the differential conductivity is the possibility of boron-rich regions due to

inhomogeneous boron-doping.” 18’ 20 This will produce regions of higher charge

133

carrier concentration (i.e., holes). Consistent with this is previous Raman imaging
work that revealed an inhomogeneous distribution of boron dopant in mechanically
polished polycrystalline diamond.20 Another possible reason for the differential
conductivity, and one consistent with the electron emission studies published recently,
is a variable work function across the surface.22 It has been reported that the work
function of diamond depends on the crystallographic orientation.32 There are two
reasons for this. First, the (100) face has the highest density of dangling bonds while
the (111) face has the lowest.” 34 Second, polycrystalline ﬁlms that are hydrogen-
terminated exist with a less than full coverage on the (100) surface33 as compared to
the (111) surface,35 with the work function being dependent on the surface coverage.
In fact, hydrogen is a ubiquitous elemental component of diamond. Chemisorbed
hydrogen serves to maintain the sp3 bonding on the surface and prevents formation of
rt-bonded carbon. Therefore, its presence has a signiﬁcant inﬂuence on the local work
function. Furthermore, it has been known for some time that the electrical properties
of diamond surfaces are strongly modiﬁed by the presence of hydrogen.36'38
Hydrogenation leads to an increase in surface conductivity through increased hole
conduction as the element acts as an electron acceptor with an activation energy lower
than that for boron. ”'38 The surface conductance can also be inﬂuenced by
electrochemical transfer doping by surface adsorbates.39 Finally, hydrogen can reduce
the conductivity of boron-dOped diamond by forming acceptor-donor complexes with
substitutionally-inserted boron.40 Differential local work functions would contribute
to variable electrical conductivity across the surface (i.e., from facet-to-facet). Several
groups have reported work function variations for hydrogen and oxygen-terminated
diamond when imaged by Kelvin probe microscopy.” 42 Based on such

measurements, Brandes and Mills concluded that defects and/or oxygen on the surface

134

likely produce the surface states that create the positive electrical ﬁeld due to surface

43 This would decrease the surface

electron occupying these surface states.
conductivity. Although these measurements were performed on undoped diamond,
it’s reasonable to expect that the accumulation of boron on the surface will increase
the surface states and change the surface conductivity of boron-doped diamond.“
Wang et a]. reported, based on ﬁeld emission imaging and scanning tunneling
spectroscopy studies, that the dominant emission sites in boron-doped p-type diamond
are located at or near the grain boundaries due to a lower work function than the
surrounding grains.45 Finally, Temyak et a1. mentioned another possible reason for the
differential conductivity and that is inhomogeneous hydrogen and impurity coverage
across the surface, which could affect the surface conductivity in CP-AF M

measurements made in air.28

4.3 CP-AFM images of microcrystalline ﬁlms

Figure 4.2 shows conductivity maps of microcrystalline ﬁlms grown with 10,
5, l, 0.5, and 0 ppm B2H6 added to the source gas mixture. A bias voltage of +2.0,
rather than +0.2 V, was needed to image these ﬁlms. This is because there was a
greater tip-electrode contact resistance with the microcrystalline ﬁlms. Therefore,
instead of measuring a current from 0 to 6.5 11A, as was the case for nanocrystalline
diamond doped with 50 ppm Bsz, the current measured for microcrystalline diamond
ranged only from O to 0.65 11A. We attribute this increased resistance to a reduced tip
contact area with the rough microcrystalline surface. The images reveal a pattern of
heterogeneous conductivity similar to that of the nanocrystalline ﬁlms and similar to
that of other previously imaged microcrystalline ﬁlms.” Some of the conductive

regions are clustered together, whereas others are more isolated. The conductive

135

regions for the highly doped ﬁlms are several micrometers in lateral dimension
meaning that the sites comprise both faceted crystallites and the grain boundary
between them. As was the case for the nanocrystalline ﬁlms, the fraction of highly

conductive area decreased with decreasing boron-doping level.

    

B. 5 ppm C. 1 ppm

      

. 0.5 ppm

    

0
0 2.5 5.0 7.5 10.0
pm

 

Figure 4.2 CP-AFM images (10 X 10 umz) of microcrystalline BDD thin
ﬁlms grown with different source gas B2H5 concentrations (i.e., different doing
levels): (A) 10, (B) 5, (C) 1, (D) 0.5 and (E) 0 ppm B2H6. The bias voltage used for all
images was +2.0 V vs. the tip. The current ranges from 0 to 0.65 11A. (Reprinted with
permission from J. Phys. Chem. 111, 3986 (2007). Copyright (2007) American

Chemical Society.) (From Ref. 49).

On the basis of the conductivity data reported above, several points of
comparison can be made between the two diamond ﬁlm types, although it should be
re-emphasized that both highly doped ﬁlm types function well as electrochemical

electrodesf” 14‘ 30 First, the contact resistance between the diamond and the Au tip was

136

greater for the microcrystalline ﬁlm. Second, the fraction of the imaged area that
exhibited high electrical conductivity was less for microcrystalline diamond. As
discussed above, there are several possible reasons for the differential conductivity
across the surface of a boron-doped ﬁlm. CP-AFM is sensitive to the contact
resistance between the tip and electrode, which is proportional to pI/a". The resistivity,
p, of the ﬁlm in the vicinity of the tip is proportional to the product of the local carrier
concentration and the carrier mobility. Since both ﬁlm types have approximately the
same macroscopic resistivity (ca. 10'2 9cm), we suppose that the difference in contact
resistance for the two ﬁlm types is related to the size and shape of the tip relative to
the grain size and morphology of the ﬁlm surface. To address this issue more clearly,

a SEM image for the conductive tip used in this study is shown in Figure 4.3.

Tip apex that
contacts the
sample

 

Figure 4.3 SEM image of a gold-coated, tetrahedral-shaped silicon nitride tip
(V eeco NPG-20) with an apex radius of curvature less than 50 nm and a tip height of

~15-20 m.

137

The apex radius of curvature is approximately 50 nm, which is much smaller
than the grain size (ca. 1 pm) of the microcrystalline ﬁlm but comparable to the
nodular features (ca. 50 nm) of the nanocrystalline ﬁlm. Therefore, the contact area
between the tip and the sample for microcrystalline diamond ﬁlms is likely greater
than that for nanocrystalline diamond ﬁlms. This is schematically depicted in Figure
4.4. When the tip scans over a microcrystalline ﬁlm, the tip interacts with the surface
using a smaller contact area (Figure 4.4A) than when imaging a nanocrystalline ﬁlm
(Figure 4.43). The smaller contact area gives rise to a higher contact resistance and
lower current for a given bias voltage. The difference in contact area could also

explain why the microcrystalline ﬁlm has a less conductive area.

A. CP-AF M tip interaction with a microcrystalline BDD ﬁlm

 

 

 

Figure 4.4 Schematic representation of how the tip interacts with the (A) rougher
and faceted microcrystalline diamond ﬁlm and (B) smoother nanocrystalline diamond
ﬁlm. (Reprinted with permission from J. Phys. Chem. 111, 3986 (2007). Copyright

(2007) American Chemical Society.) (From Ref. 49).

138

4.4 CP-AFM images and topography

Figure 4.5 shows topographical (A) and conductivity (C) maps for a
nanocrystalline diamond ﬁlm grown with 10 ppm B2H6 in the source gas, along with
line proﬁles through each image (B and D). The bias voltage was +0.2 V, and the scan
size was 500 X 500 nmz. In the topographical map, nodular features of 50-100 nm in
the lateral dimension are seen. The boundaries between these features are also clearly
visible, as is a depression in the topography in the upper right quadrant of the image.
This region is about 50 nm in width and 250 nm in length. Some of the least
conducting regions in the image are actually associated with features of greatest
topographical height. These are the sites where presumably the tip would have the
least area in contact with the surface, hence a greater contact resistance. In contrast,
many of the most conducting regions are associated with low points in the topography
where the contact area with the tip would be greatest. The exception to this is point C]
where a relatively high topographical feature exhibits high conductivity. The contact
resistance at the highly conducting regions is approximately the same over the surface

as evidenced by the constant magnitude of the current. By use of the equation

R = Vbias/i (4.2)

where Vbias is the applied voltage and i is the measured current, a contact resistance of

~ 1 M0 was calculated. It is supposed that this contact resistance also reﬂects the
resistance of current ﬂow from the tip to theelectrode and then through the diamond
bulk, the diamond/ Si interface, and the conducting Si substrate before reaching the
metal current collector (see Section 2.5 and Figure 2.9). Importantly, not all of the

nodular features and surrounding grain boundaries exhibit high conductivity. For

139

example, it can be seen in Figure 4.5D that the line scan passes through a nodular
feature on the leﬁ that exhibits low conductivity. At its boundary (mark a), high
conductivity is exhibited. The line scan then goes through another nodular feature on
the right (mark d) that exhibits high conductivity. This again maybe related to

inhomogeneous boron dopant distribution in the nanocrystalline ﬁlm.4648

A

nm B

 

20.0

 

 

 

0.4

¥
3...

 

 

 

 

Figure 4.5 Topographical (A) and conductivity maps (C) for a 0.5 x 0.5 pm2

region of a nanocrystalline BDD thin ﬁhn grown with 10 ppm B2H6 in the source gas.
Corresponding line proﬁles through the image (B and D) are also presented. The bias
voltage used to obtain the conductivity map was +0.2 V vs. the tip. The Z-axis scales
are 0 to 50 nm for height image and O to 0.65 uA for the current image. (Reprinted

with permission from J. Phys. Chem. 111, 3986 (2007). Copyright (2007) American
Chemical Society.) (From Ref. 49).

140

The relationship between the conductivity and topography of microcrystalline
ﬁlms was also examined by CP-AF M. Figure 4.6 shows the conductivity (A) and
deﬂection mode (B) maps of microcrystalline ﬁlm doped with 10 ppm diborane added
to the source gas mixture. The bias voltage applied to BDD thin-ﬁlm was +0.2 V
relative to the tip and the scan size was 2 x 2 umz. For microcrystalline diamond ﬁlms,
only one or two grains can be imaged within this scan size. The grain boundaries can
be imaged clearly in this range and this helps to study the relationship between the
topography and conductivity of the sampled surface. The deﬂection image is shown
here because it more clearly reveals the grain boundaries. The deﬂection image is
gained by measuring the twisting motion of the tip as it travels across a surface.
Therefore, it is more sensitive to the grain boundary because big changes of the
surface curvature exist in boundaries causing much bigger twisting of the cantilever.
There is evidence of this in the images in Figure 4.6, where some correlation is
observed between the surface topography and conductivity, with many areas of high
conductivity being associated with grain boundaries between diamond crystallites.
Most of crystalline facets do not show high conductivity. A possible reason for the
lower conductivity of the facets is that continuous electrical pathways from these
areas to the conductive substrate might not exist In the present work, electrical
contact was made from the bottom of the silicon substrate by carbon tape and silver
paint (see Fig. 2.9), and this means that conduction is only possible through certain
pathway on the diamond ﬁlm from the substrate to the surface. Therefore, areas of
conductivity imaged by CP-AF M may correspond to sites where surface boron dopant
(or other impurity such as sp2 carbon) sites can establish electrical contact to the
silicon substrate through the ﬁlm. Another reason might be the boron aggregation in

grain boundaries, as mentioned above.“ '8’ 20

141

 

0 2.00 0 2.00
um um

Figure 4.6 Conductivity (A) and topographical maps (B) for a 2 X 2 pm2 region of
a microcrystalline BDD thin ﬁlm grown with 10 ppm B2H6 in the source gas. The bias

voltage was +0.2 V vs. the tip. The current scale was 0 to 0.65 uA.

4.5 Effect of bias voltage polarity

When a p-type semiconductor electrode is positively biased, electrons should
easily ﬂow from the conducting tip into acceptor states at the electrode surface.
However, when the electrode is negatively biased, current ﬂow should be inhibited
due to a depletion of charge carriers at the surface (this is true for the case of (I), < (Dm,
where d), is the work function of the semiconductor, i.e., diamond, and CD". is the work
function of gold, i.e., the tip used in present work). On the other hand, a conductor
will obey Ohm’s Law, which means the current ﬂow will be proportional to the bias
voltage and that the polarity only changes the direction of current ﬂow, not the
magnitude. Experiments were performed with both diamond ﬁlm types to study how
the conductivity maps changed as a function of the bias voltage polarity. Both highly
doped microcrystalline and nanocrystalline BDD showed conductive features that did
not change in size or number with the bias polarity. Figure 4.7 shows a typical

example for a nanocrystalline ﬁlm grown using 10 ppm B2H6 in the source gas.

142

Conductivity maps were recorded over the same region at +0.2 and -O.2 V (parts A
and C). Line proﬁles of the current response are also presented (parts B and D). The
conductivity maps and the current-line plots are nearly mirror images of one another.
The same regions of the ﬁlm exhibited high conductivity at both bias voltages
regardless of the polarity. This means that the conducting sites are ohmic in nature
and possess metal-like electronic properties. Furthermore, the current magnitudes at
the conducting sites were similar for a given bias voltage and no new conductive

regions appeared or disappeared with change in bias voltage.

M B

 

0.65‘

M

0.9 .
:11 WNW WW I)

 

 

 

 

 

0.65 ~

0.9: MM“) hum!

WW

 

-0.65—

 

 

 

 

0.0 ' I 2.0
pm
Figure 4.7 Conductivity maps (2 X 2 umz) obtained from CP-AFM images of the

same area of a nanocrystalline BDD thin ﬁlm grown with 10 ppm Bzﬂo in the source
gas at bias voltages of (A) +0.2 V and (C) -0.2 V vs. the tip. Corresponding line
proﬁles of the same region are presented in (B) and (D). The Z-axis scale was 0 to
0.65 pA for both A and C. (Reprinted with permission from J. Phys. Chem. 111, 3986
(2007). Copyright (2007) American Chemical Society.) (From Ref. 49).

143

The dependence of the current magnitude on the bias voltage was also
investigated. Parts A-C of Figure 4.8 show three conductivity maps at bias voltages of
+0.2, +1.0, and +2.0 V vs. the tip for the same region of a nanocrystalline BDD grown
with 10 ppm B2H6 in the source gas mixture. The circles in every panel mark the same
conductive area in each image.

The current ﬂow at the marked site increased from 0.11 to 0.54 to 1.21 uA
with increasing bias voltage. This corresponds to a relatively constant contact
resistance of 1.82, 1.85, and 1.65 MQ, respectively. The ratio of the current
magnitudes was 11:5: 1, which is close to the ratio of the applied bias voltages, 10:5: 1.
This is the expected behavior for a metallic conductor. Even with the image driﬁ,
these images also show that no new conductive sites appear with increasing bias
voltage. A ﬁxed number of conducting areas exist in each ﬁlm, particularly for the
more highly doped ones. In other words, the fraction of conducting area does not vary
with the magnitude of the bias voltage. This is different from the continuous image

tunneling spectroscopy (CITS) results (discussed later in Section 4.8).

144

 

 

' 6.00 ' 0:25” >030“ 075 "i100

Figure 4.8 CP-AF M section analysis of (1 X 1 umz) of a nanocrystalline BDD thin
ﬁlm grown with 10 ppm B2H6 in the source gas at bias voltages of (A) +0.2 , (B) +1.0
and (C) +2.0 V vs. the tip. The measurements were made on the same area of the ﬁlm.
(Reprinted with permission from J. Phys. Chem. 111, 3986 (2007). Copyright (2007)

American Chemical Society.) (From Ref. 49).

145

4.6 CP-AFM images and surface termination

Conductivity maps for “as grown”, hydrogen-terminated and oxygen-
tenninated microcrystalline and nanocrystalline boron-doped diamond thin ﬁlms were
also acquired. Figure 4.9a and c show 1 X 1 pm2 conductivity maps for “as grown”
microcrystalline and nanocrystalline diamond thin ﬁlms (stored in lab atmosphere for
weeks), respectively. These images were collected with the sample biased at 0.2 V
positive relative to the tip. This means that the electrons were ﬂowing from the tip
into the diamond ﬁlm. The map of the microcrystalline diamond ﬁlm surface reveals
considerable electrical heterogeneity. There are regions of high electrical conductivity
(the bright areas) isolated by regions of signiﬁcantly lower conductivity. It can be
seen from the scale bar that currents approaching 0.65 uA are ﬂowing through the
conducting regions (Figure 4.9b). Similar heterogeneity was observed in images over
a larger area (10 X 10 umz). Variable conductivity is also seen for the nanocrystalline
ﬁlm. However, the fraction of area that is conducting is greater for the nanocrystalline
ﬁlm. It can be seen from the proﬁle (Figure 4.9d) that the current passed through the
conducting regions of the nanocrystalline ﬁhn is lower than that for the
microcrystalline ﬁlm. This could be due to the difference in contact area between the
pyramidally-shaped tip and the rough microcrystalline surface and the smooth
nanocrystalline surface. The tip curvature is approximately 50 nm, which is much
smaller than the grain size (ca. 1 pm) of the microcrystalline ﬁlm but comparable to
the nodular features (ca. 50 nm) of the nanocrystalline ﬁlm. When the tip scans over a
microcrystalline ﬁlm, the tip may interact with the surface using a larger contact area
than for the nanocrystalline ﬁlm in some special regions (e. g., grain boundaries) as
shown in Figure 4.4.49 The higher contact area gives rise to a lower contact resistance

and a larger current for a given bias voltage.

146

 

0.65

:0ng V v V iimm

 

 

 

 

 

 

'1‘:

‘°.-

‘? .

0 025 050 075 100

um
d

v:

‘Q-

o

 

 

-0.65

0 m L00 0 0.55 0.50 0.75 1.00
pm

 

 

 

Figure 4.9 CP-AFM images (1 X l umz) for “as grown” (a) microcrystalline and
(b) nanocrystalline BDD thin ﬁlms. A bias voltage of 0.2 V vs. the tip was used for all

images. The current ranged from 0 to 0.65 uA.

Figure 4.10a and c show 1 X 1 pm2 conductivity maps for a freshly
rehydrogenated and an oxidized nanocrystalline diamond thin ﬁlm. In these images,
the sample was biased 0.2 V positive to the tip. It’s noteworthy that, compared to the
“as grown” ﬁlms (see Figure 4.9), the rehydrogenated ﬁlm is less in surface oxygen
while the oxidized ﬁlm is rich in surface oxygen. This is why these images are so
different from those in Figure. 4.9. Although the map of the H-terminated surface
reveals heterogeneous conductivity, the fraction of area that is conducting is much
greater than that for the O-terminated nanocrystalline ﬁlm. It can be seen from the

scale bar (Figure 4.10b) that the conducting regions are passing currents near 0.65 uA

147

for the H-terminated nanocrystalline ﬁlm, whereas the currents passed for the O-

terminated ﬁlm are around just 0.1 uA (factor of 6 times less, see Figure 4.10d).

 

O
O. 8
~ 0'

 

 

 

 

 

 

 

 

 

 

 

tn
‘0. -
‘ ‘ 0 9 . . .
1.00 0 0.25 0.50 0.75 1.00
pun
d
MAJ-Jr r. mumuﬂ‘.' mu ..... .. .. .. u... I... u 1....“ ma... :1 aural
"nu-"v" If"? "'11-". 'I r, w I' .- am n - I v '1' w 1 1 l 1' I' W'Yl' "Tl '
T 0 o is o 50 o is 1 00
M I in I I

Figure 4.10 CP-AFM images (1 X 1 umz) for (a) hydrogen-terminated and (b)
oxygen-terminated nanocrystalline BDD thin ﬁlms. A bias voltage of 0.2 V vs. the tip
was used for all images. The current ranged from O to 0.65 uA for H-terminatcd ﬁlm

and 0 to 0.16 uA for O-terminated ﬁlm.

Figure 4.11a and c show 1 X 1 pm2 conductivity maps for freshly
hydrogenated and oxidized microcrystalline diamond thin ﬁlms. In these
measurements, the bias potential applied was 0.2 V relative to the tip. The map of the
H—terrninated diamond surface reveals that the surface is not uniformly conducting.

There are regions of higher electrical conductivity isolated by

148

 

 

 

 

 

 

 

 

 

 

 

 

tn
N.
('1
1° Wm‘ﬂvivnwm '
m
N.
f? r l I
0 o 25 0.50 o 75 1.00
d
W
0..
o
a i“, Y m 1 #4me
In
we-
‘?
- 0 02's 0.50 oris 1.00
"m in-

Figure 4.11 CP-AFM images (1 x 1 umz) for (a) hydrogen-terminated and (b)
oxygen-terminated microcrystalline BDD thin ﬁlms. A bias voltage of 2.0 V vs. the
tip was applied for hydrogen-terminated ﬁlm and the currents at some sites were out
of range (> 3.25 uA). For the oxygen-terminated ﬁlm, a bias voltage of 0.2 V vs. the

tip was applied and the current ranged from O to 0.65 uA.

regions of lower conductivity. The image for the O-terminated microcrystalline ﬁlm
also shows isolated conducting regions. Although the currents passed through the O-
terminated ﬁlm are lower than those for the H—terminated ﬁlm, the O-terrninated
microcrystalline ﬁlm shows a lower fraction of conducting area than does the H-
terminated microcrystalline ﬁlm (Figure 4.11b and d). Comparing with the conductive
map of the O-terminated nanocrystalline ﬁlm (Figure 4.10c and d), the conductive

map of O-terminated microcrystalline ﬁlm reveals higher fraction of conductive area

149

and larger current for the former when probed at same bias potential (Figure 4.1 lc
and (1). However, these changes in conductivity are reversible. High conductivity can

be recovered after removal of the oxygen functionalities by rehydrogenation.

4. 7Special issues with CP-AFM measurements

There are several critical issues when imaging polycrystalline diamond by CP-
AFM. One of these is the tip chemistry and shape. The tip used in this work was a
gold-coated, tetrahedral-shaped silicon nitride tip (Veeco NPG-20), with an apex
radius of curvature less than 50 nm, a tip height of ~15-20 um, and a full-tip cone
angle less than 35°, mounted on a rectangular cantilever. In limited scans, this type of

tip worked well, as shown in Figure 4.12.

A

p
5."

   

Figure 4.12 CP-AF M images during 2 consecutive scans over the same area of a 10
ppm boron-doped nanocrystalline diamond thin ﬁlm (rehydrogenated). The bias

voltage applied was +0.2 V vs. the tip.

However, because of the hardness of diamond, the Au coating over the tip as

well as its tetrahedral shape is adversely affected by the mechanical action with the

150

surface. Au is removed and the sharp tip is smoothed and ﬂattened during scanning
over the hard surface. This causes the CP-AFM image to change over time. For
example, it is observed that the conducting regions of a particular ﬁlm decrease in
number and size with increasing scan number (Figure 4.13). In general, we found that
stable conductivity maps could be obtained with a tip for both ﬁlm types for up to 4
scans over a 10 X 10 um2 area. All CP-AFM maps presented herein were recorded
over an area at least twice to make sure the image features were real and were not

artifacts of the tip condition.

A

 

Figure 4.13 CP-AF M images as a function of the scan number for a 10 ppm boron-
doped nanocrystalline diamond thin ﬁlm. (A) 15‘ scan, (B) 51h scan and (C) 10th scan.
It shows that the CP-AFM images changed after 5th scan. The bias voltage applied

was +0.2 V vs. the tip.

151

Another issue is the overloaded current. The conductivity may be very high
for some highly doped diamond nanocrystalline ﬁlms (e.g. 50 ppm hydrogenated
nanoﬁlms). In this case, a bias of just 0.2 V vs. the tip produces a large current that
can damage the Au overlayer on the tip. In our work, we experienced an extremely
serious case where the current ﬂow ceased before one image could be completed.
Using SEM, we found the tip was total etched away, as shown in Figure 4.14. The

part of the tip itself and the Au coating on it might be melted and etched away at a

high temperature due to the large current.

 

Figure 4.14 SEM images for (A) an original CP-AFM tip and (B) a damaged tip
aﬁer one scan with an overloaded current. The image was taken using a JSM-6300F
(ﬁeld emission) scanning electron microscope in the Center of Advanced Microscopy

at Michigan State University.

4.8 Continuous imaging tunneling spectroscopy ( CI TS)
images

The results that were obtained by CP-AFM on boron-doped diamond thin

ﬁlms conﬂicted, to some extent, with results published in the literature.“ so For

152

example, we observed that the grain boundary area possesses special features of
electrical conductivity.49 Using the same technology (i.e. conducting probe AFM),
Macpherson and coworkers observed no difference between the grain boundaries and
other areas on a ﬁlm.”’ 50 A noteworthy fact is that their sample was a polished,
oxygen-terminated boron-doped diamond ﬁlm, which was quite different from our
samples. Second, Schottky I-V curves have been recorded for undoped,5 1’ 52
oxidized“, hydrogen-terminated53 and boron-doped diamond samples.54 Baker et al.
reported Schottky I-V curves for an oxidized synthetic boron-doped diamond ([B]
was 3 - 1500 ppm as measured by secondary ion mass spectroscopy) with a threshold
voltage greater than +3.8 V.54 Our CP-AF M measurements showed a linear
relationship between the applied bias and the recorded current. This observation held
true both from -2.0 to -0.2 V and +0.2 to +2.0 V vs. the tip. We did not test the
electrical properties of our samples between $0.2 V vs. the tip. In other words, we
observed metallic-like properties for our boron-doped diamond ﬁlms (Figure 4.6).
CITS55 is a good choice for us to check if our results reliable. CITS is a
scanning probe microscopy developed from scanning tunneling microscopy by
Hamers et al. in 1986.55 The principle of CITS is very similar to that of STM. Usually,
CITS works in a constant-current STM mode but the feedback control circuit used for
constant-current STM is gated so that it is only active in a part time (e. g., 30%). When
the feedback loop is active, a constant voltage is applied. When it is inactive, the
position of the tip is held stationary and the tunneling current is measured at various
bias voltages. Repeating this sequence at a certain frequency, a constant sample-tip
distance is maintained during all i-V measurements and the i-V characteristics of each
point along the scan are determined simultaneously with the topography. This method

can record the real-space images of the tunneling current directly reﬂecting the spatial

153

distribution of the surface states at different potentials. In other words, the technique
can be used to measure current-voltage characteristics at each point in the scanned
area. Therefore, this technique is a powerful tool to determine whether the nature of a
metal-semiconductor contact is ohmic (linear response of current to voltage) or
Schottky (nonlinear response of current to voltage).54 In fact, Iseri et a1. recorded
both ohmic and Schottky I-V curves on different sites in the same microcrystalline
diamond ﬁlm.“5 To further study the electrical properties of our diamond ﬁlms and to
conﬁrm our CP-AF M measurements, we sent a batch of microcrystalline and
nanocrystalline BDD diamond thin ﬁlms to Prof. Steven Prawer’s laboratory
(University of Melboune, Australia). All CITS images presented here were obtained

by Prof. Prawer’s group.

4. 8. 1 CI T S measurements for microcrystalline boron-doped diamond
ﬁlms

Figure 4.15 shows 2 X 2 pm2 CITS and topographic images obtained for a
microcrystalline boron-doped diamond ﬁlm grown on a highly doped Si substrate.
The bright region in Figure 4.15A indicates the high tunneling current area. It can be
seen that most of the ﬁlm, except the middle region at top right of the image emitted
electrons. Figure 4.15 C shows a superimposed image of tunneling and topography. It
clearly shows that some grain boundaries on some speciﬁc facets emit electrons
strongly.

According to the CITS measurement, two speciﬁc regions were selected: one
of high conductivity and one of low conductivity, respectively. The I-V characteristics
for these two regions were measured and shown on Figure 4.16. The I-V curves

shown in Figure 4.16 are the averaged results over the selected region, respectively.

154

Figure 4.16A shows a selected grain boundary and its I-V characteristics. An almost
perfect linear relationship between the tunneling current and the applied sample
voltage indicates that this region possesses typical ohmic (metallic) nature within the
applied voltage (+2.5 V). Figure 4.16B shows a selected facet with low electron
emission and its I-V characteristics. The I-V curve clearly shows that the tunneling
current is near to zero ﬁom -0.7 to +0.5 V for this region. When the applied voltage is
higher than +0.5 V or lower than -0.7 V, the tunneling current increases exponentially,

indicating this region possesses Schottky (semiconductor) electrical properties.

 

Figure 4.15 (A) CITS and (B) topographic images for microcrystalline boron-
doped diamond thin ﬁlm grown by CVD on highly doped Si substrate. (C) is the

superimposing image of (A) and (B).

155

 

 

Figure 4.16 Differential electrical properties of two regions of a microcrystalline

BDD thin ﬁlm. (A) a selected region on the topographic map (white square) and its 1-
V characteristics showing ohmic behavior, and (B) a selected region on topographic

map (white square) and its I-V characteristics showing schottky behavior.

