EeFomflq.“ Jinwu. .
I W A. .
XI. .wfirn a.»

25.5

r .

 

 

 

 

.33. ..
n 1%... mg
9am... . .
OK . ‘ . 4 .

w.
I...

v.2
,e v

 

 

 

 

t :17 .t»!;..
{it‘ll A v-73.

(3.99 .

g. . wfigzwgfizéigg

a. l . . .2. . . :t,.
a In « -

 

 

 

MEIR

.2 LIBRARY
4 005; Michigan State
_. University

 

 

This is to certify that the
dissertation entitled

ELECTRO-OXIDATION AND DETECTION OF
CHLORINATED PHENOLS USING BORON-DOPED THIN-
FILM DIAMOND ELECTRODES

presented by

Grace Wangechi Muna

has been accepted towards fulfillment
of the requirements for the

Ph.D degree in Chemistry/Environmental
Toxiccfiagy

 

{71024 W $441.7

V Major Professor’s Sig‘r'l'ature

7—7-05—

 

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 DEE '7 DATE DUE DATE DUE

Nov '2 8 2007

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

205W .mns‘

ELECTRO-OXIDATION AND DETECTION OF
CHLORINATED PHENOLS USING BORON-DOPED THIN-
FILM DIAMOND ELECTRODES

BY

Grace Wangechi Muna

A DISSERTATION
Submitted to
Michigan State University

In partial fulfillment of the requirements
For the degree of

DOCTOR OF PHILOSOPHY

DEPARTMENT OF CHEMISTRY AND CENTER FOR
INTEGRATIVE TOXICOLOGY

2005

ABSTRACT

ELECTRO-OXIDATION AND DETECTION OF
CHLORINATED PHENOLS USING BORON-DOPED THIN-
FILM DIAMOND ELECTRODES
By
Grace Wangechi Muna

Historically, electrochemical detection methods for the chlorinated phenol
(CP’s) analysis have been plagued by fouling of the electrode. Diamond is a new
electrode material that shows little tendency to be deactivated and fouled by
adsorbed reaction products and intermediates during the oxidation of CPS. This
new electrode reopens the door to electrochemical detection of these pollutants.
In this work cyclic voltammetry with microcrystalline and nanocrystalline diamond
electrodes was used to study the chlorophenol oxidation reaction in acidic media
and to learn how the film morphology might affect the reaction mechanism.
Amperometry with both diamond types was coupled with flow injection analysis,
high performance liquid chromatography (HPLC-EC), and capillary zone
electrophoresis (CE-EC), for the analysis of chlorinated phenols.

Diamond electrodes exhibited good responsiveness for the electro-
oxidation of these compounds and resisted fouling during extended use. The
electrode performance was evaluated in terms of the linear dynamic range, limit
of quantitation, response precision and response stability. Both diamond types

exhibited excellent detection figures of merit. For example in HPLC - EC, the

limit of quantitation for the 2-chlorophenol was 0.1 [M (13 ppb) with a response
precision of 4.1 % (n = 7) for the microcrystalline diamond electrode. In CE-EC
using diamond microelectrode, the limit of quantitation for 2-chlorophenol was 0.1
uM (13 ppb) and the response precision was 4.7% (n = 9).

Real samples were also analyzed. The HPLC — EC assay was used for
the determination of 2-chlorophenol in a contaminated soil sample. Red Cedar
water was spiked with CPS and was extracted by solid phase extraction (SPE)
prior to analysis, and then assayed for using CE-EC. This lowered the limit of
quantitation to mid to high ppt range. The results demonstrate that
electrochemical detection methods using diamond electrode can be employed for

the analysis of CP’s.

ACKNOWLEGMENTS

My sincere appreciation and thanks go to my advisor Dr. Swain who
provided academic directions and a good environment for me to grow in as a
scientist, for his patience and great support. I admire and appreciate his quest for
excellence, vast knowledge in analytical chemistry, personality, character and a
great sense of humor. Thank you Greg!

I would also like to thank the members of my committee: Dr. McGuffin, Dr.
Blanchard, Dr. Voice and Dr. Garrett. Thank you for all the comments concerning
my research, and also for your kindness. Your input and support has been
valuable for my development as a scientist. Thanks.

I would like to thank the former and present Swain group members: Dr.
Jian Wang, Dr. Malgorzata Witek, Dr. Mateuz Hupert, Dr. Shannon Haymond, Dr.
Josef Cvacka, Suzana Cvackova, Dr. Yoshiyuki Show, Karolina Peckova, Dr.
Jesse Seegmiller, Dr. Prema Sonthalia, Dr. Jason Stotter, Pushwinder Kaur, Dr.
Veronika Mocko, Jason Bennett, Ann Fischer, Audrey Martin, Jinwoo Park,
Gloria Pimienta, Doug Knigge, Yang Song, Shihua Wang, Elizabeth McGaw, Hua
Dong, Luther Schaeffer, and Yingrui Dai. Thank you all for your great support.
Special thanks to Dr. Show who fabricated the diamond planar films and Dr.
Mocko who fabricated the diamond microelectrodes. Thank you guys for your
great support. I am indebted in my heart.

I would like to thank the Department of Chemistry here at Michigan State

University for giving me an opportunity to attend graduate school.

iv

Finally and most importantly, I thank God for His abundant grace and for
giving me a wonderful and supportive family: My husband, Muna; daughter,
Nishi; and son, Ushindi. Thank you guys for your love and moral support!

