wn

‘5."

_L41.

a
:t

H‘ '
' 1‘
i) 3.7:»

Q}

r
1"
min"- ;

n

W:

.zx

.
la

"38L:

25??

W»“&

'I’

 

 

at? ' mi?“ - ,
: l. . ‘ _
flag-"m". kl. Nu
‘ x ' Ag}. if; a:

' f

if :
an“.-
3i].€-
5:“
5“

1
“a

:gfifigf"
gg’iéfizf?

I».

a

" .21.-
.. 1195-2.
t--“"" y
g}? ‘
“V

3’5

 

mtg-13 LIBRARIES
MICHIGAN STATE UNIVERSITY
\ EAST LANSING, MICH 48824-1048

0.x I

‘r‘
C/’

(0511/5

This is to certify that the
thesis entitled

THE ELECTROCHEMICAL OXIDATION AND
AMPEROMETRIC DETECTION OF CATECHOLS AND
CATECHOLAMINES AT BORON-DOPED DIAMOND THIN-
FILM ELECTRODES

presented by

GLORIA FE M. PIMIENTA

has been accepted towards fulfillment
of the requirements for the

MS. degree in Chemistry

 

 

Mai/MM

Major Professor’s Signature
A) - /é -—o 9/

Date

MSU is an Affirmative Action/Equal Opportunity Institution

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

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 c:/CIRC/DateDue.p8&p.15

 

THE ELECTROCHEMICAL OXIDATION AND AMPEROMETRIC
DETECTION OF CATECHOLS AND CATECHOLAMINES AT
BORON-DOPED DIAMOND THIN-FILM ELECTRODES
By

Gloria Fe M. Pimienta

A THESIS

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

MASTER OF SCIENCE
Department of Chemistry

2004

ABSTRACT

THE ELECTROCHEMICAL OXIDATION AND AMPEROMETRIC
DETECTION OF CATECHOLS AND CATECHOLAMINES AT BORON-DOPED
DIAMOND THIN FILM ELECTRODES
By

Gloria Fe M. Pimienta

The electrochemical oxidation and amperometric detection of catechols and
catecholamines were investigated using the new electrode material, diamond thin film.
The films were synthesized by chemical vapor deposition and made electrically
conducting by doping with boron. In general, diamond has properties superior to those of
other commonly used Sp2 carbon electrodes (e.g., glassy carbon), among which are low
background current, good response precision, stable response without pretreatment, and
high resistance to fouling. Cyclic voltarnmetric investigations were carried out to
evaluate the electrode response for dopamine, catechol, and 3,4-dihydroxyphenylacetic
acid as a function of the potential sweep rate and solution pH. The electrode performance
for the amperometric detection of these and other catecholamines in flow injection
analysis (FIA) and high performance liquid chromatography (HPLC) was evaluated in
terms of linear dynamic range, limit of quantitation, sensitivity, and response precision.
The mass limits of quantitation (S/N > 3) were 10 — 70 pg in FIA and 10 — 1200 pg in
HPLC. The detection figures of merit for untreated diamond are comparable to those for
glassy carbon. However, diamond outperformed glassy carbon in terms of response
stability and resistance to fouling. Finally, the bioanalytical applicability of the HPLC-

EC assay was demonstrated.

T 0 Ian and Angelo

iii

ACKNOWLEDGMENTS

My deepest gratitude goes out to those who have played a vital role in this
research, my graduate studies, and my life:

Dr. Greg Swain, my research advisor, for the guidance, encouragement and
inspiration that made me do my best and for the support in my research and non-research
concerns;

NIH, for the generous funding of the research;

Dr. McGuffin, Dr. Blanchard, Dr. Garrett, and Dr. Galligan, my advisorial
committee, for their critical insights that contributed to my accomplishments as a
graduate student;

Dr. Show, for the diamond electrodes and for the troubleshooting pointers;

Dr. Lookingland, for the rat plasma samples and for the assistance in the analysis
of the results; and his students Troy and Bahareh, for the unselfish help;

Maggie, Matt, Jian, Shannon, Josef, Zuzana, Veronika, Jason S., Prema, Grace,
Jason 8., Jim Woo, Anne, Doug, Yang, Elizabeth, Wang, and Audrey, my groupmates,
for the discussions that made electrochemistry more comprehensible and for the
companionship that made the lab a fun place to be in;

MSU Filipino Club, for the Philippine spirit and camaraderie;

Mariano and Pimienta families, for the values that made me the person that I am;

My husband Ian and my baby Angelo, for their love and affection and for the joy
that they bring to my heart;

And God Almighty, my provider, for everything that He has blessed me with.

iv

TABLE OF CONTENTS

LIST OF TABLES ................................................................................ vi
LIST OF FIGURES .............................................................................. vii
Chapter 1: INTRODUCTION .................................................................. 1
Chapter 2: EXPERIMENTAL METHODS ................................................ 12
2.1. Boron-Doped Diamond Thin Film Growth ......................................... 12
2.2. Boron-Doped Diamond Thin Film Characterization ............................... 13
2.2.1. Atomic Force Microscopy ................................................. 13

2.2.2. X-ray Diffraction ............................................................ 14

2.2.3. Raman Spectroscopy ....................................................... 16

2.2.4. Contact Angle Measurement .............................................. 18

2.2.5. Cyclic Voltammetry ........................................................ 19

2.3. Glassy Carbon Polishing Procedure ................................................. 21
2.4. Flow Injection Analysis ............................................................... 21
2.5. High Performance Liquid Chromatography ........................................ 23
2.6. Chemicals and Reagents ............................................................... 23

Chapter 3: BORON-DOPED DIAMOND THIN-FILM ELECTRODES FOR THE
ELECTROCHEMICAL DETECTION OF CATECHOLS AND

CATECHOLAMINES ........................................................................... 25
3.1. Cyclic Voltammetric Behavior of Catechols and Catecholamines ............... 25
3.1.1. Introduction .................................................................. 25
3.1.2. Results and Discussion ..................................................... 27

3.2. Amperometric Detection of Catechols and Catecholamines in Flow Injection
Analysis .................................................................................. 43
3.2.1. Introduction .................................................................. 43
3.2.2. Results and Discussion ..................................................... 46
3.3. Conclusions .............................................................................. 59

Chapter 4: REVERSED-PHASE HPLC SEPARATION AND AMPEROMETRIC
DETECTION OF CATECHOLAMINES AND METABOLITES AT BORON-

DOPED DIAMOND THIN-FILM ELECTRODES ....................................... 62
4.1. Introduction .............................................................................. 62
4.2. Results and Discussion ................................................................ 65
4.3. Conclusions .............................................................................. 73

Chapter 5: SUMMARY AND CONCLUSIONS .......................................... 75

BIBLIOGRAPHY ............................................................................... 78

LIST OF TABLES

Table 1.1. Detection limits for the catecholamines with ultraviolet, fluorescence and EC

detectionf’ ............................................................................................ 2
Table 1.2. Molecular structure, pK,35 and formula mass of the analytes. ................ 10

Table 2.1. Cyclic voltammetric data for the redox test systems at a microcrystalline

diamond thin-film electrode. ..................................................................... 21

Table 3.1. AEp values of the catechol compounds at pH 7 for the three electrode types
.......................................................................................................... 40

Table 3.2. Reproducibility of the cyclic voltammetric measurements using the three

electrode types being studied (n = 3). .......................................................... 43

Table 3.3. Edet and Em values for the amperometric detection of the analytes at

microcrystalline diamond, nanocrystalline diamond, and GC. .............................. 51

Table 3.4. Response precision (% RSD) for catechol and catecholamine detection at
microcrystalline diamond, nanocrystalline diamond, and GC in 0.1 M phosphate buffer,
0.1 M NaCl, pH 7.2. .............................................................................. 53

Table 3.5. FIA-EC data for catechol and catecholamine detection at microcrystalline

diamond, nanocrystalline diamond, and GC electrodes. ..................................... 56

Table 4.1. Recent publications on HPLC-EC analysis of catecholamines and

metabolites. ........................................................................................ 64

Table 4.2. Amperometric detection figures of merit for the HPLC-EC analysis of
catecholamines and metabolites at nanocrystalline and microcrystalline diamond thin-

film electrodes. .................................................................................... 69

vi

LIST OF FIGURES

Figure 1.1. The catechol/orthoquinone redox reaction. ...................................... 3

Figure 1.2. Representative functional groups that might exist on oxidized sz carbon

surfaces. .............................................................................................. 4
Figure 1.3. The cubic diamond crystal lattice. ................................................. 5

Figure 1.4. Background cyclic voltammetric i-E curves in 1.0 M KCl for a

microcrystalline diamond film. Scan rate = 100 mV/s. Electrode geometric area = 0.16

cmz. ................................................................................................... 8

Figure 2.]. AFM images (contact mode, air) of a (A) microcrystalline and (B)

nanocrystalline diamond thin film. ............................................................. 14

Figure 2.2. XRD pattern for a (A) microcrystalline and (B) nanocrystalline diamond thin
film. ................................................................................................. 15

Figure 2.3. Visible Raman spectra for a (A) microcrystalline and (B) nanocrystalline

diamond thin film. Ac, = 532 nm. Laser power = 30 mW. Integration time = 10 s. ....17
Figure 2.4. Diagram of the single compartment, glass electrochemical cell. . . . . .. 19

Figure 2.5. Cyclic voltammetric i-E curves for (A) 1.0 mM F e(CN)(,3 ’/4' and (B) 1.0 mM

Ru(NH3)63+/2+ at a microcrystalline diamond thin film. The supporting electrolyte was

1.0 M KCl. Scan rate = 100 mV/S. Electrode geometric area = 0.16 cmz. ............... 20
Figure 2.6. Block diagram Of the FIA system. ............................................... 22
Figure 2.7. Schematic of the thin-layer flow cell. ........................................... 22
Figure 3.1. Molecular structure and pKa value for DA, CAT and DOPAC. . 27

vii

Figure 3.2. Cyclic voltammetric i-E curves for (A) microcrystalline diamond,
(B) nanocrystalline diamond, and (C) GC electrodes in 0.1 M phosphate buffer, pH 2 (---)
and pH 7 (—). Scan rate = 100 mV/s. Electrode geometric area = 0.16 cmz. ........... 28

Figure 3.3. Background current measured at 0 mV vs. pH for microcrystalline diamond
(O), nanocrystalline diamond (I), and glassy carbon (A) electrodes in 0.1 M phosphate
buffer. Scan rate = 100 mV/s. Electrode geometric area = 0.16 cmz. Each data point and

error bar represent the average and standard deviation of three measurements. ......... 29

Figure 3.4. Cyclic voltammetric i-E curves for GC in 0.1 M phosphate buffer, pH 2,

4, and 7. Scan rate = 100 mV/s. Electrode geometric area = 0.16 cmz. .................. 31

Figure 3.5. Cyclic voltammetric i-E curves (background-subtracted) for
microcrystalline diamond (——) and GC (_) electrodes in 10 uM DA, CAT, and DOPAC
in 0.1 M phosphate buffer, pH 7. Scan rate = 100 mV/s. Electrode geometric area = 0.16

cmz. ................................................................................................. 33

Figure 3.6. Cyclic voltammetric i-E curves (background-subtracted) for
microcrystalline diamond (——) and GC (_) electrodes in 10 IIM DA, CAT, and DOPAC
in 0.1 M phosphate buffer, pH 2. Scan rate = 100 mV/s. Electrode geometric area = 0.16

cmz. .................................................................................................. 34

Figure 3.7. Dopamine (A) oxidizes to duroquinone (B). The 6-position is susceptible to

nucleophilic attack by the side-chain amine forming 5,6-dihydroxyindoline (C). ....... 35

Figure 3.8. Plots of oxidation peak current vs. pH for microcrystalline diamond (O),
nanocrystalline diamond (I), and GC (A) electrodes in 10 M DA, CAT, and DOPAC
in 0.1 M phosphate buffer. Scan rate = 100 mV/s. Electrode geometric area = 0.16 cmz.
The theoretical ip value is indicated by the dashed line. ...................................... 37

Figure 3.9. Plots of oxidation peak potential vs. pH for (A) microcrystalline diamond,
(B) nanocrystalline diamond, and (C) GC electrodes in 10 uM solutions of DA (I), CAT
(0), and DOPAC (A) in 0.1 M phosphate buffer. Scan rate = 100 mV/s. Electrode

geometric area = 0.16 cmz. ...................................................................... 39

viii

Figure 3.10. Surface fimctional groups on oxidized sp2 and sp3 electrodes. .............. 42
Figure 3.11. The solution flow in a thin-layer electrochemical flow cell. ............... 45

Figure 3.12. FIA-EC hydrodynamic voltammetric background i-E curves for
microcrystalline diamond (O), nanocrystalline diamond (I), and GC (A) electrodes in
0.1 M phosphate buffer, 0.1 M NaCl, pH 7.2. Flow rate = 1.0 mL/min. Electrode

geometric area = 0.12 cmz. ...................................................................... 46

Figure 3.13. Peak-to-peak noise at microcrystalline diamond (I), nanocrystalline
diamond (B), and GC (I) electrodes in 0.1 M phosphate buffer, 0.1 M NaCl, pH 7.2, as
a function of applied potential. Flow rate = 1.0 mL/min. Electrode geometric area = 0.12

cmz. .................................................................................................. 48

Figure 3.14. Hydrodynamic voltammetric i-E curves for 20 uL injections of 30 uM
catechols and catecholamines in microcrystalline diamond (O), nanocrystalline diamond
(I), and GC (A). The carrier solution was 0.1 M phosphate buffer, 0.1 M NaCl, pH 7.2.

Flow rate = 1.0 mL/min. Electrode geometric area = 0.12 cmz. ............................ 50

Figure 3.15. Calibration curves for the catechols and catecholamines obtained at (A)
microcrystalline diamond, (B) nanocrystalline diamond, and (C) GC electrodes. Note:
At least 5 concentrations were used for each curve but some plots appear to have fewer
points because more measurements were made at the low uM level. The carrier solution
was 0.1 M phosphate buffer, 0.1 M NaCl, pH 7.2. Flow rate = 1.0 mL/min. Injected

volume = 20 uL. Electrode geometric area = 0.12 cmz. ..................................... 55

Figure 3.16. FIA-EC response for ZO-uL injections of 2 mM NE at microcrystalline
diamond (O), nanocrystalline diamond (I), and GC (A) recorded over a 7-h period.
The carrier solution was a Kreb’s buffer, pH 7.4, flowing at 1.0 mL/min. Electrode
geometric area = 0.12 cmz. Each data point shown represents an average of 10 injections

made at 1.0 min intervals. ....................................................................... 57

ix

Figure 3.17. F IA-EC response for ZO-uL injections of 1 pM NE (injections 1-10) and 1
uM NE with 1 mM BSA (injections 11-20) at microcrystalline diamond (O),
nanocrystalline diamond (I), and GC (A) electrodes. The carrier solution was a Kreb’s
buffer, pH 7.4, flowing at 1.0 mL/min. Injections were made at 1.0 min intervals.

Electrode geometric area = 0.12 cmz. .......................................................... 59

Figure 4.1. Steps in the metabolism of catecholamines. ................................... 63

Figure 4.2. Reversed-phase HPLC separation Of 20 uM DOPEG, NE, EP, VMA, NMN
and MN on a C13 column (Alltech Adsorbosphere HS, 5 pm particle size, 4.6 mm x 150
mm). The mobile phase was 95/5 (v/v) 0.05 M NaHzPO4/MeOH, pH 2.5, 0.475 mM
HSA. The working electrode was a nanocrystalline diamond thin film and the detection
potential was 950 mV vs. Ag/AgCl. Flow rate = 0.8 mL/min. Injection volume = 20 uL.

Electrode geometric area = 0.12 cmz. ........................................................ 66

Figure 4.3. Calibration curves obtained at a (A) nanocrystalline and (B) microcrystalline
diamond thin-film electrode for a series of mixtures of the catecholamines and
metabolites. The separation was carried out using the same conditions as in Figure 4.2.