More interesting images were acquired for a microcrystalline diamond thin
ﬁlm with higher boron doping level. Figure 4.17 shows the topographic images (10 X
10 umz) and I-V curves (averaged) for selected regions on this ﬁlm. In a grain

boundary region (Figure 4.17A), the I-V curve shows linear relationship over 12.0 V

156

except when the polarity of the testing voltage changes. The current jumped from ca. -
l nA to +2.0 nA as the polarity of voltage changed from negative to positive. On the
other hand, the current was near zero over this voltage range at a crystalline facet but
increased at more positive and more negative bias voltages. There is clearly a gap
where current ﬂow is negligible. It’s noteworthy to mention that the current scale for

Figure 4.17B was 2 orders of magnitude smaller than that for Figure 4.17A.

 

 

 

Figure 4.17 Differential electrical properties of two regions of a microcrystalline
BDD thin ﬁlm. (A) a selected region with high electron emission on the topographic
map and its I-V characteristics, and (B) a selected region with low electron emission

on the topographic map and its I-V characteristics. Scanned size: 5 x 5 umz.

157

All of the above results suggest that some grain boundary regions are much
more electrically conductive than some of the crystals of the microcrystalline boron-

doped diamond thin ﬁlm.

4. 8.2 CI TS measurements for nanocrystalline boron-doped diamond
thin ﬁlms

Nanocrystalline boron-doped diamond thin ﬁlms were also imaged by CITS to
study the relationship between the topography and electronic properties. Figure 4.18
shows 1 X 1 pm2 topographic images for a nanocrystalline boronodoped diamond ﬁlm
grown on highly doped Si substrate. Two representative regions were selected
(regions within white rectangle) and their I-V curves (averaged results over the
selected region) were recorded and also shown in Figure 4.18. From the topographic
images, nodular features with different sizes (50 ~ 250 nm) are clearly seen. Both
selected regions consisted of nodular surfaces and boundaries, expected a
nanocrystalline diamond ﬁlm grown from an Ar-rich source gas. But the CITS
measurements revealed quite different electron tunneling in these two regions. For the
selected region shown in Figure 4.18A, a linear relationship between the tunneling
current and the applied sample voltage was recorded indicating a typical ohmic
(metallic) nature within the applied voltage (+2.0 V). Figure 4.18B shows a different
image of electron emission on this diamond ﬁlm. The I-V curve clearly shows that the
tunneling current is near to zero from -0.7 to +0.5 V. When the applied voltage
becomes greater than +0.5 V or less than -0.7 V, the tunneling current increases
exponentially, indicating this region possesses Schottky (semiconductor) electrical
properties. Different from what was observed for microcrystalline diamond ﬁlm

(Figure 4.16), where different current scales were recorded for different regions

158

indicating high or low electron emission. The currents for two different selected
regions at same bias voltage were similar except for the inhibiting voltage. This
means that these two regions both possess high electron emission but they emitted

electrons through different mechanisms.

 

 

Figure 4.18 Different electrical properties of two representative regions on a

nanocrystalline boron-doped diamond thin ﬁlm. (A) a selected region with ohmic
behavior on the topographic image and its I-V characteristics, and (B) a selected

region with Schottky behavior on the topographic image and its I-V characteristics.

159

Prof. Prawer’s group also probed our nanocrystalline ﬁlms in contact mode
with a conductive cantilever (W2C). Just like CP-AFM, this technique can investigate
the electrical conductivity and topography spatially over the sample surface. Figure
4.19 shows (A) a current image, (B) a topographic image and (C) a merging of these

two images. In the current image, the current level is indicated by the brightness.

“CITS” data were acquired with bias voltage from -2 to +2V.

 

Figure 4.19 CITS images probed in contact mode with a conductive tip for a
nanocrystalline boron-doped diamond thin ﬁlm. (A) current image, (B) topographic

image, and (C) superimposing of (A) and (B).

160

As seen in Fig. 4.19A, current was observed over most regions of the surface
indicating a pretty high fraction of conductive area. However, levels differed from site
to site. At the top right comer, an area of 100 X 30 nm2 corresponding to part of a
nodular structure has only a few conductive points. Low conductive or insulating area
is also seen in the bottom of left comer including several nodular boundaries. The
relationship between electrical conductivity and topography can be seen more clearly
in Figure 4.19C. Most boundary and grain regions are conductive, but low conductive
or insulating areas are also seen both in nodular boundary and grain areas.

A section analysis is shown in Figure 4.20. Figure 4.20A and C show the
topographic and current images with a diagonal line (white line) selected for section
analysis. Figure 4.20B shows the height proﬁle and Figure 4.20D shows the current
proﬁle for the selected line. 4 nodular structures were probed in this measurement. As
seen in Figure 4.20, neither the nodule nor the boundary in the upper left possessed
any conductivity. On the other hand, a near-saturated current was recorded for the
second nodular structure (on the top left, next to the ﬁrst one) and boundary. The third
nodular structure in the center of the scanned area did not show high conductivity
except for several individual points. The forth nodular structure on the right bottom
comer showed good conductivity although the conductive regions were discrete and

the recorded currents were half of the saturated current.

161

3

.53
.8

 

Figure 4.20 Topographic and current images obtain by CTIS in contact mode for a
nanocrystalline highly boron-doped diamond thin ﬁlm. (A) height image with a
marked line selected for section analysis, (B) height proﬁle for the selected line, (C)
current image with a marked line selected for section analysis, (D) current proﬁle for

the selected line.

The results presented here indicated that the electrical properties of
nanocrystalline boron-doped diamond thin ﬁlm are different from those of
microcrystalline boron-doped diamond thin ﬁlm. For microcrystalline diamond,

almost all grain boundary regions are conductive with many of the crystals possessing

162

low conductivity. However, most regions of the nanocrystalline diamond ﬁlm exhibit
high electrical conductivity. There are some nodules and boundaries in the
nanocrystalline ﬁlm that possess low conductivity or are insulating, but the ﬁaction of
conductive area is greater than that for microcrystalline ﬁlms. More importantly, all
these results are consistent with the observation we got from our CP-AF M
measurements although CITS revealed the local work function of the sample instead
of the electrical conductivity directly. In fact, the local work function can be
associated with the local conductivity because lower work function means electrons
can escape or cross the surface of interface of a contact easier, which corresponds

higher electrical conductivity.

4.9 Conclusions

CP-AF M was used to investigate the spatial heterogeneity of the electrical
conductivity of microcrystalline and nanocrystalline BDD thin ﬁlms as a function of
the boron-doping level, surface morphology, polarity of the applied voltage and
surface terminations. All ﬁlms were characterized by a ﬁxed number of isolated
conductive regions separated from one other by less conducting zones. The fraction of
highly conductive area increased with increasing boron—doping level. The current at
the conducting sites increased proportionally with the bias voltage, and was the same
magnitude and independent of the bias voltage polarity indicating the conductive sites
possess metal-like electronic properties. Each ﬁlm was characterized by a ﬁxed
number of conductive sites as no new sites emerged with increasing bias voltage of
both polarities. Similar observations were made for microcrystalline BDD. CP-AF M
measurements also show that the fraction of highly conductive area depends on the

surface termination and diamond particle size. The results clearly show that the

163

fraction of highly conductive area for nanocrystalline BDD thin ﬁlm is higher than
that of microcrystalline BDD thin-ﬁlm. The morphology of CP-AFM images for
nanocrystalline diamond ﬁlms is also a little different from that for microcrystalline
diamond ﬁlms. For microcrystalline diamond ﬁlms, most highly conductive regions
are located at the grain boundary area but this association is much weaker for
nanocrystalline diamond ﬁlms. In other words, the highly conductive area is
somewhat independent on the topographic features for nanocrystalline diamond ﬁlms.

CITS measurements were also applied to our microcrystalline and
nanocrystalline boron—doped diamond thin ﬁlms. CITS measurements revealed both
microcrystalline and nanocrystalline diamond ﬁlms with both metallic and
semiconducting properties. This is an important observation. These results partially
conﬁrmed our CP-AF M measurements. For example, Schottky behavior was recorded
on some grain facets within +2 V (see Figure 4.16). These facets are poorly
conductive or insulating when probed by CP-AF M because of the Schottky effect.
CITS measurements revealed that these facets emitted electrons after the applied
voltage increased over 2. 0 V. Therefore, these facets may be non conductive when

probed using CP-AF M with a bias voltage less than 2.0 V.

164

4. 1 0 Reference

(1)

(2)

(3)

(4)

(5)

(6)

(7)

(8)
(9)

(10)
(11)

(12)

(13)

(14)

(15)

(16)

Pleskov, Y. V.; Sakharova, A. Y.; Krotova, M. D.; Bouilov, L. L.; Spitsyn, B.
P. J. Electroanal. Chem. 1987, 228, 19.

Janssen, G.; Enckevort, W. J. P. v.; Vollenberg, W.; Giling, L. J. Diamond
Relat. Mater. 1992, I, 789-800.

Swain, G. M.; Ramesham, R. Anal. Chem. 1993, 65, 345-351.

Rao, T. N.; Ivandini, T. A.; Terashima, C.; Sarada, B. V.; Fujishima, A. New
Diamond and Frontier Carbon Technology 2003, 13, 79-88.

Angus, J. C.; Martin, H. B.; Landau, U.; Evstefeeva, Y. E.; Miller, B.; Vinokur,
N. New Diamond and Frontier Carbon Technology 1999, 9, 175-187.

Alehashem, S.; Chambers, F.; Strojek, J. W.; Swain, G. M.; Ramesham, R.
Anal. Chem. 1995, 67, 2812-2821.

Swain, G. M. In Electroanalytical Chemistry; Bard, A. J., Rubinstein, I., Eds.;
Marcel Dekker, 2003; Vol. 22, pp 181-277.

Compton, R. G.; Foord, J. S.; Marken, F. Electroanalysis 2003, 15, 1349-1363.

F erreira, N. G.; Silva, L. L. G.; Corat, E. J .; Trava-Airoldi, V. J. Diamond and
Relat. Mater. 2002, I], 1523-1531.

Becker, D.; Jiittner, K. J. Appl. Electrochem. 2003, 33, 959-967.
Ferro, S.; De Battisti, A. Electrochimica Acta 2002, 4 7, 1641-1649.

Lagemaat, J. v. d.; Vanmaekelbergh, D.; Kelly, J. J. J. Electroanal. Chem.
1999, 475, 139-151.

Granger, M. C.; Witek, M.; Xu, J .; Wang, J .; Hupert, M.; Hanks, A.; Koppang,
M. D.; Butler, J. E.; Lucazeau, G.; Mermoux, M.; Strojek, J. W.; Swain, G. M.
Anal. Chem. 2000, 72, 3793-3804.

Fischer, A. E.; Show, Y.; Swain, G. M. Anal. Chem. 2004, 76, 2553-2560.

Kim, Y. D.; Choi, W.; Wakimoto, H.; Usami, S.; Tomokage, H.; Ando, T.
Diamond Relat. Mater. 2000, 9, 1096-1099.

Iseri, Y.; Honda, M.; Kim, Y.-D.; Ando, T.; Choi, W.; Tomokage, H. J. Phys. .'
Condens. Matter 2004, 16, 8171-3179.

165

(17)

(18)

(19)

(20)

(21)

(22)

(23)

(24)

(25)

(26)

(27)

(28)

(29)

(30)

(31)

(32)

Ushizawa, K.; Watanabe, K.; Ando, T.; Sakaguchi, 1.; Nishitani-Gamo, M.;
Sato, Y.; Kanda, H. Diamond Relat. Mater. 1998, 7, 1719-1722.

Ushizawa, K.; Nishitani-Gamo, M.; Watanabe, K.; Sakagushi, 1.; Sato, Y.;
Ando, T. .1. Raman Spectrosc. 1999, 30, 957-961.

Wang, J .; Swain, G. M.; Mermoux, M.; Lucazeau, G.; Zak, J .; Strojek, J. W.
New Diamond Front. Carbon T echnol. 1999, 9, 317-343.

Mermoux, M.; Marcus, B.; Swain, G. M.; Butler, J. E. J. Phys. Chem. B 2002,
106, 10816-10827.

Honda, K.; Noda, T.; Yoshimura, M.; Nakagawa, K.; Fujishima, A. J. Phys.
Chem. B 2004, 108, 16117-16127.

Robinson, V. S.; Show, Y.; Swain, G. M.; Reifenberger, R. G.; Fisher, T. S.
Diamond Relat. Mater. 2006, 15, 1601-1608.

Loiacono, M. J .; Granstrom, E. L.; Frisbie, C. D. J. Phys. Chem. 1998, 102,
1679-1688.

Kelley, T. W.; Granstrom, E. L.; Frisbie, C. D. Adv. Mater. 2005, 11 , 261-264.

Macpherson, J. V.; Mussy, J. P. G. d.; Delplancke, J. L. Electrochem. Solid St.
Lett. 2001, 4, E33-E36.

Holt, K. B.; Bard, A. J .; Show, Y.; Swain, G. M. J. Phys. Chem. B 2004, 108,
15117-15127.

Wilson, N. R.; Clewes, S. L.; Newton, M. E.; Unwin, P. R.; Macpherson, J. V.
J. Phys. Chem. B 2006, 110, 5639-5646.

Temyak, 0.; Cimmino, A. A.; Prawer, S.; Hoffman, A. Diamond Relat. Mater.
2005, 14, 272-278.

Houzé, F.; Meyer, R.; Schneegans, 0.; Boyer, L. Applied Physics Letters 1996,

69, 1975-1977.

Show, Y.; Witek, M. A.; Sonthalia, P.; Swain, G. M. Chem. Mater. 2003, 15,
879-888.

Hu, X. J.; Dai, Y. B.; Li, R. B.; Shen, H. 8.; He, X. C. Physica B 2003, 327,
39-42.

Yater, J. 13.; Shih, A. J. Appl. Phys. 2000,87, 8103-8112.

166

(33)

(34)

(35)

(36)

(37)

(38)

(39)

(40)

(41)

(42)

(43)

(44)

(45)

(46)

(47)

(48)

Zangwell, A. Physics at Surfaces Cambridge University Press: Cambridge,
1988.

Thomas, R. E.; Humphreys, T. P.; Pettenkofer, C.; Malta, D. P.; Posthill, J. B.;
Mantini, M. J .; Rudder, R. A.; Hudson, G. C.; Markunas, R. J. Mater. Res. Soc.
Symp. Proc. 1996, 416, 263-274.

Thoms, B. D.; Pehrsson, P. E.; Butler, J. E. J. Appl. Phys. 1994, 75, 1804-
1810.

Loci, H. J.; Pang, L. Y. S.; Molloy, A. B.; Jones, F .; Foord, J. S.; Jackman, R.
B. Diamond Relat. Mater. 1998, 7, 550-555.

Denisenko, A.; Aleksov, A.; Pribil, A.; Gluche, P.; Ebert, W.; Kohn, E.
Diamond Relat. Mater. 2000, 9, 1138-1142.

Maier, F .; Riedel, M.; Mantel, B.; Ristein, J .; Ley, L. Phys. Rev. Lett. 2000, 85,
3472-3475.

Chakrapani, V.; Eaton, S. C.; Anderson, A. B.; Tabib-Azar, M.; Angus, J. C.
Electrochem. Solid St. Lett. 2005, 8, E4-E8.

Chevallier, J.; Theys, B.; Lusson, A.; Grattepain, C .; Deneuville, A.; Gheeraert,
E. Phys. Rev. B 1998, 58, 7966-7969.

Rezek, B.; Nebel, C. E. Diamond Relat. Mat. 2005, 14, 466-469.

Tachiki, M.; Kaibara, Y.; Sumikawa, Y.; Shigeno, M.; Kanazawa, H.; Banno,
T.; Song, K. S.; Umezawa, H.; Kawarada, H. Surf Sci. 2005, 581, 207-212.

Brandes, G. R.; A. P. Mills, J. Phys. Rev. B 1998, 58, 4952-4962.

Huang, J. T.; Hu, C. S.; Hwang, J.; Chang, H.; Lee, L. J. Appl. Phys. Lett.
1995, 6 7, 2382—2384.

Wang, W. N.; Fox, N. A.; Steeds, J. W.; Lin, S. R.; Butler, J. E. J. Appl. Phys.
1996, 80, 6809-6812.

Sahli, S.; Aslama, D. M. Appl. Phys. Lett. 1997, 70, 2129-2131.

Bennett, J. A.; Show, Y.; Wang, S.; Swain, G. M. J. Electrochem. Soc. 2005,
152, E184-E192.

Wang, J .; Swain, G. M. J. Electrochem. Soc. 2003, 149, E24-E32.

167

(49) Wang, S.; Swain, G. M. J. Phys. Chem. C2007, 111, 3986-3995.

(50) Colley, A. L.; Williams, C. G.; Johansson, U. D. H.; Newton, M. E.; Unwin, P.
R.; Wilson, N. R.; Macpherson, J. V. Anal. Chem. 2006, 78, 2539-2548.

(51) Hayashi, K.; Yamanaka, S.; Watanabe, H.; Sekiguchi, T.; Okushi, H.;
Kajimura, K. J. Appl. Phys. 1997, 81 , 744-753.

(52) Huang, B. R.; Wu, C. H.; Sheu, R. F. Diamond Relat. Mater. 2000, 9, 73-81.

(53) Houzé, F .; Alvarez, J .; Kleider, J .-P.; Bergonzo, P.; Snidero, E.; Tromson, D.
Diamond Relat. Mater. 2006, 15, 618-621.

(54) Baker, S. M.; Rossman, G. R.; Baldeschwieler, J. D. J. Appl. Phys. 1993, 74,
4015-4019.

(55) Hamers, R. J .; Tromp, R. M.; Demuth, J. E. Phys. Rev. Lett. 1986, 56, 1972-
1977.

168

Chapter 5

Heterogeneous Electrochemical Activity of
Polycrystalline Boron-Doped Diamond Films

5.1 Introduction

Several authors have investigated the charge transfer kinetics for different
redox couples at BDD electrodes using a variety of electrochemical methods.HS
Much has been learned about how different factors control the electrochemical
properties of BDD. Depending on the redox molecule, factors such as the boron-
doping level, surface chemistry, nondiamond sp2 carbon impurity and ﬁlm
morphology can be inﬂuential.

One model of electron transfer at BDD involves mediation via impurity states
present within the band gap. These states could arise from the boron doping or from
impurities present, such as spz-bonded carbon,l so the rate of the charge transfer is
dominated by the available density of states near the formal potential of the redox
couple. More accurately, the heterogeneous electron-transfer rate constant will depend
on the availability of states of the correct energy relative to the potential of the redox
couple. Such states could arise from the overlapping of wave functions of neighboring
boron atoms and hence result in the formation of an impurity band. In fact, van de
Lagemaat et al.5 suggested that interfacial states (at or near the surface), distributed in
energy throughout the band gap, mediate electron transfer between the valence band
and the redox system at a boron-doped, single-crystal diamond electrode. The surface

states are supposed to be associated with defects or impurity sites located very close

169

to the surface, perhaps caused by hydrogen, boron, oxygen, or carbon inclusions.
Such an exchange may take place by multiple-electron-hopping (a model developed
by Mott7 to describe electrical conduction in strongly disordered systems with
localized states). The act of electron-transfer is usually considered as tunneling of the
electron between states of the electrode and those of the reactant. Therefore,
heterogeneous electron-transfer kinetics are highly dependent on the rate of tunneling
of electrons to and from the redox species in solution. Electron tunneling typically

follows an expression of the form:8

pm «I exm-BX) (51)

where p(i) is the probability of tunneling, x is the distance over which tunneling
occurs, and B is the height of the energy barrier and the nature of the medium between
the states. Thus, if deep boron dopant sites are involved in the electron-transfer, then
the kinetics may be slow and limited by the rate at which the carrier can migrate to the
surface. The electron transfer rate will also be sensitive to the state of the surface. For
example, the heterogeneous electron-transfer rate constants for BDD electrodes are
generally lower than those for noble metals. An explanation for this is that there is a
lower density of electronic states (DOS) of an appropriate energy. The available DOS
and carrier migration will depend on the dopant concentration. Thus, as the boron
content is increased, a greater density of states may be available for electron transfer,
although the charge carrier mobility will decrease as the charge carrier concentration
increases with the concentration of boron.

In the second model, polycrystalline boron-doped diamond surfaces exhibit

heterogeneity in electron—transfer rates over the surface due to nonuniform

170

distribution of dopant throughout the diamond ﬁlm. In other words, the surface may
consist of sites of fast (reversible) kinetics where the dopant level is high and sites of
very slow kinetics (irreversible) where there is less dopant. This model arises due to
the observation that boron incorporates into the diamond structure in different
concentrations depending on the crystal face of the diamond. For example, the boron
concentration in {111} growth sectors is ﬁve or more times greater than that found in
{100} sectors.9 Additionally, dopant may also accumulate in the grain boundaries,
crystal edges, and other defects, but these atoms are not expected to be electronically
active. Macroscopically, different crystal faces are exposed on the polycrystalline
surface, resulting in a heterogeneous electron-transfer rate that is an average value for
the whole electrode area. In other words, any heterogeneity in the electron transfer
kinetics can not be directly observed in electrochemical measurements. Becker et a1.
studied diamond ﬁlms with doping levels between 200 and 6000 ppm using cyclic
voltammetry and electrochemical impedance spectroscopy.3 They observed strong
deviations from reversible charge transfer for “outer-sphere” reactions (F e(CN)63 44‘
and Ru(NH3)63+’2+) with decreasing doping level. Limiting current densities that were
lower than predicted and the existence of an additional capacitive element in the
impedance spectra at higher frequencies led to the suggestion of a geometric surface-
blocking model, where only limited areas of the diamond ﬁlm are believed to be
electrically conducting. Hence this model suggests that the rate of electron transfer is
not limited by the migration of carriers to the electrode surface but more by the
availability of active surface sites for reaction. These may be sites where boron dopant
has accumulated and where conductive pathways are available through the ﬁlm. .
Both of these models suggest that the rate of heterogeneous electron transfer

will depend on boron concentration within the ﬁlm. However, the mechanism for

171

electron transfer might be different at different doping levels. For example, an
electron-hopping mechanism may dominate at low boron levels, whereas conduction
via the impurity band or surface states may be the main mechanism at higher doping
levels.

Some redox couples, such as Fe(CN)63‘/4'

, which are traditionally treated as
outer-sphere, behave anomalously at carbon-based electrodes and may change to an
inner-sphere electron-transfer mechanism at diamond electrodes, where some degree
of interaction occurs between the redox species and the electrode surface.“ '0 In these
cases, factors such as the termination of the surface and presence of oxygen
functionalities can inﬂuence the reaction kinetics.

As the redox systems, such as Ru(NH3)53+/2+, IrC162'/3' and methyl viologen,
are relatively insensitive to surface structure,ll the most important factor affecting the
heterogeneous electron-transfer rate constant is the electronic properties of the
electrode, which depend on the dopant concentration and distribution.

This chapter describes the application of scanning electrochemical microscopy
(SECM) to image diamond thin-ﬁlm electrode surfaces, doped with varying gas-phase
B2H6 concentrations from 0 to 10 ppm. These measurements produced maps of the
electrochemical activity across the surface as a function of the boron-doping level.
Combined with results from electrochemical characterization and CP-AF M, a

microscopic partial blocking model is applied to interpret the electrochemical

properties of boron-doped diamond thin ﬁlm electrode.

5.2 Details of the scanning electrochemical microscopy
measurements
5. 2.] Scanning electrochemical microscopy (SE CM)

172

Scanning electrochemical microscopy (SECM) is type of scanning probe
microscopy (SPM). SECM makes use of a microelectrode tip that approaches to
within a few micrometers of a substrate and probes the electrochemical activity across
the surface.”20 This technique was developed by the Bard group at the University of
Texas at Austin in the late 19805.”l4 SECM has been used in a variety of ways to
study heterogeneous and homogeneous electrochemical reactions. A comprehensive
review of the technique was published in 2001.2' The principle of SECM and all
concepts and terms used in this chapter can be found therein.

Typically, the tip potential is chosen to oxidize or reduce a mediator species in
a solution, with the measured current. being controlled by the ﬂux of the mediator.
This ﬂux depends on the properties of the substrate and the distance between the
microelectrode and the substrate. Keeping the distance constant, the microelectrode
current will be controlled by the substrate only. In other words, monitoring the current
of the microelectrode reveals information about the activity of the substrate for the
redox probe. There are several imaging modes with the two most common being the
tip generation and tip collection modes. The tip generation mode involves holding the
tip at a potential where an electrode reaction occurs. If the substrate is conductive and
held at a different potential where a product of the tip reaction will react (be collected),
the product of the substrate reaction will diffuse to the tip and cause an increase of the
tip current. This is the so called positive feedback tip-generation/substrate-collection
(TG/SC) mode. Figure 5.1 shows the general principle of this mode. On the other
hand, if the tip is near an electrically insulating substrate, the substrate will hurdle
diffusion of redox species to the tip and causes a decrease of the tip current (negative

feedback). Therefore, scanning over the surface of a substrate and recording the tip

173

current, one can obtain a current map revealing the electrochemical activity of the

 

 

 

 

 

 

 

 

 

 

substrate.
a
Glass Pt tip Glass
sheath sheath
l ..
O R
d Diffusion of 0 Feedback diffusion of 0
ll
0 R
e'

Conductive substrate

Figure 5.1 General principle of the positive feedback tip-generation/substrate-
collection mode for SECM. The tip is near a conductive substrate, feedback diffusion

leads to a larger tip current than when the tip is far away from the substrate.

5.2.2 Tip cleaning and activation

As mentioned before SECM involves the measurements of the current at a
microelectrode probe. The properties of the substrate (working electrode) are revealed
by the current changes resulting from the perturbation of the substrate. It’s reasonable
to expect that the microelectrode itself will strongly inﬂuence the current too. In fact,
the electrochemical behavior of the microelectrode mostly depends on its
conﬁguration. Therefore, the roughness of the microelectrode surface, the thickness of

the glass sheath and the isolation (sealing) of the microelectrode are very important

174

factors that inﬂuence the SECM measurements. One can ﬁnd details about this in the
book edited by Bard and Mirkin?‘ Different materials including Pt, Au, carbon ﬁber
etc. have been used as microelectrode tip with different conﬁgurations in SECM. All
microelectrodes used in the present work were purchased from a commercial source
(2 pm diam. Pt microelectrodes, CH Instruments, Texas). Figure 5.2 shows the

schematic diagram of the microelectrode probe tip.

 

 

 

 

 

Copper Wire
‘ ' Glass Sheath
. , ~2pm
a -—II It—
Torr Seal Epoxy = / \
i‘ J /
x
/
/
. /
/
_ r /
‘ /
Silver Paint 1 -
x/ I a, " / |‘——"i i’t Wire
Pthl'e r.\ -1” ’ u' .t-r >20“

 

Figure 5.2 Schematic of the Pt microelectrode probe tip used for imaging. All Pt

microelectrodes were commercial products (CH Instruments Inc., Austin, TX). The
nominal diameter of the microelectrode (Pt wire) was ca. 2 pm sealed in glass sheath.

The end of the assembly was treated by a special polishing procedure to make a ﬂat

disk shape.

175

All microelectrodes were cleaned before use. The method used to clean or
activate the microelectrodes is discussed in Section 2.6. Below is shown a
representative cyclic voltammetric i-E curve for a Pt microelectrode aﬁer cleaning.
Figure 5.3 shows the last cycle of a series of repetitive cyclic voltammograms used to
clean the Pt microelectrode. The cycles were performed in 0.1 mol/L HClO4 at 0.1
V/s. All characteristic features of Pt are present, particularly the H-

adsorption/desorption peaks near -0.1 V. This is consistent with the Pt being clean.

   
 
   

 

 

3.00E-010 -
2.005-010 —
2‘ 1005010 ‘ Pt Oxides Formation
E I Hydrogen Desorption
0,005+ooo —
D
0
4.005010 _ Pt Oxrdes Reduction
-2.00E-010 _ i Hydrogen Adsorption
' T ' I ' l ' I
-o.4 0.0 0.4 0.8 1.2

Potential (V vs. AglAgCl)

Figure 5.3 Typical cyclic voltammetric i-E curve for the Pt microelectrode after
potential cycle cleaning in 0.1 mol/L HC104 at 0.1 V/s. All features of Pt are present

indicative of a clean surface.

5.2.3 Calibration of the microelectrode
The nominal diameter of the microelectrode (Pt wire) is not known precisely

due to variability in the manufacturing. In practice, the diameter of a microelectrode

176

needs to be determined before it can be used for SECM imaging. In the present work,
three redox couples (Fe(CN)63'/4', Ru(NH3)¢53+/2+ and IrC152"3 ') were used for SECM
imaging; therefore, the microelectrode diameters were also determined using these
three redox couples.

22, 23

Aoki and Osteryoung showed that the current—time curve for diffusion

limited current at a disc microelectrode followed the following expression:

i_ 4nFADC'
7rro

f (r) (5.2)

where r0 is the radius of the disc microelectrode, A is the area of the disc, D and C.
are the diffusion coefﬁcient and bulk concentration of the redox species, respectively.