I dedicate this dissertation to my wonderful husband, Muna Kangethe who
has faithfully encouraged and supported me over the years to achieve my

academic dreams, and for his unending love.

CW Hm

Table of Contents

 

 

List of figures ..................................................................................... x
List of tables ..................................................................................... xiv
CHAPTER 1 ......................................................................................... 1
Background and Motivations ............................................................................. 1
1.1 Chlorinated Phenols ..................................................................... 1
1.1.1 Environmental Importance ...................................................... 1

1.1.2 Chlorinated Phenols Analysis ................................................... 3

1.2. Conductive Boron-Doped Diamond Electrodes .................................. 8

1.3 Diamond Electrode Deposition and Doping ...................................... 12

1.4 Outline of the Dissertation ............................................................. 18

1.5 References ............................................................................... 20

CHAPTER 2-- ................................................. 98
Experimental Procedures and Instrumentation ....................................... 28
2.1 Diamond Thin-Film Deposition .................................................. 28

2.1.1 Planar Film Electrodes ..................................................... 28

2.1.1.1 Microcrystalline Diamond Thin-Films Electrodes...........28

2.1.1.2 Nanocrystalline Diamond Thin-Film Electrodes ............ 31

2.1.2 Boron-Doped Diamond Thin-Film Microelectrodes .................. 33
2.2 Glassy Carbon Electrode Preparation ......................................... 35
2.3 Characterization of the Diamond Thin-Film Electrodes ................... 36
2.3.1 Scanning Electron Microscopy (SEM) .................................... 36
2.3.2 Atomic Force Microscopy (AFM) .......................................... 36

vi

2.3.3 Raman Spectroscopy ........................................................... 36

 

2.3.4 X-ray Diffraction (XRD) ......................................................... 37
2.3.4 Electrical Resistivity ............................................................ 37
2.4.5 Contact Angle Measurements ................................................ 38
2.4 Electrochemical Measurements ......................................................... 38
2.5 Flow Injection Analysis .................................................................... 40
2.6 High Performance Liquid Chromatography ........................................... 42
2.7 Capillary Electrophoresis .................................................................. 43
2.8 Solid Phase Extraction ..................................................................... 46
2.9 Materials and Chemicals .................................................................. 47
2.10 References .................................................................................. 50
CHAPTER 3 ....................................................................................... 51
The Physical and Electrochemical Characterization of Boron-Doped
Diamond Thin-Film Electrodes ............................................................. 51
3.1 Introduction ........................................................................... 51
3.2 Results and Discussion ........................................................... 53
3.2.1. Morphology and Microstructure .......................................... 53
3.2.2 Electrochemical Responsiveness ...................................... 58
3.3 Conclusions ........................................................................... 70
3.4 References ........................................................................... 71
CHAPTER 4 ....................................................................................... 74
Cyclic Voltammetric Studies of the Electra-oxidation of Chlorinated
Phenols on Boron-Doped Diamond Thin-Film Electrodes __ 74
4.1 Introduction ........................................................................... 74

vii

4.2 Results and Discussions .......................................................... 79

4.3 Conclusions .......................................................................... 93
4.4 References ........................................................................... 95
CHAPTER 5 ....................................................................................... 97

Detection of Chlorinated Phenols by Flow Injection Analysis and High
Performance Liquid Chromatography Using Amperometric Detection with a

Boron-Doped Diamond Thin-Film Electrode .......................................... 97
5.1 Introduction .......................................................................... 97

5.2 Results and Discussion ......................................................... 100

5.2.1 Flow Injection Analysis with Electrochemical Detection ......... 100

5.2.2 High Performance Liquid Chromatography with Electrochemical

Detection ................................................................................. 1 1 1
5.2.3 Soil Sample Analysis .......................................................... 117
5.3 Conclusions ........................................................................ 120
5.4 References ......................................................................... 121
CHAPTER 6 ...................................................................................... 123
Capillary Zone Electrophoresis Separation Chlorinated Phenols with
Amperometric Detection Using a Diamond Microelectrode .................... 123
6.1 Introduction ......................................................................... 123
6.2 Results and Discussions ........................................................ 126

6.2.1 Surface morphology and microstructure of diamond

microelectrodes ........................................................................ 126
6.2.2. Basic Voltammetric Response ............................................. 128
6.2.3 Direct Amperometric Detection of Chlorinated Phenols ......... 132
6.2.4 Indirect Amperometric Detection of Chlorinated Phenols ....... 143

viii

6.3 Conclusions .......................................................................... 151

6.4 References ........................................................................... 153
CHAPTER 7 ...................................................................................... 156
Chlorinated Phenol Analysis Using Solid Phase Extraction Coupled with
Amperometric Detection .................................................................... 156

7.1 Introduction ......................................................................... 156
7.2. Results and Discussion ......................................................... 159

7.2.1 Separation and Amperometric Detection of Chlorinated

Phenols ................................................................................... 159