The detection potentials used were 950 mV and 1000 mV vs. Ag/AgCl for A and B,

respectively. Injection volume = 20 pL. Electrode geometric area = 0.12 cmz. ....... 68

Figure 4.4. Reversed-phase chromatograms for the analysis of plasma samples from (A)
a normotensive rat and (B) a hypertensive rat on a C .3 column (Alltech Adsorbosphere
HS, 5 pm particle size, 4.6 mm x 150 mm). The mobile phase was 96/4 (v/v) 0.05 M
NaHzPO4/MeOH, pH 2.5, 0.48 mM HSA. The working electrode was a nanocrystalline
diamond thin film and the detection potential was 950 mV vs. Ag/AgCI. Flow rate = 0.8

mL/min. Injection volume = 20 uL. Eletrode geometric area = 0.12 cmz. ............. 71

Chapter 1

INTRODUCTION

Catechols are aromatic compounds with a single benzene ring structure and two
substituent hydroxy groups at 3- and 4- positions, and are widely distributed in living
systems. For example, they comprise the toxic substance found in poison ivy that causes
allergic contact dermatitis.1 Catecholamines, substituted catechols with an amino group
in a side chain, function as neurotransmitters in the central nervous system. They play a
vital role in human disorders, such as depression and schizophrenia. They are involved in
a variety of regulatory systems, such as emotion, stress, learning and sleep. They also
take part in many metabolic processes like hormone secretion.2

Abnormal levels of catecholamines and their metabolites in biological fluids can
be used for the diagnosis Of certain diseases. For example, patients with
pheochromocytoma, a tumor of chromaffin cells in the adrenal gland, have elevated
levels of norepinephrine and epinephrine in urine.3 Abnormal catecholamine content in
the heart, a principal location of catecholamine synthesis and storage, is associated with
cardiovascular disease. Congestive heart failure has been linked to reduced myocardial
norepinephrine and elevated urinary norepinephrine. Catecholamine levels in plasma
have been used to diagnose hypertension.4

The development of techniques for the quantification of catechols and

catecholamines is essential due to the wide range of functions these molecules serve. The

catecholamine content in urine, plasma, and cerebrospinal fluid is of clinical interest as it
can serve as a marker for a variety of metabolic, cardiovascular, and neurological
abnormalities.5 Several methods have been used to measure catecholamines including
absorbance and fluorescence detection, radioenzymatic methods, mass Spectrometry and
electrochemical (EC) detection. One advantage of EC detection over other analytical
methods for catecholamine measurement is that there is no need for a derivatization
procedure prior to analysis. EC detection is also noted for its superior sensitivity
compared to other methods. Table 1.1 presents the detection limits Obtained for the
catecholamines, norepinephrine, epinephrine and dopamine, using ultraviolet,
fluorescence and EC detection coupled with liquid chromatographic separation under

6

optimal experimental conditions. Other benefits include good precision, Simple

instrumentation, and low cost.

Table 1.1. Detection limits for the catecholamines with ultraviolet, fluorescence and EC

detection.6

 

 

 

Ultraviolet Fluorescence Electrochemical
Norepinephrine 5 ng 2 ng 0.015 ng
Epinephrine 5 ng 2 ng 0.15 ng
Dopamine 5 ng 5 ng 0.20 ng

 

 

 

 

 

Many catechols and catecholamines are electrochemically-active and are thus
ideally suited for EC detection. They can be readily oxidized to generate the
corresponding orthoquinone via a two-proton, two-electron reversible redox reaction, as

Shown in Figure 1.1

HO R O R

‘14-— + 2H+ + 26'
HO 0

Figure 1.1. The catechol/orthoquinone redox reaction.

In an electrochemical measurement, the potential applied to a working electrode
with respect to a reference electrode serves as the driving force for the redox reaction.
The working electrode then acts as an electron sink or source, depending on the
magnitude of the applied potential. Measurement of the current that flows in response to
the applied potential can be used to determine the redox analyte concentration in
solution.7

A limitation with electroanalytical measurements is that the molecular specificity
is often lacking due to the fact that different analytes undergo electrolysis at similar
potentials. For example, many of the catechols and catecholamines can undergo electron
transfer at similar potentials, making it difficult to extract qualitative information from a
measurement of a solution containing multiple analytes.7’8 In many cases, a separation
step such as reversed-phase liquid chromatography or capillary electrophoresis is coupled

to EC detection.”

High performance liquid chromatography with electrochemical
detection (HPLC-EC) still remains one of the most widely used methods for
catecholamine and related compound analysis since the technique was introduced in the
19705.9‘lo It is the method of choice in most clinical laboratories and physiological

research for its high selectivity, good reproducibility, simple instrumentation, short

analysis time, and low cost.

Another main shortcoming of EC detection is the deactivation that is Often observed
with electrodes during use. Deactivation refers to attenuated current and decreased
electrode reaction kinetics. Carbon electrodes, in particular, are subject to a gradual loss
of activity due to changes in the electrode microstructure and chemistry and electrode

fouling by adsorbed reaction intermediates and products.ll

The presence of electroactive
and ionizable surface carbon-oxygen functionalities on the commonly used sp2 carbon
electrodes, as illustrated in Figure 1.2, present additional complications.” These include
increased background current, pH-dependence Of background and faradaic signals,

susceptibility to fouling due to strong molecular adsorption, need for frequent

pretreatment, large response variability, and poor response precision.‘2

\ OH Phenol

O Carbonyl

Lactone

Carboxylic acid

o-Quinone

p-Quinone

 

Figure 1.2. Representative functional groups that might exist on oxidized Sp2 carbon

surfaces.

Given the above mentioned issues with sp2 carbon materials, it would be
advantageous to have an electrode that is devoid of surface carbon-oxygen functionalities
but is still responsive to the analytes of interest. A promising new, oxygen-free electrode
material is electrically conducting, boron-doped diamond. The atomic structure of
diamond, as shown in Figure 1.3, consists of an extended network of sp3-bonded carbon
atoms in a tetrahedral arrangement. The surface carbon atoms are stabilized by strong
covalent bonds with hydrogen for films deposited from hydrogen-rich source gas
mixtures. Diamond is the stable form of carbon at high pressure and temperature.” It is
a unique material with exceptional properties like extreme hardness, high thermal
conductivity, high chemical inertness, and low compressibility. Diamond is sometimes
referred to as the “ultimate engineering material” because of its exceptional properties.
There is interest in using diamond in a variety Of applications, such as optics, electronics,

13,14

and thermal management. One application that has gained prominence in recent

years is conductive diamond electrodes for electrochemistry.

“
y!“
as-

Figure 1.3. The cubic diamond crystal lattice.

The first reports of diamond electrodes for electrochemistry were in the mid-
1980s.”‘16 Although diamond is a very good electrical insulator, it can possess
semimetal electronic properties when doped with boron and thus be utilized in
electrochemical measurements. In boron-doped, thin-film diamond electrodes, boron
atoms substitute in place of carbon atoms during film growth.”'8 Boron functions as an
electron acceptor and promotes p-type semiconductivity with an acceptor activation
energy of 0.37 eV above the valence band. Valence band electrons are thermally
promoted to the boron acceptors, leaving free electrons in the dopant band, and holes in
the valence band to support current flow. Electrical conductivity is influenced not only
by the boron-doping level, but also by the grain boundaries at the surface and the
nondiamond Sp2 carbon impurity that might be present.17

High quality, hydrogen-terminated diamond possesses several interesting and
electroanalytically important properties that distinguish it from other sp2 carbon
electrodes. Untreated diamond films exhibit good electrochemical activity for a number
of redox systems without conventional pretreatment. One remarkable property is its
chemical stability in harsh or corrosive environments. Due to weak molecular
adsorption, the hydrogen-terminated electrode resists deactivation during long-term
exposure to air or electrolyte solutions. Conventional Sp2 carbon electrodes, on the other
hand, ofien deactivate quickly during exposure to electrolyte solution or to air.‘8 For this
reason, frequent pretreatment such as polishing must be performed and utmost care must

be taken in order to achieve a clean surface with reproducible properties.

Diamond thin films exhibit a large working potential window in aqueous media.
The material exhibits large overpotentials for oxygen and hydrogen evolution presumably
because there are few low energy sites available for the adsorbed reaction

- ' 9-21
Intermedlates. I

Figure 1.4 shows a cyclic voltammetric i-E curve for a diamond thin
film in 1.0 M KCl. The working potential window, defined as the potential range
between the points at which anodic and cathodic currents reach iSO uA, is about 3.5 V.
The wide working potential window permits the electrochemistry of redox couples with
large positive or negative standard reduction potentials to be studied, something that
would otherwise be imposssible due to interference from solvent electrolysis. Diamond
also exhibits a low and featureless background current. This results in enhanced signal-
to-background (S/B) ratios in electroanalytical measurements.l7

Hydrogen-terminated diamond electrodes, in contrast with oxygen-containing sp2
carbon electrodes, are relatively inert to the adsorption of polar molecules as shown by
experiments conducted on the adsorption of anthraquinone-Z,6-disulfonate (AQDS) to

glassy carbon, graphite and diamond electrodes.22

Physisorption Of AQDS was not
observed on diamond presumably because of the relative absence of polar carbon-oxygen
functional groups which promote strong dipole-dipole and ion-dipole interactions with
the adsorbate. Diamond electrodes are thus quite useful for electroanalytical

measurements in which the analyte of interest exists in a complex mixture containing

contaminants that tend to adsorb and foul conventional carbon electrodes.17

60 ,____ ~ ~e~~~—-—w-—~——~ —-

 

|
l
404 l

20*

Current (uA)
e

 

-60 7 1 T T r i
-2000 ~1500 -1000 -500 0 500 1000 1500
Potential (mV vs. Ag/AgCI)

 

Figure 1.4. Background cyclic voltammetric i-E curves in 1.0 M KCl for a

microcrystalline diamond film. Scan rate = 100 mV/s. Electrode geometric area = 0.16

cmz.

Polycrystalline diamond thin-film electrodes have been shown to outperform glassy

carbon, a conventional sp2 carbon electrode, for the electrochemical detection of

azide,”24 trace levels of metal ions,23 chlorpromazine,23 aliphatic polyamines,”27 and

chlorinated phenols28 using flow injection analysis. The detection of carbamate
pesticides,29 NADH,3O uric acid,31 and histamine and serotonin32 has also been reported.
Untreated diamond was shown to have a slightly lower sensitivity for 4-methyl catechol
and dopamine as compared to freshly polished glassy carbon, but the detection limit and

response precision are superior due to lower background current and noise for the

23

former. Diamond microelectrodes have been used for the electrochemical detection of

catechol and dopamine in capillary electrophoresis, and exhibited low background

current and noise, good response precision, wide linear dynamic range, and limits of

detection at fmol level.33

The modification of diamond microfiber electrodes with
overoxidized polypyrrole (OPPy) has been shown to enhance the response selectivity,
sensitivity and stability for the amperometric detection of dopamine in the presence of in
vivo interfering compounds, such as ascorbic acid and 3,4-dihydroxyphenylacetic acid.34

The goals of the research described herein were (i) to study the electrochemical
behavior of various catecholamines and metabolites at microcrystalline and
nanocrystalline diamond electrodes as a function of the solution pH, and (ii) to evaluate
the performance of both diamond electrode types for the amperometric detection of these
analytes when coupled with HPLC. This research was motivated by our desire to better
understand how the surface microstructure and chemistry of diamond electrodes affect
the electrochemical response for these compounds and to improve the detection figures of
merit in HPLC, as compared to what can be achieved with commonly used sp2 carbon
electrodes. A table of the molecular structure and pKa of the analytes used in this
research is presented in Table 1.2.35

Cyclic voltammetry was used to investigate the electrochemical behavior of various
catechols and catecholamines at two diamond types - microcrystalline and
nanocrystalline diamond thin films. Comparison measurements were made with glassy
carbon. One main issue addressed was whether or not the redox chemistry of these
analytes proceeds through an adsorbed state as in the case for sp2 carbon electrodes.

Another issue studied was the pH-dependence of the background current and the

voltammetric response of the analytes at the hydrogen-terminated diamond electrode.

Table 1.2. Molecular structure, pKa35 and formula mass of the analytes.

 

 

 

 

 

 

 

 

 

 

 

 

 

H,CO NH2

pK. Formula
Analyte Molecular structure values Mass (g/mol)
catechol (CAT) ”OI: 9.50 1 10.1 1
HO
4-methyl catechol (MC) ”OD/“‘3 9.69 124.14
HO
4-t-butyl catechol (BC) “OD/“(“193 9.66 166.22
HO
dopamine (DA) “0m N”2 9.41, 9.99 153.18
HO
3,4—dihydroxyphenylacetic “0 COO” 4.42 168. 1 5
acid (DOPAC) HO
norepinephrine (NE) 0” 9.57, 8.30 169.18
HO
epinephrine (EP) 0” 9.60, 9.16 183.20
Hofi/ NHCH3
HO
3,4-dihydroxyphenylglycol 0” 9.55 1 70. l 6
(DOPEG) Homo”
HO
vanillylmandelic acidl 0” 3.42 198.17
(V MA) HJCOUKCOOH
HO
metanephrine (MN) 0” 9.78, 9.13 197.23
H,COUI\/ NHCH3
HO
normetanephrine (NMN) 0| H 9.76, 8.25 183.20

 

:1:
O

 

 

 

10

 

Amperometric detection using both diamond types was employed with flow
injection analysis and reversed-phase liquid chromatography to assay for the analytes.
The electrochemical performance of diamond was compared with that of glassy carbon.
The goal was to determine if diamond electrodes provide significantly improved
detection figures Of merit (e.g., linear dynamic range, sensitivity, limit of detection,
response precision and response stability) for the catecholamines and metabolites as
compared to glassy carbon. The bioanalytical applicability of the assay was
demonstrated in a preliminary way using plasma samples from norrnotensive and
hypertensive rats. The work was our initial effort at applying HPLC-EC with diamond

electrodes for monitoring catecholamine levels in biological fluids.

11

Chapter 2

EXPERIMENTAL METHODS

2.1. Boron-Doped Diamond Thin-Film Growth

The procedure for diamond film growth has been described elsewhere.”’36 The
diamond thin films were deposited on p-type Si(100) substrates (~10'3 fl-cm, Virginia
Semiconductor Inc.) using a commercial microwave-assisted chemical vapor deposition
(CVD) system (1.5 kW, Astex, Inc.). The substrate surface was mechanically scratched
on a felt polishing pad with 1 pm diameter diamond powder (GE Superabrasives). The
scratched substrate was then extensively washed with ultrapure water, isopropyl alcohol
(IPA), acetone, IPA, and ultrapure water to remove the polishing debris. During the
scratching process, some Of the diamond particles are embedded into the surface. Both
the clean polishing striations and the embedded diamond particles serve as nucleation
sites for film growth. The cleaned substrate was then placed in the CVD reactor.

The microcrystalline diamond films were grown using a methane (CH4) / hydrogen
(H2) volumetric ratio of 0.50% at a total flow of 200 sccm, a microwave power of 1000
W, a growth pressure of 45 torr, an estimated substrate temperature Of 800 °C, and a
growth time of 10 h. Ultrahigh purity (99.999%) CH4 and H2 were used as the source
gases. Diborane (Bsz) diluted in hydrogen (0.1%) was used for boron doping. The
concentration of BZH6 added to the source gas was 1 ppm. The plasma was ignited with

all gases flowing to initiate the deposition. After growth, the CH4 and Bsz flows were

12

stopped and the films remained exposed to the H2 plasma for an additional 10 min. After
this period, the microwave power and system pressure was gradually reduced down to
400 W and 10 mtorr over a lS-min period to cool the samples in the presence of atomic
hydrogen. Post—growth treatment in atomic hydrogen is essential for removing
adventitious non-diamond carbon impurity and ensuring full hydrogen surface
termination. The film thickness was 4 - 6 pm and the resistivity was 0.01 D-cm, or less.
The nanocrystalline diamond films were synthesized using ultrahigh purity CH4,
Ar, and H2 (99.999%) at flow rates of 1, 94, and 5 sccm, respectively. B2H6, diluted in
hydrogen (0.1%), was used for boron doping. The concentration added to the source gas
was 1 ppm. The microwave power and deposition pressure were maintained at 800 W
and 140 torr, respectively. The substrate temperature was estimated by optical pyrometry
to be about 800 °C. The deposition time was 2 h. The plasma was ignited with all gases
flowing to initiate the deposition. At the end of the deposition period, the CH4 and B2H6
flows were stopped and the Ar and H2 flows continued. The film was then exposed to an
H2/Ar plasma for 10 min. The substrate was then cooled in the presence of atomic
hydrogen to an estimated temperature Of less than 400 °C by slowly reducing the power
and pressure over a 4-min period. The resulting diamond films were approximately 4 - 7

pm thick with a resistivity of 0.3 Q-cm, or less.

2.2. Boron-Doped Diamond Thin-Film Characterization

2.2.1. Atomic Force Microscopy

Atomic force microscopy (AFM) was used to examine the diamond film

morphology.23 The measurements were performed using a Nanoscope III instrument

13

(Digital Instruments) in the contact mode. Si3N4 probe tips, mounted on Au cantilevers
(100 um) with spring constants ranging from 0.54 to 0.06 N/m, were used to acquire
topographical images of the films in air.

Typical AFM images of a (A) microcrystalline and (B) nanocrystalline diamond
film are presented in Figure 2.1. These images show that the nanocrystalline film has
smoother surface, smaller grain size, and greater fraction of grain boundary compared to
the microcrystalline film. Microcrystalline films are well-faceted with a predominance of
octahedral and cube-octahedral crystallites. The nominal crystallite size is in the range of
1 to 3 pm.28 Nanocrystalline films consist of nodular features of ~50 — 100 nm in
diameter. These features are actually aggregates of ~15 nm diameter grains of diamond

that form as a result of the high rate of nucleation.36

 

Figure 2.1. AFM images (contact mode, air) of a (A) microcrystalline and (B)

nanocrystalline diamond thin film.