I is a dimensionless parameter deﬁned as

r = 413‘ (5.3)

r0

 

And at long times, f(‘t) z 1, equation (5.2) can be simpliﬁed to:“

in .0 = 4nFDCro (5.4)

in, is the steady-state current for a disk-shaped microelectrode. Equation (5.4)
indicates that the limiting current is related to the radius of the tip, the diffusion
coefﬁcient and concentration of the redox species.

Applying equation (5.4), the diameter of the microelectrode can be determined
by acquiring a cyclic voltammogram when the microelectrode is far away from the

substrate and the concentration of the redox species is known. Figure 5.4 shows

177

typical cyclic voltammetric i-E curves for the different redox species recorded at a Pt
microelectrode. Sigmoidal-shaped curves were observed for all three. The small size
of microelectrode makes diffusional mass transport extremely efﬁcient. At relatively
long experimental time scales, the dimensions of the diffusion layer exceed the radius
of the microelectrode, and the planar diffusion ﬁeld transforms into a spherical
diffusion ﬁeld. Consequently, the ﬂux of electroactive species to the microelectrode is
substantially higher than that for the pure planar diffusion case that is typical of a
macroelectrode. This efficient mass transport allows one to observe steady-state
responses when the applied potential is slowly scanned in cyclic voltammetry. The
sigmoidal-shaped responses observed in these experiments are analogous to the
polarograms obtained using a dropping mercury electrode, or a rotating disk electrode,
but they are observed under entirely quiescent conditions. In fact, mass transport rates
to a microelectrode are comparable with those of a conventional macroelectrode that
is being rotated at several thousand r.p.m. The high ﬂux at a microelectrode means
that one does not observe a reverse wave under steady-state conditions (Figure 5.4),
because the electrolysis product diffuses completely away from the electrode on the
time scale of the measurement.

The steady-state limiting current is directly proportional to the analyte
concentration (Equation 5.4), making it useful for analysis and determination of the
microelectrode diameter. Table 5.1 summarizes the calibration results with different
redox couples for a microelectrode with a nominal diameter of 2 pm. The results
show that the measured radius is pretty close to its nominal radius, although the

values depend on the redox species.

178

l

OBOE-+000 —4

-2.00E-01O —

Cuurent (A)

-4.00E-010 -«

 

 

l

2.00E-O1O —

1 .OOE-O1 O -

Content (A)

0.00E+OOO ~

 

I’

-o'.4 -o'.2 oio
Potential (V vs. AglAgCl)

 

 

 

moors-010 1

0.00E+OOO —

Current (A)

-1.00E-O1O _

 

oio oiz . oi4
Potential (V vs. AglAgCl)

 

 

 

r v

' I ' l

I r r
0.7 0.8 0.9 1 .0

Potential (V vs. AglAgCl)

'—
0.5 0.6

Figure 5.4 Cyclic voltammetric i—E curves for (a) 1 mmol/L Ru(NH3)63+/2+, (b) 1

mmol/L Fe(CN)63'/4', and (c) 1 mmol/L IrCl62'l3', all in 1 mol/L KCl, at a Pt

microelectrode. Scan rate = 0.1 V/s.

179

Table 5.1 Limiting current and calculated microelectrode radius determined by

different redox systems.

 

 

 

 

Redox i (pA) D‘ ( 10‘ cm/s) r0 (um)
Ru(NH3)53+’2+ 23.15 5.50 1.09
Fe(CN)63"4' 24.40 7.60 . 0.83
IrCI,m 20.42 6.80 0.78

 

 

 

 

 

 

* Data from reference 6.

5. 2.4 Tip approach curves

A plot of the experimental tip current vs. tip-substrate separation is called an
approach curve and it can provide information about the nature of the substrate.
Approach to an insulating surface typically results in negative feedback, where
diffusion of fresh mediator species to the tip is blocked as it moves within a few tip
radii of the substrate. In this case, the current rapidly drops to zero as the tip
approaches the surface. If the substrate is a conductor, its potential can be selected so
that it re-oxidizes or re-reduces the species being produced at the tip. In this case, as
the tip approaches the surface, positive feedback results and the current increases as
the tip approaches the surface. Therefore, when the microelectrode is several tip
diameters from the surface, a steady-state current i T, a, is observed. When the tip
approaches a surface, the tip current tr differs from i r. w, and it depends on both the
distance of separation, and the electrochemical activity of the surface. If the surface is
an electronic insulator, i z/ipm will be less than unity. If the surface is an electronic
conductor, then the ﬂux of the redox species to the tip will be enhanced, and iy/ipw
will be larger than unity. The actual current-distance relationship depends on the tip
shape and on the heterogeneous electron-transfer rate constant of the ‘feedback’

reaction that occurs at the substrate surface.

180

The SECM instrument can record the current and the displacement of the
microelectrode as it approaches a substrate. These data need to be treated and
converted to a current-distance (between the microelectrode and the substrate) curve
(approach curve).

Experimental approach curves can be compared to theory in order to
accurately determine the height of the tip above the substrate surface. In fact, before
performing SECM imaging of a substrate, one needs to generate the tip approach
curve mode to accurately determine the microelectrode position near the substrate
surface.

Figure 5.5 shows four approach curves recorded at the same position on a
nanocrystalline BDD deposited with 10 ppm B2H6 in the source gas mixture. The
curves were acquired with different engaging currents (approach stopped
automatically according to a prescribed set-point). Ru(NH3)(53+/2+ was used as the
redox system. The potential of the tip was -0.3 V vs. Ag/AgCl (mass-transfer limited
reduction of Ru(NH3)63‘), and the potential of the diamond electrode was 0 V vs.
AglAgCl (mass-transfer limited oxidation of Ru(NH3)62+). The approach curves were
acquired until the limiting current reached a value that was 2.3 times the value
measured with the tip far away from the surface. At this point, the distance between

the tip and the substrate was calculated to be ~1.0 urn using the following equation'4

(min...) = (7d4)(a/d) + 0.901 + 0.099 exp[-0.l6(a/d)] -0.29 exp[-0.39(d/a)] (5.5)

where i r is the tip current at any position, i 7,... is the limiting current measured when
the tip is positioned far from the surface, a is the tip radius, and d is the distance

between the tip and the substrate.

181

    

 

 

 

 

3.0 a l /Theoretical curve

2 5 _ 3rd Appeoach curve

2.0 -

2"“I Appeoach curve

.49
ll 1 5 _

1.0 -

4th Appeoach curve
05 l I T T l r I I r l l l l

L = d/a
Figure 5.5 SECM tip approach curves at the same position on a “as grown”
nanocrystalline BDD thin ﬁlm electrode in 1 mM Ru(NH3)63+’2+ + 1 M KCl. The
nominal diameter of the microelectrode was 2 pm. The microelectrode was held at -
0.3 V, and the diamond electrode potential was held at 0 V vs. Ag/AgCl. if is the tip
current, i 7,... is the limiting current, and L = d/a, where d is the distance between the tip
and substrate, a is the radius of the tip. The engaging current ratio for 1m to 4th

approach was 1.5, 1.8, 2.5 and 2.2, respectively.
The curves are normalized to the limiting current in order to more clearly

reveal the trends. The current at the tip is a function of the solution composition, the

distance between the tip and the electrode, as well as the morphology and electronic

182

properties of the diamond electrode.24 With the tip many diameters away from the
diamond electrode, a steady-state current was recorded at -0.3 V that increased
proportionally with the analyte concentration. Far away from the diamond electrode,
the diffusion layer near the tip is hemispherical, but this changes as the tip approaches
the surface. If the diamond ﬁlm is a conductor (boron-doped) and if the potential
applied is sufﬁcient to oxidize Ru(NH3)62+ back to Ru(NH3)63+ (0 V vs. Ag/AgCl in
our case), then the current at the tip will increase due to recycling of the redox system
(positive feedback). The opposite trend (negative feedback) will be observed if
diamond is an insulator (e.g., undoped diamond). The positive increase in current as
the tip approaches the diamond electrode is due to redox system recycling at the tip
and means that region of the diamond electrode is electrochemically active (i.e., rapid
rate of electron transfer) for this redox system. A theoretical approach curve for this
redox system with positive feedback is shown, for comparison.

Once the tip current increases to the ratio of iT/iT,no > 2.3, which means the

distance between the tip and the substrate is about 1 pm for a 2 pm tip, the scan was

initiated in the constant height mode.

5.3 SE CM images for boron-doped diamond thin ﬁlms

5. 3.1 Reproducibility of SE CM imaging

The SECM is capable of measuring very fast heterogeneous electron-transfer
reactions at a substrate surface by moving the tip close to the substrate. This SECM
method of measuring heterogeneous electron-transfer rates relies on a steady-state
response, and is therefore less sensitive to double-layer charging effects and

adsorption of electroactive species, than transient techniques. To make sure our

183

SECM measurements were reliable, the reproducibility of SECM imaging was
examined ﬁrst.

Figure 5.6 shows two consecutive SECM images acquired for a 10 ppm
microcrystalline boron-doped diamond thin ﬁlm probed with 5 mmol/L Ru(NH3)63+’2+ '
+ 1 mol/L KCl. The tip was scanned at a distance of about 1 pm above the surface of
each ﬁlm (determined by approach curve). The bright and brown regions represent the
regions showing high electrochemical activity on the surface of the diamond ﬁlm.
These images clearly reveal that most of areas of the diamond ﬁlm exhibit
electrochemical activity but there are several regions with extremely high
electroactivity (one possibility is that the microelectrode was closer to the substrate at
these regions because of the roughness of the diamond ﬁlm). These highlight areas
are repeated in consecutive images indicating good measurement reproducibility.
These images reveal that small drifts (several microns) existed in both x and y

directions when the microelectrode was scanned over the surface of the substrate.

11.9 IA”

  

5.97 M-
3.00 IA

0.28M

Figure 5.6 SECM images (25 x 25 m2) of a microcrystalline BDD thin ﬁlm
grown with 10 ppm B7116 in the source gas in 5 mmol/L Ru(NH3)63+’2+ + 1 mol/L KCl.
The Pt diameter was 2 pm. The tip potential was -0.3 V and the diamond electrode

potential was 0 V vs. AglAgCl. The two images were recorded consecutively.

184

5. 3.2 Effect of m orphologv of diamond thin-ﬁlm

It has been reported that the electrochemical activity of a boron-doped
microcrystalline diamond ﬁlm for Ru(NH3)63+’2+ is heterogeneous across the surface.”
26 In the present work, we sought to learn if there are factors that affect this
heterogeneity for boron-doped diamond ﬁlms. The ﬁrst factor considered was the
morphology of the diamond thin ﬁlm. To assess this, the electrochemical activity of
microcrystalline and nanocrystalline boron-doped diamond thin ﬁlms was probed in
TG/SC mode. A series of tip-approach curves were ﬁrst recorded at different regions
on the ﬁlm in the feedback mode. Ru(NH3)63+/2+ was used as the test redox system.
The approach curves were acquired until the limiting current reached a value that was
2.3 times the value measured with the tip far away from the surface. At this point, the
distance between the tip and the substrate was calculated to be ~1.0 pm, as discussed
above.

Figure 5.7A and B show 10 x 10 pm2 maps of Ru(NH3)t,3+/2+ electrochemical
activity across microcrystalline and nanocrystalline BDD electrodes, respectively. The
maps were obtained when a 2 pm diameter Pt tip was rastered across the surface at 50
pm s". Both ﬁlms were deposited with 10 ppm B2H6 in the source gas mixture. The
tip potential was -0.3 V and the substrate potential was 0 V vs. Ag/AgCl. The images
reveal that most areas of both electrodes exhibit electrochemical activity for this redox
system as the current is greater than the 1.1 nA limiting current measured when the tip
was many diameters away from the surface. However, there are some regions that
exhibit a higher level of activity than others with positive feedback currents as high as
11 nA. There are very few regions of negative feedback where the current is less than
1.1 nA. The nanocrystalline diamond ﬁlm possesses a greater fraction of active areas

than does the microcrystalline ﬁlm. On the basis of the CP-AF M images, one may

185

on!

 

8.94e-9A -

5383-8 A

 

3.02e-OA
it.» __.....
(~1.1 nA)
8.07e-1 1 A 10 till
10 pm
1.18e-8 A 0 pm
8.97e-9 A —

  

. ' 10m-
01m 10 pm

Figure 5.7 SECM images (10 x 10 umz) of a (A) microcrystalline and (B)

nanocrystalline BDD electrode, both grown with 10 ppm BzI-Is in the source gas, in 5

mM Ru(NH3)(,3+’2+ + l M KC1. In all cases, the Pt tip diameter was 2 pm. The tip

potential was —0.3 V and the diamond electrode potential was 0 V vs. Ag/AgCl.

(Reprinted with permission from .1. Phys. Chem. 111, 3986 (2007). Copyright (2007)

American Chemical Society.) (From Ref. 26).

186

conclude that the majority of the current ﬂow through the diamond ﬁlms occurs
primarily at a ﬁxed number of highly conducting sites. For the highly doped ﬁlm, the
spacing between the conductive sites is approximately a micrometer or less. This
same level of heterogeneity is not seen in the SECM image because the tip diameter is
greater than the lateral spacing of the conducting sites. In other words, this tip does
not afford the sub-micrometer spatial resolution that would be needed to probe the
electrochemical activity of the individual conducting sites. The nonuniform
electrochemical activity is, however, consistent with the differential electrical

conductivity.

5. 3.3 Effect of boron doping level

Figure 5.8A-C shows SECM images for nanocrystalline boron-doped diamond
ﬁlms grown with 10, 5 and 1 ppm diborane added to the source gas mixture. The
boron-doping level decreased with decreasing B2H6 concentration in the source gas.
In all of these images, the microelectrode potential was -0.3 V and the diamond
electrode potential was 0 V vs. Ag/AgCI. The mediator used for imaging was 1
mmoVL Ru(NH3)53+"2+ in 1 moi/L KCl. Importantly, as was observed in recent work
with hydrogen-terminated microcrystalline ﬁlms,”’ 26 the images reveal variable
electrochemical activity across the ﬁlm surface with regions of high activity isolated
by regions of lower activity. The ability of the ﬁlm to support the electrochemical
reaction (i.e., current ﬂow) increased proportionally with the boron-doping level as
the fraction of highly active regions was highest for the 10 ppm ﬁlm and decreased
progressively with decreased doping level, although, the difference between the 5 and
1 ppm ﬁlms is not so obvious. Some of the more active regions are clustered together,

whereas others are more isolated. Some of more active regions have lateral

187

dimensions of 5 to 10 pm, which may result from the low resolution of SECM. The
more active zones are separated from one another, in many cases, by tens of

micrometers or less.

 

 

(0.21 11A)

Figure 5.8 SECM images (100 x 100 ml) of nanocrystalline BDD thin ﬁlms
grown with (A) 10 ppm, (B) 5 ppm and (C) 1 ppm B2H6 in the source gas mixture
using 1 mmol/L Ru(NH3)5W2’r + 1 mol/L KC] as the test redox system. The Pt tip
diameter was 2 pm. The microelectrode potential was -0.3 V and the diamond

electrode potential was 0 V vs. AglAgCl. Images were recorded in the TG-SC mode.

188

Similar results were obtained when probed with 1 mmol/L Fe(CN)63"4' or
IrC16m' in 1 moi/L KC], as shown in Figure 5.9. The images reveal a pattern of
heterogeneous electrochemical activity similar to that probed by 1 mmol/L
Ru(NH3)(,3+/2+ in 1 mol/L KCl. Some of the electrochemically active regions are
clustered together, whereas others are more isolated. The electrochemically active
regions for the highly doped ﬁlms are 5 to 10 micrometers in lateral dimension. As
was the case probed by Ru(NH3)63+’2+, the fraction of highly conductive area

decreased with decreasing boron-doping level.

 

100 0
pm

Figure 5.9 SECM images (100 x 100 pmz) for nanocrystalline BDD thin ﬁlms
grown with 10 ppm (A and D), 5 ppm (B and E), and 1 ppm (C and F) BzHo in the
source gas mixture using either 1 mmol/L Fe(CN)63‘/4' ( A, B and C) or 0.5 mmol/L
IrC16m' ( D, E and F) as the test redox system. The supporting solution was 1 mol/L
KC]. The Pt diameter was 2 pm. E“, = -0.3 V and Esubmte = 0 V vs. AglAgCl when
tested with Fe(CN)63"4'. 15,, = 0.8 V and 15mm = 1.0 v vs. AglAgCl when tested

with er162" 3‘. Images were recorded in the TG-SC mode.

189

5. 3.4 E jfect of surface termination

The above results suggest that the boron-doped diamond electrode is
heterogeneous in terms its electrochemical activity and that redox reactions may be
limited by the availability of these reactive sites, rather than by slow heterogeneous
electron-transfer caused by limitation in either the carrier concentration or the
migration of carriers to the electrode surface. Electrochemical measurements revealed
that the surface chemistry of boron-doped diamond thin ﬁlm can also strongly
inﬂuence its electrochemical activity (see Section 4.4.1 and 4.4.3). Therefore, to
further evaluate this, the electrochemical activity of both ﬁlm types was studied using
different redox systems: Ru(NH3)63+/2+ and Fe(CN)(,3"4'.

Figure 5.10 shows SECM images for hydrogen- and oxygen-terminated
nanocrystalline BDD electrodes recorded in 1 mmol/L Ru(NH3)63’+’2+ (Figure 5.10A
and B) and Fe(CN)63"4‘ (Figure 5.10C and D). The supporting electrolyte solution was
1 mol/L KC]. The maps were generated using a 2 pm diameter Pt tip that was rastered
across the surface at 50 pm 5". The tip was scanned at a distance of about 1 pm above
the surface of each ﬁlm, using a tip potential of -0.3 V and a substrate potential of 0 V
vs. Ag/AgCl when probed with Ru(NH3),3*’2*. When probed by Fe(CN)63"4', the tip
potential was 0.5 V, and the substrate potential was 0 V vs. Ag/AgCl. (Both are
working in tip generation-substrate collection mode). The images clearly show that
the H-terminated BDD electrode possesses a greater number of active regions in
comparison with the O-terminated BDD electrode regardless of the redox probe.
Similar measurements were performed using H— and O-terminated microcrystalline
BDD electrodes, with slightly different results being obtained (Figure 5.11). It’s

noteworthy that the engaging current was set at iT/im = 1.8, which meant the distance

between the tip and the substrate was 1.25 pm according to equation 5.5. The purpose

190

of this change of the engaging current is to protect the tip because the surface
roughness of microcrystalline BDD ﬁlm is ca. 1 pm. More active regions were found
on H-terminated BDD electrode but the fraction of active area for the O-terminated
microcrystalline diamond electrode was greater than that for the O-terminated
nanocrystalline electrode. These results matched our electrochemical measurements
very well and clearly show that the redox processes are limited by the availability of

sites (or density of active sites) for electron transfer on the surface of boron-doped

 

Figure 5.10 SECM images (50 x 50 umz) for 1 mmol/L Ru(NT-l3)63+/2+ + 1 mol/L
KC] at nanocrystalline BDD thin ﬁlms: (a) hydrogen-terminated and (b) oxygen-
terminated (Em, = -0.3 V and Esubsme = 0 V vs. AglAgCl). SECM images (50 x 50
ml) for 1 mmol/L Fe(CN)6344- + 1 mol/L KC] at nanocrystalline BDD thin ﬁlms: (c)
hydrogen-terminated and (d) oxygen-terminated (Em, = 0.5 V and Emma,e = 0 V vs.
Ag/AgCl). In all cases, the Pt tip diameter was 2 pm. The z-axis scale shows the tip

current. Images were recorded in the TG-SC mode.

191

diamond. More active sites were mapped out on H-terminated than on O-terminated
boron—doped diamond when imaged by SECM, while much smaller peak separations
(faster kinetics) were recorded for the H-terminated boron-doped diamond electrode
when characterized by cyclic voltammetry. This suggests that faster kinetics of
electron-transfer on boron-doped diamond electrode may be related to a higher

number of active sites on the electrode surface.

 

Figure 5.11 SECM images (100 x 100 umz) for 1 mmol/L Ru(NH3)63+/2+ + 1 molfL
KC] at microcrystalline BDD thin ﬁlms: (a) hydrogen-terminated and (b) oxygen-
terminated (Em, = -0.3 V and E,,u(,mte = 0 V vs. AglAgCl). SECM images (100 x 100
ml) for 1 mmol/L Fe(CN)6344- + 1 mol/L KC] at micocrystalline BDD thin ﬁlms: (c)
hydrogen-terminated and (d) oxygen-terminated (Em, = 0.5 V and Emmate = 0 V vs.
Ag/AgCl). In all cases, the Pt tip diameter was 2 pm. The z-axis scale shows the tip

current. Images were recorded in the TG-SC mode.

192

5. 3. 5 Effect of redox system

The electrochemical activity of both ﬁlm types, highly boron-doped and H-
terminated, were further evaluated using different redox systems: Ru(NH3)63+/2+,
Fe(CN)63‘/4', and IrClsz'B'. The cyclic voltammetric response for these redox systems
at both types of conducting diamond has been discussed in detail elsewhere.6’ 27'”
Ru(NH3)63+/2+ and IrC16H3 ' are so-called outer-sphere redox systems on diamond with
heterogeneous electron-transfer rate constants that depend primarily on the density of
electronic states. 6’ 27'” Fe(CN)63‘/4' is a more complicated, so-called inner-sphere
redox system on carbon electrodes, with a heterogeneous electron-transfer rate
constant that strongly depends not only on the density of states but also the surface
cleanliness and surface oxygen coverage. 6’ '0’ ”'29 Figure 5.12 shows electrochemical
activity maps for a nanocrystalline BDD grown with 10 ppm B2H6 in the source gas
mixture. In all cases, the image features were reproducible with scanning and they
cover the same region of the electrode. Figure 5.]2A shows an image recorded in 1
mmol/L Ru(NH3)63+’2+ and 1 mol/L KC]. Much of the electrode exhibits high activity
for this redox analyte, as evidenced by the positive feedback current. The active
regions are distributed randomly over the surface with some of them being close
enough together to create a rather large zone (multiple micrometers in dimension).
Figure 5.12B shows a map of the same electrode area for 1 mmol/L Fe(CN)63'/4° and 1
mol/L KC]. The fraction of the surface exhibiting high electrochemical activity for
this redox system appears greater than that for Ru(NH3)63+fz+ or IrClom' (see below).
Figure 5.12C shows a map of the same electrode area for 1 mmol/L IrClgz'B' and 1
mol/L KC]. It can be seen that over half the surface exhibits high electrochemical

activity for this redox system. The fraction of active area in this image appears less

than that for the Fe(CN)63"4' probe but larger than that for the Ru(NH3)63+/2+ probe.

193

l.le€~8A “533:;

 

    
     

887139.41 —
594e$A -
(Mlle-SA
- (0.21 nA)
km _._s
assault “ ‘ ‘-' -
Gum 10pm
Mlle-816133.} ,-- ., 0pm
893e-9A —
5,98e-9A -
11‘,
10
man: pm
Own
592e-9A -
1100
' - ' tom
0 pm 10w);

Figure 5.12 SECM images (10 X 10 pmz) of a nanocrystalline BDD thin ﬁlm grown
with 10 ppm 82115 in the source gas in (A) 1 mmol/L RuCNH3)63+/2+, (B) 1 mmol/L
Fe(CN)6M” and (C) 1 mmol/L IrClsm'. The supporting electrolyte in all cases was 1
mol/L KC]. The Pt tip diameter was 2 pm. The z-axis scale in all three images is the
same. Images were recorded in the TG-SC mode. (Reprinted with permission from J.
Phys. Chem. 11], 3986 (2007). Copyright (2007) American Chemical Society.) (From
Ref. 26).

194

For Fe(CN)63'/4' and ITC162-/3-, as was the case for Ru(NH3)63+/2+, nearly all of
the imaged area exhibit positive feedback current indicating that all regions are
electrochemically active. Noteworthy, though, is the fact that the most of the active
regions are different for each redox system even though the same area of the electrode
was imaged. The exact reason for this is unclear at present but may be related to

different electric double layer structures locally on the surface.

The results clearly show that the electron-transfer mechanism of these
commonly used redox systems at diamond, and other carbon electrodes, is more
complex than previously thought. In the case of diamond, these three redox systems

undergo electron-transfer at different sites on the surface.

5. 3. 6 Comparison with other electrode materials

The SECM images acquired for the nanocrystalline diamond ﬁlm revealed
different electrochemically-active areas for the three redox systems. These results
suggested the mechanism of electron transfer on diamond electrode might be different
for Ru(NH3)63+’2+, Fe(CN)63'/4' and IrClsz'B, even though they are considered as outer-
sphere redox couples on other electrodes,30 the exception being Fe(CN)63"4'.

Our SECM images were acquired when a microelectrode scanned over the
surface of a substrate in the constant distance mode. In other words, the height of the
microelectrode is kept constant relative to the starting point (where approach curve
was taken), hence the distance between the microelectrode and the point or region
being measured varied according to the roughness of the substrate surface. From the

theory of SECM, we know that this distance change will strongly affect the tip

current,‘3 therefore, the ﬁlm morphology could inﬂuence the SECM image although

195

this is less likely for the smooth nanocrystalline diamond electrodes (~ 100 nm nns).
One possibility is that the different images obtained for the same area of the diamond
electrode with different redox couples may result from changes in the surface
roughness. Therefore, to better understand the variations in electrochemical activity,
studies were performed to learn what role the electrode roughness has on the image
features. To study the effect of surface roughness, we applied SECM measurements
on smoother, non-diamond electrodes using Ru(NH3)63+/2+, Fe(CN)53'/4' and IrClbz'B'
as probes. Glassy carbon has more uniform electronic properties and smoother surface
than does diamond after special polishing procedure. Therefore, the ﬁrst material
studied was glassy carbon (GC-30, Tokai Ltd. Japan) prepared by polishing with
successively smaller grades of alumina powder slurried in ultrapure water (1.0, 0.3,
and 0.05 pm) on separate felt pads. The electrode was rinsed thoroughly with and
ultrasonically cleaned for 5 min in ultrapure water after each polishing step.31
Representative electrochemical activity maps are shown in Figure 5.13. A 10
x 10 pm2 image recorded in 1 mmol/L F e(CN)(53 44‘and 1 mol/L KCl is shown in
Figure 5.]3A. Much of the electrode exhibits high activity for this redox analyte, as
evidenced by positive feedback current. The active regions are distributed randomly
over the surface with few of them being close enough together to create an active
zone (less than 1 micrometer in dimension). There are some noisy signals in the
horizontal (scanning) direction that may result from the movement of the
microelectrode or the polishing residue. Figure 5.13B shows a 100 x 100 pm2 map
recorded in 1 mmol/L Ru(NH3)63+’2+ and 1 mol/L KC]. Only several points exhibit
high activity. Most of the area is green color, which means that the current is close to
the current limit of the measurement. It can be seen from the current scale bar that the

lowest current limit is 0.14 nA, which is about the limiting current for the

196

microelectrode positioned far away from the substrate. It’s hard to ﬁgure out if the
regions showing green color in this image are electrochemically active or non-active
because of the low color resolution of SECM image. Figure 5.13C shows a map of 10
x 10 pm2 recorded in 0.5 mmol/L IrClgz'B‘ and 1 mol/L KC]. It can be seen that much
of the surface exhibits high electrochemical activity for this redox system. The
fraction of highly active area in this image is greatest for the three probes. The results
shown in Figure 5.13 indicate that glassy carbon exhibits similar heterogeneous
electrochemical activity as diamond, which is reﬂected in the SECM images. This is
different from the SECM image of GC reported by Hoéevar et a1.32 It’s noteworthy
that they performed SECM imaging on GC with a 25 pm diam. Pt microelectrode at a
distance of 10 pm.

As expected, very uniform images were obtained, as shown in Figure 5.]3D, E
and F when we reduced all current scales with a factor of 1/10. Figure 5.13D shows

that current larger than 1.19 nA (>> 0.2 nA, the limiting current) was recorded over

344-

the scanned area when probed using Fe(CN)6 except some regions near the center

of the image that show smaller current (ca. 0.94 nA). Figure 5.]3E shows that current
larger than 0.82 nA (>> 0.2 nA, the limiting current) was recorded over half of the

3+/2+

scanned area when probed using Ru(NH3)6 . Half of the scanned area (right part in

Figure 5.13B) shows smaller current (ca. 0.6 nA). More uniform image was recorded

when probed using IrClém'

and only a few points show lower current (Figure 5.13F).
These images clearly show that the whole scanned area exhibits high activity for all

redox couples, as evidenced by positive feedback current.

197

 

10pm

 

Figure 5.13 SECM images of glassy carbon in (A) 1 mmol/L Ru(NI-13)63+’2+, (B) 1
mmol/L Fe(CN)63"/4' and (C) 0.5 mrnol/L IrClsz'B'. The supporting electrolyte in all
cases was 1 moi/L KC]. The Pt tip diameter was 2 pm. Images were recorded in the
TG-SC mode. The left column shows the original images recorded by SECM. These
images were replotted in right column with a reduced current scale by a factor of 1/10.