7.2.2. Off-line Coupling of SPE with CE-EC ................................ 162

7.3 Conclusions ......................................................................... 169

7.4 References ......................................................................... 171
CHAPTER 8 ...................................................................................... 173
Summary ......................................................................................... 173

ix

List of Figures

Figure 1.1 A schematic for the band structure of boron-doped diamond ......... 17
Figure 2.1 Microwave-assisted CVD apparatus ......................................... 29
Figure 2.2 Schematic of a boron-doped diamond microelectrode sealed in a
pipette tip .......................................................................................... 35
Figure 2.3 Design of the three-electrode, single compartment electrochemical
cell .................................................................................................... 39
Figure 2.4 A schematic of the FIA system ................................................. 40
Figure 2.5 Design of the thin-layer flow cell ............................................... 41
Figure 2.6 Capillary electrophoresis set up ................................................ 44
Figure 2.7 Design of the end-column electrochemical detection
cell .................................................................................................... 45
Figure 2.8 SPE extraction steps ............................................................... 47

Figure 3.1 AFM images showing a (A) microcrystalline and (B) nanocrystalline
diamond thin film deposited on Si.. ..................................................................... 53

Figure 3.2 Visible Raman spectra for a (A) microcrystalline and (B)
nanocrystalline diamond thin film ............................................................. 56

Figure 3.3 XRD patterns for a (A) microcrystalline and (B) nanocrystalline
diamond thin-film electrode ..................................................................... 58

Figure 3.4 Background cyclic voltammetric i-E curve in 0.1 M HCIO4 for a
microcrystalline diamond film deposited on Si ............................................. 60

Figure 3.5 Cyclic voltammetric i-E curves for (A) ) Fe(CN)6'3’4 (B) Ru(NH3)6*2’*3
(o) "org“ (0) methyl viologen in 1 M KCI, and (E) Fe+2’+3, (F) tert-butyl catechol
in 0.1 M HCIO4 at a boron-doped microcrystalline diamond thin-film
electrode ............................................................................................ 64

Figure 3.6 Cyclic voltammetric i-E curves for (A) ) Fe(CN)6'3“ (B) Ru(NH3)5*2’*3
(o) erle'Z’a (D) methyl viologen in 1 M KCI, and (E) Fem“, (F) tert-butyl catechol
in 0.1 M HCIO4 at a boron-doped nanocrystalline diamond thin-film
electrode ........................................................................................... 65

 

Figure 3.7 (A) Plots of i °" vs. scan rate"2 for Fe(CN)6’3"‘, Ru(NH3)6*3’*2, "C's-21.3,
methyl viologen, Fem , tert-butyl catechol at a boron-doped microcrystalline
diamond thin-film electrode. (B) Experimental and theoretical comparison for
Fe(CN)5'3"‘ curves ................................................................................ 69

Figure 4.1 Phenoxy radical formation ....................................................... 74

Figure 4.2 The different possible reaction pathways for phenol oxidation
........................................................................................................ 75

Figure 4.3 Cyclic voltammetric i-E curves for (A) microcrystalline diamond, (B)
nanocrystalline diamond, and (C) GC in 0.05 M phosphate buffer, pH 3.5 ........ 80

Figure 4.4 Cyclic voltammetric i-E curves for 2-CP, 4-CP, and phenol at (A)
microcrystalline diamond, (B) nanocrystalline diamond electrodes in 0.05 M
phosphate buffer, pH 3.5 ....................................................................... 83

Figure 4.5 Cyclic voltammetric i-E curves for phenol, 2-CP, and 4-CP at (A)
microcrystalline and, (B) nanocrystalline diamond electrodes in 0.05 M
phosphate buffer, pH 3.5 ....................................................................... 85

Figure 4.6 Cyclic voltammetric i-E curves for 0.1 mM PCP at (A)
microcrystalline and (B) nanocrystalline diamond electrodes in 0.05 M phosphate
buffer, pH 3.5 ..................................................................................... 88

Figure 4.7 AFM images after 25 cycles of 0.1 mM PCP in 0.05 M phosphate
buffer, pH 3.5, at a microcrystalline diamond electrode ................................ 90

Figure 4.8 Cyclic voltammetric i-E curves for 0.1 mM 3-CP at (A)
microcrystalline and (B) nanocrystalline diamond electrodes in 0.05 M phosphate
buffer, pH 3.5 ...................................................................................... 91

Figure 4.9 Cyclic voltammetric i-E curves for 0.1 mM 2-CP at 60 in 0.05 M
phosphate buffer, pH 3.5 ....................................................................... 93

Figure 5.1 FIA-EC hydrodynamic voltammetric background i-E curves for
microcrystalline and nanocrystalline diamond, and GC in 0.05 M phosphate
buffer, pH 3.5 .................................................................................... 101

Figure 5.2 Hydrodynamic voltammetric i-E curves for 40 uM concentrations of
each phenolic compound in 50 mM phosphate buffer, pH 3.5, at a (A)
microcrystalline and (B) nanocrystalline diamond thin-film electrode .............. 104

Figure 5.3 8/3 vs potential plots for 40 uM concentrations of each phenolic

compound in 50 mM phosphate buffer, pH 3.5, at a (A) microcrystalline and (B)
nanocrystalline diamond thin-film electrode .............................................. 105

xi

Figure 5.4 Calibration curves for phenol and chlorinated phenols obtained at
microcrystalline diamond ...................................................................... 107

Figure 5.5 FIA-EC response for multiple injections of 40 (1M 2-CP. ................ 110

Figure 5.6 FIA-EC responses for multiple injections of 40 (M phenol, 2-CP, 3-CP,
4-CP, and PCP on GC ......................................................................... 111

Figure 5.7 Reversed-phase chromatogram for phenol and the chlorinated
phenols with amperometric detection using a microcrystalline diamond thin-film
electrode .......................................................................................... 114