2.2.2. X-ray Diffraction
X-ray diffraction (XRD) was employed to characterize the bulk crystallinity of

the diamond thin films.“ A Rigaku Rotaflex diffractometer equipped with a rotating

anode and a Cu K-alpha radiation source was used to Obtain the XRD patterns. The thin
film rotates by an angle, 0, with respect to a fixed x-ray source, and the detector rotates
by 20 to measure the reflected light. The measurements were made from 20 Of 30° to
100°.

Figure 2.2 shows the XRD spectra for a (A) microcrystalline and (B)
nanocrystalline diamond thin film. For both films, three reflections are observed at 20
values of 44°, 76°, and 92°, and are assigned to the (111), (220), and (311) planes of
cubic diamond, respectively.36 The peaks for the nanocrystalline film are broader than

those for the microcrystalline film due to the smaller grain size. This result confirms that

both film types possess a crystalline diamond structure.

 

 

A (111)

E (220)

15 (311)
E“

3 B

E.

 

 

 

J1 --JL 1”

1 l
I

 

30 40 50 60 70 80 90100
2-Theta (degrees)

Figure 2.2. XRD pattern for a (A) microcrystalline and (B) nanocrystalline diamond thin
film.

15

2.2.3. Raman Spectroscopy

Raman spectroscopy was used to evaluate the film microstructure and to detect the
presence of nondiamond spz-bonded carbon impurity.36 Raman Spectra were obtained
with a Raman 2000 spectrometer equipped with a microprobe attachment (2080) and a
direct video module (2090) (Chromex, Inc.) in a 180° back-scattered collection mode.
The spectra were obtained using a 1200 groove/mm holographic grating. Excitation was
from a frequency-doubled NszAG (532 nm, 30 mW) laser. The laser beam was focused
to a spot size of approximately 5 pm resulting in an estimated power density of 150
kW/cmz. Spectra were collected with a 10 s acquisition time. The spectrometer was
calibrated with a piece of single crystal (Type Ila) diamond standard (100 orientation,
Harris diamond) prior to use.

Figure 2.3 shows Raman spectra for microcrystalline and nanocrystalline diamond.
A sharp peak at 1332 cm'1 is seen for the microcrystalline film and is characteristic of
high quality diamond originating from the sp3—bonded carbon structure. It is the first-
order phonon for diamond. For nanocrystalline diamond, this peak has a broader peak
width and sits atop a large background signal. There are two possible explanations for
the line broadening. The first is the surface confinement model which states that the
smaller the domain size, the larger the range of phonon modes that are allowed to
participate in the Raman process. The broad line width is due to the spread in phonon
energy. Another explanation is phonon scattering by impurities and defects (i.e., grain
boundaries). The scattering shortens the phonon lifetime resulting to the broadening of
the Raman line.37 Birrell et al. recently proposed that the origin of this peak is spz-

bonded carbon in the grain boundary and not sp3-bonded carbon.38 The peak at 1150

16

cm'1 is a signature for high quality nanocrystalline diamond, but its origin has not yet
been confirmed. Ferrari and Robertson assigned this peak to spz-bonded carbon

39 The 1550 cm'I peak is assigned to

transpolyacetylene segments at grain boundaries.
disordered sp2 carbon in the grain boundaries. It is downshifted from the expected 1580
cm’1 position as in graphite because the spz-bonded carbon is amorphous and is mixed
with sp3-bonded carbon. In the nanocrystalline film, the spz-bonded carbon is confined to

the abrupt grain boundaries which consist of a network of 3- and 4-fold coordinated

carbon atoms.40

 

 

13326Ei1

 

 

 

 

 

M K

I

800 1000 1200 1400 1600 1800 2000

 

B 1332 cm"1
I 1550ch1

I

Intensity (Arbi. Units)

1150 cm"

I

 

 

 

I T T

800 1000 1200 1400 1600 1800 2000
Raman Shift (cm“)

Figure 2.3. Visible Raman spectra for a (A) microcrystalline and (B) nanocrystalline

diamond thin film. Ac, = 532 nm. Laser power = 30 mW. Integration time = 10 s.

2.2.4. Contact Angle Measurement

The contact angle for water was measured to determine the relative
hydrophobicity of the diamond thin-film surface. A hydrogen-terminated surface is more
hydrophobic than an oxygen-terminated one and should exhibit a larger contact angle for
clean water. For example, Ostrovskaya and coworkers found that hydrogenation of CVD
diamond films in an H2 plasma increases the hydrophobic nature of the surface, as the
contact angle for water increases from 0 = 32° for an oxidized surface to up to 0 = 93°
after hydrogenation.“

An automated contact angle apparatus (First Ten Angstroms) was used to deliver
a 10 uL drop of deionized water from a 50-mL syringe on to the diamond surface. The
drop was formed at the end of the blunt, stainless steel syringe needle and the diamond
film was then raised Slowly to touch the drop. Images of the static drop on the film
surface were captured with a video camera. The contact angle was obtained using image
analysis software which fitted a curve to the drop image and calculated the line tangent to
the curve. The angle between the tangent line and the electrode surface was measured at
the left and right sides of the drop. The right and left angles were averaged to give one
contact angle value. Five measurements were made on a single film, with the electrode
rinsed with acetone and dried under N2 between measurements. Four microcrystalline

diamond thin-film electrodes were tested and gave contact angles of 69.2 at 55°, 64.0 i

30°, 72.8 :1: 23°, and 75.3 :I: 5.1°.

18

2.2.5. Cyclic Voltammetry

Electrochemical characterization of the films was carried out by cyclic voltammetry
(CV) using Fe(CN)63’/4' and Ru(NH3)(,3+/2+ as redox test systems. The physical, chemical
and electronic properties of a diamond thin-film electrode influence the electrode reaction
kinetics and mechanisms for these systems.

Solutions of 1.0 mM potassium ferrocyanide (K4Fe(CN)6, Aldrich) and 1.0 mM
hexaamine ruthenium(III) chloride (Ru(NH3)6C1, Aldrich) were freshly prepared in 1.0 M
potassium chloride (KCl, Fisher) solution. CV was performed with a CS-1090 digital
potentiostat (Cypress Systems) using a single compartment, three-electrode glass cell.

The cell design is Shown in Figure 2.4.

- _ _ _Auxiliary
electrode

   
  
 

Reference _ _ _

electrode -- - - N2 purge

Clamp

 

Working
electrode

Figure 2.4. Diagram of the single compartment, glass electrochemical cell.

The planar working electrode was pressed against a smooth glass joint at the bottom
of the cell. A VitonTM o-ring was used to define the geometric area exposed to the
solution. A large-area carbon rod served as the auxiliary electrode and an Ag/AgCl
electrode (3 M NaCl, Cypress Systems) was used as the reference. All potentials are
reported versus this reference.

Typical CV curves for the two redox analytes are shown in Figure 2.5. A summary of

the CV data is presented in Table 2.1. The CV data noted are the oxidation and reduction

19

peak potentials (Ep0x and Epred) and the oxidation and reduction peak currents (ip0x and
if”). Of particular importance is the peak potential separation (AEp = Epox - Epred), which
reflects the rate of the redox reaction. The electrode kinetics for F e(CN)(,3 "/4' system are
strongly influenced by the physicochemical properties of diamond. AEp is very sensitive
to the surface chemistry with the lowest values seen for clean surfaces with low oxide
coverage. Surface oxygen apparently blocks a surface site that is involved in the reaction
on the hydrogen-terminated surface.42 The kinetics of the Ru(NH3)(,3+/2+ system are much
less sensitive to the physicochemical properties of both sp2 carbon and diamond
electrodes. AEp, for example, is insensitive of surface modification indicating that
electron transfer step does not involve an interaction with a surface site or functional
group.43 AEp is most influenced by the electronic properties of the electrode particularly

the density of electronic states near the formal potential of the redox system. AEp

increases as the boron doping level of the diamond electrode decreases.44

 

 

 

 

 

 

so - » 60 “— ‘* ’*‘ _ ' T.
A B a
a 40 — 30 ‘ i
V l
*5 0 0 “
O
t .40 S
5 -30 l
-80 . . . " -60 . . ,
.400 0 400 800 -700 -400 -100 200

Potential (mV vs. Ag/AgCl)
Figure 2.5. Cyclic voltammetric i-E curves for (A) 1.0 mM Fe(CN),3"4' and (B) 1.0 mM

Ru(NH3)63+/2+ at a microcrystalline diamond thin film. The supporting electrolyte was

1.0 M KC1. Scan rate = 100 mV/s. Electrode geometric area = 0.16 cmz.

20

Table 2.1. Cyclic voltammetric data for the redox test systems at a microcrystalline

diamond thin-film electrode.

 

 

 

Analyte AEP (mV) E9” (mV) ip‘“ (uA) i.,'”‘/ipred
1 mM Fe(CN),3"“/1 M KCI 91 323 62.2 1.00
1 mM Ru(NH3),3+’2+/1 M KCl 66 -131 45.8 1.02

 

 

 

 

 

 

Note: Experimental conditions as described in Figure 2.5.

2.3. Glassy Carbon Polishing Procedure

The glassy carbon electrode (GC-30, Tokai) was polished with alumina powder
(Buehler) slurried in ultrapure water (~18 MQ-cm, Bamstead E-Pure) on a felt polishing
cloth (Buehler). Alumina of decreasing grit size was used: 1.0, 0.3 and 0.05 um. The
electrode was sonicated in ultrapure water for 10 min after each polishing step. At the
end of the polishing sequence, the electrode was thoroughly rinsed with water and
allowed to soak in distilled IPA for 10 min prior to use. The electrode was used in

electrochemical measurements immediately after polishing.

2.4. Flow Injection Analysis

Flow injection analysis (FIA) was performed using a home-made thin-layer
electrochemical flow cell with a cross-flow geometry. The experimental set-up has been
described elsewhere.25 A block diagram of the FIA system and a schematic of the thin-

layer electrochemical flow cell are shown in Figures 2.6 and 2.7.

21

Pulse Injection Grounded
dampener valve cage

 

 

 

 

 

 

 

    
  
 
  

@— Flow
cell I
Waste
N2 Eluent
Figure 2.6. Block diagram of the FIA system.
Reference
electrode 0111.191 /
Auxrlrary
electrode
Inlet
Kel-F
. Rubber gasket
Working . 5
electrode g-_____-g‘ Cu foil

Kel-F

 

Screw clamp

Figure 2.7. Schematic of the thin-layer flow cell.

The flow rate of the carrier solution was regulated with a consta-Metric III
metering pump (Milton Roy), through a model 7125 injector valve (Rheodyne) with a 20-
uL injection loop, and into the flow cell. The carrier solution, flowing at 1.0 mL/min,
was constantly Sparged with nitrogen. Pump noise was reduced with a Model LP-21
pulse dampener (Scientific Systems). Amperometric detection was used and the current
was recorded in analog form using a strip chart recorder. The entire set-up was grounded
and the flow cell was housed in a Faraday cage to reduce the pick-up Of electrical noise.
The flow cell was constructed with two pieces of Kel-F. The top piece contained the
inlet and outlet for the fluid flow and a place for the reference electrode. The outlet port
was fitted with a short piece of stainless steel tubing that served as the counter electrode.
The bottom piece supported the working electrode. Electrical contact was made by

pressing a piece of copper foil against the back side of the planar working electrode. A

22

rubber gasket separated the surface of the working electrode from the top piece of the
cell. A rectangular groove (0.80 cm x 0.15 cm x 0.10 cm) was cut into the gasket and
this defined the flow channel. Assuming a 25% compression of the gasket when the top
and bottom pieces of the cell were clamped together, the cell volume was estimated to be
~9 uL. An Omni-90 analog potentiostat (Cypress Systems) was employed to control the
potential applied to the working electrode and to measure the current during
amperometric detection. The potential was applied with respect to a commercial

Ag/AgCl mini-electrode (3 M NaCl, Cypress Systems).

2.5. High Performance Liquid Chromatography

Reversed-phase high performance liquid chromatography (HPLC) was performed
using a commercial system (Shimadzu Corp.) which consisted of a controller (SCL-
10AVP), a solvent delivery module (LC-10ADVP), a degasser (DGU-14A), and an auto-
injector (SIL-lOADVP). The mobile phase flow rate was maintained at 0.8 mL/min. The
samples (20 uL) were injected onto a 5 pm, 4.6 mm x 150 mm Alltech Adsorbosphere
HS C13 column. The same thin-layer flow cell described above was used for
electrochemical detection. An Omni-90 analog potentiostat (Cypress Systems Inc.)
controlled the potential applied to the working electrode and measured the current during

detection. Data were acquired in digital form with Class VP 7.0 software (Shimadzu).

2.6. Chemicals and Reagents

Reagent-grade quality chemicals and ultrapure distilled and deionized water (~18

MQ-cm, Bamstead E—Pure) were used to prepare all solutions. Sodium phosphate

23

monobasic (monohydrate, NaH2PO4, CCI), sodium phosphate dibasic (heptahydrate,
Na2HPO4, Spectrum), sodium chloride (NaCl, Spectrum), sodium hydroxide (NaOH,
Spectrum), phosphoric acid (H3PO4, 99.999+%, Aldrich), methanol (MeOH, HPLC
grade, Spectrum) and l-heptanesulfonic acid (sodium salt, 98%, Aldrich) were used to
prepare the carrier solution/mobile phase for F IA and LC. For some F IA measurements,
a Kreb’s buffer was used as the carrier solution and was prepared as follows: 117 mM
NaCl, 4.7 mM KCl, 2.5 mM calcium chloride (anhydrous, CaCl2 Mallinckrodt) , 1.2 mM
magnesium chloride (MgCl2, 98%, Aldrich), 25 mM sodium bicarbonate (N3HCO3,
Spectrum), 1.2 mM NaH2PO4, and 11 mM glucose (anhydrous, Mallinckrodt). Bovine
albumin (98%, Sigma) was used as an interfering biomolecule.

The catechol (99+%, Aldrich), 4-methyl catechol (95+%, Aldrich), 4-tert—butyl
catechol (97%, Aldrich), dopamine (hydrochloride salt, Sigma), norepinephrine (L-
bitartrate hydrate salt, 99%, Aldrich), epinephrine (99%, Aldrich), 3,4-dihyroxyphenyl
acetic acid (Sigma), normetanephrine (hydrochloride salt, 98%, Sigma), metanephrine
(DL-m-O-methylepinephrine hydrochloride, Sigma), vanillylmandelic acid (a,4-
dihydroxy-3-methoxybenzeneacetic acid, Sigma), DL-3,4-dihydroxyphenylglycol
(Sigma) were used without any additional purification. Stock solutions were prepared
monthly in 0.1 M perchloric acid (HClO4, 99.999%, Aldrich) and stored in the
refrigerator. Dilution of the stock solution with the appropriate solvent to the desired
concentration was done daily.

All glassware was cleaned by rinsing in a KOH/ethanol bath, then in an Alconox

solution, and finally with ultrapure water.