Note that the scan size for (B) and (E) is 100 x 100 11.1112 and the other are 10 x 10 mz.

198

To better compare the glassy carbon and diamond electrode results, the
diamond images were replotted with an enhanced current axis. The surprising results
are shown in Figure 5.14. Figure 5.14A shows a 10 x 10 pm2 image of highly doped
nanocrystalline diamond ﬁlm recorded in 1 mmol/L Fe(CN)53'/4'and 1 mol/L KC].
Much of the electrode exhibits high activity for this redox analyte, as evidenced by
positive feedback current. The active regions are distributed randomly over the
surface with some of them being close enough together to create a rather large zone
(multiple micrometers in dimension). When the current scale was reduced by a factor
of 1/10, a very similar image was obtained, as shown in Figure 5.14D. There were no
new active regions evident except the weak, streamlined wave-lines resulting from the
scanning of the microelectrode. This is very different from the result for glassy carbon.
These results revealed that there were certain regions on the doped diamond electrode
that show very high electrochemical activity. These regions were so active that the
current reached a very high level (11.9 nA in this case) when the microelectrode was
scanned over the surface. The activity of other areas was very different with less tip
current being observed. Because of the low color resolution, when a high current scale
was used, the SECM image showed lots of bright brown islands (high active regions)
surrounded by green ocean (low active regions). When the current scale reduced an
order of magnitude, the glassy carbon electrode showed uniform activity. On the other
hand, diamond electrode showed non-uniform activity (Figure 5.14D). These results
demonstrate that glassy carbon electrode has a greater fraction of electrochemically
active area than does the diamond electrode, even though the most active regions

possessed the same level of current ﬂow for both electrodes.

199

 

Figure 5.14 SECM images of a nanocrystalline BDD thin ﬁlm grown with 10 ppm
BzH6 in the source gas mixture in (A) 1 rmnol/L Fe(CN)63'/4', (B) 1 mmol/L
Ru(NH3)63+’2+ and (C) 1 mmol/L IIC16243-. The supporting electrolyte in all cases was
1 mol/L KC]. The Pt tip diameter was 2 pm. Images were recorded in the TG-SC
mode. The left column shows the original images recorded by SECM. These images

were replotted in the right column with a reduced current scale by a factor of 1/10.

200

This observation was further conﬁrmed by SECM measurements using
manner-““2“ and Ierz"3‘ as probes (Figure 5.1413 and F). The SECM images with
10 times lower current scales still show similar images to the original ones, which
were different from the same measurements on glassy carbon electrode. The images
are consistent with lower electrochemical activity at diamond electrode than at glassy
carbon for these redox systems.

These results also suggest that the surface roughness of diamond does not
strongly contribute to the image features. The glassy carbon electrode used here was
polished by a standard procedure,3 ' so it’s reasonable to assume the surface roughness
was at 50 nm level, which is comparable to the roughness of nanocrystalline diamond
thin ﬁlm. If the non-uniformity of the SECM image essentially resulted from the
roughness of the electrode surface, glassy carbon should show similar properties to
diamond and this was not the case.

To further validate our observations, we recorded electrochemical activity
maps of a thin Au-ﬁlm electrode. The Au was sputter-coated onto a clean Si substrate.
Ohmic contact was made directly though the Au ﬁlm by contacting with copper foil.
The resulting images are shown in Figure 5.15.

A 100 x 100 um2 image recorded in 1 mmol/L Fe(CN)63'/4'and 1 mol/L KC] is
shown in Figure 5.15A. Much of the electrode exhibits high activity for this redox
analyte, as evidenced by positive feedback current. The active regions are distributed
randomly over the surface with some of them being close enough together to create
active zones. Figure 5.15B shows a 10 x 10 um2 map recorded in 1 mmol/L
Ru(NH3)63+/2+ and 1 mol/L KC]. Only several points exhibit high activity that can be
shown. Most of the area shows a green color indicating the current close to the steady-

state current of the measurement. It can be seen from the current scale bar that the

201

least current limit is 0.17 nA that is close to the limiting current of the microelectrode
far away from the substrate. It’s hard to distinguish the area showing green color as
electrochemically active or non-active because of the low color resolution. In general,

heterogeneity is seen for Au as was the case for diamond and glassy carbon.

      

imam-pg "V' ._ . .(¢1.19nA% [C

Figure 5.15 SECM images of a gold ﬁlm deposited on a conducting Si substrate in
(A) 1 mmol/L Fe(CN)?” and (B) 1 mmol/L Ru(NH3)53+’2+. The supporting
electrolyte was 1 mol/L KC]. The Pt tip diameter was 2 pm. Images were recorded in
the TG-SC mode. The leﬁ column shows the original images recorded by SECM.
These images were replotted in the right column with a reducd current scale by a

factor of 1/10.

202

As expected, very uniform images were obtained, as shown in Figure 5.15C
and D when we reduced all current scales with a factor of 1/10. Figure 5.15C shows
that current larger than 1.19 nA (>> 0.2 nA, the limiting current) was recorded over
the scanned area when probed using Fe(CN)63'/4' except some regions near the right
comer of the bottom that show smaller current (ca. 0.4 nA). Figure 5.15D shows that
current near 1.1] nA (>> 0.2 nA, the limiting current) was recorded over half of the

3+/2+. Half of the scanned area show lower

scanned area when probed using Ru(NH3)6
current shows smaller current (ca. 0.8 nA). Different from the SECM image for GC,
this image shows that the more active regions intermixed the less active regions each

other.

5.4 Mechanism of electron-transfer on polycrystalline boron-

doped diamond

Both the CP-AF M and SECM images revealed that there are randomly
distributed electrically- and electrochemically-active sites on the ﬁlms that are
surrounded by less active regions. We suppose that electron-transfer reactions at
diamond can be considered to occur at a partially blocked electrode. In fact, a
macroscopic, partially-blocking model was proposed several decades ago.”35
However, partial coverage of the electrode surface may result in non-linear diffusion
effects. This effect may result in the extension of analysis of macroscopic blocking to
microscopic inhomogeneities unsatisfactory. To solve this problem, Amatore et al.36
proposed a microscopic model. In this section, we applied this model to boron-doped

diamond electrode. We sought to theoretically interpret the electrochemical properties

of boron-doped diamond thin ﬁlm electrodes.

203

In this model, it is assumed that a microscopic conﬁguration for the boron-
doped diamond electrode consists of active sites in a disc-like shape of the same size
evenly distributed over the surface. This assumption simpliﬁes the theoretical
calculations even though CP—AFM and SECM images showed active sites with
different shapes and sizes randomly distributed over boron-doped diamond surface.
This model and the diffusion volume unit are shown in Figure 5.16. In this model, the
radius of an active site is R, and the distance between the centers of two adjacent
active sites is 2R0, which are assumed to be small compared to the thickness of the
diffusion layer. It is assumed that beyond a distance L z 2R0 from the electrode
surface, diffusion can be regarded as linear and semi-inﬁnite. Within a distance less
than L, diffusion is non-linear and a stationary situation exists because L is much

smaller than the thickness of the diffusion layer.

2R3 2R0
H

oUe o
0000

Figure 5.16 Model of a boron—doped diamond electrode surface and a diffusion

 

 

 

 

   

 

 

 

 

 

volume unit for an active site. Black dots represent active sites.

Consider the single, reversible electron-transfer reaction
O + e <—) R (5.6)
and assign Co as the concentration of O at any distance from the electrode surface

with Co. being the bulk concentration of 0. There is no R present at the beginning of

204

the reaction. E0 is the standard reduction potential, or is the transfer coefﬁcient, k0 is
the standard heterogeneous electron-transfer rate constant, tc is the time of the
electrolysis (tc = RT/Fu for cyclic voltammetry with a scan rate of u), and D is the
diffusion coefﬁcient. It is assumed D = D0 = DR. 0 is the fraction of the electrode
surface occupied by non-active sites and S is the total area of the electrode. To

simplify the calculations, the following dimensionless parameters were used:

c = Co/Co' (5.7)
‘t = t/tc (5.8)

x = X(Dt.,)'“2 (5.9)
\p = i/[Fsco’(D/t,)"’2] (5.10)
A = Ide,/D)"2 (5.11)
g = -(F/RT)(E-E°) (5.12)

Under these conditions, for a normal planar electrode (i.e., linear and semi-

inﬁnite diffusion apply), integration of the linear and semi-inﬁnite diffusion equation

(Fick’s Law) gives

c(0, t) = 1— a“ Igy/(tXt — t)dt =1—I‘l’ (5.13,

where I '1’ is the convolution integral.8 On the other hand, the Butler-Volmer equation

can be expressed as

_a_c
W 6x

 

x=0 = A exp(acf)[ca — CR e"PG-5 )1 (5.14)

and because there is no R present initially,

Co+CR=1 (5.15)

Substituting (5.15) to (5.14) gives

205

w = A exr3(aé‘) {coll + elm-6)] - exp(-é‘)} (5.16)
and thus to
w = A exp(a~f)[1 - I ‘1’ - 1‘? exp(—5)] (5.17)

Now returns to the partially blocking model, at x = p = 2R0, linear and semi-

inﬁnite diffusion still apply and similarly, we have36

 

I do
_ _ —1/ 2 __ _
9-1” L&harwr (m)
The conservation of the concentration ﬂux in the stationary diffusion region (x s p)
leadsto
ac dc
W=—|#=_|0(1—6) 519
ax dx (')
60

where ax '0 is the average value of the dimensionless current density at the active

sites. At the active sites, the Butler-Volmer relationship still applies; therefore,

if 0 = Aexp(a5)[€—E;exp(—§)] (5.20)

again,
25+; =1 (5.21)

combine (5.20), (5.21) with (5.19), one gets

91 =A(1-6>exp(ar>{2311+exp<—5)l-exp(-r)l (5.22)

206

The problem then becomes ﬁguring out what a is. Amatore et a]. solved this

problem in their paper.” They gave the relationship between a and c,l for disc-type

active sites as follows,

_ 1 °°
CO : Cy —‘UW[1 +EZBn(77)] (5.23)
n=l
where Bum) is the Bessel function with n = (1-0)m. In our case, p = 2R0 and assume
0 is close to unity, then the sum of the Bessel function can be approximated by 0.3(1-

(9)4”, hence,

Co = Cp " 2R0w[0.3(l — 49)-“2] (5.24)

Combining (5.18), (5.22) and (5.24) leads to the ﬁnal expression of the current-

potential curves:

W = A(1- 6’) eXP(acf){ [1 - 1‘? -0-6R0W(1-9)"’2]-U‘P + 0-6Row(1-9)"’216XP(-§)}
(5.25)

where I‘I’ is the convolution integral as deﬁned by (5.13).

The signiﬁcance of equation (5.25) can be identiﬁed by comparing the
polarization curves having a reaction scheme as follows

1

1 é“ o (K=-lf-‘—)

k2 k2
0+e€ R (E°,a,k°))

k2 k (5.26)
R e J (K=-—‘-)

k1 k2 ,

 

207

Saveant et al.37'39 studied this reaction scheme and gave a general expression for the

“pure kinetic” polarization curves
‘1’ = A[K/(1+ K)] exp(a§) {[1 - I‘i’ — (‘P / Kh'/2)]-[I‘P + (‘l’ / K/i'”)]exp(—.f)}

(5.27)

with
x1 = (k1 + k2 )tc (5.28)

the form of equation (5.25) is the same as equation (5.27).

Based on the similarity of equation (5.25) and equation (5.27), it’s possible to
characterize the electrochemical properties of boron-doped diamond electrodes as
follows:

1. The “pure kinetic” conditions for reaction scheme (5.26) means the
chemical reactions are so fast that the two equilibria are perturbed only in a thin
reaction layer adjacent to the electrode surface.”39 On BDD electrode, there is a
stationary non-linear diffusion layer that corresponds to this thin layer. In fact, similar
33-35

results were obtained for partially blocked electrode system in macroscopic view.

2. When 0.6Rot|1(1-0)'”2 << 1, equation (5.25) becomes

w = A(1- €)eXP(a§)[l - I‘I’ - [‘1’ elm-6)] (5.29)

Comparing equation (5.29) with equation (5.17), we conclude that the boron-
doped diamond electrode simply results in an apparent decrease of the standard rate
constant of electron-transfer. In other words, the voltammograms for boron-doped
diamond electrodes are the same as those obtained on a perfect electrode of same
surface area S (no non-active sites, e.g. a noble metal electrode) with a decreased

standard rate constant kapp0 = k°(1-0). Without question, the decrease is dependent on

0, i.e., the larger 0, the larger peak separation, which is exactly the cases for boron-

208

doped diamond with decreasing doping level or the surface termination changed from
hydrogen to oxygen. 0 can be calculated from an SECM image using a graphical
software and km,0 for a certain BDD ﬁlm can be determined from SECM
measurements. Assuming GC as a uniformly electroactive electrode, i.e. using kO
obtained from GC as the rate constant of an ideal carbon electrode without blocking,
one can validate the partial blocking model for BDD thin-ﬁlm electrodes with
different doping levels. This will be a future work.

3. As 0 increases, which corresponds to the cases for boron-doped diamond
electrode at low doping level (e.g., < 1 ppm or undoped) and oxidized nanocrystalline

BDD, R0 will increase greatly. All these will make 0.6RollJ(l-0)'”2 can not be

neglected and kg”,0 = k°(l-0) tends to 0. Under these conditions, the shape of
voltammogram changes as the peak separation increases and peak current decreases.
These have been conﬁrmed by our electrochemical measurements, CP-AF M and
SECM images for low boron-doped or oxidized nanocrystalline diamond electrodes.
4. SECM images clearly show the partially blocked electrochemically-active
surface of BDD thin-ﬁlm electrode and the coverage of blocked area decreases as the
doping level increases. Besides SECM image, the tip-approach curve can also reveal a
certain region electrochemically-active or not. Approach curves recorded at
nanocrystalline BDD thin-ﬁlm grown with 1 ppm diborane in the source gas
conﬁrmed this, as shown in Figure 5.17. Figure 5.17A and B show the tip-approach

curves recorded in 1 mmol/L Ru(NH3)(33+’2+

and 1 mol/L KC] at different positions of
the same BDD ﬁlm. The curves are normalized to the limiting current in order to
more clearly reveal the trends. With the tip many diameters away from the diamond

electrode, a steady-state current was recorded at -0.3 V that increased proportionally

with the analyte concentration. Far away from the diamond electrode, the diffusion

209

layer near the tip is hemispherical, but this changes as the tip approaches the surface.
The positive increase in current as the tip approaches the diamond electrode is due to
redox system recycling at the tip and means that region is electrochemically-active
(i.e., rapid rate of electron transfer, Figure 5.17A) for this redox system. On the other
hand, at another position (Figure 5.]7B), the tip current decreases as the tip
approaches the diamond electrode due to the blocking effect of the substrate and the
absence of the redox system recycling at the tip, which means that region of the
diamond electrode is not electrochemically-active for this redox system. Theoretical
approach curves for this redox system with positive and negative feedback are shown
in both Figure 5.17A and B, for comparison.

Similar results were recorded at a nanocrystalline BDD ﬁlm grown with 5
ppm diborane in the source gas but we did not recorded negative feedback approach
curve at the diamond ﬁlm grown with 10 ppm diborane in the source gas. A possible
reason is that the coverage of the blocked area is too low at this doping level. Another
possible reason might be the microelectrode probe tip used in this study (2 pm in

diam.) is too big that can not distinguish the non-active area at this ﬁlm.

210

 

:hsl ‘\
,l
3.0 -
. \
2-5 d Theoretical curve

Experimental curve

 

 

 

 

 

 

05 I l I l l I l I T I
02468101214161820
L=d/a
1.04 3
0.8- .
Experimental curve
8 /
“R I
I. 0.6-1 ..°
II I"
‘ ;/ Theoretical curve
6
0.4-
l
o.2.,.,...,.,.,..e,..
o 2 4 e 8 1o 12 14 16 18

L = d/a

Figure 5.17 Tip-approach curves at (A) a conductive site and (B) a non-conductive

site of a nanocrystalline BDD thin ﬁlm electrode in 1 mM Ru(NH3)63+/2++ 1 M KC1.

The nominal diameter of the microelectrode was 2 pm. The microelectrode was held

at -0.3 V, and the diamond electrode potential was held at 0 V vs. Ag/AgCl. ir is the

tip current, in, is the limiting current, and L = d/a, where d is the distance between the

tip and substrate, a is the radius of the tip. Theoretical curves are calculated according

to the models given in Ref. 21.

211

5.5 Conclusions

SECM was used to investigate the spatial heterogeneity of the electrochemical
activity across the microcrystalline and nanocrystalline BDD thin ﬁlms as a function
of the boron-doping level, surface morphology and chemistry, and the redox system.
All ﬁlms were characterized by a ﬁxed number of isolated electrochemically-active
regions separated from one other by less active zones. The fraction of highly active
area increased with increasing boron-doping level. Similar observations were made
for microcrystalline and nanocrystalline BDD electrodes. SECM measurements also
showed that the fraction of highly active area depends on the surface termination and
diamond crystal size, which are consistent with the results of electrochemical
measurements (nanocrystalline BDD electrode has lightly greater peak current that
microcrystalline BDD electrode with same geometric area) and CP-AF M images. The
results clearly show that the fraction of highly active area for the nanocrystalline BDD
thin ﬁlm is larger than that for microcrystalline BDD thin ﬁlm. The pattern of SECM
images for nanocrystalline diamond ﬁlms is also a little different from that for
microcrystalline diamond ﬁlms. For nanocrystalline diamond ﬁlms, most highly
active regions are more evenly distributed in smaller cluster size compared with
microcrystalline diamond electrode. The morphology of SECM images for
nanocrystalline diamond ﬁhns also show that the redox systems used as mediators
have strong inﬂuence on the SECM images.

A microscopic partial blocking model was applied to the boronedoped
diamond electrode. The theoretical results deduced from this model matched well
with our results obtained from electrochemical measurements and SECM images for
both microcrystalline and nanocrystalline boron-doped diamond thin ﬁlm electrodes.

Especially, both the effect of surface termination and the inﬂuence of doping level on

212

the electrochemical properties of boron-doped diamond electrode can be explained by

this model.

213

5.6

(1)

(2)

(3)
(4)

(5)

(6)
(7)

(8)

(9)

(10)
(11)
(12)
(13)
(14)

(15)

(16)
(17)

Reference

Granger, M. C.; Witek, M.; Xu, J.; Wang, J.; Hupert, M.; Hanks, A.; Koppang,
M. D.; Butler, J. E.; Lucazeau, G.; Mermoux, M.; Strojek, J. W.; Swain, G. M.
Anal. Chem. 2000, 72, 3793-3804.

Ferreira, N. G.; Silva, L. L. G.; Corat, E. J .; Trava-Airoldi, V. J. Diamond and
Relat. Mater. 2002, II, 1523-1531.

Becker, D.; Jiittner, K. J. Appl. Electrochem. 2003, 33, 959-967.
F erro, S.; De Battisti, A. Electrochimica Acta 2002, 4 7, 1641-1649.

Lagemaat, J. v. d.; Vanmaekelbergh, D.; Kelly, J. J. J. Electroanal. Chem.
1999, 475, 139-151.

Fischer, A. E.; Show, Y.; Swain, G. M. Anal. Chem. 2004, 76, 2553-2560.
Mott, N. F. Phil. Mag. 1969, 19, 835.

Bard, A. J .; Faulkner, L. R. Electrochemical Methods F undationals and
Applications, Second ed.; John Wiley & Sons Inc., 2001.

Kolber, T.; Piplits, K.; Haubner, R.; Hutter, H. Fresenius'J. Anal. Chem. 1999,
365, 636.

Granger, M. C.; Swain, G. M. J. Electrochem. Soc. 1999, 146, 4551-4558.
Chen, P.; McCreery, R. L. Anal. Chem. 1996, 68, 3958-3965.

Bard, A. J.; Fan, F .-R. F.; Kwak, J.; Lev, 0. Anal. Chem. 1989, 61 , 132.
Kwak, J.; Bard, A. J. Anal. Chem. 1989, 61, 1221-1227.

Kwak, J.; Bard, A. J. Anal. Chem. 1989, 61 , 1794-1799.

Bard, A. J .; Mirkin, M. V.; Unwin, P. R.; Wipf, D. O. J. Phys. Chem. 1992, 96,
1861-1868.

Mirkin, M. V.; Bard, A. J. J. Electroanal. Chem. 1992, 323, 29.

Bard, A. J.; Mirkin, M. V.; Unwin, P. R.; Widf, D. O. J. Phys. Chem. 1992, 96,
1861.

214

(18)
(19)
(20)

(21)

(22)
(23)

(24)

(25)

(26)

(27)

(28)

(29)

(30)

(31)

(32)

(33)
(34)

(35)

Unwin, P. R.; Bard, A. J. J. Phys. Chem. 1991, 95, 7814.
Martin, R. D.; Unwin, P. R. J. Electroanal. Chem. 1997, 439, 123.
Shao, Y.; Mirkin, M. V. J. Phys. Chem. B 1998, 102, 9915.

Bard, A. J .; Mirkin, M. V., Eds. Scanning Electrochemical Microscopy;
Marcel Dekker, Inc.: New York, 2001.

Aoki, K.; Osteryoung, J. J. Electroanal. Chem. 1981, 122, 19.
Aoki, K.; Osteryoung, J. J. Electroanal. Chem. 1984, I60, 335.

Bard, A. J .; Denuault, G.; Lee, C.; Mandler, D.; Wipf, D. 0. Acc. Chem. Res.
1990, 23, 357-363.

Holt, K. B.; Bard, A. J .; Show, Y.; Swain, G. M. J. Phys. Chem. B 2004, 108,
15117-15127.

Wang, 8.; Swain, G. M. J. Phys. Chem. C 2007, 111, 3986-3995.

Alehashem, 8.; Chambers, F.; Strojek, J. W.; Swain, G. M.; Ramesham, R.
Anal. Chem. 1995, 67, 2812-2821.

Swain, G. M. In Electroanalytical Chemistry; Bard, A. J ., Rubinstein, I., Eds;
Marcel Dekker, 2004; Vol. 22, pp 181-277.

Show, Y.; Witek, M. A.; Sonthalia, P.; Swain, G. M. Chem. Mater. 2003, 15,
879-888.

McCreery, R. L.; Cline, K K.; McDermott, C. A.; McDermott, M. T. Colloids
and Surfaces A: Physicochemical and Engineering Aspects 1994, 93, 211-219.

Xu, J.; Chen, Q.; Swain, G. M. Anal. Chem. 1998, 70, 3146.

Hocevar, S. B.; Daniele, S.; Bragato, C.; Ogorevc, B. Electrochimica Acta
2007, 53, 555-560.

Gueshi, T.; Tokuda, K.; Matsuda, H. J. Electroanal. Chem. 1978, 89, 247-260.
Gueshi, T.; Tokuda, K.; Matsuda, H. J. Electroanal. Chem. 1979, 101, 29-38.

Tokuda, K.; Gueshi, T.; Matsuda, H. J. Electroanal. Chem. 1979, 102, 41-48.

215

(36)

(37)
(38)

(39)

Amatore, C.; Saveant, J. M.; Tessier, D. J. Electroanal. Chem. 1983, I 4 7, 39-
5].

Saveant, J. M.; Vianello, E. Electrochimica Acta 1963, 8, 905-923.
Saveant, J. M.; Vianello, E. Electrochimica Acta 1967, 12, 629-646.

Nadjo, L.; Saveant, J. M. J. Electroanal. Chem. 1973, 48, 113.

216

Chapter 6

The Physical and Electrochemical Properties of Two
Types of Boron-Doped Nanocrystalline Diamond Thin-
Film Electrodes

6. I Introduction

It has been reported that diamond ﬁlms from different sources exhibit quite
reproducible electrochemical behavior." 2 Therefore, it has been shown that bomo-
doped microcrystalline and nanocrystalline diamond ﬁlm electrodes possess similar
properties. Several factors can inﬂuence the electrochernica] response of a
polycrystalline diamond ﬁlm: (i) the presence of any non-diamond sp2 carbon
impurity phases, (ii) the surface termination (H vs. 0), (iii) the dopant type, level,
and distribution, (iv) grain boundaries and other morphological defects, and (v) the
primary crystallographic orientation. The extent to which any one of these factors
inﬂuences the response will depend on the charge-transfer mechanism of a particular
redox analyte.3'7 In 2000, our group published a paper titled “Standard Electrochemical
Behavior of High-Quality, Boron-Doped Polycrystalline Diamond Thin-Film
Electrodes ” and suggested that the data presented serve as a benchmark for research
groups evaluating the electrochemical properties of polycrystalline, hydrogen-
terrninated, boron-doped diamond electrodes.1

To better understand the relationship between the microstructure of boron-
doped diamond thin ﬁlm, and its electrical and electrochemical properties, we used
CP-AF M and SECM to study the spatial heterogeneity in the electrical and

electrochemical properties. In previous chapters, it was shown that these two

217

techniques, combined with SEM, Raman spectroscopy and electrochemical
measurements, provide researchers with a more complete picture of the structure-
function relationship at diamond thin-ﬁlm electrodes. In this chapter, we report on the
use of these methods to investigate the properties of boron-doped, nanocrystalline dia-
mond thin ﬁlms produced by two different methods. One high-quality nanocrystalline
ﬁlm was produced by the Naval Research Laboratory (NRL) group using a
conventional CH4/H2 source gas mixture. The other was produced using a
CHa/Hz/Ar source gas mixture, a method pioneered by the Argonne group. CP-AF M
was used to examine the morphology and electrical conductivity of both ﬁlm types.
SECM and visible-Raman spectroscopy were employed to probe the ﬁlm
morphology and microstructure. Electrochemical measurements were used to

assess the electrochemical properties.

6.2 Results and discussion

6. 2.1 Morphology

The MSU diamond ﬁlm was deposited by microwave-assisted chemical vapor
deposition (CVD), as described in Chapter 2. The NRL ﬁlm (grown on 3 in Si wafer)
was provided by Dr. James Butler. This ﬁlm was cut into small pieces (1 cm2 in area)
for all tests described herein. Figure 6.] shows representative AFM images (contact
mode in air) of the two ﬁhns. Clearly, the morphology of the MSU ﬁlm is quite
different from that of the NRL ﬁlm. The image of the MSU ﬁlm shows cauliﬂower-like
structures consisting of many nodular features. TEM images (not shown here) revealed
that these nodular features are aggregates of smaller diamond grains (grain size is 15 -
20 nm).8’ 9 On the other hand, the image of the NRL ﬁlm shows more uniform

morphology, consisting of crystal-like grains with different grain sizes (i.e. more

218

nanofaceted). The grain size is 50 - 200 nm which is of the same order as the
cauliﬂower-like structures found in the MSU ﬁlm. The biggest difference between the

two ﬁlms is the faceted crystallites of the NRL ﬁhn.

0.25

 

Figure 6.1 AFM mages for MSU (left) and NRL (right) boron-doped

nanocrystalline diamond ﬁlms.
Because of the cauliﬂower-like structure, the roughness of the MSU ﬁlm is
greater than that of the NRL ﬁlm as evidenced by the large nominal roughness factor

and maximum feature height. These data are summarized in Table 6.1.

Table 6.1 Rouglmess analysis of MSU and NRL ﬁlms.

 

Film RMS* (Avg. :t SD, um) Rmu“ (Avg. :1: SD, um)
MSU (n=4) 19.9 i 2.3 128.4 :1: 18.9
NRL (n=3) 8.3 i 0.9 53.8 i 2.2

 

 

 

 

 

 

 

* RMS = Root mean square of surface roughness over a scan area of 1 x 1 pmz.

"Rm = Maximum height.

The nanocrystalline diamond ﬁhn grown from an Ar-rich source gas results

from a high rate of renucleation during growth.9 C2 dimer is the primary growth species.

219

In the Ar environment, large crystals never have an opportunity to grow as new nuclei
frequently form. The nanocrystalline diamond ﬁlm gown from the Hz-rich source gas
was achieved by a special seeding process that produces a high initial nucleation density.
Therefore, a large number of nuclei initially form and a continuous ﬁlm is achieved
within a short gowth period.lo Even that faceted crystals are present, one can conclude
that the rate of renucleation is low and the rate of crystal growth is fairly uniform across
the surface.

Representative images of the two ﬁlm types are presented in Figure 6.2. The
MSU ﬁlm has feature sizes of around 50 nm, which is consistent with the AFM
images. These nodular features are actually aggegates of diamond grainsz‘ 8 The
smaller feature and gain size is typical of ﬁlms deposited from an Ar-rich source gas
mixtures and results from a high renucleation rate during gowth.9 On the other hand,
the NRL ﬁlm shows a much larger gain size with some secondary nucleation in the
gain boundaries. Different from the MSU ﬁlm, crystal facets are clearly visible on
the NRL ﬁlm. These features are typical of ﬁlms deposited from hydrogen-rich

CH4/H2 source gas mixtures and result from a low renucleation rate during deposition.

 

Figure 6.2 SEM images of the (A) MSU and (B) NRL nanocrystalline boron-

doped diamond ﬁlms.