Figure 5.8 Calibration curves for phenol and chlorinated phenols at a (A)
microcrystalline and (B) nanocrystalline diamond thin-film electrode .............. 115

Figure 5.9 Reversed-phase chromatogram for the separation and amperometric
detection of 2-CP in a contaminated soil sample at a microcrystalline diamond
thin-film electrode ............................................................................... 119

Figure 6.1 SEM images of an electrochemically etched 76-um sharpened
platinum wire coated with a polycrystalline diamond film ............................. 126

Figure 6.2 Raman spectrum for a microcrystalline boron-doped diamond film
deposited on a 76-um platinum wire ....................................................... 127

Figure 6.3 (A) Background cyclic voltammeter l-E curve for a boron-doped
diamond microelectrode in 1 M ACL. Scan rate = 100 mV/s. (B) Cyclic
voltammeter I-E curve for 1 mom Fe(CN)6'3’“‘ in 1 M ACL at a boron-doped
diamond microelectrode ....................................................................... 130

Figure 6.4 Cyclic voltammeter l-E curve for 50 uM 2-chlorophenol in a 10/20
mom mixed-borate/phosphate run buffer, pH 8.4 at a boron-doped diamond
microelectrode ................................................................................... 132

Figure 6.5 Hydrodynamic voltammeter l-E curves for the six chlorinated phenols
at a boron-doped diamond microelectrode recorded in a 10/20 mom
borate/phosphate run buffer, pH 8.4 ....................................................... 135

Figure 6.6 A typical electropherogram for phenol and six chlorinated phenols
using direct amperometry recorded in a 10/20 mM borate/phosphate run buffer,
pH 8.4 .............................................................................................. 139

Figure 6.7 Cyclic voltammetric i-E curve for 0.8 mM ferrocene carboxylic acid in
10 mM phosphate buffer, pH 8.1, at a boron-doped diamond microelectrode...145

xii

Figure 6.8 An electropherogram for 3-chlorophenol, 2-chlorophenol, and 2,4-
dichlorophenol using indirect amperometry recorded in a 10 mM phosphate run
buffer, pH 8.1 .................................................................................... 147

Figure 7.1 An electropherogram for a standard solution of chlorinated phenols
using direct amperometry recorded in a 10/20 mM borate/phosphate run buffer,
pH 8.4 ............................................................................................. 161

Figure 7.2 Calibration curve for a river water sample spiked with different
concentrations of 2-CP ........................................................................ 164

Figure 7.3 An electropherogram for a river water sample spiked with 0.1-0.2 ppm
of chlorinated phenols after extraction, recorded in a 10/20 mM borate/phosphate
run buffer, pH 8.4 ............................................................................... 165

xiii

List of Tables

Table 1.1 EPA current national recommended water quality criteria .................. 4
Table 2.1 Microcrystalline Diamond Thin Film Deposition Parameters ............. 31
Table 2.2 Nanocrystalline Diamond Thin Film Deposition Parameters ............. 33
Table 2.3 Properties of Phenol and the Chlorinated phenols ......................... 49
Table 3.1 Analytical Techniques Used for Film Characterization .................... 52

Table 3.2 Cyclic Voltammetric Data for the Microcrystalline Boron-Doped
Diamond Electrode ............................................................................... 66

Table 3.3 Cyclic Voltammetric Data for the Nanocrystalline Boron-Doped
Diamond Electrode ............................................................................... 66

Table 5.1 FIA-EC Data for Phenol and Chlorinated Phenols at a Microcrystalline
Diamond Thin-Film Electrode ................................................................ 108

Table 5.2 FIA-EC Data for Phenol and Chlorinated Phenols at a Nanocrystalline
Diamond Thin-Film Electrode ................................................................ 108

Table 5.3 Response precision for phenol and chlorinated phenols at
microcrystalline and nanocrystalline diamond, and 60 in 50 mM phosphate
buffer, pH 3.5 .................................................................................... 110

Table 5.4 HPLC-EC Detection Figures of Merit for Phenol and the Chlorinated
Phenols at a Microcrystalline Diamond Thin-Film Electrode ......................... 116

Table 5.5 HPLC-EC Detection Figures of Merit for Phenol and the Chlorinated
Phenols at a Nanocrystalline Diamond Thin-Film Electrode ......................... 116

Table 6.1 pKa Values and Elution Order for Phenol and the Chlorinated
Phenols ............................................................................................ 135

Table 6.2 Detection Figures of Merit for Phenol and the Chlorinated Phenols in
Direct Amperometry Coupled with Capillary Electrophoresis ........................ 140

Table 6.3 Detection Figures of Merit for Chlorinated Phenols in CZE with Indirect
Amperometric Detection ....................................................................... 151

xiv

Table 7.1 Recovery of chlorinated phenols from 50 mL river water (pH 2.5)
spiked with 0.1 - 0.2 ppm of the pollutants .............................................. 165

Table 7.2 Detection Figures of Merit for Chlorinated Phenols using off-line SPE
with Capillary Electrophoresis and Amperometric Detection ........................ 167