24

Chapter 3

BORON-DOPED DIAMOND THIN-FILM ELECTRODES
FOR THE ELECTROCHEMICAL DETECTION OF
CATECHOLS AND CATECHOLAMINES

3.1. Cyclic Voltammetric Behavior of Catechols and Catecholamines

3.1.1. Introduction

Some molecules, such as catecholamines, that serve as messengers between
biological cells, are easily oxidized and therefore can be electrochemically detected.
Adams and coworkers pioneered the use of electrochemistry to monitor neurotransmitters
in the 19705.45 Their motivations were to develop a sensitive catecholamine assay for use
with brain tissue and cerebrospinal fluid, and to use carbon microelectrodes for the in
vivo monitoring of catecholamines and metabolites. Over the years, electrochemical
methods have been used extensively for neurochemical monitoring with much emphasis
on enhancing the response sensitivity and selectivity.“2 Fast-scan cyclic voltammetry
(CV) and chronoamperometry are two most widely used techniques.46

CV is a powerful tool for generating kinetic and mechanistic information about an
electroactive compound. The technique involves linearly scanning the potential applied
to an electrode immersed in an unstirred solution, from an initial value to a final or
switching value and then back to the initial value. A potentiostat controls the potential
applied to the working electrode with respect to a reference electrode, and measures the

current that flows between the working and counter electrodes. The resulting data is a

25

plot of current versus potential called a cyclic voltammogram.” 48 CV is a very versatile
technique that can be used to study the thermodynamics Of redox processes, to evaluate
the effects of various electrode surface conditions on a redox reaction’s kinetics and
mechanism, and to investigate coupled chemical reactions and adsorption processes. The
technique can also be used to determine the redox potential of the electroactive species.49

It can be difficult to use CV to analyze complex mixtures because of the limited
selectivity the technique affords, especially when multiple analytes that undergo redox
reaction at Similar potentials are present . Despite this limitation, CV is routinely
employed for the analysis of neurotransmitters and related compounds using
microelectrodes at fast scan rates.50 Carbon fiber is the most commonly used electrode
type for these measurements. The technique offers good temporal resolution to follow
the rapid concentration dynamics of the analyte.50 Identification of a redox-active species
is possible based on the shape of the cyclic voltammogram. It should be noted that the
response time, selectivity, and sensitivity are critically dependent on the condition of the
electrode surface.46 The control of electrode surface properties has been investigated
extensively in order to optimize the electrode performance for the electrochemical
detection of catecholamines. For example, selectivity is enhanced by coating electrodes
with Nafion, which acts as an electrostatic gate that preconcentrates cationic species
(catecholamines) and repels anionic interferents (ascorbic acid).51 The effects of surface
pretreatments such as electrochemical polarization, vacuum heat treatment, anaerobic
polishing and chemisorption of molecular adlayers on the electrode reaction kinetics for
catechols and catecholamines at glassy carbon (GC) has been studied thoroughly by

McCreery and coworkers. Fast electron transfer has been correlated with enhanced

26

catechol adsorption on the electrode surface with a “self-catalysis” mechanism being
proposed.52

In this Chapter, cyclic voltammetric studies of the electrochemical behavior Of
dopamine (DA), catechol (CAT), and 3,4-dihydroxyphenylacetic acid (DOPAC) at
microcrystalline and nanocrystalline diamond thin-fihn electrodes are discussed. The
molecular structure and pKa value” of each analyte are shown in Figure 3.1. These
analytes were selected in order to probe the electrode response for a cationic, neutral, and
anionic redox system at pH 7. The CV curves were used to assess whether or not
adsorption is a part of the electrode reaction mechanism at the two diamond types. The
behavior for CC is also presented for comparison. The pH-dependence of the
electrochemical response was also studied with focus on understanding the order of H+
and e' transfer for the redox analytes and probing the effects of the electrode surface

chemistry on the electrode response. Again, studies with GC are reported, for

comparison.
HO NHZ HO HO

:./\COOH
HO HO HO

DA (pKa = 9.99) CAT (pKa = 9.50) DOPAC (pKa = 4.42)

Figure 3.1. Molecular structure and pKa value for DA, CAT and DOPAC.

3.1.2. Results and Discussion

Cyclic voltammetric background i-E curves were recorded for microcrystalline
diamond, nanocrystalline diamond and freshly polished GC in 0.1 M phosphate buffer

solutions (0.1 M NaH2PO4 + 0.1 M Na2HPO4) of pH 2, 3, 4, 5, 6, and 7. The electrode

27

geometric area was 0.16 cm2 and the potential was scanned from -500 to +500 mV vs.

Ag/AgCl at a scan rate (1)) Of 100 mV/s. Three measurements were made with each

electrode type. A different electrode was used for each measurement in the case of the

microcrystalline and nanocrystalline diamond films, while only one piece of GC was

used that was mechanically polished between measurements. The diamond films were

used as is without any activation pretreatment. Representative cyclic voltammograms at

pH 2 and 7 are shown in Figure 3.2. The anodic background current was measured at 0

mV and is plotted vs. pH in Figure 3.3.

 

 

 

 

 

 

 

 

 

 

 

4 2 -—- ~— —— ~ a ~ —
A

2 .4

-2 4

.4 I H I I I 1
2 -600 .400 -200 o 200 400 600
3. 4 — - rrrrrrr
e B !
ii". 2 ‘ i
5 .-
° M *
'2 0 “ .4571?- --— """ l
= ’ l
e l
.92 ‘2 ‘ E
O
a l
m -4 1 I I Y j 1

-600 .400 -200 0 200 400 600

4 1 — e

2 A

o ..

-2 -4

4 I r I I 1

-600 .400 -200 o 200 400 600

Potential (mV vs. Ag/AgCl)
Figure 3.2. Cyclic voltammetric i-E curves for (A) microcrystalline diamond,

(B) nanocrystalline diamond, and (C) GC electrodes in 0.1 M phosphate buffer, pH 2 (-—-)

and pH 7 (—). Scan rate = 100 mV/s. Electrode geometric area = 0.16 cmz.

28

2.0 l _,,-_ , _,--_,_,_ L ~ ELL -— .2..-- Wm.-- _

 

 

E 15 ~ I { i I I l
a I
g i
O 1.0 « l
3 i
E
0.5 ~
‘93 I i i i i i J
I: O o e O O O
0.0 r T .
0 2 4 6 3
pH

Figure 3.3. Background current measured at 0 mV vs. pH for microcrystalline diamond

(O), nanocrystalline diamond (I), and GC (A) electrodes in 0.1 M phosphate buffer.

2

Scan rate = 100 mV/s. Electrode geometric area = 0.16 cm . Each data point and error

bar represent the average and standard deviation of three measurements.

The magnitude of the background current for GC is about ten times greater than
that for microcrystalline diamond and about five times greater than that for
nanocrystalline diamond. A low background current is a common property of diamond
electrodes and lead to improved S/B and signal-to-noise (S/N) ratios in electroanalytical
measurements. There are three causes for the low background current at diamond. First,
compared to CC, there is much less contribution to the background current from
electroactive carbon-oxygen functionalities (pseudocapactitance) since the diamond
surface is hydrogen—terminated. GC contains surface carbon-oxygen functionalities,
some of which are electroactive (e. g., quinones/hydroquinones).53 The increased
background current for CC is also due to ionizable surface carbon-oxygen fimctionalities
(e. g., carboxylic acid), which are charged at pH above their pKa and may interact with

electrolyte ions.5 ' Second, the semimetal-semiconductor nature of the diamond electrode

29

causes a slightly lower density of surface electronic states near the Fermi level. The
lower surface charge density leads to a reduced accumulation of charge on the solution
side of the interface (e.g., double-layer charging). Third, the diamond surface may be
structured as an array of microelectrodes, with conductive regions isolated from more
insulating regions.17

The background current for both diamond types is independent of the solution pH,
consistent with the absence of ionizable surface carbon-oxygen functionalities. The
background current for nanocrystalline diamond is larger than that for microcrystalline
diamond presumably because of the increased fraction of exposed grain boundary. The
n-bonded carbon in the grain boundaries is a source of charge carriers which can increase
the capacitive component of the background current. Some of these carbons may contain
electrochemically active surface carbon—oxygen functionalities which would give rise to a
faradaic component in the background current.36

As evident in Figure 3.2, the cyclic voltammetric i-E curves for both diamond
types are featureless with no evidence for any surface redox chemistry. For GC at pH 2,
broad oxidation and reduction peaks are present between 200 and 400 mV. These peaks
are associated with redox-active surface carbon-oxygen functionalities. During
polishing, carbon-carbon bonds are cleaved, leaving dangling bonds on the surface.
Upon air exposure, atmospheric oxygen or water may react with the dangling carbon
bonds to form surface carbon-oxygen functional groups.54 Some of the resulting surface
functional groups (e. g., hydroquinones/quinones) are electroactive and cause the
background voltammetric features. The midpoint potential for the redox peaks shift

negative with increasing pH, but the peaks are not as prominent as they are in acidic

30

media. The peaks shift negatively with increasing pH with a charge of ca. -59 mV/pH
unit. This is consistent with equal numbers of protons and electrons being transferred.
The data in Figure 3.3 illustrate the greater variability in the background current for CC
than for diamond over this pH range. This difference is attributed to the pH-dependent
electrochemical behavior of the electroactive carbon-oxygen surface functionalities. It is
also evident that the standard deviation of the current for GC at all potentials is much
larger than the deviation for the two diamond types. It can be difficult to reproduce the
surface properties of GC during polishing, hence there is inherently some variability in
the response. The fact that diamond electrodes do not require pretreatment for activation
is an advantage for electroanalytical measurements. Figure 3.4 shows the cyclic
voltammograms for GC in phosphate buffer at pH 2, 4, and 7 during a potential sweep
from -200 to +1200 mV. Clearly, the redox peaks Shift to more positive potentials with

decreasing pH, and are most prominent at pH 2.

 

 

 

 

 

A H 2
i 2 pH 7 pH 4 p l I
E $31 '72'25';’;l /
I. 41:;12/ MT.”
5 0 ’x’ /
'U / _,-— : T "' ’3';
= /-"’:,7 T
5 1
E0 Va
:1 /
8 -2 ,
m /
400 0 400 800 1200

Potential (mV vs. Ag/AgCl)

Figure 3.4. Cyclic voltammetric i-E curves for GC in 0.1 M phosphate buffer, pH 2,

4, and 7. Scan rate = 100 mV/s. Electrode geometric area = 0.16 cmz.

31

The influential nature of the surface oxides on the magnitude of the background
current and on the electrode reaction kinetics at sp2 carbon electrodes has been discussed
by other research groups.54’55 For example, Deakin et al. Showed that the k°app value for
several outer-sphere redox system changed with pH.56 The rate was found to decrease
with increasing pH for Fe(CN)(,'3/“1 and IIC16'2/'3, and the opposite effect was observed for
Ru(NH3)(,+2/+3 . The behavior was consistent with the presence of functional groups on the
GC surface. Runnels et al. reported on the effect of solution pH and surface functional
groups on the voltammetric behavior of carbon fiber toward catecholamines.S4 The
authors observed redox waves at the background voltammetric response that shifted and
decreased in magnitude with increasing pH. The background voltammetric peaks for the
surface oxides on carbon fiber overlap those for neurotransmitters like dopamine.
Additionally, the electrode is quite responsive to the pH of the solution in contact with it.
This is a problem in neurochemical investigations in vivo because simultaneous changes
Of dopamine after exocytosis and pH occur in brain tissue. Hydrogen-terminated
diamond microelectrodes may be ideal for these types of applications because the
background current is low and insensitive to pH changes.

The cyclic voltammmetric response for 10 uM solutions of DA, CAT, and
DOPAC at microcrystalline diamond and GC are presented in Figures 3.5 and 3.6.
Measurements were made at pH 2 to 7, with the electrodes being soaked in distilled IPA
then rinsed copiously with ultrapure water between measurements. Data are presented

for two different pH values, 7 and 2.

32

2.0 It...” -- -— _,-._ ———#—~-~——~-- _-. _L________

1.0 i

0.0 -‘

-1.0 “

 

 

-2.0

 

 

 

 

 

 

 

 

-800 -400 0 400 800 1200
2.0 ,_._,__., —— —-- ~—— -
i 1.0 A
El
0 0.0 -«
L:
i-
:5
U -1.0 ~
-2.0 I H I T J
-800 -400 0 400 800 1200
2.0 ~ ————-- -—-~_—— ---— — ——— —————— — —— j
i
1.0 -< l
0.0 ~
-1.0 2 l
l
-2.0 I I 1 I i
-800 -400 0 400 800 1200
Potential (mV vs. Ag/AgCl)
Figure 3.5. Cyclic voltammetric i-E curves (background-subtracted) for

microcrystalline diamond (—) and GC (_) electrodes in 10 uM DA, CAT, and DOPAC

in 0.1 M phosphate buffer, pH 7. Scan rate = 100 mV/s. Electrode geometric area = 0.16

01112.

33

 

 

 

 

Current, uA

 

 

 

 

 

 

 

-2.0 I 1 1* fit
~800 -400 0 400 800 1200
2.0 _ -
1.0 —
0.0 .
-1.0 2 j
l
-2.0 1 T r 4
-800 .400 0 400 800 1200

Potential (mV vs. Ag/AgCl)

Figure 3.6. Cyclic voltammetric i-E curves (background-subtracted) for
microcrystalline diamond (——) and GC (_) electrodes in 10 uM DA, CAT, and DOPAC

in 0.1 M phosphate buffer, pH 2. Scan rate = 100 mV/s. Electrode geometric area = 0.16

C1112.

34

The curves for DA at the diamond electrodes at pH 7 show peaks besides those
that correspond to the dopamine quinone/hydroquinone redox reaction (Ep0x ~ 350 mV
and Epred ~ 50 mV). An extra pair of peaks at Ep0x ~ -225 mV and Epred ~ -260 mV (data
not shown) is also Observed for GC when the potential is scanned more negative. The
additional peaks and the quite low peak currents for duroquinone mean that there are
other stable Species formed during the electrode cycling. Duroquinone is a highly reactive
species, and is susceptible to nucleophilic attack in the 6-position. The nucleophilic
attack of aniline, cysteine, and glutathione has been observed.“58 The most common
nucleophilic addition is that of its own side-chain amine, producing 5,6-
dihydroxyindoline which is also oxidizable.59 Pertinent molecular structures are shown
in Figure 3.7. The extra peaks are attributed to 5,6-dihydroxyindoline and its oxidation
product. This intramolecular cyclization may only occur when there is sufficient
unprotonated orthoquinone available to allow the side-chain amine to act as a

nucleophile. Thus, the extra peaks were observed at pH 7, and to a lesser extent at pH 6.

 

(A) (C)

Figure 3.7. Dopamine (A) oxidizes to duroquinone (B). The 6-position is susceptible to

nucleophilic attack by the side-chain amine forming 5,6-dihydroxyindoline (C).

The cyclic voltammetric peak current for reversible and irreversible systems is

described by the Randles-Sevcik equation:60

1, = (2.69 x 105)n-"‘°AD‘"2C0”2

35

in which ip is peak current (A), n is electron stoichiometry, A is electrode area (cmz), D is
diffusion coefficient (cmz/s), C is concentration (mOl/cm3), and U is scan rate (V/s). This
equation relates the maximum current to various parameters when the reaction rate is
limited by semi-infinite linear diffusion of reactant to the interfacial reaction zone. The
peak currents at the three electrode types were found to vary linearly with the square root
of the scan rate, Om (R2 > 0.99) for all the analytes, which is indicative of diffusion

control. Using n = 2, A = 0.16 cmz, D = 6 x 10'6 cmz/s, C = 1 x 10'8 mol/cm3, and U = 0.1

X

V/s, ip was calculated to be 0.96 uA. Figure 3.8 shows plots of ip° vs. pH for

microcrystalline diamond, nanocrystalline diamond, and GC in 10 11M analyte solutions,
the theoretical iID value indicated by the dashed line. The ip0x values are lower at the two
diamond types compared to that at CC over the entire pH range investigated. The larger
current is attributed to preconcentration or adsorption of the analyte on GC. The catechol
redox reaction on GC involves both adsorption and diffusion transport.

The present findings also indicate higher ip0x values on nanocrystalline than on
microcrystalline diamond, for all the analytes at all pH, indicating some preconcentration
of the analyte occurring on the electrode surface. Nanocrystalline diamond has more
grain boundaries and thus a greater fraction of exposed spz—bonded carbon which was
reported to have a correlation with surface coverage of adsorbed catechols.

Nanocrystalline diamond exhibits greater adsorption than microcrystalline diamond.36

36

 

 

 

 

 

 

1.82 DA ‘
A ‘ ‘ A I
A I
1.32 -
A .
- . l
....... '.‘,
0.8— ' I l
0.3 1
1 2 3 4 s 6 7
A ... i
a 1.8i CAT
H
1:
E . .
a 1.3" ‘ ‘ ‘ ‘
U l
g _______ .U ...... _.- ....... .‘ ....... '.' ................... l
O
A. 0.8‘ ° 0 _ Q
5 i
a:
.g 0.3 T . . . . fl
5 1 2 3 4 s 6 7
18 DOPAC
‘ A
A A
1.34 ‘ l
i l
....... ‘2
. I 2
0.8 9 .
I
O
0.3 .
1 2 3 4 s 6 7
pH

Figure 3.8. Plots of oxidation peak current vs. pH for microcrystalline diamond (O),
nanocrystalline diamond (I), and GC (A) electrodes in 10 uM DA, CAT, and DOPAC
in 0.1 M phosphate buffer. Scan rate = 100 mV/s. Electrode geometric area = 0.16 cmz.

The theoretical ip value is indicated by the dashed line.

37

Figure 3.9 shows plots of E,” vs. pH for DA, CAT and DOPAC at
microcrystalline diamond, nanocrystalline diamond and GC. Such plots are Often used to
study the mechanism of an electrochemical redox reaction. Cyclic voltammetric
investigations as a function of pH and in conjunction with mathematical treatment permit
the determination of the order of the proton (H+) and electron (e') transfers on platinum“

and carbon55’62‘63

electrodes. For example, data on the kinetic studies of catechol
oxidation at a carbon paste electrode at neutral pH were found to be consistent with an
HeHe mechanism.55 Such a comprehensive task was not pursued in this research but it is
certainly interesting to make similar investigations with diamond electrodes.

Ep0x values reflect the ease of oxidation, that is, high Ep0x means that more driving
force is needed for the electron transfer reaction between the solution analyte and the
electrode. The higher Ep0x values of the catechol compounds at diamond, as compared to
that at GC, mean that the electrode reaction kinetics are more sluggish at the former.
This means that more positive potentials are required to oxidize the catechols and
catecholamines at diamond at a mass transfer-limited rate. This is not expected to hinder
the performance of diamond as a detection electrode due tO its wide working potential
window and low background current.