220

6. 2.2 Ram an Spectra

Raman spectroscopy was used to probe the ﬁlm microstructure. Spectra for the
MSU and NRL ﬁlms are displayed in Figure 6.3. The spectrum for each ﬁlm is quite
different from the typical spectrum for single crystal diamond, which consists of
single, narrow band at 1332 cm". This band is attributed to the ﬁrst-order lattice
phonon of cubic diamond.ll A broad peak centered at 1331 cm‘1 is seen in Figure
6.3A, which is typical of nanocrystalline boron-doped diamond deposited from an Ar-
rich CH4/H2+Ar source gas mixture. This broad peak is likely a convolution of the
1332 cm'1 diamond phonon peak with scattering by the spz-bonded carbon in the grain
boundaries. The peak is broadened in energy because of the small size of the diamond
grains. '2 Other researchers argued for a different assignment for this peak.13 Birrell et
a]. assigned this broad peak to spZ-bonded carbon (i.e., D-band Raman peak).'4 The
argument for this assignment is based on the observation that the peak is absent from
the Raman spectrum when using UV excitation.” The D-peak is believed to arise
from a double resonant Raman process, where the intensity is highest for low energy
excitation and decreases strongly with increasing excitation energy. 13

The shoulder at 1155 cm", often observed for nanocrystalline diamond, has
been assigned to transpolyacetylene segments at the grain boundaries.’2 The peak at
1550 cm"1 is clearly the G-mode Raman peak, which is well known for disordered
carbon ﬁlms and arises from the in-plane stretching modes of the spz-bonded carbon

at the gain boundaries of nanocrystalline diamond. '3

221

 

 

 

 

 

    
 

 

 

 

_A 1331 an"\‘ 1550 cm“
3
'E
D
:5
I-
:5
a
=
O
o
A _ B
3 477 cm"
E " /
3
IE
I-
3
3 1312 cm“
H
5 ..
O
O L
F
400 800 1200 1600 2000

Raman Shift (cm‘)

Figure 6.3 Raman spectra (visible excitation, 514.4 nm) for the (A) MSU and (B)

NRL boron-doped diamond ﬁlms.

The Raman spectrum for the NRL ﬁlm is very different from that for the MSU
ﬁlm, as shown in Figure 6.3B. An asymmetric peak at 1312 cm'1 is seen that can be
attributed to the one-phonon Raman diamond line. The asymmetry of the one-phonon
band is caused by the Fano interference induced by quantum mechanical interference
between the discrete phonon state and electronic continuum. The band width, Raman
frequency shift and asymmetry of the peak are all affected by the Fano interference.ls

The broad band around 1215 cm'l and the broad and intense band around 500
cm'1 are two other typical bands observed for highly boron-doped diamond ﬁlms. The

origins of these two bands are still open to debate. To date, several assignments have

222

been proposed by different research groups. First, the band at 1215 cm'1 was proposed

to arise from boron atoms by Ager III et a1.16

However, a similar band along with a
band around 500 cm'1 are often observed for undoped diamond ﬁlms, as reported by
Buckley and co-workers.17 Therefore, these bands are not directly related to boron
bonding in the ﬁlm. Second, Gonon et a]. suggested that these bands are related to
maxima in the phonon density of states of diamond and arise from scattering by
phonons outside the center of the Brillouin zone. '8 This is based on the theoretical
phonon density of states for diamond,19 which has a maximum around 1200 cm“1 and
matches quite closely the band observed in the spectrum. May et a]. gave another
explanation for these bands. They noted that the positions of these two bands ageed
with the two maxima in the phonon density of states (PDOS) and claimed that these
bands may be connected with a relaxation of the wavevector selection rules.
Therefore, they may be associated with the actual boron incorporation in the lattice,
rather than the hole concentration.20 More recently, Vlasov et a]. studied the
vibrational modes of heavily boron-doped diamond using isotopic substitution of
boron and carbon. They found none of the bands in the Raman spectra shifted upon
10B-substitution, whereas shifts to lower frequency were observed for all bands upon
13C-substitution, as compared to a sample with natural isotope abundancies. Their
results supported the assignment of these bands to perturbed diamond lattice phonons.
Therefore, they ascribed these bands to vibration modes in maxima of the phonon
DOS of diamond.21 The broad peak at 477 cm'1 is the one-phonon mode of Si '8' 22
and the broad peak centered at 1010 cm'1 is the second-order scattering from the
silicon substrate. 16

As evidenced by the AFM and SEM images, the morphology of the MSU and

NRL ﬁlms is very different. There are also microstructural differences, as conﬁrmed

223

by Raman spectroscopy. The Raman spectrum for the MSU ﬁlm is strongly
inﬂuenced by sp2 bonded carbon in the gain boundaries but the Raman spectrum for
NRL ﬁlm is identical to that for highly boron—doped microcrystalline diamond. The
morphology of NRL ﬁlm is very similar to that for a highly boron-doped
microcrystalline diamond ﬁlm except the grain size is much smaller. This may
explain why the Raman spectrum for NRL ﬁlm is more similar to that for a highly

boron-doped microcrystalline rather than a nanocrystalline diamond ﬁlm.

6. 2.3 Electrochemical Characterization

Background cyclic voltammetric i—E curves are useful for examining the
diamond ﬁlm quality because the electrochemical response is highly sensitive to
physicochemical properties of the surface. The magnitude of the background current,
the working potential window, and the voltammetric features present within the working
potential window are all sensitive to the presence of nondiamond (amorphous or
graphitic) sp2 carbon impurity.”27 Figure 6.4 shows representative background cyclic
voltammetric i-E curves for MSU and NRL nanocrystalline BDD thin-ﬁlm electrodes
in 1 mol/L KC] at 0.1 V/s. The geometric area of both electrodes was the same (0.183
cmz), although the currents in Figure 6.4 are not normalized to this area. The
background current for the NRL ﬁlm is low and featureless between -0.8 and 2.0 V.
There is no evidence for any surface redox processes; therefore, the ﬁlm appears to be
ideally polarizable in this potential region. A similarly low and featureless background
current is observed for the MSU ﬁlm between -1.3 and 1.8 V. The MSU ﬁlm shows
an anodic peak on the forward sweep at 1.5 V. This peak is assigned to the oxidation

of the gain boundary sp2 carbon, where electrochemically-active carbon-oxygen

224

functionalities can form (e. g., quinine/hydroquinone couple).23’ “'28 The MSU ﬁlm
has more of this type of sp2 carbon than does the NRL ﬁlm.

The backgound currents for both electrodes are a factor of ~10 lower than
freshly polished glassy carbon of the same geometric area. No change in the shape
of the traces was observed during multiple cycles suggestive of a very stable surface
structure. The working potential window (3:500 1.1.A/cm2 geometric area) for these
two ﬁlms is 3.0 (NRL ﬁhn) and 3.4 V (MSU ﬁlm), respectively. The NRL ﬁlm has a
less negative onset potential for hydrogen evolution but more positive onset potential
for oxygen evolution, in comparison with the MSU ﬁlm. We suppose this is due, in
part, to more conductive sites on NRL ﬁhn maybe because of higher doping level
(refer to the CP-AF M images in next section), which would increase the density of
electronic states at negative potentials. But we can’t explain why NRL ﬁhn has a
slightly more positive onset potential (oxygen evolution) at present. The potential
windows for the MSU and NRL ﬁlms are within the range of 3——4 V typical of

good quality boron-dOped polycrystalline diamond.23'27

500 _. ,- w

0- MSU\‘ —-——"'—'=£

-500 ‘ \NRL

-1000 -

 

 

Kim)

-1500 -

-2000 -

 

 

.2500 . . . . . .
-2 -15 -1 -0.5 o 0.5 1 1.5 2

Potential (V vs. AglAgCl)
Figure 6.4 Background cyclic voltammetric i-E curves for the MSU and NRL

nanocrystalline boron-doped diamond ﬁlms in 1 mol/L KC] at 0.1 V/s.

 

225

 

Panels A-F of Figure 6.5 show cyclic voltammetric i—E curves for Fe(CN)63’/4',
Ru(NH3)63+’2+,IrC162‘/3', methyl viologen, dopamine, and Fey/2+, respectively, at the two
ﬁlms. The voltammetric data are summarized in Table 6.2. As an example, Figure 6.6
shows the cathodic current vs. em for the two ﬁhns in 0.1 mmol/L Ru(NH3)63+/2+ + 1
mol/L KC] (scan rate 1): 0.1 to 0.9 V/s). These redox systems were selected for
testing because of the known sensitivity or insensitivity of each to the electronic

properties, surface microstructure, and surface chemistry of sp2 and sp3 carbon

electrodes (HOPG, glassy carbon, and diamond).3'7’ 23‘ 2"

 

 

 

 

A 40, D MSU
20.4 .1
2 0 0]
3 . l
"' -20~ .
NRL MSU
40 -4o~ NRL
-03 0.0 0.3 0.6 0.9 -1'.2 -0:9 -036 -03
l B L E
20 . 40.‘
:5 cl 1
r? ‘ MSU 07
'20- : MSU
. NRL ‘ NRL

 

 

 

 

 

 

 

 

 

0.9 1.2 -0.3 0.0 0.3 0.6 0 9
Potential (V vs. Ag/AgCl) Potential (V vs. AglAgCl)

Figure 6.5 Cyclic voltammetric i-E curves for (A) 1 mmol/L Fe(CN)63"4' in 1
mol/L KC], (B) 1 mmol/L Ru(NH3).,3*’2+ in 1 mol/L KC], (C) 1 mmol/L er16243'+ 1
mol/L KC], (D) 1 mmol/L methyl viologen in 1 mol/L KC], (E) 1 mmol/L dopamine

in 0.1 mol/L HClo4, and (F) 1 mmol/L lie”2+ in 0.1 mol/L HC104. Scan rate: 0.1 V/s.

226

Table 6.2 Peak separations of the MSU and NRL ﬁlms in different redox

 

 

 

 

 

 

 

systems“.
NRL MSU
(mV) (mV)
Fe(CN)63""' 61 r 3 64 r 3
Ru(NH3)63+’2+ 57 r 1 59 a 2
IerM" 60 r 2 57 r 5
W 59 r 2 66 a 5
DA 513_+_15 577:18
ram“ 603 r 25 663 a 22

 

 

 

 

 

* Scan rate: 0.1 V/s, Solution: 1 mmol/L analyte + 1 mol/L KC] (for Fe(CN)63'/4' ,

Ru(NH3)63+’2+ , er16243' and MV) and 0.1 mol/L HClO4 (for DA and Few”). The data

are presented as average value 1r standard deviation with n =3.

 

 

 

 

 

 

 

 

25 -
- NRL
o MSU
2° ‘ Linera ﬁt with R=0.9997
J Linera ﬁt with R=1.0000
A 15 ..
15.
'8
m 10 -
.3“.
5 ..
o 1 l ' l ' I ' l ' l ' l T
0 5 10 15 20 25 30

U1I2

Figure 6.6 Relationship between the cathodic peak current and the square root of
the scan rate for the two BDD electrodes in 0.1 mmol/L Ru(NH3)63+/2+ + 1 mol/L KC].

The scan rate was valued from 0.1 to 0.9 V/s.

227

Fe(CN)63'/4' was used as a redox probe because the electrode reaction kinetics
for this couple are strongly inﬂuenced by the fraction of exposed edge plane on spz-
bonded carbon electrode, as well as the surface cleanliness.s Therefore, its
heterogeneous electron-transfer rate constant is quite sensitive to the surface
cleanliness and chemistry of sp2 and sp3 carbon electrodes." 5’ 25’ 29’ 30 Compared to
other outer-sphere systems, this redox system often behaves anomalously on carbon
electrodes and should not be considered outer-sphere. The physicochemical properties
of boron-doped diamond also strongly inﬂuence the electrode kinetics of Fe(CN)63"4‘.
For example, the cyclic voltammetric AEp is very sensitive to the surface termination
with the smallest value observed at the clean, hydro gen-terminated surface. After
oxygen termination, AEp increases by over 125 mV but is reversibly reduced back to
the original value after removal of the oxygen functionalities by hydrogen plasma
treatment.31 The AEp for the curves in Figure 6.5 are 64 mV for the MSU ﬁlm and 61
mV for the NRL ﬁlm. The near-Nemstian ABp indicates that both boron-doped
diamond electrodes possess the requisite surface structure and properties to support
rapid charge transfer.

Ru(NH3)(53+’2+ involves simple electron transfer on most electrodes including
diamond23'26' 32 and sp2 carbon.3'6 The electrode kinetics are relatively insensitive to
the surface microstructure, surface oxides, and adsorbed monolayers on sp2 carbon
electrodes.345 The fact that the electron-transfer rate constant is relatively insensitive
to surface chemistry indicates the electrode reaction does not involve a speciﬁc
interaction with a surface site or functional group.5 The most important factor
affecting the kinetics is the electronic properties of the electrode, speciﬁcally the
density of electronic states near the formal potential of the redox system. The AEp

values are nearly the same, 57 and 59 mV, for the NRL and MSU ﬁlm, respectively.

228

Similar to Ru(NH3)63+/2+, IrClsz'B' involves simple electron transfer at most

electrodes including diamondzo‘ 33

and sp2 carbon.5 The electron-transfer rate constant
is relatively insensitive to the surface microstructure, surface oxides, and adsorbed
monolayers on sp2 carbon electrodes.3‘6 Surface cleanliness is important, although to a
lesser degree than for Fe(CN)63"4'. The most inﬂuential factor is the electronic
structure of the electrode. As shown in Figure 6.5C, both the MSU and NRL
electrodes showed near-Nemstian responses for this redox couple with a AEp of 60
and 57 mV for the NRL and MSU ﬁlm, respectively.

Methyl viologen involves simple electron transfer at diamond and most other
electrodes.7’ 3" 34 The rate of reaction at diamond is relatively insensitive to surface
oxides, gain boundaries, defect density, and the presence of nondiamond carbon
impurity" 7’ 3" 34. Like Ru(NH3)53+’2+, the most important factor inﬂuencing the rate is
the density of electronic states near the formal potential. Near-Nerstian cyclic
voltammetric i-E curves are seen for both electrodes with a AEp of 59 and 66 mV for
the NRL and MSU ﬁlm, respectively. The near-Nemstian AF.p values indicate that
both ﬁlms possess similar electronic properties and similar activity for this redox
system.

The near-Nemstian responses for Fe(CN)63'/4", Ru(NH3)53+/2+, IrCl62'/3' and
methyl viologen all indicate that ohmic resistance within the ﬁlm is minimal even at
very negative potentials.

Dopamine exhibits much more electrochemical irreversibility, as evidenced by
the relatively large AEp. This more irreversible behavior is characteristic of all the
catechols and catecholamines investigated so far at diamond.25’26‘33’ 3’5 AEp is 513 mV for

the NRL ﬁhn and 577 mV for the MSU ﬁhn. For comparison, AEp, at polished glassy

carbon under identical conditions is in the range of 125-175 mV.36 The formal

229

potential is positive of that for Fe(CN)6344-

so a low density of electronic states is not the
reason for the larger AEp. We suppose that minimal adsorption on diamond is the
reason for the sluggish kinetics. Recent work in our group with several catechols
and catecholamines has revealed no evidence for adsorption at solution
concentrations as low as 1 x 10'6 mol/L. Hunt-Duvall and McCreery have presented a
detailed study showing that low A153p values correlate with catechol and catecholamine
adsorption on glassy carbon, and surface treatments that decreased adsorption also
increased AB,”

ABp for DA is quite sensitive to the presence of nondiamond sp2 carbon
impurity at which adsorption can occur.38 Therefore, this is a good redox system to test
the MSU ﬁhn against the NRL ﬁlm. The MSU ﬁlm has a high fraction of sp2 bonded
carbon in gain boundary, more than the NRL ﬁlm. If grain boundaries dominate the
electrochemical properties of the ﬁlm, then one might expect a smaller AEp for DA at
the MSU ﬁlm. This is not true for the results present here, which suggest that
electrochemical activity of the MSU ﬁlm is dominated by boron-doping rather than sp2
carbon in grain boundary.

Very sluggish electrode kinetics are also observed for Few2+

on diamond, as
evidenced by the large AEp of 603 mV for NRL and 663 mV for MSU ﬁlm. The
electrode kinetics for F e3 ”2+ are highly sensitive to the presence of oxides on sp2
carbon electrodes. The McCreery group has shown that electron transfer for this
system is catalyzed by a speciﬁc chemical interaction with surface carbonyl function-
alities."’5 Therefore, we conclude that the sluggish electrode kinetics at diamond are

due to the absence of mediating surface carbon-oxygen functionalities, speciﬁcally

carbonyl groups.

230

The voltammetric data presented above reveal that for Fe(CNXM',
Ru(NH3)63+’2+, IrClém' and MV, the clean, hydrogen-terminated NRL and MSU ﬁlms
are very active without any conventional pretreatment prior to use. On the other hand,
the relatively large peak separation observed for F e3 ”2+ at hydrogen-terminated diamond
is presumably because of the absence of surface carbonyl functional groups, which
are known to catalyze this reaction.39 Relatively slow electrode reaction kinetics are
also observed for dopamine, presumably due to weak adsorption on hydrogen-

terminated diamond.28

6.2.4 CP-AFM measurements

Figure 6.7 shows the topographical (A) and conductivity (B) maps for the
MSU boron-doped diamond nanocrystalline thin ﬁlm obtained by CP-AF M. The bias
voltage was +0.2 V (sample vs. tip), and the scan size was 200 x 200 nmz. In the
topogaphica] map, nodular features of 50-100 nm in the lateral dimension are seen.
The boundaries between these features are also clearly visible. In the conductive map,
nonuniformly distributed conductive sites are seen. Although the most of conductive
sites appear to be associated with gain boundaries (most of small conductive regions
in Figure 6.7B are associated with gain boundaries in Figure 6.7A), Figure 6.78
shows evidence that several nodular features are conductive (e. g., the particle in the
middle right and the nodular feature in the lower left quadrant of the image). On the
other hand, some of the least conducting regions in the image are actually associated
with nodular features and even grain boundaries (see the upper edge area of Figure

6.78).

231

200.0

100.0

0.0 nm

   

Figure 6.7 CP-AFM images for the MSU nanocrystalline boron-doped diamond

thin ﬁlm. (A) Height image and (B) Conductivity map.

Generally, not all of the nodular features and surrounding gain boundaries
exhibit high conductivity. This may be related to inhomogeneous boron dopant
distribution in the ﬁlm,4°‘42 as was discussed in Chapter 4.

Figure 6.8 shows the topographical (A) and conductivity (B) maps for the
NRL boron-doped diamond nanocrystalline thin ﬁlm. The bias voltage was +0.2 V
(sample vs. tip), and the scan size was 200 X 200 nmz. In the topographical map, a
few of the small crystals are seen . In the conductive map, nonuniformly distributed
conductive regions are seen. Most of gains are conductive, which is a little different
from the MSU ﬁlm. Therefore, the fraction area that is conductive appears higher for
the NRL ﬁlm. This is one reason why the nominal AEp values were smaller for the
NRL ﬁlm when probed with Fe(CN),“T, dopamine and Fe3+fz+ (refer to section 6.2.3).
Similar to the MSU ﬁlm, most of the imaged area, including crystal facets and
boundary areas, are conductive. There are some grains and boundaries showing lower
conductivity (see the gains at the middle of the upper edge and lower left comer of

Figure 6.8A and B). We suppose there are two possible reasons for this observation.

232

First, there may be less boron aggegates in these regions, which results in a lower
charge carrier concentration. Second, the contact area between the CP-AFM tip and
the sample may play a role because of the bigger gain size as we discussed in

Chapter 4.

 

Figure 6.8 CP-AF M images of an NRL nanocrystalline boron-doped diamond thin

ﬁlm. (A) Height image and (B) Conductivity map.

6.3 Conclusions

The electrochemical properties of high-quality, boron-doped nanocrystalline
diamond thin ﬁlms produced by two different gowth methods were compared.
Several factors, such as the gain size, morphology, microstructure and electrical
conductivity that inﬂuence the electrode kinetics for several aqueous-based redox
systems, were investigated. The similarity of the voltammetric data for the NRL and
MSU ﬁlms indicates the high quality and electrical conductivity of both ﬁlms. Several
important conclusions regarding the structure-reactivity relationship of these

nanocrystalline ﬁlms can be made. First, similar CVs for Ru(NH3)63+a+, IrClgm‘ and

233

methyl viologen were obtained for both ﬁlms. These redox systems appear to
involve simple electron-transfer that is relatively insensitive to surface oxides, grain
boundaries and other defects including nondiamond carbon impurities present in the
polycrystalline ﬁlms. Therefore, our results suggest that the electrochemical
properties of the MSU ﬁlm are similar to those of the NRL ﬁlm, even the two ﬁlms
show different microstructures because of the different growth conditions.
Ru(NH3)(r,3+’2’r and methyl viologen have formal potentials that fall well into the midgap
region of diamond. The near-Nerstian responses for both redox systems suggest a
sufﬁcient density of electronic states exists near the formal potentials of each. These
midgap states result mainly from the boron doping and the lattice hydrogen, rather
than from non-diamond carbon impurity phases or gain boundaries and other
defects. Second, Fe(CN)63'/4' involves more complex, surface-sensitive electron-
transfer that is highly sensitive to the presence of surface oxides but insensitive to
non-diamond carbon impurity phases. The fastest kinetics were observed for the
hydrogen-terminated surface. Third, Few2+ undergoes sluggish electron-transfer at
both the NRL and MSU ﬁlms consistent with the absence of carbonyl groups on the

surface as electron-transfer kinetic of Few2+

are known to be strongly inﬂuenced by
these groups.39 Fourth, the slow electrode reaction kinetics for dopamine may be due a
lack of adsorption on the relatively non-polar surfaces. Both SEM and AF M images
show the difference of the morphologies of two ﬁlms, which may result from
different growth conditions. CP-AF M images revealed that although most of the
imaged area was conductive for both ﬁlms, there were some gains and boundaries

showing low conductivity. The fraction of highly conductive regions for the NRL

ﬁlm was greater than that for the MSU ﬁlm.

234

The key ﬁnding was that nanocrystalline boron-doped diamond ﬁlms
produced by two different growth procedures possess similar electrochemical and

electrical properties.

235

(£41

(1)

(2)
(3)
(4)
(5)
(6)

(7)

(8)

(9)
(10)

(11)

(12)
(13)
(14)

(15)

(16)

(17)

Ilegﬁareatcz:

Granger, M. C.; Witek, M.; Xu, J .; Wang, J .; Hupert, M.; Hanks, A.; Koppang,
M. D.; Butler, J. E.; Lucazeau, G.; Mermoux, M.; Strojek, J. W.; Swain, G. M.
Anal. Chem. 2000, 72, 3793-3804.

Fischer, A. E.; Show, Y.; Swain, G. M. Anal. Chem. 2004, 76, 2553-2560.
Kneten, K. R.; McCreery, R. L. Anal. Chem. 1992, 64, 2518-2524.

Chen, P.; Fryling, M. A.; McCreery, R. L. Anal. Chem. 1995, 67, 3115-3122.
Chen, P.; McCreery, R. L. Anal. Chem. 1996, 68, 3958-3965.

Ranganathan, S.; Kuo, T.-C.; McCreery, R. L. Anal. Chem. 1999, 71, 3574-
3580.

Yang, H.-H.; McCreery, R. L. Anal. Chem. 1999, 71 , 4081-4087.

Show, Y.; Witek, M. A.; Sonthalia, P.; Swain, G. M. Chem. Mater. 2003, 15,
879-888.

Gruen, D. M. Annu. Rev. Mater. Sci. 1999, 29, 211-259.

Butler, J. E.; Woodin, R. L.; Brown, L. M.; Fallon, P. Phil. Trans. R. Soc.
Lond. A 1993, 342, 209-224.

May, P. W.; Ludlow, W. J.; Hannaway, M.; Heard, P. J.; Smith, J. A.; Rosser,
K. N. Chem. Phys. Lett. 2007, 446, 103-108.

Ferrari, A. C.; Robertson, J. Phys. Rev. B 2001, 63, 121405R.
Ferrari, A. C.; Robertson, J. Phys. Rev. B 2000, 61 , 14095.

Birrell, J.; Gerbi, J. E.; Auciello, 0.; Gibson, J. M.; Johnson, J.; Carlisle, J. A.
Diamond Relat. Mater. 2005, 14, 86-92.

Ushizawa, K.; Watanabe, K; Ando, T.; Sakaguchi, 1.; Nishitani-Gamo, M.;
Sato, Y.; Kanda, H. Diamond Relat. Mater. 1998, 7, 1719-1722.

Ager 111, J. W.; Walukiewicz, W.; McCluskey, M.; Plano, M. A.; Landstrass,
M. 1. Appl. Phys. Lett. 1995, 66, 616-618.

Buckley, R. G.; Moustakas, T. D.; Ye, L.; Varon, J. J. Appl. Phys. 1989, 66,
3595-3599.

236

(18)

(19)

(20)

(21)

(22)

(23)

(24)

(25)

(26)

(27)

(28)
(29)
(30)
(31)

(32)

(33)

(34)

Gonon, P.; Gheeraert, E.; Deneuville, A.; Fontaine, F.; Abello, L.; Lucazeau,
G. J. Appl. Phys. 1995, 78, 7059-7062.

Tubino, R.; Piseri, L.; Zerbi, G. J. Chem. Phys. 1972, 56, 1022.

May, P. W.; Ashfold, M. N. R.; Mankelevich, Y. A. J. Appl. Phys. 2007, 101 ,
053115.

Vlasov, I. I.; Ekimov, E. A.; Basov, A. A.; Goovaerts, E.; Zoteev, A. V.; Los
Alamos National Laboratory, Condensed Matter, 2008, pp 1-7.

Ricci, P. C.; Anedda, A.; Carbonaro, C. M.; Clemente, F .; Corpino, R. Thin
Solid Films 2005, 482, 311-317.

Xu, J.; Granger, M. C.; Chen, Q.; Strojek, J. W.; Lister, T. E.; Swain, G. M.
Anal. Chem. 1997, 69, 591A-597A.

Swain, G. M.; Anderson, A. B.; Angus, J. C. MRS Bulletin 1998, 56.

Wang, J.; Swain, G. M.; Mermoux, M.; Lucazeau, G.; Zak, J.; Strojek, J. W.
New Diamond Front. Carbon Technol. 1999, 9, 317-343.

Granger, M. C.; Xu, J.; Strojek, J. W.; Swain, G. M. Anal. Chim. Acta 1999,
397, 145-161.

Martin, H. B.; Argoitia, A.; Landau, U.; Anderson, A. B.; Angus, J. C. J.
Electrochem. Soc. 1996, I43, L133-Ll36.

Swain, G. M. J. Electrochem. Soc. 1994, 141, 3382.

Hu, 1. F .; Karweik, D. H.; Kuwana, T. J. Electroanal. Chem. 1985, 188, 59-72.
Kuo, T.-C.; McCreery, R. L. Anal. Chem. 1999, 71, 1553-1560.

Granger, M. C.; SWain, G. M. J. Electrochem. Soc. 1999, 146, 4551-4558.

Vinokur, N.; Miller, B.; Avyigal, Y.; Kalish, R. J. Electrochem. Soc. 1996,
143, L238.

Alehashem, 8.; Chambers, F .; Strojek, J. W.; Swain, G. M.; Ramesham, R.
Anal. Chem. 1995, 67, 2812-2821.

Cline, K. K.; McDermott, M. T.; McCreery, R. L. J. Phys. Chem. 1994, 98,
5314-5319.

237

(35)

(36)
(37)

(38)

(39)

(40)

(41)

(42)

Popa, E.; Notsu, H.; Miwa, T.; Tryk, D. A.; Fujishima, A. Electrochem. Solid
State. Lett. 1999, 2, 49.

Chen, Q.; Swain, G. M. Langmuir 1998, I4, 7017.
Hunt-DuVall, S.; McCreery, R. L. Anal. Chem. 1999, 71 , 4594.

Bennett, J. A.; Wang, J.; Show, Y.; Swain, G. M. J. Electrochem. Soc. 2004,
I51, E306-E313.

McCreery, R. L.; Cline, K. K.; McDermott, C. A.; McDermott, M. T. Colloids
and Surfaces A: Physicochemical and Engineering Aspects 1994, 93 , 211-219.

Sahli, S.; Aslama, D. M. Appl. Phys. Lett. 1997, 70, 2129-2131.

Bennett, J. A.; Show, Y.; Wang, S.; Swain, G. M. J. Electrochem. Soc. 2005,
I52, E184-E192.

Wang, J .; Swain, G. M. J. Electrochem. Soc. 2003, I49, E24-E32.