XV

CHAPTER 1

Background and Motivations

1.1. Chlorinated Phenols

1.1.1. Environmental Importance

Chlorinated phenols are ubiquitous pollutants present in the wastewater
from several industrial processes including the manufacture of dyes, plastics,
pesticides, antioxidants, paper and petroleum products.” Leaching of these
priority pollutants from point sources, such as landfills, is another source of
contamination in ground and surface water. Pentachlorophenol and 2,4,6-
trichlorophenol have been used as wood preservatives and pesticides, whereas
2,4-dichlorophenol and 2,4,5-trichlorophenol are used as precursors for the
synthesis of herbicides.4 Chlorinated phenols can also be formed by biotic and
abiotic transformation of pesticides in agriculture. Photolytic degradation of
substituted phenols is connected with the formation of lower chlorinated phenols
and chlorinated catechols, dihydrobenzenes, hydroxybiphenyls and

chlorophenoxyphenols.3

Chlorinated phenols are of particular importance in the environment
because of their toxicity to most aquatic organisms and tendency to
bioaccumulate in the food chain.”7 The more highly chlorinated phenols, such
as trichlorophenols and pentachlorophenol (PCP), are also persistent. Large-
scale use of chlorinated phenols over the years has led to the contamination of

terrestrial and aquatic ecosystems.

Chlorinated phenols are readily absorbed when administered by oral,
inhalation or dermal contact.8'9 They accumulate mostly in the liver and kidney of
experimental animals and to a lesser degree in the brain, muscle and fat.8 PCP is
enriched easily in adipose tissue of humans and animals. Chlorinated phenols
are bound to glucuronide or sulphate in the liver. Their elimination from the
organism occurs via the urine either in the free form or as sulphate and
glucuronide conjugates.1o The toxic effects of chlorinated phenols are directly
proportional to the degree of chlorination. PCP is the most toxic of these
compounds, presumably as a result of increased Iipophilicity. An increase in the
number of chloro—substituents can also promote the accummulation of these
compounds in fish and other organisms, thus increasing their toxic effects.11 The
major mode of action of chlorinated phenols involves the uncoupling of oxidative
phosphorylation and the inhibition of the electron transport system. The
chlorophenolate ion is evidently responsible for the uncoupling reaction, while the

undissociated molecule causes convulsions.8

Acute exposure to lesser chlorinated phenols in humans results in muscular
twitching, spasms, tremors, weakness, ataxia, convulsions and collapse. Acute
poisoning by PCP is characterized by general weakness, fatigue, ataxia,
headache, anorexia, sweating, hyperpyrexia, sweating, nausea, vomiting,
tachycardia, abdominal pain, terminal spasms and death. Soft-tissue sarcomas,
Hodgkin's disease, and leukaemia have been reported in epidemiological studies
of occupational groups exposed to chlorinated phenols and phenoxy acids.12
Some chlorinated phenols have been suspected of embryotoxicity and
tetarogenicity properties.8 Studies dealing with chronic toxicity of some phenol
congeners have confirmed their carcinogenic (hepatocellular tumors, leukaemia)
and immunosuppressive properties.3'8 For example, PCP has been classified as
a 82 probable carcinogen for humans on the basis of a large body of evidence
from animal toxicity studies and human clinical data.13 PCP can also provoke

damage to the eyes, pharynx and larynx.9

1.1.2. Chlorinated Phenols Analysis

Development of sensitive, reproducible and stable detection methods for
chlorinated phenols is an important area of research. Several chlorinated
phenols, such as 2-chlorophenol, 2,4-dichlorophenol, 2,4,6-trichlorophenol, and
pentachlorophenol, have been classified as priority pollutants by the
Environmental Protection Agency.“ For example, PCP is regulated in drinking
water with a maximum allowable contaminant level (MCL) of 1 uglL (1 ppb).15

The EPA has also set the maximum limit in drinking water of 2-chlorophenol at

40 pg/L and 2,4-dichlorophenol at 20 ug/L for “lifetime exposure” which is defined
as the concentration of a chemical in drinking water that is not expected to cause
any adverse noncarcinogenic effects for a lifetime exposure.16 Chlorinated
phenols can also impart foul taste and odor to drinking water and food products
(fish), even at low concentrations?"7 Table 1.1 shows the current EPA
recommended water quality criteria for organoleptic effect of some chlorinated

phenols.17

Table 1.1 EPA Recommended Water Quality Criteria

 

 

 

 

 

 

 

 

 

 

 

Compound Organoleptic effect
criteria (pg/L)“

Phenol 300

2-chlorophenol 0.1

3-chlorophenol 0.1

4-chlorophenol 0.1

2,4-dichlorophenol 0.3

2,4,6-trichlorophenol 2

Pentachlorophenol 30

 

a maximum allowed concentration in water that does not cause any organoleptic
effect in water

The most commonly used separation and detection methods for phenol
and chlorinated phenol analysis are reversed-phase liquid chromatography
(HPLC) and gas Chromatography (GC).""28 In LC, UV detection is usually
employed. The sensitivity for UV-Vis for chlorinated phenols is low leading to
detection limits in low ppm range.1M1 In GC, various detection techniques are

employed, such as flame ionization (FID), electron-capture (ECD), atomic

emission, or mass spectrometry (MS). In order to enhance the volatility and
detectability of phenols, derivatization is usually necessary.”28 Derivatization
procedures increase the complexity of the assay, the analysis time, and the risk
for possible indeterminate error. The EPA-recommended protocols (methods 604
and 625) for phenol and chlorinated phenol speciation and detection involve
liquid-liquid extraction (LLE) or solid phase extraction (SPE) followed by either
GC with electron capture or mass spectrometric (MS) detection of derivatized
phenols?“0 Typical detection limits are in the range of 0.2 — 5 pg/L (0.2 -5 ppb).
These methods utilize expensive instnlmentation and complex sample treatment,
which involves preconcentration and derivatization of phenols to
pentafluorobenzoyloxy derivatives. Another protocol (method 8041) involves
methylation of the chlorinated phenols using diazomethane prior to GO
analysis.31 Unfortunately, this derivatization agent is carcinogenic and
explosive.32 Even with these accepted protocols, there remains a need for an
inexpensive, reliable, fast, and sensitive method for monitoring chlorinated

phenols in the aqueous environments.