Another important parameter to consider is the peak separation (AEp = E],0x —
Epmd), which is related to the heterogeneous rate constant. A large AEp value is indicative
Of sluggish redox reaction kinetics. Table 3.1 shows the (IE, values for DA, CAT, and

DOPAC at pH 7. Figure 3.9 and Table 3.1 show that EPO" and AEp values follow the

trend: microcrystalline diamond > nanocrystalline diamond > GC. The Slow redox

reaction rate on the diamond electrodes is presumably due to the relative absence of

38

The Ep0x values of catechol compounds at diamond electrodes decrease with
increasing pH. In addition, a trend based on the charge of the analyte emerges, in
particular, higher E100" values are observed for negatively-charged analytes and lower for
those which are positively-charged. At pH 2, 3, and 4, when DOPAC is predominantly
protonated (neutral), Ep0x is lowest for DA followed by DOPAC then by CAT. The Ep0x
of DOPAC and CAT are almost equal at pH 5, right above the pKa Of DOPAC. At pH 6
and 7, when the unprotonated form Of DOPAC (negatively-charged) is dominant, the Er,0x
values follow the order DA < CAT < DOPAC. The AEp values for DA, CAT, and
DOPAC at pH 7 also Show charge discrimination on the three electrodes. Fastest
electron transfer, as evidenced by the least positive Epo", is observed for DA and slowest
electron transfer is seen for DOPAC. The general trends in the Epm‘ values at varying pH
that are observed on GC are also seen on diamond. Considering the disparity in their
electrode surface chemistries, a greater dissimilarity of the trends in the electron transfer
kinetics of the analytes on GC and on diamond is expected.

A trend based on analyte charge is an indication that an electrostatic mechanism is
in effect. The results presented for GC agree with previously published workf’5 that is,
that the rate of electron transfer is dependent on the analyte charge. The redox reaction is
fastest for the positively-charged DA and slowest for the negatively-charged DOPAC.
The electron transfer kinetics are governed by electrostatic interactions between the
analytes and the electrode surface. DA kinetics are enhanced while that of DOPAC are
inhibited since the surface oxides are neutral or anionic, and never cationic.“ On the

diamond surface, an electrostatic mechanism is unanticipated but is apparent.

41

adsorption. An increased electron transfer rate (lower Epox and AEp values) at
nanocrystalline diamond compared to microcrystalline diamond is observed. This is
another evidence of the occurrence of adsorption through the spz-bonded carbon on the

grain boundaries Of the nanocrystalline diamond surface.

 

 

 

 

 

800 — A
A
3 _ x
I . A
600 2 I
I
400 ~ g I
G I
To 200 T . . . . . I
g) 1 2 3 4 s 6 7
< _-_______ ._ , __ _ M. -_L_ I
"3 B
> 800 - i
> a 2 l
E I x A ‘
V I
'23 600 « ' °
0'; l
E :
I
c .
h 400 I
a: I
N
0
m 200 T 7 T T T If I
S l 2 3 4 s 6 7
'5
1, 300- - -——--
o: C
O ;
600 ~ 3 . l
a 0 .
- ° 1
I 6 ,
200 ~ ' l
0 r

 

 

pH
Figure 3.9. Plots of oxidation peak potential vs. pH for (A) microcrystalline diamond,
(B) nanocrystalline diamond, and (C) GC electrodes in 10 uM solutions of DA (I), CAT
(9), and DOPAC (A) in 0.1 M phosphate buffer. Scan rate = 100 mV/s. Electrode

geometric area = 0.16 cmz.

39

Table 3.]. AB], values of the catechol compounds at pH 7 at the three electrode types.

 

 

 

 

 

 

 

AED (mV)
DA CAT DOPAC
Microcrystalline Diamond 353 547 711
Nanocrystalline Diamond 233 422 516
Glassy Carbon 81 168 233

 

 

Note: 10 uM solutions of the analytes in 0.1 M phosphate buffer, pH 7 was used. Scan

rate = 100 mV/s. Electrode geometric area = 0.16 cmz.

Investigations done by McCreery’s group indicate that adsorption of the catechols

to the carbon electrode surface is necessary for an increased electron transfer rate. It was

found that surface pretreatments that inhibit adsorption result in high AB], and those that

64

promote adsorption lead to low AEp values. Their findings led to a self-catalysis

mechanism in which a layer of adsorbed quinones (physisorbed or chemisorbed quinone
or catechol) catalyzes the electron transfer between catechols and GC electrodes.
Hydrogen-bonding between a surface quinone oxygen and the catechol is a possible
catalytic step that can accelerate the overall redox process.52 Adsorption and fast electron
transfer translate to sharper peaks and higher peak currents.51

For sp2 carbon electrodes, a trade-off exists between the electrode’s enhanced
sensitivity through fast electron transfer kinetics and its susceptibility to polar adsorbates
that lead to fouling and decreased activity. Little is known about electron transfer
reaction mechanism of these compounds. A better understanding of the redox reaction
mechanism and the factors that affect the redox reaction rate on diamond would be
beneficial for improving the sensitivity and fouling resistance of this material in

electrochemical detection.

4O

It is possible that surface oxides may have also played a role in the case of
diamond electrodes. Electrochemically generated carbon-oxygen functional groups may
have resulted from the wide potential sweeps used to drive the redox reaction on
diamond. Change in diamond surface termination from hydrogen to oxygen have been
reported to occur at applied potentials as low as 1.0 V vs. SCE in acidic media."6 Angus
and coworkers have attributed a redox couple observed at 1.7 V vs. SHE to surface

oxidation.2|

Fujishima et al. used anodic treatment to replace surface hydrogen with
oxygen functionalities to achieve improved selectivity between DA and ascorbic
acid.67 Figure 3.10 shows possible oxygen functional groups found on sp2 and sp3
electrode surfaces."8 It is important to note that the oxidation of diamond differs from
that of Sp2 carbon electrodes in that the reactions are surface-confined and do not produce

any morphological and microstructural damage on diamond owing to the strong three-

dimensional bonding and high atomic density of diamond.66

2 3

SP

Sp
aromatic-bonded ' —-—-—-C—H aliphatic-bonded .
hydrogen \ hydrogen
————-C—OH
imOH phenols . ‘ alcohols
:Szo quinones ' C=O ketones
OH i //O
C/ aromatic acids —C carboxylic acids
\
\O OH

Figure 3.10. Surface functional groups on oxidized sp2 and sp3 electrodes.

42

Three trials were performed to determine the reproducibility of the cyclic
voltammetric measurements of the analytes on the three electrode types. Cyclic
voltammetryI was performed at U = 100 mV/s, and A = 0.16 cmz, with 0.1 M phosphate
buffer as the supporting electrolyte. The RSD values on the Ep"x and ip0x values are
tabulated in Table 3.2. The high RSD values for CC shows the difficulty in obtaining
polished GC surfaces with reproducible electrode characteristics. The microcrystalline
diamond films used were grown in one batch resulting to lower RSD values compared to

that of the nanocrystalline diamond films which were grown individually.

Table 3.2. Reproducibility Of the cyclic voltammetric measurements using the three

electrode types being studied (n = 3).

 

 

 

 

 

 

 

 

RSD Analyte
EDOX ipOX
Microcrystalline Diamond 1.4% 1.3% 5 uM CAT, pH 2
Nanocrystalline Diamond 3.9% 10.9% 5 11M CAT, pH 2
Glassy Carbon 8.5% 17.1% 10 11M DA, pH 7

 

The supporting electrolyte used was 0.1 M phosphate buffer. Scan rate = 100 mV/s.

Electrode geometric area = 0.16 cmz.

3.2. Amperometric Detection of Catechols and Catecholamines in
Flow Injection Analysis

3.2.1. Introduction

It was shown in the previous section that catechols and catecholamines can be
electrooxidized at boron-doped microcrystalline and nanocrystalline diamond thin films,

albeit with higher overpotential than GC. The results presented in this section

43

demonstrate the capability of diamond electrodes for the oxidative detection
(amperometric mode) and quantification of these analytes when coupled with flow
injection analysis (FIA). The goal of the work was to compare the performance Of both
microcrystalline and nanocrystalline diamond thin films with that of GC. A
comprehensive reporting of the detection figures of merit is presented.

FIA was conceived in the mid—19705 in order to process large numbers of samples
for analysis. FIA was developed as a solution to the slow productivity resulting from the
use of segmented flow analyzers (SFA).69 The FIA technique is based on the injection Of
a sample in a continuously flowing, nonsegmented carrier solution propelled by a
peristaltic pump. The injected sample forms a well-defined zone, which is then
transported through the detector. In one mode of operation, a detector continuously
records a physical parameter, such as absorbance (fixed wavelength) or electrode current
(fixed potential), and its change due to passage of the sample zone. The output is a
detector signal as a function of time, which can be used for quantitative analysis (peak

height or area).70‘71

The benefits of FIA include automated sample processing, high
repeatability, short analysis time, adaptability to microminiaturization, containment of
chemicals, and waste reductionn’73 FIA is more analogous to modern HPLC than to
SF A, such that it has been called "HPLC without the column and high pressures".73
Electrochemical detection relies on a compound undergoing oxidation or
reduction at an electrode to which a potential has been applied. The amperometric
detection mode provides the advantage of minimizing the charging current by operating

at a fixed potential as opposed to voltammetric techniques. The electrochemical flow cell

acts as the transducer between the potentiostat and the FIA or HPLC system. A wide

44

range of electrochemical cell designs have been used for flowing streams. The cell
design must have a high SfN ratio, low dead volume, well-defined hydrodynamics, small
Ohmic drop, and ease of construction and maintenance.49

The thin-layer cell is probably the most popular detector cell configuration. A
schematic of the homemade thin-layer flow cell used in this work was presented in
Chapter 2. A simple diagram is presented in Figure 3.11 to illustrate the solution flow
with respect to the working electrode. The carrier solution in FIA-EC and the mobile
phase in HPLC-EC act as the supporting electrolyte. Electrochemistry proceeds while a
thin layer of solution flows parallel across the planar electrode surface, which is
imbedded in a rectangular channel. Mass transfer occurs not only through diffusional
transport to the electrode surface, but also through forced convection due to the fluid
flow, hence electrochemistry applied to flowing streams is classified as a hydrodynamic
technique. The enhanced mass transfer translates to higher currents and thus higher S/N

49.74

ratios. The conversion efficiency, that is the amount of analyte electrolyzed as it is

swept past the electrode, is typically less than 5% in this design.75

Inlet Outlet

 

 

W orking electrode

Figure 3.11. The solution flow in a thin-layer electrochemical flow cell.

FIA coupled with amperometric detection was used to assess the performance of
boron-doped diamond for the measurement of catechols and catecholamines. The figures

of merit used for this evaluation are the linear dynamic range, sensitivity, limit of

45

quantitation, and response precision. In addition, the response stability and the resistance

to fouling by a biomolecule were evaluated.

3.2.2. Results and Discussion

Hydrodynamic voltammetric background i-E curves for microcrystalline
diamond, nanocrystalline diamond and freshly polished GC in 0.1 M phosphate buffer
with 0.1 M NaCl, pH 7.2, are shown in Figure 3.12. The carrier solution volumetric flow
rate (U) was 1.0 mL/min. The curves were obtained by changing the applied potential in
100-mV increments starting at 200 mV. The current magnitude and the peak-to-peak

noise were recorded once a stable signal was reached.

 

 

 

 

Background Current (nA)

200 400 600 800 1000
Potential (mV vs. Ag/AgCl)

Figure 3.12. F IA-EC hydrodynamic voltammetric background i-E curves for
microcrystalline diamond (O), nanocrystalline diamond (I), and GC (A) electrodes in

0.1 M phosphate buffer, 0.1 M NaCl, pH 7.2. Flow rate = 1.0 mL/min. Electrode

geometric area = 0.12 cmz.

After application of a detection potential, the background current for the diamond

films stabilized much faster than did the current for GC. For example, a stable

46

background current was easily reached within 10 min for the diamond films, while 40
min was required for GC at 700 mV. The quick stabilization time for diamond is
expected due, in part, to the low double-layer capacitance. Additionally, the diamond
surface is morphologically and microstructurally stable at anodic potentials so no
dynamic oxidation processes or microstructural alteration occur on the surface. Surface
oxidation does occur on diamond but is confined to a simple functional group exchange
on the surface.66 The rapid stabilization time of diamond leads to shorter overall analysis
times.

As was the case in the cyclic voltammetric measurements, the background current
is larger for GC than for the two diamond types. A low background current, as discussed
in the previous section, is a characteristic feature of diamond electrodes. For example,
the anodic background current at 400 mV was 0.24, 0.41, and 3.40 nA for
microcrystalline diamond, nanocrystalline diamond, and GC, respectively. All electrodes
had the same exposed geometric area. The rapid rise in background at potentials above
700 mV for GC is likely due to a combination of surface oxidation and onset of oxygen
evolution. The low background current on diamond leads to an enhanced S/B ratio.

Figure 3.13 shows the peak-to—peak noise for the three electrode types as a
function of applied potential. The noise levels are significantly lower for both diamond
types at potentials less than 800 mV, with the lowest noise seen for microcrystalline
diamond. At potentials positive of 800 mV, the noise levels start to be more comparable,
especially for the nanocrystalline diamond and GC electrodes. The lower noise levels for
diamond lead to improved S/N ratios in electroanalytical measurements and lower limits

of detection.

47

 

 

 

 

 

 

 

 

 

 

 

 

 

 

280‘ ~
3:

8

.5604

z

$540+

0

Ch

6

T20~

;:

re

0

In 0- A L

200 300 400 500 600 700 800 900 1000

Potential (mV vs. Ag/AgCl)

Figure 3.13. Peak-to-peak noise at microcrystalline diamond (I), nanocrystalline
diamond (B), and GC (El) electrodes in 0.1 M phosphate buffer, 0.1 M NaCl, pH 7.2, as

a function of applied potential. Flow rate = 1.0 mL/min. Electrode geometric area = 0.12

2
cm.

Figure 3.14 shows hydrodynamic voltammetric i-E curves for 20 [AL injections of
CAT, MC, BC, NE, EP, DA, and DOPAC, all 30 11M, at microcrystalline diamond,
nanocrystalline diamond, and GC electrodes. These curves were obtained by making
multiple injections (n 2 4) of each analyte at different applied potentials. The error bars
are within the size of the marker. The resulting hydrodynamic i-E curves generally have
well-defined, sigrnoidal profiles with mass transport limited currents. The limiting

current for a thin-layer flow cell is described by the equation:49

in... = 1.47nFC(DA/b)2/3U”3

48

where ih-m is the limiting current (A), n is electron stoichiometry, F is the Faraday
constant, C is concentration (mOl/cm3), D is diffusion coefficient (cmz/s), A is electrode
area (cmz), b is channel height (cm), and U is average volume flow rate (cm3/s).
Calculation ofihm, using 11 = 2, C = 3 x 10’8 mol/cm3, D = 6 x 10'6 cmz/S, A = 0.12 cmz, b
= 0.075 cm, and U = 0.017 cm3/s, gives a value of 0.98 11A. The experimental results
reveal currents all lower than this theoretical value. One possible reason for this is the
error in the value of b in the flow cell. A rubber gasket compression of 25% that was
used in the calculation is an estimate. There is always some uncertainty in b because the
actual channel height is a factor of the degree of compression, which is a function of how
tightly the screw clamp is made to hold the top and bottom pieces of the flow cell
together.

The optimum detection potentials (Eda) and oxidation half-wave potentials (E172)
are tabulated in Table 3.3. The Ede, is selected to be in the limiting current region.
Potentials in this region yield the most intense current and at these potentials the
electrode sensitivity is not greatly affected if the potentials happen to drift. Higher
potentials are required to detect the catechols and catecholamines at diamond, as
previously discussed. Fortunately, the more positive detection potential required for
diamond is not a significant detriment to the electrode performance due to the fact that
the microstructure and morphology are stable, which results in the maintenance of low

and stable background currents.

49

Current (ILA)

0.8 i

 

1.0, 77V , W

0.81 CAT
0.6
...I
all
0.0 ‘-
0 200 400
l
MC
200 400 600 800 1000 1200
NE
l
I
200 400 600 800 1000 1200

 

400

600 800

1000 1200

1000 1200

1.0 —— —-~
0.8 BC

0.6

0.4

0.2

 

800 1000 1200

1.0 .__
0.8
0.6
0.4
0.2

 

—‘r'_—.

400 600 800 1000 1200

 

0 200
Potential (mV vs. Ag/AgCl)

400 600 800 1000 1200

Figure 3.14. Hydrodynamic voltammetric i-E curves for 20 11L injections of 30 11M

catechols and catecholamines in microcrystalline diamond (O), nanocrystalline diamond
(I), and GC (A). The carrier solution was 0.1 M phosphate buffer, 0.1 M NaCl, pH 7.2.