238

Chapter 7
Diamond Microelectrodes for Continuous Amperometric

Recording at Blood Vessel Surfaces

7.1 Introduction

The peripheral nervous system (PNS) consists of sensory nervous system, which
sends information to the central nervous system (CNS) from internal organs and from
external stimuli, and the motor nervous system, which carries information from the CNS
to organs, muscles, and glands. The motor nervous system is divided into the somatic
nervous system and the autonomic nervous system. The somatic nervous system controls
skeletal muscle as well as external sensory organs, such as the skin. The autonomic
nervous system controls involuntary muscles, such as smooth and cardiac muscle. The
autonomic nervous system can further be divided into the parasympathetic and
sympathetic divisions. The sympathetic nervous system is concerned especially with
preparing the body to react to situations of stress or emergency, which contains chieﬂy
adrenergic ﬁbers and tends to depress secretion, decrease the tone and contractility of
smooth muscle, increase heart rate and blood pressure. The parasympathetic nervous
system, on the other hand, is evident when a person is resting and feels relaxed, and is
responsible for such things as the constriction of the pupil, the slowing of the heart, the
dilation of the blood vessels, and the stimulation of the digestive and genitourinary

systems.

239

Sympathetic nerves are responsible for acute changes in blood pressure and they
contribute to long-term blood pressure regulation. Peripheral sympathetic nerves travel
along arteries and veins, and are present near the adventitial-medial border to make close
contact with vascular smooth muscle cells. The nerves communicate with the effector
cells of the resistance (arteries) and capacitance (veins) vessels through the release of
norepinephrine (NE), a vasoconstrictor neurotransmitter, as shown in Figure 7.1. This is
the typical pattern for the sympathetic innervation of blood vessels.1 Only the smooth
muscle cells located at the adventitial/medial border are directly innervated; 3 The
appearance and arrangement of sympathetic nerves associated with the veins is similar to
that of arteries, although nerve density in veins is generally less than that observed in the

adjacent arteriesf” 5

A B C

Varicosity

    

Sensory
nerves

Blood —->
vessel

Post-junction (Artery) Post-Junction (vein)

Sympathetic
nerves

Figure 7.1 Schematics for (A) the distribution of sensory and sympathetic nerves
around a blood vessel, (B) an arterial neuroeffector junction, and (C) a venous

neuroeffector junction.

240

Sensory nerves arising from neurons in the dorsal root ganglia (DRG) supply
most blood vessels and these nerve ﬁbers contain and release the neuropeptides,
substance P (SP) and calcitonin gene related peptide (CGRP).6 Activation of
perivascular sensory nerves causes arterial dilation.7 While sympathetic nerve activity is
an important determinant of systemic blood pressure, sensory nerves may also contribute
to the regulation of vascular tone. For example, CGRP levels are increased in DRG of
deoxycorticosterone acetate (DOCA)-salt rats and treatment of these animals with a CGRP
receptor antagonist increases blood pressure.8

The communication between sensory and sympathetic nerves is accomplished
along a pathway called the reﬂex arc. A normal reﬂex are, as schematized in Figure 7 .2,
involves the afferent nerve carrying nerve impulses from receptors or sense organs
toward the CNS and the efferent nerve carrying nerve impulses away from the CNS to
effectors, such as muscles or glands. It has been known that repetitive activation of
afferent nerve ﬁbers can cause cross-excitation of neighbors not stimulated directly.9 This
cross-excitation has been demonstrated at sites of peripheral nerve injury and sensory
gangliaw’ 11 However, it is still unclear if sensory nerves directly communicate with
sympathetic nerves locally near the neuroeffector junctions. One of the speciﬁc aims of
our research was to determine if there is local (direct) communication between sensory
and sympathetic nerves that innervate mesenteric blood vessels of rats when the
connections from both sensory and sympathetic nerves to the CNS are severed (no reﬂex
arc).

In this chapter, progress toward this speciﬁc aim is discussed including

determining the distribution of sensory and sympathetic nerves along blood vessels,

241

understanding the effects of drugs used in the work on the diamond electrode’s response

for norepinephrine.

Afferent neuron

 

Receptor

Efferent neuron

  

intemeuron

Effector

Figure 7.2 An illustration of a simple reﬂex are.

7.2 E ﬂeets of drugs on the electrochemical responses of carbon

ﬁber and diamond microelectrodes

7. 2.1 Norepinephrine and drug effect

Norepinephrine (NE) functions as both a hormone and neurotransmitter. It is
secreted by the adrenal medulla and released ﬁom sympathetic nerves. It plays a key role
in the regulation of vascular tone, heart rate, blood pressure, and blood sugar level.”’ ’3
NE levels have been measured in tissue and biological ﬂuids because of its involvement

in stress related diseases, such as, hypertension, depression and cardiovascular disease.“

242

‘4 In the past, measurements of NE in plasma were done by high performance liquid
chromatography (HPLC) or with radioactive labeling methods. These so-called
“overﬂow” measurements provide little information about real-time, local release of
NEH—16

Since NE can be oxidized via a 2H+/2e redox reaction, it can be detected

electrochemically as an oxidation current. Since the pioneer work of Gonon et al., 17’ '8

on
NB release from rat tail artery as measured by continuous amperometry, electrochemical
measurements with carbon ﬁber microelectrodes have been widely employed for in vitro
studies of NE release and clearance from sympathetic nerves innervating mesenteric
arteries and veins.”23 Unlike the “overﬂow” methods, the electrochemical methods
provide information on dynamic changes in neurotransmitter concentrations near the sites
of release.

Two types of carbon electrodes have been used: carbon ﬁber and boron-doped
diamond. Carbon ﬁber has been routinely used for in vivo and in vitro electrochemical

measurements since the l970’s.2"'27

Recently, diamond has been used because of its
excellent response sensitivity, reproducibility, pH-independent current response and its
resistance to fouling.”32 Drugs are routinely used with the electrochemical
measurements to pharrnacologically characterize the neuroeffector junction. It is unclear
how these drugs might affect the microelectrode response for NE .21’22’31'34 There have
been reports in the dopamine studies that some drugs decrease the carbon ﬁber
microelectrode sensitivity, which leads to decreased dopamine current measured in

vitro.” 36 Previous studies have discussed the inﬂuence of some drugs on electrochemical

measurement.” 37 Among the drugs that show electroactivity, some of them don’t have

243

signiﬁcant effects on the sensitivity of the microelectrode, however, some other drugs,
like GBR 12909 and nomifensine, the effect of the drug is quite strong that it needs to be
considered if the current change is due to both pharmacological effect and the change of
microelectrode sensitivity.36 Non-electroactive drugs may also change the responsivity of
the microelectrode.17 The mechanisms that cause electrode fouling drop are not fully

understood.

We recently found that in addition to the pharmacological effect, some commonly
used drugs also changed the surface chemistry of the microelectrode and thus changed
the NE oxidation current. For some drugs, even after washing out, the NE current could
not be totally recovered. In this work, we systematically “studied if and how drugs
commonly used to study sympathetic neuroeffector transmission affect the electrode
response for NE oxidation. Seven drugs were investigated: prazosin, an at adrenergic
receptor antagonist; yohimbine, an a; adrenergic receptor antagonist; idazoxan, an 012
adrenergic antagonist; UK, 14,304, and a2 adrenergic agonist; cocaine, NE transporter
antagonist; PPADS, a P2X-purinoceptor antagonist and capsaicin, a drug used for TRPV
1 channel.” 34’ 33' 39 The electroactivity of these drugs and their effect on the electrode

response for NE were studied using both carbon ﬁber and diamond microelectrodes.

Table 7.1 shows the molecular structure and function of the neuropharrnacologica]
agents. Two types of carbon ﬁber were used: Thornel P55 pitch-based ﬁbers and Thornel
T650 PAN-based ﬁbers. Both types have been widely used to investigate

neurotransmission in biological environments.” 4'

244

 

 

 

 

 

 

Table 7.1 The structures and functions for NE and seven studied drugs.
Chemicals Function Structure
Neurotransmitter in OH
NE sympathetic nervous ”0 ””2
system
HO
P 1 l”
Capsaicin Agonist for TRPV l b
r 3C\c V\CH3
channel
C
o
Cocaine Antagonist for NE O/CH3
transporter '
O
Antagonist for el-2 / >
Idazoxan autoreceptor 1"
H
Antagonist for P2X CHO
PPADS ' ' t “0 P Na
purrnerglc recep or \ o/ 03 2
/ / N
H3C N N/
NaO3S O3Na

 

 

 

 

245

 

Table 7.1 The structures and functions for NE and seven studied drugs (continued).

 

Chemicals Function Structure

 

Prazosin Antagonist for \ /

ﬁ 0\
01-1 receptor N\"/ O CH 3
C H
/ O/ 3

N

“*2

 

Agonist for 01-2 H

UK 14,304 autoreceptor

(L
1
if

 

Antagonist
Yohimbine for 01-2

autoreceptor

 

 

 

 

 

 

Refer to Chapter 2 for more information about the drugs.

7.1.2 Results

A. Raman spectroscopy
The carbon microelectrodes were ﬁrst characterized by Raman spectroscopy.
Figure 7.3 shows characteristic spectra for diamond and two types of carbon ﬁber

microelectrode. Figure 7.3a shows an intense one-phonon diamond line at 1332 cm”.4244

246

There is minimal scattering intensity observed between 1500 and 1600 cm], which arises
from amorphous nondiamond carbon (i.e., sp2 bonded carbon) irnpurity.31’43’ 45 Figure
7.3b shows P55 carbon ﬁber and Figure 730 shows T650 carbon ﬁber microelectrode
microstructure. The exterior of the carbon ﬁbers is an amorphous carbon network with a
majority of sp2 hybridized bonds and of nanocrystalline graphite regions. The G band
peak, as the peak around 1583 cm'1 shown in Figure 7.3b and the peak around 1587 cm’|
in Figure 7.3c, represent the crystal graphite structure."6'48 The peaks shown around 1350
cm'1 in Figure 7.3b and 1370 cm'1 in Figure 7.3b are D band peaks, which are good
indicators of disorder in the crystal structure. The higher ID/IG ratio suggests the more
disordered structure of T650 carbon ﬁber, of which we suppose to have higher
electroactivity.46’ 47 This may explain why the Elm values of T650 carbon ﬁber
microelectrode for NE redox systems are lower comparing with P555 carbon ﬁber and

diamond microelectrodes (Table 7.2).

 

 

 

 

 

 

 

 

 

Table 7.2 E )2 values of seven drugs, Fe(CN)63'/4' and Ru(l\lH3,).53+/2+ on three types of
microelectrodes.

E172 (mV) vs. SCE PSSS T650 Diamond
FeCN 255 :1: 5 229 i 5 283 :1: 13
Ruhex -240 :1: 2 -223 :1: 2 -251 i 4

NE 137 i 4 98 :1: 7 294 :1: 10
Capsaicin 563 :1: 16 494 i 19 548 :1: 21
PPADS N/A N/A 805 :1: 10
Prazosin 772 i 10 794 :t 17 751 :1: 2
UK 14,304 893 :1: 6 877 i 9 876 :1: 4
Yohimbine 703 i 12 696 :1: 12 637 :1: 13

 

 

 

 

 

 

247

 

1.8! a

 

 

 

«— 1332cm'1

1.5-
=5
5 1.2-
E’
m
c 0.9‘
3 i V
E .

0.0- p - '

0 Y 500 7 1000 ' 15$ 7 20110 . 25'00

Raman Shift (cm'1)

*— 1583 cm'

 

 

 

 

 

=1
5.
E
or
C
3
5
'1
0.1 r I ' I ' I ' I ' I
0 500 1 000 1 500 2000 2500
Raman Shift (cm-1)
‘ c
0.40 - 1
1370 cm ___.. 158., cm"
a“. 0.35 "‘
q .
S 0.30 -
8 _
c 0.25
g .
— 0.20 4
i
0-15 ' I ' 7 ‘ I Y I ' I
0 500 1000 1500 2000 2500

Raman Shift (cm'1)

Figure 7.3 Raman spectra of (a) a boron-doped diamond thin ﬁlm deposited on a Pt

wire, (b) a P55 carbon ﬁber, and (c) a T650 carbon ﬁber.

248

B. Electrochemistry
Figure 7.4 shows cyclic voltammetric i-E curves for a 0.1 mmol/L concentration

of the drug in Krebs’ buffer (0.1 V/s) at P55 (Figure 7.4a and b), T650 (Figure 7.4c and d)
and diamond (Figure 7.4e and f) microelectrodes. The curves indicate that capsaicin,
prazosin, yohimbine and UK 14,304 are electrochemically-active at both carbon ﬁber and
diamond microelectrodes. Capsaicin can be oxidized between 0.5 and 0.8 V (vs. Ag/AgCl)
on both carbon ﬁber and diamond. However, the oxidation current is signiﬁcantly lower
than that for the other three drugs at equal concentration. Prazosin, UK 14,304 and
yohimbine can be oxidized between 0.7 and 1.0 V (vs. Ag/AgCl) on carbon ﬁber and
between 0.8 and 1.1 V (vs. Ag/AgCl) on diamond. Figure 7.4 also shows that cocaine and
idazoxan are electrochemically inactive on carbon ﬁber and diamond electrodes within 0
- 1.2 V (vs. Ag/AgCl) window and PPADS is electrochemically inactive on carbon ﬁber
electrodes but can be oxidized on diamond electrode between 0.9 and 1.1 V (vs.

Ag/AgCl).

Eﬂ'ects of Capsaicin and Prazosin. Figure 7 .5 shows the cyclic voltammetric i-E
curves for 0.02 mmolL'l NE, 0.02 mmolL'1 NE + 0.1 mmolL'1 capsaicin or prazosin, and
0.02 mmolL'1 NE after capsaicin or prazosin exposure at P55 (Figure 7.5a and b), T650
(Figure 7.5c and d) and diamond (Figure 7.5e and f) microelectrodes. These data show
that the oxidation current peak of NE could not be distinguished when capsaicin or
prazosin was present except on T650 carbon ﬁber electrode, on which the oxidation
current peak of NE could be seen but signiﬁcantly decreased when capsaicin was present.

249

 

and diamond microelectrodes.
P55 carbon ﬁber electrode. (b) Idazoxin, PPADS and cocaine on P55 carbon ﬁber
electrode. (0) Capsaicin, prazosin, UK 14,304 and yohimbine on T650 carbon ﬁber
electrode. ((1) Idazoxin, PPADS and cocaine on T650 carbon ﬁber electrode. (e)
Capsaicin, prazosin, UK 14,304 and yohimbine on diamond microelectrode. (f)
Idazoxin, PPADS and cocaine on diamond rrricroelectrode. All curves were acquired in

Krebs solution at 0.1 V/s. The concentration of the drug was 0.1 mmol/L.

 

  

 

 

 

  

   

 

 

    

 

40.04 a
300‘
g 20.0-
‘8'
t 10.0'
3
o .
0.0‘
‘ Capsaicin
0:0 3 0:3 ' 0:0 ' 0:9 ' 1:2
Potential (V vs. AglAgCl)
30.0~ c
g 20.0‘ Prazosln
E
93 10.0
s- Yohlmblne
3
0
0.0- -
UK 14,304/Capsaicin
etc of: ' ole ' 0:9 - 11:
Potential (V vs. AglAgCl)
150.04 e
A , UK14,304
< 100.0
5
E
at, 50.02 Prazosln
3
L) Ca sakﬁn
p \l
0'0 d I Yohimbine
0.0 013 0:0 0:0 71
Potential (V vs. AglAgCl)
Figure 7.4

Cyclic voltammetric i-E curves for the drugs at carbon ﬁber (PSS, T650)

Current (nA) Current (nA)

Current (nA)

20.01

 

 

 

 

 

   
 

 

 

 

 

 

 

 

10.0"
0.0“
PPADS
0:0 7 0:3 I 0.7' 019 ' 11:
Potential (V vs. AglAgCl)
30.0l d
1
20.0
Cocaine—-
10.0-
) Idazoxan
0.0‘ :—
Franz
0:0 ' 033 ' 0:0 ' 0:0 1.?
Potential (V vs. AglAgCl)
40.0-
30.0j
20.0-
10.0,
1
0.01
010 0:3 0:0 0:9 7 1:2

250

Potential (V vs. AglAgCl)

(a) Capsaicin, prazosin, UK 14,304 and yohimbine on

 

   

 
    

  

 

 

 

 

    

 

a
20.0-
‘ NE
E NE after Capsalcln \ /.
:1 10.0:
s
t . '
3
O
0.04 /
NE with Capsalcln
0:0 03 ' 0:0 0:9 7 17.2
Potential (V vs. AglAgCl)
c
NE with Capsaicin
NE
10.0 - \
‘8'
S
0 0.0 1
NE after Capsaicin
010 of: 030 0:0 ' 1:2
Potential (V vs. AglAgCl)
e
1
A 10.0.
78
l:
3
0
0,0 . NE with Capsaicin
0:0 7 0:3 0:0 ' 0:9 ' 1:2
Potential (V vs. AglAgCl)
Figure 7.5

Current (nA)

Current (nA)

Current (nA)

 

 

    

 

 

     
   

 

 

  
   

 

40.0 - b
30-0 ‘ NE with Prazosin
20.0 -
10.0 -
0.0 -
NE after Prazosin
0:0 0:3 0:0 0:0 1T2
Potential (V vs. AglAgCI)
30.0’ d
NE with Prazosin
20.0‘1
1 NE
10.0 ‘ 1 I
0-0 I A 1 NE after Prazosin
0:0 013 01.0 0:0 11:
Potential (V vs. AglAgCl)
80.0 f
60.0 4
’ NE with Prazosin
40.0
20.0
’" 7 T
0.0 1/E alter Prazosin
0:0 0:3 0:0 0:9 1T2

Potential (V vs. AglAgCl)

Cyclic voltammetric i-E curves for NE at P55 (a), T650 (c) carbon ﬁber
electrode and diamond microelectrodes (c) with and without capsaicin and Cyclic
voltammetric i-E curves for NE at P55 (b), T650 ((1) carbon ﬁber electrode and diamond
microelectrodes (f) with and without prazosin. All curves were acquired in Krebs’ buffer

at 0.1 We The concentration of NE was 0.02 mmol/L and the concentration of each drug

was 0.1 mmol/L.

251

The data in Figure 7.6 summarize how the NE oxidation response at all three
electrodes was affected by exposure to capsaicin and prazosin. Results are presented as a
shift in Em and a ratio of ip°"(w,-th mane, drug) /ip°" (before drug). Both capsaicin and prazosin
signiﬁcantly affect the detection of NE at their presence at carbon ﬁber and diamond
electrodes (Figure 7.6a and b). The peak current of NE could not be determined using
P553 and diamond microelectrode with these two drugs. NE could be measured using
T650 microelectrode, however, the current amplitude has been signiﬁcantly diminished.
After washing out drugs (Figure 7 .50 and d), the effects of capsaicin on the detection of
NE with three types of electrodes were reversed, however, the inhibitory effect of
prazosin on P55s and diamond microelectrodes were hard to be reversed (E1 /2 of NE

right-shifted and the current decreased).

To test if the adsorption of the drug species to the electrode surface is physically
or not, we used two redox systems, Ru(NH3)(53+/2+ and Fe(CN)63'/4' to characterize the
drug effect on the microelectrode sensitivity. The mechanisms that how these two redox
systems work will be discussed later. Data are shown in Table 7.3. These results show
that capsaicin signiﬁcantly deactivates the diamond microelectrode (E1 /2 for Fe(CN)63'/4'
is right-shifted and Ru(NH3)63+’2+ is left-shifted signiﬁcantly) but they have negligible
effect on carbon ﬁber electrodes (El/2 is shifted only slightly). Prazosin deactivated all
three types of microelectrode. The effects of capsaicin on three types of electrode were
reversed when the drug was washed out but not for prazosin, suggesting that prazosin

changed the electronic state of the microelectrode.

252

 

Table 7.3 Statistical data (n=3) for Em values and current changes for Fe(CN)?“

and Ru(NH3)(5+3/+2 before and after exposure to capsaicin and prazosin at P55, T650

carbon ﬁber and diamond microelectrodes.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

      
  

 

   

 

P55 T650 Diamond
Drug Probe
AE1/2(mV) AE1/2(mV) AE1/2(mV)
C Fe(CN)63-/4- 18.3 i 9.7 -7.5 r 20.7 89.7 :t 9.5 *
‘ Ru(NH3)63+/2+ 1.0 i: 6.0 0 r 0.6 -127 r 2.9
Fe(CN)63-/4- 69.7 r 13.5 * 7.0 r 5.0 251.3 i 30.3 *
Prazosin
RutNH3)63+/2+ -6.7 i- l.9 -5.7 r 1.2 -290 r 10.0
3“ £15555" " 'L
1‘0 - T650 )
25° _ - Diamond‘
.5
E 200 E
150
E 10 {a
u‘: ° 2
<1 50 g
_N
0
Capsaicin Prazosin 0 Capsaicin Prazosin
10 c
.1 .. . :28:
- Diamond[
S‘ .10— 3 TOT
g a
In 410- a
< -Psss '3
4°” -T850 3
«w -eon _.
-60- 0.0
Capsaicin Prazosin Capsaicin Prazosin

Figure 7.6 Statistical results (11 = 3) for the effect of capsaicin and prazosin. (a) AE1 ,2
before and after drug on P55, T650 and diamond electrode for NE detection. (b) Peak
current changes before and after drug for NE detection on three types of electrode. (c)
AE1/2 before and after drug was applied for NE detection. ((1) Peak current changes before
and after drug was applied for NE detection using three types of electrode.

253

Effects of UK 14,304 and Yohimbine. Figure 7.7 shows cyclic voltammetric i-E
curves for 0.02 mmolL'1 NE, 0.02 mmol/L NE + 0.1 mmol/L UK 14,304 or yohimbine,
and 0.02 mmol/L NE after UK 14,304 or yohimbine was applied on P55 (Figure 7.7a and
b), T650 (Figure 7.7c and d) and diamond (Figure 7.7e and f) microelectrodes. These data
show that the oxidation current peak of NE was reduced when UK 14,304 or yohimbine
was present. It seems that UK 14,304 and yohimbine can be reversibly adsorbed on the
surface of the electrode and occupied the active sites because the oxidation current peak
of NE can be reversed after the drug is washed out.

The statistic data shown in Figure 7.8a and b indicate that at the presence of UK
14,304 and yohimbine, NE could be determined in most situations but the current was
signiﬁcantly decreased. Figure 7.8c and (1 show that the effects of UK 14,304 and
yohimbine on the determination of NE at three types of microelectrodes were reversed to

some extent after the drug was washed out.

254

     
 
 
  
   

 

 

 

    

 

   
   

a
30.0 ~ NE with UK 14.304
2 20.0 ~ NE after UK 14,304 E,
5 E
.. o
E 10.0 « g
3 i O
0
0.0 ‘
0.0 7 0:3 . 0:6 ' ois ' 11:
Potential (V vs. AglAgCl)
60.0 "i C
NE with UK 14,304 A
i 40.0 - \ g
r - i=3
: NE after UK 14,304
g 20.0 - 5
3 O
O
0.0 -l NE
o.o - 013 i ofe ' 0:9 ' :2
Potential (V vs. AglAgCl)
‘ 9
120.0-
NE with UK 14,304
7g 80.0. g
: NE after UK 14,304 I:
a a
g 40.0 - g
0 O
0.0.

 

 

0.0 . 0:3 ' oie ' 0:9 . 1:2
Potential (V vs. AglAgCl)

 

 

 

 

 
 
  
 
 
    
  

  
 

 

 
  
   

20.0 ‘ NE with Yohimbine
10.0‘
0.0-
NE afterYoh bins
010 ' 013 f 016 F 019 ' 1:2
Potential (V vs. AglAgCl)
‘ d
NE with Yohimbine
20.0-
‘ NE after Yohimbine
NE
10.0‘
1 /
0.0. /
oio ' 0:3 . tire - 0:9 . 1:2
Potential (V vs. AglAgCl)
80.0] .f
J
60.0 NE with Yohimbine
40. 0 NE after Yohimbine
20.0
0.0‘

 

 

 

j,

010 . 0:3 ' 010 0:9 ' 1:2
Potential (V vs. AglAgCi)

Figure 7.7 Cyclic voltammetric i-E curves for NE at P555 (a), T650 (c) carbon ﬁber
electrode and diamond microelectrodes (6) with and without UK 14,304. Cyclic
voltammetric i-E curves for NE at P55 (b), T650 (d) carbon ﬁber electrode and diamond

microelectrodes (t) with and without Yohimbine. All curves were acquired in Krebs’

buffer at 0.1 V/s. The concentration of NE was 0.02 mmol/L and the concentration of the

drug was 0.1 mmol/L.

255

 

  
 

 

  
     

 

 

 

 

25° a - P555 b P553
- T850
200~ - Diamond
S~ 150- V :5
E :2
ﬁ 100- ‘tn
"‘1' E
'5
< 50- g
o l-N
.0—
UK 14,304 Yohimbine UK 14,304 Yohimbine
c — p’sss ' l
0' d - T650
-7l:liamond
.10-
s ”*3
.2oJ 3:
E 3
q - P558 2
‘- _ - T650 3
”.3 "30‘ - Diamond g
.9
40- _
UK 14,304 Yohimbine UK 14,304 Yohimbine

 

Figure 7.8 Statistical results (n = 3) for the effects of UK 14,304 and yohimbine. (a)
AEm before and with drug on P55, T650 and diamond electrodes for NE. (b) Peak
current changes before and after drug on NB detection using three types of
microelectrode. (c) AEm before and after drug was applied for NE. ((1) Peak current
changes before and after drug was applied for NE detection using three types of

microelectrode.

Eﬁ’ects of Cocaine, Idazoxan, and PPADS. As shown in Figure 7.4, idazoxan and
PPADS are electrochemically inactive on carbon ﬁber and diamond electrodes within the
voltage range studied. PPADS is electrochemically-inactive at carbon ﬁber but has an
oxidation peak around 805 mV vs. AglAgCl at diamond. This may be due to the different

adsorption ability of PPADS to three types of microelectrode surface. We anticipate that

256

these three drugs should not aﬁ‘ect the determination of NE too much, at least for carbon
ﬁber microelectrode. Our measurements conﬁrmed this. Figure 7.9 shows the cyclic
voltammetric i-E curves for 0.02 mM NE, 0.02 mM NE + 0.1 mM idazoxan or PPADS
and 0.02 mM NE after idazoxan or PPADS was applied on P558 (Figure 7.93 and b),
T650 (Figure 7.9c and d) and BDD (Figure 7.9e and t) electrodes. As expected, idazoxan
and PPADS show little effects on the oxidation current peak of NE on all electrodes.
Figure 7.9 also shows that the cyclic voltammetric i-E curves of NE, NE with and after
these drugs are not exactly identical, therefore, calibrations should be performed before

and after the drug is applied for the determination of NE.

Cocaine is also electrochemically inactive on carbon ﬁber and diamond electrodes
as shown in Figure 7 .4d, e and f. Different from idazoxan and PPADS, cocaine activated
all electrodes in our experiments. The oxidation current peak of NE increased slightly
after cocaine was applied as shown in Fig 7.10 (except for BDD microelectrode). This
activation effect was also conﬁrmed when probed by Fe(CN)63'/4’ and Ru(NH3)63+’2+ as
summarized in Table 7.4. Both of the oxidation peak current Fe(CN)63'/4' and the
reduction peak current of Ru(NH3)63+’2+ increased slightly after cocaine was applied. It’s
noteworthy that these increases are not signiﬁcant when checked with AN OVA.

Figure 7.10 also shows that the cyclic voltammetric i-E curves of NE, NE with
and after cocaine are not exactly identical, therefore, calibrations should be performed

before and after the drug is applied for the determination of NE.