Separation science coupled with electrochemical detection is a practical
approach for the speciation and detection of chemical contaminants
(electroactive ones) in water supplies. The instrumentation required is relatively
inexpensive, easy to use, and field deployable. Electrochemical detection can be
readily combined with separation techniques (e.g., HPLC, CE). Due to the high

sensitivity of electrochemical detection, detection limits in mid to high ppb range

are possible.”36 The limiting factor in implementing electrochemical detection for
water quality monitoring is often the electrode material in terms of the stability of
its structure and properties, and its electrochemical responsiveness. This is
particularly true for the detection of phenol and chlorinated phenols. Phenols are
electro-oxidizable at positive potentials (+ 1.0 V vs. Ag/AgCI) in aqueous media,
making electrochemistry a viable detection method. Unfortunately,
electrochemical detection of this class of pollutants is often plagued by the strong
irreversible adsorption (i.e. fouling) of reaction intermediates and products.37
Mechanistic studies by several groups reveal that the phenol oxidation reaction
proceeds initially through the formation of a phenoxy radical species, which can
subsequently undergo radical-radical coupling to form polymeric and oligomeric
species.3742 Fouling by oligomeric and/or polymeric (e.g. polyphenol) leads to
poor electrode response reproducibility and stability, and sometimes complete
response attenuation. In such cases, the electrode must be reconditioned (e.g.

mechanical polishing). recalibrated and revalidated.

Most of the reported electrochemical detection studies of chlorinated
phenol analysis have used modified electrodes due to the complications with
direct detection on bare electrodes.”36 Lu and co-workers modified 3 Pt
electrode with the conducting polymers, polyaniline and polypyrrole, to alleviated
electrode fouling during phenol oxidation.35 The phenol oxidation response was
unchanged during multiple injections, and low detection limits (10 ppb) were

achieved in flow injection analysis with amperometric detection (FIA-EC). In

another report, Saby and co-workers immobilized glucose oxidase on a glassy
carbon electrode for the detection of chlorinated phenols.33 The detection limits
were in nanomolar range (4 — 65 nM) for several chlorinated phenols. Given the
electrode fouling issue and the need to have an inexpensive, sensitive, and
reproducible analysis method for the chlorinated phenols, it would be
advantageous to have an electrode that is resistant to fouling but is still
responsive to the analytes of interest. Boron-doped, hydrogen-tenninated,
polycrystalline diamond thin-film is an advanced electrode material that is more
microstnlcturally stable and exhibits less tendency to foul than do other carbon
materials.”44 This is due, at least in part, to the fact that the electrodes are
relatively non-polar when hydrogen terminated and contain no extended 1r-
electron system. Adsorption studies conducted in our laboratory using
anthraquinone-2,6-disulfonate (AQDS) on glassy carbon, graphite and diamond
electrodes showed that 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.44

More important is that diamond exhibits superb response stability. The
advantage of this property for electroanalysis is the fact many analytes are
detectable via electrooxidation at positive potentials. A problem with spz-bonded
carbon electrodes for example, is the alteration in the surface microstructure and

chemistry that occurs over a wide potential range. These changes lead to

progressive increases in the background current (i.e., decreasing SIB), pH-
dependence of background and faradaic signals, a greater susceptibility to
fouling due to molecular adsorption, and sometimes an attenuation in the analyte
response. While the diamond surface chemistry can easily be changed, the
microstructure and morphology are extremely stable, leading to excellent long-

term response precision and stability.”46

The objective for this research was to investigate the use of the boron-
doped diamond electrodes for the electrochemical detection of chlorinated
phenols in water supplies. More specifically, this research sought (i) to gain a
fundamental understanding of how the chlorinated phenols oxidation reaction
kinetics and mechanism are affected by the physicochemical properties of the
diamond electrode and (ii) to develop a sensitive and stable monitoring method
for chlorinated phenols by coupling separation techniques, such as liquid

chromatography and capillary electrophoresis with electrochemical detection.

2. Conductive Boron-Doped Diamond Electrodes

sp2 bonded carbon materials in the forms of highly ordered pyrolytic
graphite (HOPG), glassy carbon (GC), carbon paste, and carbon fibers, have
been used for over four decades in electrochemical measurements. Carbon
electrodes are used in electrosynthesis, electroanalysis, and electrochemical-
based energy storage and conversion devices.47 sp2 carbon electrodes have

several important properties such as being (i) available in a variety of forms, (ii)

inexpensive (iii) chemically modifiable to control electron-transfer and molecular

adsorption and (iv) stable over a wide potential range.“53

Single crystal graphite and HOPG are benchmark carbon materials that
possess an anisotropic structure and properties. These well ordered materials
have dimensions on the order of a few pm’s (> 1 pm) compared to tens of A for
more disordered materials, like GC and carbon black. HOPG consists of stacked
graphite sheets with an interplane distance of 3.354 A. GC resembles a twisted,
bent, interlocked mass of graphite ribbons. The interplane spacing is 3.48 A,
slightly larger than that of HOPG. The graphite ribbons are composed of basic
structural units made up of 10-20 rings stacked up more or less in parallel by two
to four layers. GC contains graphite like domains ranging from 25 to 100 A. GC is
one of the most widely used sp2 carbon electrode materials for electroanalysis.53
It has low porosity, a reversible surface and is chemically inert in most common

solvents.