Scan rate = 1.0 mL/min. Electrode geometric area = 0.12 cmz.

Table 3.3. Ede, and E1/2 values for the amperometric detection of the analytes at

microcrystalline diamond, nanocrystalline diamond, and GC.

 

 

 

Ede, (mV vs. Ag/AgCl) E1/2(mV vs. Ag/AgCl)
Micro- Nano— Micro- Nano-
Analyte crystalline crystalline GC crystalline crystalline GC
Dramond Diamond Diamond Diamond

CAT 1 100 1000 500 700 650 250
MC 1000 950 350 750 650 150
BC 1150 1050 350 800 650 150
NE 950 900 550 550 550 350
EP 1150 850 450 850 600 250
DA 1 150 850 400 700 550 200
DOPAC 1150 850 450 900 800 350

 

 

 

 

 

 

 

 

 

There are several trends evident in the hydrodynamic voltammetric data. First,
there is a difference in the electrode reaction kinetics for these catechols and
catecholamines as the E1/2 becomes more positive in the following order: GC <
nanocrystalline diamond < microcrystalline diamond. The more sluggish kinetics for this
class of redox analytes at diamond is well known. The electrode reaction kinetics are
presumed to be slower on the diamond electrodes because of a lack of adsorption.
Second, sigrnoidal voltammetric peaks are seen for all the analytes with limiting currents
that are Similar for a given analyte at all three electrodes. This observation is different
from the cyclic voltammetric data in which the oxidation and reduction peak currents
were consistently larger for GC, attributed to preconcentration or adsorption of the
analyte on the GC surface. Due to the short residence time in the flow cell brought about

by convection, there may be less adsorption occurring. Also, the additional convective

51

mass transport leads to more uniform currents for all these electrodes. The data presented
in Table 3.3 show that it would be difficult to distinguish the analytes from each other
because the Edct and E1/2 values are similar for each electrode type. For analysis of
solutions containing these analytes, a separation step, such as liquid chromatography or
capillary electrophoresis, is required prior to electrochemical detection for speciation.

In order to evaluate the effectiveness Of diamond for the analysis of catechols and
catecholamines, the analytical detection figures of merit (i.e., response precision,
sensitivity, linear dynamic range, limit of quantitation, and response stability) were
obtained and compared with those for GC. The response precision was evaluated from a
run of twenty consecutive 20-uL injections of 30 11M analyte at 1.0-min intervals. Three
runs were made with each electrode type. For the two diamond types, one electrode was
used for each run, while for GC, one piece was used for all three runs with polishing done
prior to each run. Table 3.4 presents the data for the response precision reported as the %
RSD, averaged from the three runs on each electrode type. The RSD values range from
1.4 to 2.4% for microcrystalline diamond, 1.2 to 2.0% for nanocrystalline diamond, and
1.5 to 5.0% for GC. The % RSD values between runs were also calculated and the values
ranged from 5 to 50%. The variance in the degree of compression which affects the
channel height is likely to be a major contributing factor in the run-tO-run variability as
the flow cell had to be disassembled and reassembled between runs to put a different test

electrode in place.

52

 

Table 3.4. Response precision (% RSD) for catechol and catecholamine detection at
microcrystalline diamond, nanocrystalline diamond, and GC in 0.1 M phosphate buffer,

0.1 M NaCl, pH 7.2.

 

 

Analyte Mrcggzc‘rnyzfghne N33223:?“ Glassy Carbon
CAT 1.4 1.3 1.7
MC 1 .5 1 .3 1.8
BC 1 .9 1 .2 1.7
NE 2.4 1 .4 2.9
EP 1 .5 1 .2 1.5
DA 1.5 1 .9 2.5
DOPAC 2.1 2.0 5.0

 

 

 

 

 

 

Note: The response precision was evaluated from 20 consecutive injections at 1.0-min
intervals. The data are reported as a mean for 3 runs. Analyte concentration = 30 11M.
Injection volume = 20 11L. F low rate = 1.0 mL/min. Electrode geometric area = 0.12

cmz. The detection potential used is as tabulated in Table 3.3.

Calibration curves for each of the analytes were generated from the detector
response as a function of the concentration in order to determine the response sensitivity.
Characteristic plots are shown in Figure 3.15. The curves are constructed from at least 5
concentrations with the lowest being the limit of quantitation (LOQ). The injected
volume was 20 uL and the carrier solution flow rate was 1.0 mL/min. The calibration
curves indicate that the current response increases linearly with analyte concentration for
all the catechols and catecholamines tested at all three electrode types, each plot having
R2 > 0.99 and a near-zero y-axis intercept. The slope of these curves is the electrode
sensitivity. The sensitivity is influenced, among other things, by the number of electrons

involved in the redox reaction and electrode area.l2

53

The linear dynamic range, LOQ, and the sensitivity for the catechols and
catecholamines at all three electrode types are summarized in Table 3.5. The linear
dynamic range for all the analytes studied, at all three electrode types, is 6 - 7 orders of
magnitude ranging from the nM to mM range. The mass LOQ for all the analytes are
similar, at all three electrode types, ranging from 10 to 80 pg. The sensitivity values for
all the analytes at all three electrode types are also similar with values in the 10 - 30

nA/uM range.

The detection figures of merit presented in Tables 3.4 and 3.5 indicate that the
diamond electrode performance is comparable to that of GC, albeit without any time-
consuming petreatment. The advantages of diamond in this electrochemical assay are (i)
the rapid stabilization time after detector tum-on which leads to shorter overall analysis
times, (ii) low background current and peak-to-peak noise resulting in high S/B and S/N
ratios, and (iii) good electrode performance without the need for pretreatment for
reactivation. The frequent polishing needed for GC is not only a tedious and lengthy
procedure but is likely a cause of increased variability between measurements.
Dismantling of the flow cell is necessary to gain access to the electrode for pretreatment.
During reassembly of the flow cell, as discussed previously, differences in the channel

height result from the different degrees of the compression of the rubber gasket.

54

 

 

 

 

 

Current (ILA)

 

 

 

 

 

 

 

 

 

Concentration (mM)

Figure 3.15. Calibration curves for the catechols and catecholamines obtained at (A)
microcrystalline diamond, (B) nanocrystalline diamond, and (C) GC electrodes. Note:
At least 5 concentrations were used for each curve but some plots appear to have fewer
points because more measurements were made at the low uM level. The carrier solution

was 0.1 M phosphate buffer, 0.1 M NaCl, pH 7.2. Flow rate = 1.0 mL/min. Injected

volume = 20 11L. Electrode geometric area = 0.12 cmz.

55

Table 3.5. FIA-EC data for catechol and catecholamine detection at microcrystalline

diamond, nanocrystalline diamond, and GC electrodes.

 

 

 

 

1...“. “negafiggam'c 8mg)” mg .. sm .3
Microcrystalline Diamond
CAT 10nM-3mM 19:4 22pg
MC 10nM-5mM 17:3 25pg
BC 10 nM-ZmM 17: 33 pg
NE 10nM-3mM 15: 34pg
EP 20nM-2mM 27: 73 pg
DA 10nM-3mM 17: 38pg
DOPAC 20 nM - 5 mM 11 : 67pg
Nanocrystalline Diamond
CAT 5nM-9mM 19:7 llpg
MC 20nM-6mM 21:9 50pg
BC 10nM—10mM 17:6 33pg
NE 10nM—7mM 17:7 34pg
EP 10nM—7mM 24:3 37pg
DA 10nM—5mM 16:9 38pg
DOPAC 10nM—4mM 9:4 34pg
Glassy Carbon
CAT 5nM-1mM 20:5 llpg
MC 20nM-2mM 131.7 50pg
BC 10nM-4mM 12:9 33pg
NE 10nM-4mM 171.1 34pg
EP 10nM-3mM 19:4 37pg
DA lOnM-lmM 17:4 38pg
DOPAC 10nM-2mM 10:6 34 pg

 

I The data are based on the response of three electrodes and are reported at 95%

confidence level. Correlation coefficients, R2, on the response curves are > 0.99.

56

Another series of experiments was performed to evaluate the diamond electrode
response stability and resistance to fouling. The FIA-EC response for 20 uL injections of
2 mM NE at microcrystalline diamond, nanocrystalline diamond and GC electrodes was
recorded over a 7-h period. Ten injections were made each hour with a Kreb’s buffer, pH
7.4, flowing continuously at 1.0 mL/min for the entire duration of the experiment. The
results are shown in Figure 3.16. Each data point represents the average of 10 injections.

The error bars are within the size of the marker.

U
U!

 

W
O

 

 

N
O
.l

Current (11A)

Time (h)

Figure 3.16. FIA-EC response for 20-uL injections of 2 mM NE at microcrystalline
diamond (O), nanocrystalline diamond (I), and GC (A) recorded over a 7-h period.
The carrier solution was a Kreb’s buffer, pH 7.4, flowing at 1.0 mL/min. Electrode
geometric area = 0.12 cmz. Each data point shown represents an average of 10 injections

made at 1.0 min intervals.

The current response over the period is quite stable for the two diamond types,
while the response for GC decreases over the period before reaching a relatively constant
value after approximately 4 h. At the end of the 7-h period, the current for GC decreased
by 11% while that of microcrystalline diamond and nanocrystalline diamond changed by
only 1% and 2%, respectively. Diamond possesses a better response stability than GC, in

part, because of the morphological and microstructural stability Of the material at the

57

detection potentials. The nonpolar, hydrogenated surface of diamond is also more
resistive than the polar, oxidized GC surface to contaminant adsorption. As a result,
diamond electrodes can be exposed to the laboratory atmosphere for days without any
response attenuation while GC needs to be polished prior to each use in order to remove
adsorbed contaminates and activate the electrode for electron transfer. The diamond
electrodes are copiously rinsed with ultrapure water prior to placing in the flow cell. No
response attenuation was Observed during the course of an electrode performance
evaluation which usually lasts for 2 — 3 days.

Figure 3.16 shows a GC current response markedly lower than that for the two
diamond types. This is unexpected based on the comparable sensitivity values on the
three electrodes as presented in Table 3.4. Possible explanations for this Observation are
the varying electrode activity that may result from each polishing and the varying channel
height.

The resistance of diamond electrodes to fouling was further investigated by
comparing the response for NE with and without added bovine serum albumin (BSA).
Serum albumin is a good test interferent because it makes up 60% of the total plasma
proteins in humans76 and it readily and irreversibly adsorbs to oxidized carbon surfaces.
Results for the two diamond types and GC are shown in Figure 3.17. The current
responses for ten 20-uL injections of 1 11M NE were recorded at 1.0-min intervals and are
represented by injection numbers 1 - 10. This was followed by another set of ten 20-uL
injections with 1 mM BSA added and are represented by injection numbers 11 - 20. A
17% decrease in current response Of GC is observed when BSA is present. The GC

response exhibits a gradual loss of signal with increasing injection number Of the BSA-

58

containing solution. The response attenuation is about two times greater than that
observed for the two diamond types, both of which showed an 8% decrease in response.
Although there is some fouling for the two diamond types, the extent of fouling is less
than that for GC. Adsorbed contaminants can be easily removed from the diamond

surface by rinsing with ultrapure water, or in severe cases, by soaking in distilled IPA.

 

 

 

0.05 4 «we» - 7- — _-,-_,_-__,,,,..,L A, ,,,_,
A
€0,044‘AAAAAA‘AA
v A ‘ A A A A ‘ ‘ .L
*5 0.032 .
o i
'- l
'5002“IIII..IIII.....-...I—
U 0 O O o O O o 0 O 9 o O O O O o o o 09
0.01“ i
0 Without BSA With BSA 1
0 5 10 15 20

Injection Number

Figure 3.17. FIA-EC response for 20-pL injections of 1 11M NE (injections 1-10) and 1
uM NE with 1 mM BSA (injections 11-20) at microcrystalline diamond (O),
nanocrystalline diamond (I), and GC (A) electrodes. The carrier solution was a Kreb’s

buffer, pH 7.4, flowing at 1.0 mL/min. Injections were made at 1.0 min intervals.

Electrode geometric area = 0.12 cmz.

In summary, the response stability of both diamond electrode types was shown to
be superior to that of GC. This property is beneficial particularly for electroanalytical
applications that require longer term use of an electrode with minimal loss of activity and

those that necessitate exposure of the electrode to interfering biomolecules.

3.3. Conclusions

Cyclic voltammetry was used to investigate the electrochemical behavior of

catechols and catecholamines at microcrystalline diamond, nanocrystalline diamond, and

59

GC electrodes. Both diamond types exhibited a lower and more stable background
current and noise than did GC. The diamond electrodes also showed no evidence for
electrochemically-active surface carbon-oxygen functionalities. The background
voltammetric current for both diamond types exhibited very little variation with solution
pH. These results are in contrast to those for GC. The pH-insensitivity of the
background current for diamond is clearly an advantage of this electrode material,
particularly for in vivo measurement of catecholamines.

Extra peaks in the cyclic voltammograms of DA on the three electrode types at
pH 7 demonstrate that coupled intramolecular cyclization reaction occurs when the 6-
position undergoes nucleophilic attack by the amine side-chain. Preconcentration of the
analytes on the GC electrode surface due to adsorption is evident from the experimental ip
values that are higher than the theoretical value. Because the ip°" values demonstrated a
linear relationship with OW, mixed diffusion and adsorption control is presumed. The
AEp values for the two diamond types are larger than those for GC. For example, the
AEp values for DA at pH 7 are 353, 233, and 81 mV, for microcrystalline diamond,
nanocrystalline diamond, and GC, respectively. A high AEp value is indicative Of
sluggish redox reaction kinetics. The slow electron transfer kinetics of the
catechol/orthoquinone system is attributed to the absence of adsorption. Hydrogen-
bonding between the catechol and the adsorbed quinone or catechol layer adsorbed on the
electrode surface is believed to be a catalytic step. The faster redox reaction rate at
nanocrystalline compared to microcrystalline diamond is attributed to the greater fraction
of sz-bonded carbon which has been previously shown to give rise to an increased

coverage of adsorbed catechols.

60

The values of Ep0x at varying pH exhibited trends that are charge-based indicating
electrostatic effects for all three electrode types. This outcome is unexpected on the
hydrogen-terminated diamond surface, so the presence of surface oxides was suggested.
Oxidation of the diamond surface may have occurred because wide potential sweeps were
necessary to drive the redox reaction.

FIA-EC was used to evaluate the performance of diamond as electrodes in the
electrochemical detection of catechols and catecholamines. There were many similarities
in the detection figures of merit for both diamond and GC electrodes. Linear dynamic
ranges from nM to mM levels were seen (R2 > 0.99). The response sensitivity values
were in the range of 10 — 30 nA/uM while the LOQ values were in the 10 — 60 pg range.
Noteworthy is the fact that these figures of merit were Obtained for diamond electrodes
without any pretreatment while those for CC were obtained only after activation by
mechanical polishing. The response precisions for diamond (1 — 3%) were slightly better
than those for GC (1 — 5%), and the response stability, at least for NE, was superior for
diamond. Diamond electrodes resisted deactivation during 7-h use period and resisted
fouling in the presence of BSA more so than GC. While comparable in many respects,
diamond electrodes afford advantages over GC for the analysis of catechols and
catecholamines in terms of (i) more rapid response stabilization time after detector tum-
on, (ii) low background current and peak-to-peak noise, and (iii) good figures of merit
without pretreatment. Diamond is attractive for those applications that require
continuous measurements over a long period of time and exposure to complex matrices

such as those that contain various molecules that cause electrode fouling.

61

n'aa

 

 

 

Chapter 4

REVERSED-PHASE HPLC SEPARATION AND
AMPEROMETRIC DETECTION OF CATECHOLAMINES
AND METABOLITES AT BORON-DOPED DIAMOND
THIN-FILM ELECTRODES

4.1. Introduction

Catecholamine neurotransmitters are involved in the control of a variety of
regulatory systems, such as psychomotor activity, memory, and hormone secretion.2
They are found in the brain, chromaffin cells, sympathetic nerves, and sympathetic
ganglia.3 The metabolic pathway of the catecholamines is shown in Figure 4.1.77 Two
major metabolic processes occur. One involves the oxidation of the aliphatic amine side
chain by the enzyme monoamine oxidase (MAO). The aldehyde formed is unstable and
may either be reduced to the corresponding alcohol (neutral metabolites) or be further
oxidized to an acid (acidic metabolites). The second process involves an O-methylation
in the 3-position via the enzyme catechol-O-methyl transferase (COMT).58 Since the
neurotransmitters are integrated with all bodily systems, changes in catecholamine

metabolism are likely to be good markers for various disease states and disorders.4

62

 

 

OH OH

OH
NH; MAO HOW MAO HOUK/NH(CH3)
HO O HO

DOPGAL

COM 7/ A V AD E\COMT

'43::sz ::I:/K/OH 1::m/OH H3C:U‘\/NH(CH3)

 

 

 

 

DOPEG DOMA
cowl j COMT
H3:::©/l\/OH H3C:mOH
MOPEG /VMA
A .D
OH
MA 0 HOD/S 4 MA 0
> I V
O
HO
MOPGAL

Abbreviations: NE, norepinephrine; DOPGAL, 3,4-dihydroxyphenylglycoaldehyde; EP,
epinephrine; NMN, normetanephrine; DOPEG, 3,4-dihydroxyphenylethyleneglycol; DOMA,

3,4-dihydroxymandelic acid; MN, metanephrine; MOPEG, 3-methoxy-4-
hydroxyphenylethylene glycol; VMA, vanillylmandelic acid; MOPGAL, 3-methoxy-4-
hydroxyphenylglycoaldehyde; MAO, monoamine oxidase; COMT, catechol-O-

methyltransferase; AR, aldehyde reductase; AD, aldehyde dehydrogenase

Figure 4.1. Steps in the metabolism of catecholamines.