257

 

Table 7.4

properties of carbon

Effects

of Idazoxan, PPADS and Cocaine on the electrochemical

ﬁber and BDD microelectrodes probed by Fe(CN)63'/4' and

 

 

 

 

 

 

 

 

 

 

 

 

Ru(NH3k3+/2+-
P558 T650 Diamond
Drug Probe*
AEm iratio AEm iratio AEm iratio
(mV) (mV) (mV)

F 29.5 :1: 0.89 :1: 10.8 :i: 0.93 i 79.0 :t 1.00 :t

I dazoxan 26.4 0.06 7.2 0.04 4.6 0.08
R -3.8 d: 1.00 :1: 1.3 :i: 1.00 :t -4.3 :t 0.97 :i:

3.7 0.03 0.6 0.01 4.3 0.02
F 6.3 d: 0.96 :l: 7.7 :1: 0.88 :l: 8.3 i 0.98 :1:

9.4 0.05 2.7 0.03 16.2 0.04
PPADS R 0.3 :1: 1.02 :1: -0.7 :1: 0.99 :i: 7.0 :t 1.05 :t

1.7 0.02 14 0.03 3.8 0.03
F -15.3 :b 1.07 i 4.3 :l: 1.08 :I: 16.5 :1: 1.03 :1:

Cocaine 9.2 0.1 1 3.5 0.05 2.5 0.05
R 6.8 :1: 1.05 :i: 0.6 :i: 1.03 i -5.7 :1: 1.01 :i;

4.8 0.05 0.9 0.01 27 0.01

 

 

 

 

 

F: Fe(CN)63"4'

R: Rum3)63+/2+

258

 

       

 

    

 

 

 

 

     
      
 

 

 

 

 

 
 
  
   

 

 

 

 

   

 

a b
20.0.
NE with Idazoxan
10.0 .
:3 10.0 NE , g
C
:3 l / E '
c / m
m t
t 3
a 1 o .
0 0'0 ' NE atterldazoxan 0'0 X
NE with PADS
oio ' 0:3 ' oie - 0.3 ' 112 oio ' 0:3 ' 0:6 - 01.9 V 11:
Potential (V vs. AglAgCl) Potential (V vs. AglAgCl)
c 10.0- d NE
NE with ldazoxan
2‘ NE after ldazoxan A
5100 ‘E 5.0:
E NE :E”
‘1’ a:
S t 1
3
0 o
0.0 - 00‘ NE after PPADS
NE with PPADS
oio ' 0:3 ' oie - 0:9 ' 112 0.0 ' oi: ' ofo ' oft F 1f2
Potential (v vs. AglAgCl) Potential (V vs- AglAgCl)
60.0: e 90'01 f
NE with Idazoxan . ‘ NE with PPADS \
A 400‘ €600 .
E, 35:: NE afterPPADS
g 2
g 20.0i a 30.0
0
0,0 - NE after ldazoxan 0 O-
oio * 0:3 ' oie ' 0:9 ' 1.3 0.0 ' C3 7 016 ' oi; ' 1:2
Potential (V VS- AQIAQC') Potential (V vs. AglAgCl)

Figure 7.9 Cyclic voltammetric i-E curves for norepinephrine at P55 (a), T650 (c)
carbon ﬁber and diamond microelectrodes (c) with and without idazoxan. Cyclic
voltammetric i-E curves for norepinephrine at P55 (b), T650 (d) carbon ﬁber and
diamond microelectrodes (f) with and without PPADS. All curves were acquired in
Krebs’ buffer at 0.1 V/s. The concentration of norepinephrine was 0.02 mmol/L and the

concentration of each drug was 0.1 mmol/L.

259

   
   

 

 

 

 

 

 
     
    

a
20.0 ‘
NE with Cocaine
2; NE after Cocain
T: 10.0“
C
93
‘5
o
0.0 ‘
NE
0:0 0:3 oie 0:9 1:2
Potential (V vs. AglAgCi)
40.0 ' b
30.0 «
NE with Cocaine
2:; 20.0 - NE after Cocaln
E
92
5 10.0-
o
0.0 7
010 ' 013 f oie ' 0:9 1 1:2
Potential (V vs. AglAgCl)
‘ c
20'0 ‘ NE with Cocaine\
3?
5
‘5 10.0 -
93
‘5
o
0.0 . NE after Cocain

 

 

f r

0.0 0:3 016 ' 0:9 ' 11:
Potential (V vs. AglAgCl)

Figure 7.10 Cyclic voltammetric i-E curves for norepinephrine at P55 (a), T650 (b)

carbon ﬁber and diamond microelectrodes (c) with and without cocaine. All curves were

acquired in Krebs’ buffer at 0.1 V/s. The concentration of norepinephrine was 0.02

mmol/L and the concentration of each drug was 0.1 mmol/L.

260

7. 1.3 Discussion

As shown in Table 7.1, the chemical structure of the neuropharmacological agents
can be divided into three categories: 1) the ——NH2 group in prazosin, —OH group in
capsaicin, and —OH group in yohimbine make these drugs oxidizable; 2) although PPADS
also have —OH group in a pyridine structure, N=N group which connects with this
pyridine make —OH group really difﬁcult to be oxidized; and the — NH group in ﬁve
member ring both in UK 14,304 and idazoxan make these two drugs also not easy to be
oxidized; 3) there is no function group in cocaine which could be oxidized. Since these
drugs may possess different electrochemistry property, we predict that they may have
different effects on the responsivity of the microelectrode when used in biological
environment. Normally, in the electrochemical measurement of NE in vitro, the electrode
used should possess good biocompatibility, good sensitivity, low limit of detection, stable
and low background current which is not affected by the changes in surrounding solution
and resistance to the adsorption, either by the interested redox species and their products
or the components of the background solution (e.g. Krebs’ buffer) 22’ 34’49’50. However,
when there are various redox system in the biological buffer (e. g. drug application), there
are changes, more or less, to the sensitivity of microelectrode. The concern that if and
how these commonly used drugs affect the NE oxidation current measurement and their
effect on the microelectrode sensitivity drives us doing this study.

Our results show that according to the effect on the determination of NE on
carbon ﬁber and boron-doped diamond microelectrodes, the seven drugs can be classiﬁed
into 3 categories: 1) capsaicin and prazosin, which voltammagrams immerge the

oxidation current peak of NE; 2) UK 14,304 and yohimbine, which voltammagrarns

261

partially overlap the oxidation current peak of NE; 3) Cocaine, Idazoxin, and PPADS
which have no signiﬁcant effect on the determination of NE.

Capsaicin and prazosin can be oxidized, as shown in Figure 7.4. At the presence
of capsaicin, the oxidation current peak of NE could not be distinguished except on T650
carbon ﬁber microelectrode, on which the oxidation current peak of NE could be seen but
signiﬁcantly decreased (Figure 7.5a, c and e). But after drug washing out, the effects of
capsaicin on three types of microelectrode were reversed (Figure 7.6a and b). One reason
for this might be capsaicin has a phenol structure, which is really similar to the structure
of NB. This makes capsaicin behave similarly as NE and has similar Em value (Table
7.2). In this case, the competition of being oxidized for these two species depends largely
on their concentration. In our experiment, we tried to mimic the situation in in vitro NE
measurement, the ratio of capsaicin and NE is around 5, which suggest more capsaicin is
prone to be oxidized on the surface of microelectrode. Thus, there would be a great
interference on NB measurement at the presence of capsaicin and our data proved this.
Figure 7.5b, d and f show that the oxidation current peak of NE could neither be
distinguished at the presence of prazosin. However, we think the reason for this is not
exactly the same as what we saw from capsaicin. Because after drug washing out, the
inhibitory effect of prazosin on all 3 electrodes were difficult to be reversed (Figure 7.6c
and d) and the response to NE was also diminished (Figure 7.5b, d and f) except for
T65 0.

The mechanism of how these drugs affect the NE detection is not clear. However,
we think one potential reason is the adsorbed drug occupied the activation points on the

surface of the microelectrode which inﬂuenced the efficacy of NE detection. The

262

recovered currents after capsaicin washed out are consistent with this hypothesis.
McCreery and DuVall have shown that the speciﬁc molecular features present at the
carbon surface plays an important role in determining the kinetics of heterogeneous
reactions. The adsorption may involve a distribution of interaction strengths between
each type of carbon surface site and adsorbed analyte species 51’52. While there is only
NE in the solution, NE bind with those activation sites, and reodx reaction happens at the
surface of the electrode. However, at the presence of adsorbing electroactive drugs, the
efficacy of NE being oxidized depends on the adsorption competition between the drug
and NE. If the adsorption ability of the drug to the electrode surface is comparable with
NE, then at the presence of this drug, how much NE would be oxidized depends on the
kinetics, which is related with the concentration of both drug and NE; and after drug
washing out, the response to NE could recover, e.g. capsaicin. However, if the drug could
be strongly adsorbed on the surface of the electrode, then NE has less chance to bind with
activation sites. As a consequence, once equilibrium is established, the strongly held
drugs remain on the surface, and only the weaker adsorbing carbon surface sites remain
exposed for NE adsorption, and thus there is a right shift NE Em value and a decrease of
current amplitude. To the maximum, the NE current is decreased to the extent that it is
totally merged and could not be distinguished from the drug peak, 6. g. prazosin. And
what’s more, since the adsorption to the electrode are so strong, even aﬁer washing out,
some drug species are still adsorbed on the electrode surface. As a consequence, the
response to NE could not get back, no matter for E1 /2 or oxidation current.

In order to verify our hypothesis, we also used two redox systems, Ru(NH3)63+’2+

263

and Fe(CN)63'/4' to characterize the drug effect on the microelectrode sensitivity which
has a great inﬂuence on NB measurement. That is because how the chemical and physical
properties affect the electrochemical response of sp2 (e.g. carbon ﬁber) and sp3 (e.g.
diamond) carbon electrodes depends to some extent on the electron transfer mechanism
for the speciﬁc redox system 5 1'53. Ru(NH3)63+/2+ is an outer-sphere redox systems on both
carbon ﬁber and diamond with heterogeneous electron-transfer rate constant depend
primarily on the density of electronic states and is less affected by the physical and
chemical properties of the electrode surface 43’44’51'53. F e(CN)63 " 4’ is a more complicated,
inner sphere redox system, with a heterogeneous electron-transfer rate constant highly
sensitive to the surface cleanliness, surface oxygen coverage and microstructure, as well
as the density of electronic states 3" 43’ 44’ 51. Thus, these two redox systems could
characterize both the change of density of electronic states and surface cleanliness of the
microelectrode under the effect of drugs.

After capsaicin was washed out, both the E1 /2 and oxidation current recovered
suggest that neither the electronic structure nor the surface of the electrode had been
changed aﬁer drug application. So the adsorption of capsaicin on the electrode surface
was reversible and the interference in NE detection at its presence was mainly caused by
the competition between these two species. While after prazosin washed out, there was
only a current decrease for Fe(CN)63"4’ but not for Ru(NH3)53+/2+ on T650
microelectrode indicated that prazosin molecules only bind and contaminate the surface
of the electrode but didn’t change electrode electronic structure. However, the current
signiﬁcantly decreased for both Ru(NH3)63+/2+ and F e(CN)63 44' comparing to original at

P55 carbon ﬁber and diamond miecroelectrode demonstrated that prazosin molecules

264

were not only physically adsorbed on the electrode , but also changed the electrode
electronic density of states. This process was not reversible.

The differences of the inﬂuence on the NE detection with the same drug among
these three types of microelectrode are mostly caused by their different microstructures.
Carbon ﬁber surface is polar and hydrophilic due to its sp2 bonded carbon structure with
a n-electron system and also a variety of oxygen functional groups. As a consequence,
these properties promote the adsorption of polar molecules by hydrogen bonding,
dipole-dipole and ion-dipole interactions 52’ 53. As shown in Figure 7.3, T650 carbon ﬁber
possesses a higher ID/IG ratio suggesting a more disordered structure, which was
supposed to have stronger electroactivity. This may explain why the E1 /2 value for NE at
T650 carbon ﬁber microelectrode is lower comparing with P55s carbon ﬁber
microelectrodes (Table 7.2). And also because of the disordered structure, the adsorption
of drugs to the electrode surface is not that stable, which may be the reason of why T650
carbon ﬁber microelectrode could recover better from drug application (Figure 7 .6c and
d). Sp3 bonded carbon microstructure and little n-electron density make diamond more
stable and not prone to be adsorbed, especially by catecholamines, e.g. NE 30’33’54. The
Em value for NE is around 150 mV more positive for diamond than for carbon ﬁber due
to the more sluggish electrode reaction kinetics, which might be caused by the weak NE
adsorption on the diamond surface. However, the By; value for each drug is comparable
at these three types of microelectrodes (Table 7.2). And our data suggest that drug species
we studies here are easier to be adsorbed on diamond comparing with NE and that is
maybe why applied drug inﬂuence diamond microelectrode more comparing with T 650

carbon ﬁber microelectrode (Figure 7.60 and d).

265

Both UK 14,304 and yohimbine have a —- NH (or —- OH ) group within and out of
a ﬁve member ring which could be oxidized (Table 7.1). So we predicted that these two
drugs would have some effects on NB detection and our results veriﬁed this. The data
shown in Figure 7.7 and Figure 7.8 suggested that the adsorption ability for both UK
14,304 and yohimbine are comparable with NE at the surface of diamond and T650
carbon ﬁber microelectrodes. So NE could be determined at the presence of these drugs
although the oxidation current peak of NE was ﬂattened and reduced. This is not true
with P55s carbon ﬁber microelectrode. The almost non-distinguishable NE current peak
at the presence of UK 14,304 and yohimbine indicate the adsorption to the electrode
surface for these two drugs are stronger than that for NE. It seems that UK 14,304 and
yohimbine could be reversibly adsorbed on the surface of the electrode and occupy the
active sites because the oxidation current peak of NE can be reversed aﬁer the drug was
washed out, although there was still current decreases (Figure 7 .8). Data from

Ru(NI-I3)63+’2+ and F e(CN)63 ’/ 4’ are consistent with this conclusion (not shown here).

Cocaine and idazoxan have very stable structures which make them almost
non-electroactive, at least in the potential range we employed (Figure 7.4). Although
PPADS has a —OH group which could be oxidized, it needs relatively high voltage, where
its oxidation peak may have negligible effect on the oxidation peak of NE. In this case,
the adsorption to electrode competition between drugs and NE are mild and NE is prone
to adsorb ﬁrst on the electrode surface and undergo the redox reaction at the detection
potential area. As expected, idazoxan and PPADS showed little effects on the oxidation

current peak of NE on all electrodes no matter with or after their present (Table 7.4).

266

 

And the fouling effect was much weaker as the consequence of the less efﬁcient binding
between drug and microelectrode. Data from Ru(NH3)¢53+’2+ and Fe(CN)63'/4' are consistent
with this conclusion (not shown here). A little different ﬁ'om idazoxan and PPADS,
cocaine activated all electrodes in our experiments although not all of the statistic data
mean signiﬁcantly different (Table 7.4). The reason why cocaine may potentially activate
the electrode might be that it could increase the activation sites while binding with the
electrode surface or there are high electron density points in cocaine molecule which
could make the electron transfer between NE and electrode more efﬁciently.

To conclude, we have found capsaisin and prazosin have severe effect on the
measurement of NE at the surface of both carbon ﬁber and diamond microelectrode at
their presence; UK 14,304 and yohimbine could reduce the response of NE to the
electrode, but washing out will reverse this effect; cocaine, idazoxan and PPADS do not
inﬂuence the detection of NE. These data suggest that, in voltammetric and amporometric
experiments, consideration should be make wherever possible a drug is electroactive.
Direct effect of a drug on electrode response sensitivity should be detemiined before
measuring NE level. Therefore, at the similar situation, it may be more secure to use the
drug which is not electroactive or at least electrochemically inactive with the potential
range of NE detection. Calibrations should be performed before and after that drug is

applied for the precise determination of NE level.

7.2 Distribution of sympathetic and sensory nerves

7. 2. I Glyoxylic acid ﬂuorescence histoch em istiy

267

 

Glyoxylic acid (GA) is a reagent for the ﬂuorescent histochemical visualization of
biogenic monoamines.55 GA reacts with indolylethylamines and catecholamines to
produce a ﬂuorescence product that can be used for histochemical visualization. NE can
be converted into the intensely ﬂuorescent 2-carboxymethyl-dihydroisoquinoline
derivative by GA,55 as shown in Figure 7.10. Therefore, the distribution of GA-induced
ﬂuorescence of neuronal stores of NE was used to assess the distribution of periarterial

and perivenous sympathetic nerves.

H o

+ Hooccno—%>
./\ILHZ H / N
HjY//

Hal COOH
N/E——coou <— + HOOCCHO
\ are
OH

 
  

COOH coon
2 COOH
M:$5 __— "‘<;" -
9 4:55 (ﬁtCOCi
_ C02
C00

Figure 7.11 Reaction mechanism for GA and norepinephrine. The circled groups

represent the ﬂuorophore in the molecule.

Figure 7.12 shows GA-induced ﬂuorescence images of mesenteric arteries (MA)

268

and veins (MV), which reveal a network of varicose nerve ﬁbers. The periarterial nerve
plexus consists of bundles of several axons arranged in a net-like manner.4’ 56 For a Sham
or nonnotensive rat, the nerve plexus for MA consists of bundles of axons arranged in
network-like distribution with nerve ﬁbers equally likely to pass parallel or perpendicular
to the longitudinal axis of the vessel. The perivenous plexus is less dense than in arteries
and consists of individual varicose axons largely oriented circumferentially around the
vessel. This does not necessarily mean that the density of innervation of MV is less than
that of MA. MV is thinner than MA so the number of axons per smooth muscle cell could
be similar. This suggestion is supported by the fact that ultrastructural studies have shown
that in guinea pigs, the density of neuromuscular junctions is similar for MA and MV.57
Similarly, for DOCA-salt rat, the nerve plexus for MA consists of bundles of axons
arranged in a mesh-like network with nerve ﬁbers parallel or perpendicular to the
longitudinal axis of the vessel. On the other hand, the plexus for MV consists of single
axons with more of a circumferential arrangement and few axons are localized along the

longitudinal axis of the blood vessel.

269

 

Figure 7.12 Glyoxylic acid-induced ﬂuorescence images of catecholamines in the
perivascular sympathetic nerves of Sham MA (A), Sham MV (B), DOCA-salt MA (C)
and DOCA-salt MV (D). The nerve plexus in arteries has a mesh—like arrangement with
ﬁbers oriented along both the longitudinal and circular axes of the blood vessel. The

plexus in veins has a circular arrangement with very few ﬁbers oriented longitudinally.

7.2.2 ' ' ' ‘ ' ' ‘ y

Glyoxylic acid-induced ﬂuorescence imaging can be used to determine the
distribution of sympathetic nerves but not sensory nerves. A different procedure was
needed to visualize the sensory nerves arrangement along arteries and veins. This is

described below.

270

Tyrosine hydroxylase (TH) is the ﬁrst enzyme involved in catecholamine (CA)
biosynthesis. It catalyses the conversion of L—tyrosine to L-3,4-dihydroxyphenylalanine
(L-DOPA). Tyrosine hydroxylase is, therefore, a useful marker of all CA—containing
neurons.57’ 58 An Immunohistochemical imaging method was used that involved bonding
an anti-TH antibody with TH. The resulting complex was then reacted with a secondary
antibody conjugated with a special ﬂuorescent dye. The TH complex becomes
ﬂuorophoric allowing one to spatially image the ﬂuorescence locations by microscopy.

In a similar way, the occurrence of substance P (SP) and calcitonin gene-related
peptide (CGRP) immunoreactive varicose ﬁbers were localized by double
immunohistochemistry using anti-substance P and anti-CGRP antibody conjugated with
different ﬂuorescent dyes, respectively.

Theoretically, because of the high selectivity of antigen-antibody reaction, the
distribution of TH, SP and CGRP containing nerve ﬁbers can be localized in the same
preparation, respectively, by choosing different ﬂuorescent dyes and excitation
wavelengths. In this study, we used three different secondary antibodies with two
different ﬂuorescent dyes to localize the distribution of TH-containing nerve ﬁbers
(sympathetic nerves) and SP- or CGRP-containing nerve ﬁbers (sensory nerves) around
the mesenteric blood vessels of rats. The mechanisms behind these measurements are
diagramed in Figure 7.13. We were speciﬁcally interested in the arrangement of TH-, SP-
and CGRP-containing nerve ﬁbers associated with mesenteric arteries and veins from

Sham and DOCA-salt rats.

271

 

Sym oathetic Nerve

   

Blood Vessel
‘ v V ' V '
Red Fluorescence Green Fluorescence Red Fluorescence

‘

CGRP-IgG-
TH—lgG-GtotRat-Cy3

GtaMouse-FITC GtaRabbit-Cy3

Figure 7.13 Schematic of immunohistochemical staining of TH, SP and CGRP. In the
sympathetic nerve varicosity, anti-TH antibody (primary antibody) binds to TH, followed
by binding of a secondary antibody with a ﬂuorescent dye, which generated ﬂuorophoric
complex. In the sensory nerve varicosity, anti-SP antibody binds to SP, followed by
binding of a secondary antibody with a ﬂuorescent dye, which generated ﬂuorophoric
complex. Also in the sensory nerve varicosity, anti-CGRP antibody binds to CGRP,
followed by binding of a secondary antibody with a ﬂuorescent dye, which generated
ﬂuorophoric complex. The ﬂuorescence intensity depends on the concentration of

ﬂuorescent complex at a particular excitation wavelength.

Examination of whole mount preparations from Sham rats revealed a dense

plexus of varicose nerve ﬁbers containing TH-immunoreactivity (TH-ir) was associated

272

with mesenteric arteries (Figure 7.14A and Figure 7.15A) and veins (Figure 7.14D and
Figure 7.15 D). Also, a dense plexus of varicose nerve ﬁbers containing
CGRP-immunoreactivity (CGRP-ir) was associated with mesenteric arteries (Figure
7.14B and Figure 7.15B) and veins (Figure 7.14B and Figure 7.15B). Overlay of these
images reveal that the sensory nerves have a similar net-like distribution and there is
signiﬁcant co-localization of sensory and sympathetic nerves (right column in Figure

7.14 and Figure 7.15).

 

Figure 7.14 Immunohistochemical localization of TH and CGRP in whole mount

preparations of mesenteric artery and vein from a Sham rat. (A) Localization of TH-ir in
Sham MA. (B) Localization of CGRP-ir in the same tissue. While there are much higher
dense TH-ir nerve ﬁbers, CGRP-ir are less dense. (C) Overlay of (A) and (B). (D)
Localization of TH-ir in Sham MV. (E) Localization of CGRP-ir in the same tissue. (F)

Overlay of (C) and (D). Scale bars in all images are 50 pm.

273

 

Figure 7.15 Immunohistochemical localization of TH and CGRP in whole mount

preparations of mesenteric artery and vein from a DOCA-salt rat. (A) Localization of
T H-ir in MA. (B) Localization of CGRP-ir in the same tissue. While there are much
higher dense TH-ir nerve ﬁbers, CGRP-ir are less dense. (C) Overlay of (A) and (B). (D)
Localization of TH-ir in MV. (E) Localization of CGRP-ir in the same tissue. (F) Overlay

of (C) and (D). Scale bars in all images are 50 pm.

Our results show that the distribution of the TH-containing nerve ﬁbers
(sympathetic nerves) for the artery is different from that for the vein, which is similar to
the GA histochemical results. There is no signiﬁcant difference in the distribution of
CGRP-containing nerve ﬁbers (sensory nerves) between arteries and veins except that the
density of CGRP in DOCA-salt MV (Figure 7.15B) seems less than that in Sham MV
(Figure 7.14E).

Figure 7.16 shows the immunohistochemical localization of TH and SP in whole

mount preparations of mesenteric vein from a DOCA rat. There are very weak

274

ﬂuorescent lines shown in Figure 7.16 indicating that rare SP can be detected in
DOCA-salt mesenteric vein. In fact, we did not get obvious SP-ir image from Sham
mesenteric blood vessels. This is consistent with the literature reports. Galligan et al.
published their immunohistochemical research on Sham and DOCA-salt mesenteric
blood vessels. They found very little SP-ir in the nerve ﬁbers in arteries or veins in

tissues from Sham rats, however, there was clear evidence that the intensity of nerve ﬁber

staining for SP was higher in veins from DOCA-salt rats.59

 

Figure 7.16 Immunohistochemical localization of TH and SP in whole mount
preparations of mesenteric vein from a DOCA-salt rat. (A) Localization of TH-ir. (B)
Localization of SP-ir in the same tissue. (C) Overlay of (A) and (B). Scale bars in all

images are 200 pm.

275

Figure 7.17 shows the immunohistochemical localization of TH and SP in whole
mount preparations of mesenteric artery from a DOCA rat. Although the dense of SP-ir is
much lower that that of CGRP-ir, the overlay image clearly shows that the SP-containing

varicosities are exactly consistent with those CGRP containing varicosities, which

demonstrates SP and CGRP exist in the same nerve ﬁbers (Figure 7.17C).

 

Figure 7.17 Immunohistochemical localization of SP and CGRP in whole mount
preparations of DOCA-salt mesenteric arteries. (A) Localization of SP-ir. (B)
Localization of CGRP-ir in the same tissue. While there are numerous CGRP-ir nerve
ﬁbers, SP-ir is rare. (C) Overlay of (A) and (B). Note the yellow lines which are
combinations of the red lines (SP-ir) in (A) and green lines (CGRP-it) in (B). Scale bars

in all images are 50 pm.

276

7. 3 Vasoconstriction and norepinephrine oxidation current at
MA and M V

7. 3.1 Determination of norepinephrine

As we mentioned before, one of the speciﬁc aims in our group’s research is to
determine if there is local communication between sensory and sympathetic nerves at
mesenteric blood vessels of rats when the connections ﬁ'om both sensory and sympathetic
nerves to the CNS were severed. Researches have found that repetitive activation of
afferent nerve ﬁbers can cause cross-excitation of neighbors not stimulated directly.9'" It
was hypothesized that local communication may exist between sensory and sympathetic
nerves using the SP or CGRP neurotransmitters released by sensory nerves. Therefore, if
release of SP and CGRP is inhibited then the response of mesenteric blood vessels to
stimulations may change and this change may be recovered by exogenous SP application.
The effects of substance P and CGRP on NE release from sympathetic nerves were
assessed using continuous amperometry with a diamond microelectrode.

Norepinephrine (NE) functions as both a hormone and neurotransmitter. It is
secreted by chromafﬁn cells produced in the adrenal medulla and at the sympathetic
nerve endings. It is a vasoconstrictor and increases in heart rate, blood pressure, and the
sugar level in the blood.1 NE levels has been measured in tissue and biological ﬂuids
because of its involvement in the stress related-diseases, such as hypertension, depression
and cardiovascular disease.“62 Previous measurements of NE in plasma have been
performed using high performance liquid chromatography (HPLC)63’ 64 or radioactive
labeling methods.“’ 66 These overﬂow techniques provide little information about

real-time local release of NE from sympathetic nerves. NE can be oxidized via a 2H+/2e

277

redox reaction, as shown below.

OH on
H0 NH; 0 NH:
O _..._. 0 + 2H+ + 2c-
H0 0 '
NE NE-o-quinone

In other words, norepinephrine is electrochemically active and can be detected
electrochemically. To date, electrochemical measurements using fast scan cyclic
voltammetry and continuous amperometry with carbon ﬁber microelectrodes have been
widely used in vitro to study the release and clearance mechanisms for NE in the
sympathetic nervous system because of the high spatial and temporal resolution this
technique provides.”70 Previous work in our group demonstrated that boron-doped
diamond microelectrodes could be used for sensitive, reproducible and stable
electrochemical determination of NE.”76 In this work, all measurements of NE were

performed using 76 um Pt wire microelectrodes coated with boron-doped diamond.

7.3.2 Sensory nerve denervation by capsaicin

A. Hydrodynamic and calibration curves for NE determination

To determine an optimal detection potential for the determination of NE,
hydrodynamic i-E curves for NE, substance P (SP) and capsaicin at a BDD
microelectrode were ﬁrst recorded, as shown in Figure 7 .18. The curves were generated
by measuring the current of a BDD microelectrode in 0.1 mmol/L analyte (NE, SP or

capsaicin) in Krebs’ buffer with a 50 mV potential step starting from 0 V vs. Ag/AgCl.

278

 

 

 

 

 

 

 

1
1.2x10'° -
—I- NE
. .. -0- Capsaicin
1.0x10 -11.. Sub P
A 8.0x10‘ .
s 4
E 6.0x10" —
e .
'5 4 0x10" -
0 ' .
2.0x10" -
1 __ _. .A—4
0.0 u .A- .g—n-l- -A—nA-‘A ‘ A
F u l r l ' i ' r m i
0.0 0.2 0.4 0.6 0.8 1.0

Potential (V vs. AglAgCl)

Figure 7.18 Hydrodynamic i-E curves for NE, SP and capsaicin at a BDD
microelectrode. The measurements were performed in 0.1 mmol/L NE, SP or capsaicin in

Krebs’ buffer with a 50 mV potential step starting from 0 V vs. AglAgCl, respectively.

Figure 7.18 shows that SP is not electroactive and has no inﬂuence on the
determination of NE. Capsaicin, on the other hand, seriously affect the determination of
NE as we discussed in Section 7.1.2. Capsaicin is an agonist of vanilloid receptors
localized at sensory (SP/CGRP-containing) nerve endings and activation of these
receptors leads to SP and CGRP release.77 Perﬁising capsaicin for a long period (usually
30 min) will deplete the ﬁnite stores of SP and CGRP in sensory nerves, a chemical

denervation. After that, capsaicin can be washed out.

279

To quantitatively determine NE, all BDD microelectrode should be calibrated

before used. A representative calibration curve is shown in Figure 7.19.

 

 

60
l. ./
m d r
//
/
404 [/2/
A x /
a //
2’ 30 - I 4/
/.//
///9//‘/ Applied potential = 850 mV
20 - /// ’ vs. Ag/AgCl
- ,1 Electrode area is not calibrated
/
10 r l I l U I t r ' l ' i I I ﬁ
10 15 20 25 30 35
ONE ("W

Figure 7.19 Calibration curve for NE oxidation at a BDD microelectrode. The oxidation
currents were measured amperometrically by perfusing Krebs’ buffer with different NE

concentrations.