Diamond is a new carbon material that possesses unique properties for a
wide range of applications. It consists of a tetrahedrally bonded, face-centered
cubic (FCC) crystal structure. Diamond possesses spa-bonded carbon and has
unique properties including chemical inertness, extreme hardness, a wide
bandgap, optical transparency, low dielectric constant, high thermal conductivity,
and high carrier mobility. These properties are as a consequence of the small

interatomic distance of the carbon atoms in the lattice and the hydrogen surface

termination.5"”56 A unique property of diamond is its morphological and
microstructural stability as compared to conventional electrode materials. The
stability arises from the high density and the strong directional spa C-C bonding.
Diamond electrodes, for example, are stable during anodic polarization in harsh
acidic and alkaline media. Surface damage, pitting, grain roughing and
delamination are not normally observed even after lengthy electrolysis times

producing chlorine and ozone."""58

Diamond possesses a number of practical electrochemical properties that
distinguish it from other commonly conventional spZ-bonded carbon electrode
materials. These properties include (i) a low and stable background current that
leads to improved signal-to-background (SBR) and signal-to-noise (SNR) ratios,
(ii) a wide working potential window in aqueous and nonaqueous media, (iii)
superb microstructural and morphological stability at high temperature and
current density, (iv) good responsiveness for several aqueous and nonaqueous
redox analytes without conventional pretreatment, (v) weak adsorption of polar
molecules due to the hydrophobic and non-polar nature of the hydrogen-
terrninated surface, and (vi) long-term response stability.59 3"“ ”mm“ "mm The
wide working potential window of diamond in aqueous media stems from the
large overpotential for the hydrogen evolution (HER) and oxygen reduction
reactions (OER). The as-deposited diamond surface is hydrogen-tenninated and
is not favorable for physical and chemical interactions promoting adsorption and

stabilization of radicals. Thus, electrochemical processes that involve adsorbed

10

intermediates or radical intermediates are generally kinetically hindered at
diamond. 6°51 The poor stabilization of reaction intermediates, like those involved
in water electrolysis (oxygen or hydrogen evolution), leads to the large working
potential window in aqueous media.62 For high quality diamond, the potential
window is usually 3.5 V or more. The wide working potential window and superb
morphological and microstructural stability of the diamond surface permits
electrochemical reactions at extreme potentials that otherwise would be difficult
to accomplish. Within the working potential window, diamond generally shows a

small double layer capacitance and low background current.63

The capacitance
ranges from 1 to 5 )JF/cm2 for diamond as compared to G0 which ranges from 30
to 40 uF/cmz. The background current for diamond is generally an order of
magnitude lower than that for G0. The reasons for the low capacitance for

diamond are explained in Chapter 3. The low background current and noise

magnitudes give improved signal-to-background and signal-to-noise ratios.64

Several factors influence the electrochemical properties of polycrystalline
diamond films including: (i) the dopant type, level, and distribution, (ii) the surface
termination (H vs. 0), (iii) non-diamond carbon impurity phases, (iv) the grain
boundaries and other morphological defects, and (vi) the primary crystallographic
orientation. The complexity of the diamond electrode properties resulting from
these factors is known, but not completely understood. Several groups have
contributed to the development of and applications for diamond electrodes (i.e.

aspects of growth and nucleation mechanisms of diamond, the factors

11

influencing the electrochemical response, application in electroanalysis and
electrocatalysis), including, Fujishima et al.,94'97 (Japan), Pleskov et al.,7578
(Russia), Angus et al.,79432 (USA), Miller, et al.,97'101 (USA), Gruen et alugcge

(USA), and Compton et al.,83439 (England).

Diamond shows more versatility as an electrode than any other material. It
can be used in electroanalysis to provide sensitive detection of analytes with
superb precision and stability; for high current density electrolysis in aggressive
solution environments without any microstructural or morphological degradation,
and as a optically transparent electrode for spectrochemical measurements in
the UVNis and iR regions of the electromagnetic spectrum.‘°"* 107'115'1‘6 Diamond
electrodes have been used successively used for the amperometric detection of

44,110-113

several analytes (e.g. azides,107 chlorinated phenols, chlorpromazine,59

aliphatic amines117 carbamate pesticides“, cysteine”). Diamond has also been
used for the detection of trace metal ions by anodic stripping voltammetrym"119
In general, hydrogen-terminated diamond electrodes provide superior detection
figures of merit in terms of the linear dynamic range, response precision,

response stability, and limit of detection, as compared to all other sp2 carbon

materials.