In the previous chapter, FIA-EC measurements revealed that diamond is a reliable
electrode for the electrochemical detection of catechols and catecholamines. These
electroactive analytes are present in complex biological fluids and they all undergo
oxidation at Similar potentials. Therefore, a separation step is a prerequisite for their
detection. Separation and quantification of these analytes in biological fluids provide

valuable information on metabolic abnormalities. For example, if an altered level of a

63

catecholamine is found, the metabolite concentration may be used to confirm the

diagnosis of a disease.3

In this chapter, the utility of this new electrode for detecting
catecholamines (amperometric mode) in HPLC is demonstrated.
HPLC-EC was originally developed for the analysis of catecholamine

0 Since then, the determination of

neurotransmitters about three decades ago.9"
catecholamines and related compounds has remained among the most common
applications of the method due to their amenability for electrochemical detection and for
their biomedical importance. Some of the related reports on HPLC-EC published

recently are presented in Table 4.1 .78'83 The mass limits of detection reported and types

Of working electrodes used are given in the table.

Table 4.1. Recent publications on HPLC-EC analysis of catecholamines and

 

 

 

 

 

 

 

 

metabolites.
Ref. Electrode Analyte Detection Limit
[78] glassy carbon DOPEG 14 pg
NE 40 pg
[79] glassy carbon NE 4.0 pg
EP 3.2 pg
[80] multi-wall carbon nanotubes -COOH NE 0.85 pg
chemically-modified electrode
(MWNT-COOH CME)
[81] glassy carbon NMN 183 pg
MN 197 pg
[82] overoxidized polypyrrole (OPPy) NE 8.5 pg
film CME EP 9.2 pg
[83] self-assembled n-alkanethiol NE 676 pg
monolayer (Clo-SAM)-modified Au

 

 

 

 

 

64

In this chapter, the performance of diamond for the determination of
catecholamines and metabolites using amperometric detection coupled with HPLC is
described. The objectives for the work were: (i) to develop a protocol for separating the
catecholamines (NE and EP) and metabolites (DOPEG, VMA, NMN, and MN) by
reversed-phase HPLC, (ii) to comprehensively evaluate the analytical detection figures
of merit for both microcrystalline and nanocrystalline diamond thin-film electrodes, and
(iii) to demonstrate the capability of using this method for catecholamine analysis in a

biological sample (e.g., rat plasma).

4.2. Results and Discussion

The reversed-phase HPLC separation of a mixture of catecholamines and related
metabolites, coupled with amperometric detection, was investigated. Significant effort
was made to achieve an isocratic separation of the components in the mixture. The
variables adjusted were the mobile phase composition with the organic modifier
(methanol, MeOH), the concentration of the ion-pairing agent (heptanesulfonic acid,
HSA), the mobile phase flow rate, and the pH. A satisfactory separation was achieved
using 95/5 (v/v) 0.05 M NaH2PO4/MeOH, 0.475 mM HSA, pH 2.5, at a flow rate of 0.8
mL/min. A typical chromatogram for a mixture Of 20 [AM DOPEG, NE, EP, VMA,
NMN, and MN is shown in Figure 4.2. The injection volume was 20 [IL and the working

electrode was a nanocrystalline diamond thin film poised at 950 mV vs. Ag/AgCl for

detection.

65

 

 

 

 

 

 

DOPEG

H
5 VMA
r...
‘5
0 NE
5
3:: EP
«
E
x
O

 

 

 

 

 

 

 

 

 

f 1 '_"‘T'—____-—-_'

:[10 nA l NMN MN

: J l M“-
0 2 4 6 8 10 12 14 16 l 8 20
Time, min

Figure 4.2. Reversed-phase HPLC separation of 20 11M DOPEG, NE, EP, VMA, NMN
and MN on a C13 column (Alltech Adsorbosphere HS, 5 pm particle size, 4.6 mm x 150
mm). The mobile phase was 95/5 (v/v) 0.05 M NaH2PO4/MeOH, pH 2.5, 0.475 mM
HSA. The working electrode was a nanocrystalline diamond thin film and the detection
potential was 950 mV vs. Ag/AgCl. Flow rate = 0.8 mL/min. Injection volume = 20 [AL

Electrode geometric area = 0.12 cmz.

The background current and peak-to-peak noise were lower for microcrystalline
diamond (3.10 nA and 3.95 pA, respectively) than for nanocrystalline diamond (4.30 HA
and 4.35 pA, respectively), consistent with the trends observed in Chapter 3. The
expected reversed-phase elution order, based on the solute polarities in the presence of
the ion-pairing agent, was observed. The dihydroxy analytes (DOPEG, NE, and EP) are
retained less strongly on the column than are the 3- O-methyl metabolites (VMA, NMN,

and MN). This is because the hydroxy group imparts a greater polar character to the

66

molecule than does the methoxy group. The ion-pairing agent results in a greater
retention of the analytes with the amine side chain, which are extremely polar and
hydrophilic, by imparting nonpolar character and hydrophobicity. DOPEG, having an
uncharged side-chain does not form an ion-pair, and thus, elutes first. The same can be
said about VMA which elutes prior to NMN and MN. EP is less polar than NE due to
the presence of the electron-donating methyl group on its aliphatic amine side chain, and
is thus retained more strongly. The effect of the presence of methyl on the amine group
is two-fold: a reduction of the effective positive charge and a steric hindrance to the
electrostatic attraction with the ion-pairing agent. This same reasoning applies to the
retention order between MN and NMN. Excellent baseline resolution was achieved for
all six solutes with the separation taking a total of 14 min. The calculated retention
factors (k) are 1.1, 1.3, 2.1, 2.6, 3.6 and 6.2 for DOPEG, NE, EP, VMA, NMN, and MN,
respectively. The peaks are highly symmetric with no tailing. All the peaks are well-
resolved with the lowest resolution (R5) of 1.4 observed between DOPEG and NE. The
total number of theoretical plates, N, for the separation is 8210, based on the retention
time and peak width of MN, the last eluting solute.

A comprehensive set of detection figures of merit for the catecholamines and
metabolites were determined for both nanocrystalline and microcrystalline diamond thin-
film electrodes. Calibration curves for all the solutes are Shown in Figure 4.3. The

detection figures of merit are summarized in Table 4.2.

67

500

 

400

300

200

100

 

 

500

 

Current (nA)

400

300

200

100

 

 

 

 

 

.. ’ . : fi L fl
0 50 100 150
Concentration (uM)

Figure 4.3. Calibration curves obtained at a (A) nanocrystalline and (B) microcrystalline
diamond thin-film electrode for a series Of mixtures of the catecholamines and
metabolites. The separation was carried out in the same conditions as in Figure 4.2. The

detection potentials used were 950 mV and 1000 mV vs. Ag/AgCl for A and B,

respectively. Injection volume = 20 [IL Electrode geometric area = 0.12 cmz.

68

Table 4.2. Amperometric detection figures of merit for the HPLC-EC analysis of

catecholamines and metabolites at nanocrystalline and microcrystalline diamond thin-

film electrodes.

 

 

 

Response Precision Linear Dynamic Sensitivity LOQ at
Ana'yte (% RSD) Range (nA/uM) S/N 2 3
Nanocrystalline Diamond
DOPEG 0.9 0.003 11M — 150 uM 3.3 : 0.5 10 pg
NE 1.0 0.01 11M — 150 11M 2.0 : 0.2 34 pg
EP 1.0 0.01 uM -— 150 uM 1.4 : 0.2 37 pg
VMA 1.6 0.01 [AM - 150 11M 2.0 : 0.6 40 pg
NMN 2.0 0.03 uM — 150 uM 0.2 : 0.1 110 pg
MN 1.5 0.05 uM — 150 11M 0.1 : 0.0 197 pg
Microcrystalline Diamond
DOPEG 5.1 0.007 — 150 11M 1.7 : 0.90 24 pg
NE 5.0 0.03 — 150 uM 0.8 : 0.96 102 pg
EP 5.3 0.03 —- 150 11M 0.5 : 0.64 110 pg
VMA 8.8 0.03 — 150 uM 0.6 : 0.57 119 pg
NMN 6.5 0.1 — 150 11M 0.1: 0.13 366 pg
MN 7.1 0.1 — 150 uM 0.04 : 0.07 1180 pg

 

Note: The data are based on the response of three electrodes. The sensitivity values are

reported at 95% confidence level. The response precision was evaluated based on six

injections of a 20 [AM mixture. The calibration curves consisted of at least six points.

Experimental conditions are the same as those listed in Figure 4.2.

The linear dynamic range for the solutes is 4 - 6 orders of magnitude (R2 > 0.999)

for both diamond types. It should be noted that the maximum concentration of 150 uM

does not represent the upper end of the linear dynamic range but rather just the highest

concentration measured. The response precision was determined from six consecutive

injections of a 20 11M mixture Of the analytes. The response precision was superior for

69

nanocrystalline diamond with RSD values of 0.9 — 2.0%. The large response variability
for microcrystalline diamond may have been influenced by the column conditions and not
the electrode properties. Rigorous column cleaning was performed before the start of the
series of measurements with nanocrystalline diamond and may have improved the
column performance. For all the solutes, the sensitivity is 2-3 times higher for all the
solutes at nanocrystalline than at microcrystalline diamond. The LOQ values for both
diamond types are comparable to those Of the recently published results for other
electrode materials shown in Table 4.1. The LOQ values for nanocrystalline diamond are
all at the pg level and are 2-3 times lower than those for microcrystalline diamond. In
general, the figures Of merit reveal more favorable detection for nanocrystalline diamond,
in terms of LOQ, sensitivity and response precision. This is consistent with the larger ipox
values for nanocrystalline diamond observed in the voltammetric measurements,
something that appears to be caused by increased adsorption. Adsorption of catechols is
reported to be enhanced when there is a greater fraction of exposed spz-bonded carbon,
which is related to smaller grain size.36 However, the superior electrode performance of
nanocrystalline diamond in the HPLC-EC is more pronounced than in the FIA-EC
measurements.

The usefulness of the HPLC-EC assay for the analysis of a real biological sample
was also demonstrated. The catecholamine and metabolite levels in the blood plasma of
a laboratory test animal were determined. The work presented herein is a preliminary
effort to use HPLC-EC to monitor the circulation levels of catecholamines and
metabolites in normotensive and hypertensive rats. The main focus is on the main

neurotransmitter released in the sympathetic nerve system, NE, and its four major

70

metabolites: DOPEG, VMA, NMN, and MN. The rat plasma samples were Obtained
from Dr. Keith Lookingland of the Department of Pharmacology and Toxicology at
Michigan State University. Chromatograms for plasma samples from a normotensive and
a hypertensive rat are shown in Figure 4.4. In the hypertensive rat, hypertension was
induced by a procedure, introduced by Grollman, known as renal wrap, which involves

removing one kidney and then traumatizing the other kidney.84

 

 

 

 

 

 

NE

E
Q
i-
h-
=
U
u:
g [0.5 HA
G:
G
.2
H
N
.. LJJ
oi W—Jv .-
O 1

fl

0 2 4 6 8 10

Time (min)
Figure 4.4. Reversed—phase chromatograms for the analysis of plasma samples from (A)
a normotensive rat and (B) a hypertensive rat on a C13 column (Alltech Adsorbosphere
HS, 5 pm particle size, 4.6 mm x 150 mm). The mobile phase was 96/4 (v/v) 0.05 M
NaH2PO4/MeOH, pH 2.5, 0.48 mM HSA. The working electrode was a nanocrystalline
diamond thin film and the detection potential was 950 mV vs. Ag/AgCl. Flow rate = 0.8

mL/min. Injection volume = 20 uL. Electrode geometric area = 0.12 cmz.

71

Purification of the plasma samples was accomplished by alumina extraction, a
technique that is based on the selective adsorption of catecholamines onto alumina in an
alkaline medium. Adsorption occurs by the formation of cyclic complexes, as the vicinal
hydroxy groups of the catecholamines are attracted to the alumina matrix. The presence
of an amine side chain seems to be unimportant, resulting in the extraction of all
catechols, not just catecholamines. The alumina was then washed with water and the
catecholamines desorbed by acidification.4 The eluted samples were then stored in the

freezer (-20 °C) prior to measurement. When the time came for analysis, the sample was

thawed and 20 [IL was injected onto the reversed-phase column. Elution was performed
isocratically with 96/4 (v/v) 0.05 M NaH2PO4/MeOH, pH 2.5, 0.48 mM HSA. The
chromatogram for the normotensive rat plasma sample revealed seven peaks — the system
peak, four peaks with retention times not matching any of the analytes studied, and NE
and EP peaks at 4.6 min and 6.7 min, respectively. Peak identification was performed by
retention time matching. The chromatogram for the hypertensive rat plasma sample, on
the other hand, contained only four peaks — the system peak, two unknown peaks and the
NE peak. The identity of the unknown peaks was not determined. Peaks for the
metabolites DOPEG, VMA, NMN, and MN were not seen, probably because the
sensitivity of the method is not sufficient enough to detect the low concentration Of these
compounds probably present.

In order to quantitate the NE and EP present, a series Of standard solutions was
analyzed and calibration curves were constructed from the data. The standard addition
method is preferred for complex samples such as this but the quantity Of sample was too

small to use this method. The NE peak height for the hypertensive plasma sample is

72

larger than that for the normotensive one indicative of a higher concentration. Although
this is the expected result, an uncertainty exists because the NE level of 396 pmol/mL
determined for the normotensive sample is much higher than values found in literature, in

the low pmol/mL range.”86

There are a couple of possible reasons for this difference.
One is that since only two normotensive and two hypertensive samples were measured,
there is a possibility that the animals tested were not representative Of the population. A
second possibility is that the extraction efficiency Of 52% used in the calculations is just
an estimate based on internal standard measurement (3,4-dihydroxybenzylamine or
DHBA). The extraction coefficient is calculated from the known concentration Of the
internal standard added to the sample before extraction and the concentration determined
by the analysis of the extract. This may be a major source of error because the internal
standard measurement was not done on the actual sample injected. The EP level Of 4.0
pmol/mL for the normotensive sample agrees with the value of 3.76 mel/mL obtained
by the group of Y. Wang.86 Other groups have reported plasma EP levels ranging from
0.57 to 0.82 mel/mL.85‘87 These variabilities are expected to arise due to differences in
the test animal and environmental factors, such as the source, age, diet, weight and

condition of the rats, and the time of day, ambient temperature, and sampling site during

sample collection.

4.3. Conclusions

The reversed-phase HPLC separation Of several catecholamines (NE and EP) and
metabolites (DOPEG, VMA, NMN, and MN) was optimized and the performance of

microcrystalline and nanocrystalline diamond thin-film electrodes for the amperometric

73

detection of these analytes was evaluated. Nanocrystalline diamond exhibited good
response precision for all analytes with RSD values between 1 and 2%. Microcrystalline
diamond exhibited higher RSD values between 5 and 9%. The linear dynamic range was
4 to 6 orders of magnitude for both diamond types. The sensitivity ranged from 0.1 to 4
nA/uM for nanocrystalline diamond and from 0.04 to 2 nA/uM for microcrystalline
diamond. The mass LOQ values ranged from 10 to 100 pg and 20 to 1200 pg, for
nanocrystalline and microcrystalline diamond, respectively. In general, nanocrystalline
diamond outperformed microcrystalline diamond for the detection of these important
compounds. These figures Of merit were compared with the recent reports on the HPLC-
EC analysis of catechols and catecholamines presented in Table 4.1. The RSD values are
comparable to the lowest reported.82 The linear dynamic range obtained in this work is
the widest. Other groups that used OPPy CME and MWNT-COOH CME reported 3 to 4

80,82

orders of magnitude in range. Sensitivity values of l to 16 nA/uM and 8 to 36

nA/uM were reported by S. Zhang82 et a1 and W. Zhang80 et a1, respectively, indicating
that the values obtained in this research needs improvement in this aspect. The LOQ
values reported here are comparable to those of the recent works presented in Table 4.1.
A bionalytical application of the HPLC-EC method was demonstrated using rat
plasma samples. An attempt was made to correlate hypertension with an increased
plasma level of NE. The plasma concentration of epinephrine in a normotensive rat was
determined. Further improvement in the sensitivity is needed to extend the usefulness of
the detection scheme in the quantitation of biological compounds present in low

concentrations.