B. vasoconstriction and NE overﬂow from Sham and DOCA-salt mesenteric blood
vessels

We studied the relationship between sensory and sympathetic nerves of
mesenteric blood vessels following the protocol given in Section 2.8.2E. 14 Sham rats
and 15 DOCA-salt rats were examined. Figure 7.20 shows typical experimental results
recorded from a Sham mesenteric artery. The results demonstrate that the blood vessel

constriction (top curves) did not signiﬁcantly change after capsaicin treatment and wash

280

out (Figure 7.20, top panel in column B). The artery constricted reproducibly by ca. 20%.
Substance P did not inﬂuence sympathetic nerve activity either (Figure 7.20, top panel in
column C). This was conﬁrmed by the current traces as the oxidation current did not

signiﬁcantly change in all cases.

   

 

  
 

 

 

 

 

 

       

 

 

 

 

A B C D
E 320
v 300 i ' i
5 i ' i
*3 no ' : i
e i - i
3 no i : :
i
D 240 : : :
. I g ‘
r too 2603 160 Too I; 160 Foo F too zoo
: i . i
I
: : I i
. i : I
: : - :
240.: ' : '
A | I | :
i
1‘: i I ' i
:I: . I : I
g I l ' I
g : : . :
a230‘ ' : :
o I
- 1 i ' i
o 100 200 a too 260 F 160 Too 6 too foo
Time (s)

Figure 7.20 Typical vasoconstriction and NE overﬂow results recorded for a Sham
mesenteric artery. The top panels show the blood vessel constriction and the bottom
panels show the recorded NE oxidation current aﬁer electrical stimulation. The dash lines
indicate when the electrical stimulus was applied. Column A shows the diameter and
current traces for the blood vessel control. Column B shows the diameter and current
traces after 1 umol/L capsaicin was applied (30 nrin) and washed out (perfused with
Krebs’ solution for 1 h). Column C shows the diameter and current traces during
perfusion of 0.4 umol/L SP. Column D shows the diameter and current traces after SP

wash out. The time axis and baseline for each current trace were adjusted for comparison.

281

 

Similar results were obtained on DOCA-salt mesenteric artery, as shown in Figure
7 .21. The only difference is that we observed that capsaicin obviously dilated the blood

vessel (Figure 7.21, top panel in colmnn B).

N
C
O

   
 

§_.

N

N

L .°
--------------

Diameter (um)

 

 

q——-—---

 

 

too :60 too :60 too 26o too zoo

(ME

   
 

Current

.3

.a

O
L

   

‘- —---—--————-91-

_- ----------

 

 

°‘__- --___-___-__----.PT-__--_-...._.

too 260

01

too 260 o too 260
Time (s)

0 {Bo 26o

Figure 7.21 Typical results recorded from a DOCA mesenteric artery. The top panels
show the blood vessel constriction and the bottom panels show the recorded NE
oxidation current after electrical stimulation. The dash lines indicate when the electrical
stimulus was applied. Column A shows the diameter and current traces for control.
Column B shows the diameter and current traces after 1 umol/L capsaicin was applied
(30 min) and washed out (perfused with Krebs’ solution for 1 hour). Column C shows the
diameter and current traces during perfusion of 0.4 umol/L SP. Column D shows the

diameter and current traces after SP wash out. The time axis and baseline for each current

trace were adjusted for comparison.

282

Figure 7.22 and 7.23 show results for Sham and DOCA mesenteric veins.
Different from the results for the arteries, the diameters of both Sham and DOCA veins
signiﬁcantly increased after capsaicin was applied. On the other hand, when Substance P
was perfused, both tissues constricted strongly. The effects of capsaicin and substance P

were not obviously evident on the oxidation current of NE for these two tissues.

 

 

     

A zoo A B C D
E 240 : '
I. l :
g 220 : :
a m i I :
D 100 : :
I . .
(l lo (o b Jo a do «'7 in no
1 l :
I I I
I I .
I I I
250-: i :
A ' : I
g. : I i
v I I I
H I | |
C " ' 'P/\/v———w——\
2 . ' '
a . I
o l l :
230 1 ' '
0 4'0 3'0 6 in so a in so 6 Jo a'o

 

 

Time (s)
Figure 7.22 Typical results for a Sham mesenteric vein. The top panels show the blood
vessel constriction and the bottom panels show the recorded NE oxidation current after
electrical stimulation. The dash lines indicate when the electrical stimulus was applied.
Column A shows the diameter and current traces for the control. Column B shows the
case after 1 umol/L capsaicin was applied and washed out. Column C shows the case
when 0.4 umol/L Substance P was perfused. Column D shows the cases after Substance

P was washed out. The time axis and baseline current for each trace were adjusted for

comparison.

283

 

 

 

        

 

 

 

A B C D
i Tl 1 r‘
V 300% I I
s . i : . :
s 2...: i g : : ‘
§=°E : : :
° 2.... : : : -
0: 4'0 so (i 4'0 in d 4'0 so u so so
I I I I
I I I I
I I I I
I I I I
I I I I
I I I I
I I I I
I I I I
2 ‘ | I I I
360 : : : :
e-I I I I I
5 ' i i I
t ' I I
555+ : i
I I I
‘ . . .g r . )7 . . FL .
0 40 to o so oo 0 40 so o 40 no

Time (s)
Figure 7.23 Typical results for a DOCA-salt mesenteric vein. The top panels show the
blood vessel diameter and the bottom panels show the recorded NE oxidation current
after electrical stimulation. The dash lines indicate when the electrical stimulus was
applied. Column A shows the diameter and current traces for the control. Column B
shows the case after 1 umol/L capsaicin was applied and washed out. Column C shows
the case when 0.4 umol/L Substance P was perfused. Column D shows the cases after

substance P was washed out. The time axis and baseline current for each trace were

adjusted for comparison.

C. Statistical analysis of the results

The statistical analysis of the results for mesenteric arteries from 10 Sham rats
and 11 DOCA-salt rats are shown in Figure 7.24. These results are the ratio of the
constriction and oxidation current after capsaicin or substance P were applied to control

values (no any drugs were applied). Our results show that after denervation of the sensory

284

nerves, the exogenous Substance P does not inﬂuence sympathetic nerve activity. There
is no difference in the vasoconstriction or oxidation current for NE after the application
of capsaicin, and with or without Substance P perfusion for both Sham and DOCA-salt
mesenteric arteries, as shown in Figure 7.24. Figure 7.24A also shows that the NE
oxidation current for Sham artery is reduced alter capsaicin and Substance P were applied.
The reason for this is not clear yet.

The statistic results for mesenteric veins from 4 Sham rats and 4 DOCA-salt
rats are shown in Figure 7.25. Similar to mesenteric artery, aﬁer capsaicin or substance P
was applied, the constrictions and oxidation currents of NE for mesenteric veins from
both Sham and DOCA-salt rats did not change too much. But different from mesenteric
artery, when substance P was applied, the constriction and NE oxidation current for
mesenteric veins from both Sham and DOCA-salt rats strongly decreased. No difference

for this change was found between Sham rats and DOCA-salt rats.

  

    

 

 

 

 

 

 

 

 

 

. [:lAfterCaps
é B E: WIthSubP
A T Aft s bP
§1zo- A g 1201 - er u
g _.' [:lAfterCaps g .......... “—
S WIIhSubP 0 ~— ‘.
3 so— - AfterSubP f2 go. ‘
O" > ‘
é , 3 i
s 0
I? 60‘ _ I! so-
:5 g
E : g i
o 304 g 30_
5 i E ‘
2 §
8 1 ﬁ' / I I
DOCA(n=11) Sham (n=10) E DOCA(n=11) MA Sham (n=1o)
MA

Figure 7.24 ' Percent change in the NE oxidation current (left) and the blood vessel

constriction (right) for mesenteric arteries from 10 Sham rats and 11 DOCA-salt rats.

285

 

 

 

 

 

 

 

 

 

 

 

 

 

 

     

 

 

 

 

 

 

 

 

 

 

 

 

 

 

A [:1 AftorCaps s? B [:1 Amman
E 120 - Izzza WithSubP : _, WItII8ubP
E ‘ - AfterSubP g 110' i - AfterSubP

4 O

= o
O
o 90. 3 oo«
3 g
g j I; 7
3 so. at," 3 g ,0 J
I: ;‘»:3’_f.,,;p- ;’;3'.’:'..' > 31'
g 0 V" . , 2
a DOCA (II-4) Sham (nu-4) g Sham ("-4)

Figure 7.25 Percent change in the NE oxidation current (left) and the blood vessel

constriction (right) for mesenteric veins from 4 Sham rats and 4 DOCA-salt rats.

7.3.3 Discussion

In DOCA-salt hypertensive rats, venomotor tone is increased.78 Venomotor tone
is a ﬁmction of the structural properties of veins and also of constrictor input to venous
smooth muscle cells.79 The primary determinant of venomotor tone is sympathetic nerve
activity and it has been shown that sympathetic input to veins is elevated in DOCA-salt78
and spontaneously hypertensive rats (SI-IRS).80 It is known that in hypertensive rats,
CGRP released from sensory nerves acts as a vasodilator to attenuate elevated blood
pressure.81 The ﬁmction and disposition of perivascular sensory nerves can be altered in
hypertension and these changes might contribute to the development of maintenance of
elevated blood pressure.82 CGRP and Substance P are two neurotransmitters released
from sensory nerves. In this study, we focused on Substance P but not CGRP. Our results
show that Substance P affects sympathetic nerve activity at mesenteric veins but not

arteries (both constriction and oxidation current decreased when Substance P

286

exogenously administrated after senserory nerves were denervated by capsaicin, Figure
7.25). Substance P is a vasodilator peptide that acts on NK-1 receptors localized to
endothelial cells to cause release of nitric oxide (NO) and other vasodilator substances.6
There is an upregulation of SP-ir in perivascular nerves in tissues from DOCA-salt rats.59
Galligan et a1. investigated the distribution of SP in sensory nerve ﬁbers around
mesenteric blood vessels and the venoconstrictor effects of SP. They concluded that SP
acts to constrict mesenteric veins via an action at NK-3 receptors.59 This conclusion is
similar to that reached by others about venoconstrictor NK-3 receptors in rat mesenteric

veins.83

The results we obtained on mesenteric veins clearly show that substance P
affects the sympathetic nerve activity and this suggests that it’s possible that there is a
local communication between sensory and sympathetic nerves. We failed to observe any
change of sympathetic nerve activity on mesenteric artery. However, this should not serve
as a conclusion that there is no communication between sensory nerves and sympathetic
nerves because previous researchers demonstrated that Substance P only contracts veins
but not arteries.59 Positive results may be observed by exogenous administration of CGRP

on mesenteric arteries because CGRP acts as a strong vosadilator on mesenteric arteries.

This need to be conﬁrmed in the future research.

287

71‘!

(1)

(2)
(3)
(4)
(5)
(6)
(7)
(8)

(9)

(10)
(11)

(12)

(13)

(14)

(15)

(16)

(17)

Iteyﬁarewrcuz

Burnstock, G.; Costa, M. Adrenergic neurons. Their organization, ﬁmction and
development in the peripheral nervous sytem; Chapman and Hall: London, 1975.

Luff, S. E.; McLachlan, E. M.; Hirst, G. D. J. Comp. Neurol. 1987, 25 7, 578-594.
Luff, S. E.; McLachlan, E. M. Blood Vessels 1989, 26, 95-106.

Furness, J. B.; Marshall, J. M. J. Physiol. 1974, 239, 75-88.

Nilsson, H. Acta Physiol. Scand. Suppl. 1985, 541, 1-34.

Brain, S. D.; Grant, A. D. Pharmacol. Rev. 2004, 84, 903-934.

Kawasaki, H.; Takasaki, K.; Saito, A.; Goto, K. Nature 1988, 335, 164-167.

Supowit, S. C.; Zhao, H.; Hallrnan, D. M.; DiPette, D. J. Hypertension 1997, 29,
945-950.

Devor, M.; Wall, P. D. J. Neurophysio. 1990, 64, 1733-1746.
Devor, M.; Dubner, R. Brain Res. 1988, 446, 396-400.
Lisney, S. J. W.; Devor, M. Brain Res. 1987, 415, 122-136.

Shoukas, A. A.; Bohlen, H. G. The American Journal of Physiology 1990, 259,
H674-680.

Anderson, E.; Sinkey, C.; Lawton, W.; Mark, A. Hypertension 1989, 14, 177-183.

Peiper, U.; Griebel, L.; W.Wende European Journal of Physiology 1971, 330,
74-89.

Tsukada, M.; Chiba, S. Clinical and Experimental Pharmacology and Physiology
2000, 27, 876—880.

Vanhoutte, P. M.; Flavahan, N. A. Federation Proceedings 1986, 45, 2347-2354.

Gonon, F.; navarre, F.; Buda, M. Analytical Chemistry 1984, 56, 573-575.

288

(18)

(19)

(20)

(21)

(22)

(23)

(24)

(25)
(26)

(27)

(28)

(29)
(30)

(31)

(32)

(33)

Gonon, F. G.; Fombarlet, C. M.; Bud, M. J.; Pujol, J. F. Analytical Chemistry
1981, 53, 1386-1389.

Gonon, F .; Msghina, M.; L.Stjarne Neuroscience 1993, 535-538

Gonon, F.; Bao, J. X.; Msghina, M.; Suaud-Chagrry, M. F.; Stjame, L. Journal of
Neurochemistry 1993, 60, 1251-1257.

Brock, J. A.; Dunn, W. R.; Boyd, N. S. F.; Wong, D. K. Y. British Journal of
Pharmacology 2000, 131, 1507 - 1511.

Brock, J. A.; Tan, J. H. C. British Journal of Pharmacology 2004, 142, 267—274.

Dunn, W. R.; Hardy, T. A.; Brock, J. A. British Journal of Pharmacology 2003,
140, 231—238.

Ewing, A. G.; Dayton, M. A.; Wightrnan, R. M. Analytical Chemistry 1981, 53,
842-847.

Wightrnan, R. M. Science 1988, 24, 415 - 420.
Stamford, J. Journal of Neuroscience Methods 1986, I 7, 1-29.

Kawagoe, K. T.; Zimmerman, J. B.; Wightrnan, R. M. Journal of Neuroscience
Methods 1993, 48, 225-240.

Pelskov, Y. V.; Sakharova, A. Y.; Krotova, M. D.; Bouilov, L. L.; Spitsyn, B. V.
Journal of Electroanalytical Chemistry 1987, 228, 19-27

Swain, G. M.; Ramesham, R. Analytical Chemistry 1993, 65, 345-351.
Swain, G. M. Electroanalytical Chemistry 2004, 22, 181-277.

Park, J.; Show, Y.; Quaiserova, V.; Galligan, J. J.; Fink, G. D.; Swain, G. M.
Journal of Electroanalytical Chemistry 2005, 583, 56—68.

Park, J .; Galligan, J. J .; Fink, G. D.; Swain, G. M. Analytical Chemistry 2006, 78,
6756-6764.

Park, J .; Quaiserova'-Mocko, V.; Peckova, K.; Galligan, J. J .; Fink, G. D.; Swain,
G. M. Diamond & Related Materials 2006, 15, 761-772

289

 

(34)

(35)

(36)

(37)

(38)

(39)

(40)

(41)

(42)

(43)

(44)

(45)

(46)

(47)

(48)

Park, J .; Galligan, J. J .; Fink, G. D.; Swain, G. M. Journal of Physiology 2007,
584.3, 819—834.

McElvain, J. S.; Schenk, J. O. Annals of the New York Academy of Sciences 1992,
654, 480-482.

Davidson, C.; Ellinwood, E. H.; Douglas, S. B.; Lee, T. H. Journal of
Neuroscience Methods 2000, 101 , 75-83.

Stamford, J. A.; Palij, P.; Davidson, C.; Jorm, C. M.; Millarb, J. Journal of
Neuroscience Methods 1993, 50, 279-290

Luo, M.; Hess, M. C.; Fink, G. D.; Olson, L. K.; Rogers, J .; Kreulen, D. L.; Dai,
X.; Galligan, J. J. Autonomic Neuroscience: Basic and Clinical 2003, 104, 47-57.

Zhang, L. L.; Yan, L. D.; Ma, L.; Luo, Z.; Cao, T.; Zhong, J .; Yan, Z.; Wang, L.;
Zhao, Z.; Zhu, S.; Schrader, M.; Thilo, F.; Zhu, Z.; ., M. T. Circulation Research
2007, 13, 1063-1070.

Bath, B. D.; Michael, D. J.; Trafton, B. J.; Joseph, J. D.; Runnels, P. L.;
Wightman, R. M. Analytical Chemistry 2000, 72, 5994-6002.

Bath, B. D.; Martin, H. B.; Wightman, R. M.; Anderson, M. R. Langmuir 2001,
I 7, 7032-7039.

Bennett, J. A.; Show, Y.; Wang, S.; Swain, G. M. Journal of The Electrochemical
Society 2005, 152, E184-E192.

Fischer, A. E.; Show, Y.; Swain, G. M. Analytical Chemistry 2004, 76,
2553-2560.

Wang, S.; Swain, G. M. Journal of Physical Chemistry C 2007, 111, 3986-3995.

Cvacka, J.; Quaiserova, V.; Park, J.; Show, Y.; Muck, A.; Swain, G. M.
Analytical Chemistry 2003, 75, 2678-2687.

AgerIII, J. W.; Veirs, D. K.; Sharnir, J.; Rosenblatt, G. M. Journal of Applied
Physics 1990, 68, 3598-3608

F auteux, C.; Pegna, J. Applied Physics A 2004, 78, 883—888.

Leveque, D.; Auvray, M.-H. Composites Science and Technology 1996, 56,

290

(49)

(50)

(51)

(52)

(53)

(54)

(55)

(56)

(57)
(58)

(59)

(60)

(61)

(62)

(63)

(64)

749-154.

Novotny, M.; Quaiserov'a-Mocko, V.; Wehrwein, E. A.; Kreulen, D. L.; Swain,
G. M. Journal of Neuroscience Methods 2007 , I 63, 52—59.

Patel, B. A.; Bian, X.; Quaiserova'-Mocko, V.; Galliganb, J. J.; Swain, G. M.
Analyst 2007, 132, 41—47.

DuVall, S. H.; McCreery, R. L. Journal of the American Chemical Society 2000,
122, 6759-6764.

Bowling, R.; Packard, R. T.; McCreery, R. L. Langmuir 1989, 5, 683-688.

Chen, P.; Fryling, M. A.; McCreery, R. L. Analytical Chemistry 1995, 67,
3115-3122.

Yater, J. E.; Shih, A.; Butler, J. E.; Pehrsson, P. B. Journal of Applied Physics
2004, 96, 446-453.

Bjorklund, A.; Lndvall, 0.; Svensson, L. A. Histochemie 1972, 32, 113-131.

Nilsson, H.; Goldstein, M.; Nilsson, 0.; Ljung, B.; Sjoblom, N.; Wallin, B. G.
Acta Physiol. Scand. 1986, 126, 121-133.

Geffen, L. B.; Livett, B. G.; Rush, R. A. J. Physiol. 1969, 204, 593-605.
Goldstein, M.; Fuxe, K.; Hokfelt, T. Pharmacol. Rev. 1972, 24, 293-309.

Galligan, J. J.; Miller, S. B.; Katki, K.; Supowit, S.; DiPette, D.; Fink, G. D.
Regulatory Peptides 2006, 133, 97-104.

Brody, M. J .; Dorr, L. D.; Shaffer, R. A. Am J Physiol 1970, 219, 1746-1750.

Tsuda, K.; Kuchii, M.; Nishio, 1.; Masuyama, Y. Clin. Exp. Hypertens 1986, 8,
259-275.

Westfall, T. C.; Carpenter, S.; Naes, L.; Meldrum, M. J. Clin. Exp. Hypertens
1986, 8, 221-237.

Bobalova, J .; Mutafova-Yambolieva, V. N. J. Auton. Pharmacol. 2001, 21, 47-55.

Smyth, L.; Bobalova, J .; Ward, S. M.; Keef, K. D.; Mutafova-Yambolieva, V. N.

291

(65)

(66)
(67)
(68)

(69)

(70)

(71)

(72)

(73)

(74)
(75)
(76)

(77)

(78)
(79)

(80)

 

Auton. Neurosci. 2000, 86, 18-29.

Ramme, D.; Illes, P.; Spath, L.; Starke, K. Naunyn. Schmiedebergs. Arch.
Pharmacol. 1986, 334, 48-55.

Cox, S. L.; Story, D. F.; Ziogas, J. Br. J. Pharmacol. 1996, 119, 976-984.
Mermet, C.; Suaud-Chagny, M. F.; Gonon, F. Neuroscience 1990, 34, 423-432.
Gonon, F. Neuromethods 1995, 27, 153-177.

Dugast, C.; Cespuglio, R.; Suaud-Chagny, M. F. J. Neurochem. 2002, 82,
529-537.

Brock, J. A.; Tan, J. H. Br. J. Pharmacol. 2004, 142, 267-274.

Cvacka, J .; Quaiserova, V.; Park, J .; Show, Y.; Alexander Muck, J .; Swain, G. M.
Anal. Chem. 2003, 75, 2678-2687.

Park, J.; Galligan, J. J.; Fink, G. D.; Swain, G. M. Anal. Chem. 2006, 78,
6756-6764.

Park, J.; Quaiserovaia-Mocko, V.; Peckovaiéi, K.; Galligan, J. J .; Fink, G. D.;
Swain, G. M. Diamond Relat. Mater. 2006, 15, 761-772.

Park, K. H.; Lee, S.; Koh, K. H. J. Appl. Phys. 2006, 99, 034303.
Park, J .; Galligan, J. J .; Fink, G. D.; Swain, G. M. J. Physiol. 2007, 584, 819—834.

Park, J. Dissertation, Michigan State University, East Lansing, 2006.

Bartho, L.; R., B.; Patacchini, R.; Petho, G.; Holzer-Petsche, U.; Holzer, P.; Lazar,

Z.; Undi, S.; Illenyi, L.; Antal, A.; Horvath, O. P. Eur. J. Phamacol. 2004, 500,
43-57.

Fink, G. D.; Johnson, R. J .; Galligan, J. J. Hypertension 2000, 35, 464-469.
Pang, C. C. Pharmacol. T her. 2001, 90, 179-230.
Stekiel, W. J.; Contney, S. J.; Lombard, J. H. Am J Physiol 1986, 250,

C547-C556.

292

(81) Lind, H.; Edvinsson, L. J. Hum. Hypertens 2002, 16, 53-59.
(82) Kawasaki, H.; Saito, A.; Takasaki, K. Brain Res. 1990, 518, 303-307.

(83) D'Orleans-Juste, P.; Claing, A.; Telemaque, 8.; Warner, T. D.; Regoli, D. Eur. J.
Phamacol. 1991, 204, 329-334.

293

 

 

 

 

Chapter 8

Conclusions

Natural diamond is an electrical insulator with a band-gap of ~5.5 eV and a
resistivity greater than 10'2 Q-cm.l However, synthetic diamond can be made
electrically conducting via doping with a dopant, such as boron, nitrogen and
phosphorous.

The most extensively employed CVD diamond in electrochemistry is the p-
type, boron-doped diamond. The resistivity of highly boron-doped diamond thin ﬁlm
can be as low as ~0.005 Q-cm. Both high-quality microcrystalline and nanocrystalline
boron-doped diamond thin ﬁlms can be prepared by CVD. Bom-doped diamond thin
ﬁlms show low capacitive background current, a wide potential window in aqueous
media of 3 - 4V, chemical inertness, extreme hardness and robustness, and
dimensional stability at high potentials and current densities.

The electrical and electrochemical properties of boron-doped diamond tlrin-
ﬁlms are inﬂuenced by many factors, such as the growth conditions, grain size,
doping level, impurities (e. g., grain defects, sp2 carbon and adsorbed hydrogen), and
surface chemistry. The growth conditions are the primary determinant of the surface
morphology, although both microcrystalline and nanocrystalline diamond thin-ﬁlms
can be deposited in similar hydrogen-rich gas mixtures. Nanocrystalline diamond
thin-ﬁlm can be deposited from a hydrogen-rich or an Ar—rich source mixture

(Chapter 6). We demonstrated that the two types of nanocrystalline BDD thin ﬁlms

294

have very electrochemical properties, although their grain size and surface
morphology are very different as revealed by SEM images and Raman spectra.

According to the grain size, microcrystalline and nanocrystalline BDD thin-
ﬁlms can be classiﬁed. Microcrystalline BDD tlrin-ﬁhn has a different Raman
spectrum than the nanocrystalline ﬁlm grown from an Ar-rich gas but their
electrochemical properties are similar. Using CP-AFM and SECM, we found that the
electrical conductivity and electrochemical activity of both microcrystalline and
nanocrystalline BDD thin-ﬁlms are variable across the surface. Microcrystalline BDD
thin-ﬁlm possesses lower fraction of active regions but more active individual sites
than do the nanocrystalline ﬁlms. On the other hand, nanocrystalline BDD thin-ﬁlm
shows greater fraction of electrically-active regions.

The doping level inﬂuences the surface morphology of BDD thin-ﬁlm.
Increasing boron-doping level increases the surface roughness of the ﬁlm. This was
observed for both microcrystalline and nanocrystalline ﬁlms. More importantly, the
doping level inﬂuences the electrical and electrochemical properties of BDD thin ﬁlm.
CP-AF M and SECM measurements illustrated that both the fraction of electrically
conductive area and the fraction of electrochemically active regions increase as the
doping level increases for both microcrystalline and nanocrystalline ﬁlms.

The impurities in BDD thin-ﬁlm have various effects on the physical and
chemical properties. The Raman spectrum for the nanocrystalline BDD thin-ﬁlm is
very different from that for microcrystalline BDD thin ﬁlm. One reason for this is that
the content of sp2 carbon in nanocrystalline ﬁlm is much higher than that in
microcrystalline ﬁlm. CP-AFM images showed that the electrically conducting sites
are associated with the grain boundaries of microcrystalline BDD thin ﬁlm.

Comparing the electrochemical reaction of dopamine at the MSU and NRL BDD thin-

295

 

 

ﬁlm electrodes, it was found that sp2 carbon impurity does not dominate the
electrochemical activity. Other important impurities in BDD thin-ﬁlm are hydrogen
and hydrocarbon species. Hydrogen can be physisorbed and chemisorbed on the
surface or in the bulk of the thin-ﬁlm. Heating treatment at 400 °C in Ar can desorb
the physisorbed hydrogen and hydrocarbon species making the thin-ﬁlm surface more
hydrophilic (contact angle decreases). The desorption of physisorbed hydrogen and
hydrocarbon species also inﬂuence the electrochemical properties of BDD thin-ﬁlm
electrodes with a low doping level but this effect can be negligible on BDD thin-ﬁlm
electrodes with higher doping levels (> 5 ppm). We did not investigate how the
absorbed hydrogen inﬂuence the electrical conductivity of BDD ﬁlm with CP-AF M
because we do not have an oxygen-free, isolated system to perform the measurement.
Therefore, to further study what role hydrogen play for the electrical conductivity of
boron-doped diamond can be one of the future work.

The surface termination had a signiﬁcant inﬂuence on the properties of CVD
diamond. For example, the fastest electron-transfer kinetics for Fe(CN)63'/4’ has been
recorded at hydrogen-terminated BDD electrode but very sluggish reaction has been
recorded at oxygen-terminated BDD electrode. Electrochemical measurements show
that O-terminated BDD electrode has a little larger background current that H-
terminated BDD electrode. The reason for this might be the O-terminated surface is
more hydrophilic and this may increase the dielectric constant of H20 at the surface.
CP-AF M and SECM studies show that both the fraction of electrically-active regions
and the fraction of electrochemically-active regions are greater for H-terminated than
for O-terminated BDD ﬁlms.

In general, the present work answered the following questions:

296

 

1. What is the spatial uniformity of the electrical conductivity of boron-doped
microcrystalline and nanocrystalline diamond thin ﬁlms? The electrical properties of
boron-doped microcrystalline and nanocrystalline diamond thin ﬁlms are not uniform
over the surface.

2. What factors contribute to the electrical conductivity (e.g., boron-doping
level, surface hydrogen, surface oxygen and nitrogen compensators)? Factors such as
boron doping level, surface hydrogen, surface oxygen and sp2 carbon contribute to the
electrical conductivity but boron-doping level is the determinant.

3. How does the uniformity of the electrical properties correlate with the
spatial uniformity of the electrochemical reaction rates across the electrode surface for
various outer-sphere redox systems? The electrochemical properties of boron-doped
microcrystalline and nanocrystalline diamond thin ﬁlms are also not uniform over the
surface. Because the spatial resolution of the SECM used in the present work (in
micrometer level) does not match the spatial resolution of CP-AF M (in nanometer
level), we can not associate the electrical conductivity with the electrochemical
activity of BDD ﬁlms. SECM with a resolution in nanometer level is available now.
Therefore, using SECM with a higher spatial resolution and CP-AFM to investigate
the spatial relationship between the electrical conductivity and electrochemical

activity of boron-doped diamond may be another future work.

297

 

 

 

A

Illﬂlitlililiilﬂljlﬂl[ﬂjijlﬂl