74

Chapter 5

SUMMARY AND CONCLUSIONS

The results Of the research presented herein demonstrate that catechols and
catecholamines can be electrooxidized and quantitatively detected using boron-doped,
polycrystalline diamond thin—film electrodes. CV investigations have shown that the
electron transfer kinetics are more sluggish on diamond than on GC, presumably due to
the absence of adsorption. The redox reaction is faster on nanocrystalline diamond than
on microcrystalline diamond and this is attributed to enhanced adsorption due to an
increased fraction Of exposed spz-bonded carbon. The redox reaction rate for diamond
demonstrated pH-dependence, that is, the El,"x value decreases as the pH increases.
Moreover, the rate of electron transfer on diamond Show charge discrimination, that is,
faster rates are observed for positively-charged than for negatively-charged analytes.
This finding is consistent with an electrostatic effect, so the presence of surface oxides is
supposed. The surface oxides may have resulted from the wide potential sweeps applied
to drive the redox reaction.

The FIA-EC results have indicated that catechols and catecholamines can be
detected on diamond electrodes at constant potential. The performance of the diamond
electrodes is found to be comparable to that of the more commonly used GC electrode.
Diamond exhibited faster stabilization time, and lower background current and peak-to-
peak noise compared to GC. The response precision for all three electrode types is good,

with average RSD values of 1.8%, 1.5%, and 2.4% for microcrystalline diamond,

75

nanocrystalline diamond, and GC, respectively. The linear dynamic range, for all three
electrode types, is 5 —— 6 orders of magnitude, ranging from the nM to the mM level. The
sensitivity values for all analytes at the three electrodes are similar, with values ranging
from 10 - 30 nA/uM. The LOQ values are comparable, all in tens of picograms. The
diamond electrodes have demonstrated better response stability. The measured changes
in the current response within a 7-h period for microcrystalline and nanocrystalline
diamond are 1% and 2%, respectively, while the current decreased by 11% for GC.
Diamond exhibits a greater resistance to fouling by interfering biomolecules. The current
decrease in the presence of added BSA are 8% for both diamond types and 17% for GC.
The performance of diamond thin-film electrodes is comparable to that of GC in terms of
linear dynamic range, sensitivity and LOQ. Diamond showed better response stability
and lesser susceptibility to fouling. It is important to note that diamond electrodes exhibit
optimum performance without any time-consuming and tedious pretreatment steps, as
opposed to GC which has to be polished prior to measurement.

The results Of the HPLC-EC measurements provide a good starting point for the
use of diamond electrodes in physiological research and in routine bioanalytical assays.
Response precision is better at nanocrystalline diamond, with RSD values between 1 -
2%, compared to that at microcrystalline diamond, with values between 5 - 9%. The
linear dynamic range for the two diamond types is 3 - 5 orders of magnitude. The
sensitivity Of nanocrystalline diamond, ranging from 0.1 - 3 nA/uM, is slightly higher
than that of microcrystalline diamond, with values Of 0.04 - 2 nA/uM. The LOQ values
at nanocrystalline diamond are 10 — 200 pg while those at microcrystalline diamond are

20 - 1200 pg. The analytical detection figures of merit for nanocrystalline diamond are

76

better than that for microcrystalline diamond in the HPLC-EC measurements. This
presumably results from adsorption that occurs through the spz-bonded carbon at the
grain boundaries of the nanocrystalline diamond surface. The superior performance of
nanocrystalline diamond compared to microcrystalline diamond is more pronounced in
the HPLC-EC measurements than in the FIA-EC measurements.

The HPLC-EC assay for rat plasma samples showed the presence of NE and EP.
However, the calculated NE level is much higher than literature values. Moreover, there
is a need to improve the sensitivity of the method to detect the catecholamine metabolites
present in low concentrations. This may be accomplished through the manipulation of
the diamond film growth conditions to produce nanocrystalline films with a greater
fraction of grain boundaries. The use of diamond microelectrodes instead of thin-film
electrodes in HPLC-EC may lead to enhanced mass transport and may result in lower
mass detection limits. Further improvements in the detection limits and selectivity may
be achieved by employing capillary electrophoresis as the separation method. A better
knowledge and understanding of the effects of surface modifications such as OPPy-
coating,34 electrochemical anodizationf’7 and oxygen termination88 on the sensitivity and

selectivity for the detection of catecholamines and metabolites are beneficial.

77

(1)

(2)

(3)
(4)

(5)

(6)

(7)

(8)

(9)
(10)

(11)

(12)

(13)

(14)

(15)

BIBLIOGRAPHY

In http://www.lsuagcenter.com/news/July2003/GetGrow/GIGColumn(POison
Iivy)-07-25-03.asp, accessed August 2004.

J. Bergquist, A. Sciubisz, A. Kaczor, J. Silberring J. Neurosci. Methods 2002,
113, 1.

Z. Yi, P.R. Brown Biomed. Chromatogr. 1991, 5, 101.

AM. Krstulovic Quantitative Analysis of Catecholamines and Related
Compounds; John Wiley & Sons: New York, 1986.

N. Unceta, E. Rodriguez, Z.G. de Balugera, C. Sampedro, M.A. Goicolea, S.
Barrondo, J. Salles, R. Barrio Anal. Chim. Acta 2001, 444, 211.

GA. Scratchley, A.N. Masoud, S.J. Stohs, D.W. Wingard J. Chromatogr., 169,
313.

CT. Duda, P.T. Kissinger In Methods in Neurotransmitter and Neuropeptide
Research, Part 1; SH. Parvez, M. Naoi, T. Nagatsu, S. Parvez, Ed.; Elsevier: New
York, 1993; Vol. 11.

J .8. Justice, A.C. Michael, D.B. Neill In Neuromethods 2: Amines and Their
Metabolites; A. A. Boulton, G.B. Baker, Ed.; Humana Press: New Jersey, 1985.

P.T. Kissinger, C. Refshauge, R. Dreiling, R.N. Adams Anal. Lett. 1973, 6, 465.

C. Refshauge, P.T.Kissinger, R. Dreiling, C.L. Blank, R. Freeman, R.N. Adams
Life Sci. 1974, 14, 311.

K. Stulik, V. Pacakova In Quantitative Analysis of Catecholamines and Related
Compounds; A.M. Krstulovic, Ed.; John Wiley and Sons: New York, 1986.

KL. McCreery In Voltammetric Methods in Brain Systems; A.A. Boulton, G.B.
Baker, R.N. Adams, Ed.; Humana Press: New Jersey, 1995, pp 1-26.

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

P.W. May Phil. Trans. R. Soc. Lond. 2000, 358, 473.

M. Iwaki, S. Sato, K. Takahashi, H. Sakairi Nucl. Instrum. Methods 1983, 209,
1129.

78

(16)

(17)

(18)
(19)

(20)

(21)

(22)
(23)
(24)
(25)
(26)
(27)
(28)

(29)

(30)
(31)
(32)

(33)

(34)
(35)

(36)

Yu.V. Pleskov, A.Ya. Sakharova, M.D.Krotova, L.L. Bouilov, B.V. Spitsyn J.
Electraanal. Chem. 1987, 228, 19.

J. Xu, M.C. Granger, Q. Chen, T.E. Lister, J .W. Strojek, G.M. Swain Anal. Chem.
1997, 69, 591A.

G.M. Swain, A.B. Anderson, J .C. Angus MRS Bull. 1998, 23, 56.

R. DeClements, G.M. Swain, T. Dallas, M.W. Holtz, R. Herrick 111, J .L Stickney
Langmuir 1996, 12, 6578.

L. Boonma, T. Yano, D.A. Tryk, K. Hashimoto, A. FujishimaJ. Electrochem.
Soc. 1997, 144, L142.

H.B. Martin, A. Argoitia, U. Landau, A.B. Anderson, J .C. Angus J. Electrochem.
Soc. 1996, 143, L133.

J. Xu, Q. Chen, G.M. Swain Anal. Chem. 1998, 70, 3146.

MC. Granger, J. Xu, J .W. Strojek, G.M. Swain Anal. Chim. Acta 1999, 397, 145.
J. Xu, G.M. Swain Anal. Chem. 1998, 70, 1502.

S. Jolley, M. Koppang, T. Jackson, G.M. Swain Anal. Chem. 1997, 69, 4099.
MD. Koppang, M. Witek, J. Blau, G.M. Swain 1999, 71, 1188.

MA. Witek, G.M. Swain Anal. Chim. Acta 2001, 440, 119.

G.W. Muna, N. Tasheva, G.M. Swain Environ. Sci. T echnal. 2004, 38, 3674.

TN. Rao, B.H. LOO, B.V. Sarada, C. Terashima, A. Fujishima Anal. Chem. 2002,
74, 1578.

TN. Rao, 1. Yagi, T. Miwa, D.A. Tryk, A. Fujishima Anal. Chem. 1999, 71, 2506.
E. Popa, Y. Kubota, D.A. Tryk, A. Fujishima Anal. Chem. 2000, 72, 1724.
S. Sarada, T.N. Rao, D.A. Tryk, A. Fujishima Anal. Chem. 2000, 72, 1632.

J. Cvacka, V. Quaiserova, J. Park, Y. Show, A. Muck, Jr., G.M. Swain Anal.
Chem. 2003, 75, 2678.

H. Olivia, B.V. Sarada, D. Shin, T.N. Rao, A. Fujishima Analyst 2002, 127, 1572.
SciFinder Scholar; Am. Chem. Soc.: Columbus, 2002.

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

79

(37)

(38)

(39)

(40)

(41)

(42)
(43)

(44)

(45)

(46)

(47)

(48)
(49)
(50)

(51)

(52)
(53)

(54)

(55)

L. Bergman, R.J. Nemanich J. Appl. Phys. 1995, 78, 6709.

J. Birrell, J .E. Gerbi, O. Auciello, J .M. Gibson, J. Johnson, J .A. Carlisle Diam.
Relat. Mater. 2005, 14, 86.

AC. Ferrari, J. Robertson Phys. Rev. B 2001, 63, 121405.

D.M. Gruen, A.R. Krauss, C.D. Zuiker, R. Csencsits, L.T. Terminello, J .A.
Carlisle, I. Jimenez, D.G.J. Sutherland, D.K. Shu, W. Tong, F. Himpshel J. Appl.
Phys. Lett. 1996, 68, 1640.

L. Ostrovskaya, V. Perevertailo, V. Ralchenko, A. Dementjev, O. Longinova
Diam. Relat. Mater. 2002, 11, 845.

MC. Granger, G. M. SwainJ. Electrochem. Soc. 1999, 146, 4551.
P. Chen, R.L. McCreery Anal. Chem. 1996, 68, 3958.

C. Levy-Clement, F. Zenia, N. Awa Nadao, A. Deneuville New Diamond Frontier
Carbon Techno]. 1999, 9, 189.

R. Keller, 1. Mefford, A. Oke, E. Strope, J. Conti, R.M. Wightman, P. Plotsky,
R.N. Adams Modern Pharm. 1977, 10, 761.

PS. Cahill, J .M. Finnegan, GE. Mickelson, E.R. Travis, R.M. Wightman Anal.
Chem. 1996, 68, 3180.

D. Pletcher A First Course in Electrode Processes; Alresford Press: England,
1991.

P.T. Kissinger, W.R. Heineman J. Chem. Ed. 1983, 60, 702.
J. Wang Analytical Electrochemistry, 2nd ed.; Wiley: New York, 2000.
M.L.A.V. Heien, M.A. Johnson, R.M. Wightman Anal. Chem. 2004, 76, 5697.

R. L. McCreery In Voltammetric Methods in Brain Systems; A.A. Boulton, G.B.
Baker, R.N. Adams, Ed.; Humana Press: New Jersey, 1995, pp 1-26.

S.H. Duvall, R.L. McCreeryJ. Am. Chem. Soc. 2000, 122, 6759.
ON. Kamau, W.S. Willis, J .F. Rusling Anal. Chem. 1985, 5 7, 545.

PL. Runnels, J .D. Joseph, M.J. Logman, R.M. Wightman Anal. Chem. 1999, 71,
2782.

MR. Deakin, P.M. Kovach, K.J. Stutts, R. M. Wightman Anal. Chem. 1986, 58,
1474.

80

(56)
(57)

(58)

(59)

(60)

(61)
(62)

(63)

(64)

(65)

(66)

(67)

(68)

(69)

(70)

(71)
(72)

(73)

(74)

MR. Deakin, K.J. Stutts, R.M. Wightman J. Electraana]. Chem. 1985, 182, 113.
D.C.S. Tse, R.L. McCreery, R.N. Adams J. Med. Chem. 1976, 19, 37.

AW. Stemson, R. McCreery, B. Feinberg, R.N. Adams J. Electroana]. Chem.
Interfacia] Electrochem. 1973, 46, 313.

M.D. Hawley, S.V. Tatawadi, S. Piekarski, R.N. Adams J. Am. Chem. Soc. 1967,
89, 447.

A.J. Bard, L.R. Faulkner Electrochemical Methods: Fundamentals and
Applications, 2nd ed.; John Wiley and Sons, Inc.: New York, 2001.

E. Laviron J. Electraana]. Chem. 1984, I64, 213.
MR. Deakin, R.M. Wightman J. Electraana]. Chem. 1986, 206, 167.

KT. Kawagoe, P.A. Garris, R.M. Wightman J. Electraana]. Chem. 1993, 359,
193.

S.H. Duvall, R.L. McCreery Anal. Chem. 1999, 71, 4594.
CD. Allred, R.L. McCreery Anal. Chem. 1992, 64, 444.

MC. Granger, M. Witek, J. Xu, J. Wang, M. Hupert, A. Hanks, M.D. Koppang,
J .E. Butler, G. Lucazeau, M. Merrnoux, J .W. Strojek, G.M. Swain Anal. Chem.
2000, 72, 3793.

E. Popa, H. Notsu, T. Miwa, D.A. Tryk, A. Fujishima Electrochem. Solid-State
Lett. 1999, 2, 49.

1. Duo, C. Levy-Clement, A. Fujishima, C. Comninellis J. Appl. Electrochem.
2004, 34, 935.

J. Ruzicka, B.H. Hansen Anal. Chim. Acta 1975, 78, 145.

J. Ruzicka, B.H. Hansen Flaw Injection Analysis, 2nd ed.; J. Wiley: New York,
1988.

B.H. Hansen, J. Ruzicka J. Chem. Ed. 1979, 56, 677.
J. Ruzicka, B.H. Hansen Anal. Chem. 2000, 72, 212A.

In http://www.lachatinstruments.com/products/qcfia/FlAPrimer.asp., accessed
August 2004.

W.R. Lacourse Pulsed Electrochemical Detection in High Performance Liquid
Chromatography; Wiley: New York.

81

(75)

(76)

(77)

(78)

(79)

(80)

(81)
(82)
(83)
(84)
(85)
(86)

(87)

(88)

G. Dryhurst NATO ASI Sen, Ser. A 1994, 267, 171.

In http://www.thinkquest.org/library/site_sum.html?tname=C01 15080&ur1=
C0115080/%3fcmlasma, accessed August 2004.

G. Eisenhofer, T. Huynh, M. Hiroi, K. Pacak Rev. Endocr. Metab. Disord. 2001,
2, 297.

E. Sastre, A. Nicolay, B. Bruguerolle, H. Portugal J. Chromatogr. B 2004, 801,
205.

J. Ueyama, K. Kitaichi, M. Iwase, K. Takagi, K. Takagi, T. Hasegawa J.
Chromatogr. B 2003, 798, 35.

W. Zhang, Y. Xie, S. Ai, F. Wan, J. Wang, J. Jin J. Chromatogr. B 2003, 791,
217.

K. Vuorensola, H. Siren, U. Kaijalainen J. Chromatogr. B 2003, 2003, 277.

S. Zhang, Q. Xu, W. Zhang, L. Jin, J. Jin Anal. Chim. Acta 2001, 427, 45.

Z. Liu, J. Li, S. Dong, B. Wang Anal. Chem. 1996, 68, 2432.

A. Grollman Proc. Soc. Exp. Biol. Med. 1944, 57, 102.

K. Takezawa, M. Tsunoda, N. Watanabe, K. Imai Anal. Chem. 2000, 72, 4009.
Y. Wang, D.S. Fice, K.F. Yeung J. Pharm. Biomed. Anal. 1999, 21, 519.

J .W.M. Lenders, R. Kvetnansky, K. Pacak, D.S. Goldstein, I.J. Kopin, G.
EisenhoferJ. Pharmacol. Exp. Ther. 1993, 266, 288.

I. Yagi, H. Notsu, T. Kondo, D.A. Tryk, A. Fujishima J. Electroanal. Chem, 473,
173.

82

   

I"Elljjjjjlt‘lliljljl13:11