1.3. Diamond Electrode Deposition and Doping

Diamond electrodes for electrochemical applications can be formed in

several different geometries: (i) a planar thin film deposited on a conducting

12

substrate, like Si (macroelectrode) (ii) thin films deposited on sharpened metal
wire, like Pt (microelectrode), (iii) a thin film deposited on high surface area metal
meshes (gas diffusion electrode), and (iv) a pattern of electrically-isolated

diamond array elements (e.g., discs) on a planar support (microelectrode array).

Synthetic diamond can be produced by high-pressure, high temperature
(HPHT) techniques or by low pressure chemical vapor deposition (CVD). During
the HPHT process, solid carbon is heated under extreme pressure in the
presence of a molten salt catalyst. The diamond crystals produced are used for a
wide range of industrial applications that require thermal conductivity, hardness
and wear resistance. A drawback, however, is the cost and the fact that the
method produces diamond only in a single crystal.5‘"12"122 CVD techniques used
to produce diamond require activation of gas phase precursor molecules. This
activation can be accomplished thermally (e.g., a hot filament), by electric
discharge (e.g. DC, RF, or microwave) or by a combustion flame (e.g. an
oxyacetylene torch).5“"2"'121 Microwave plasma CVD (MWCVD) is the most
widely used method for diamond growth. In this method, microwave power is
coupled into the chamber via a dielectric window to create a plasma discharge.
The microwaves transfer energy into electrons in the gas phase, which in turn
transfer their energy to the gas molecules through collisions. This leads to
heating and dissociation of the gas molecules, and the formation of the active

species, that react on the substrate to form diamond.5'"""°'121

13

Diamond is deposited from a hydrogen-rich hydrocarbon source gas
mixture (CH4/H2).‘23'125 It is recognized that hydrogen, specifically atomic
hydrogen formed in the plasma, plays an essential role in diamond growth.
Hydrogen atoms react with neutral species such as CH4 to create CH3- reactive
radicals in close proximity to the substrate surface and can attach to surface
sites. Atomic hydrogen is known to gasify graphitic spz-carbon. The gasification
(formation of CH4 or Csz) rate for sp2 bonded carbon is significantly higher than
sp3 bonded carbon. This means that the non—diamond carbon that does form is
rapidly removed from the surface. Atomic hydrogen also serves to activate gas
phase reactants by hydrogen abstraction and to activate growth or attachment

sites on the surface by hydrogen abstraction.

Diamond films, grown by CVD, can be formed in two morphologies:
microcrystalline and nanocrystalline. CH4/H2 source gas mixtures are used to
produce microcrystalline diamond thin film. This film possesses a well-faceted,
polycrystalline morphology with a nominal crystallite size of ~ 2 um, or greater.
CHJHg/Ar source gas mixtures are used to produce nanocrystalline diamond thin
film.‘26'127 The fraction of atoms in the grain boundaries of microcrystalline films
relative to the bulk is extremely small. The nanocrystallinity is a result of a new
growth and nucleation mechanism that involves insertion of the carbon dimer, C2,
into surface carbon-carbon and carbon-hydrogen bonds. Very high
heterogeneous renucleation rates are observed and the resulting films consist of

randomly oriented, phase pure diamond grains with well-defined grain

14

boundaries.”128 Very simply, to grow nanograined films, the growth conditions
are adjusted such that the rate of renucleation exceeds the rate of crystal growth.
The opposite is the case for the growth of microcrystalline films. The
nanocrystalline diamond consists of randomly oriented 3 — 10 nm crystallites,
compared with the columnar microstructure of the microcrystalline film. The
difference in individual crystallite size means that secondary nucleation rates
associated with growth from hydrogen-poor plasmas are much higher than from
CH4/H2 plasmas.9‘"128 Grain boundaries of a nanocrystalline are different from
those of microcrystalline. It was found that they are mainly three fold coordinated
and form it bonds. The electronic band structure of grains was characterized by a
smaller band gap than the bulk diamond and it was shown that these states can
participate in hopping conductionf’o'128 The special nature of n—bonded grain
boundaries imparts unique electrical properties on nanocrystalline diamond films,
and the conducting grain boundaries will rule the electronic properties of
undoped film. With increasing crystallite size their number vastly increases, and
the entire film becomes electrically conducting. Changes in nanocrystallites,
therefore, can dramatically alter film properties. 90"” More detailed discussion of
the film morphology and microstructure, along with evaluation of the
electrochemical properties of boron—doped microcrystalline and nanocrystalline

diamond is presented further in Chapter 3.

Diamond is a very good electrical insulator with a band gap of 5.5 eV.

Impurities, such as boron, can be incorporated into the film during growth that

15

impart semiconducting or semimettalic electronic properties depending on the
doping level. Naturally occurring impurities that have implications for the host
diamond consists of N, H, and B. 90'1” The N incorporation gives diamond n-type
electronic character. Substitutional nitrogen is a deep donor with ionization
energy of about 1.7 ev. The donors are typically too deep to provide any useful
semiconducting properties to the material at room temperatures. Among the
common dopants, boron is the most often used and can be introduced into
diamond without significant distortion of the lattice. Boron may be introduced into
CVD diamond film from a gaseous source, like 82%, a solid state source, such
as 3203 or h-BN, or from a boron containing solution such as 8203 dissolved in
organic solvents from which vapors are carried into the reaction chamber by the
growth gases.13o Boron atoms substitutionally insert for some of the carbon
atoms in the growing diamond lattice. These boron atoms function as electron
acceptors, with an activation energy of 0.37 eV or less, depending on the doping
level, and, at room temperature, contribute to the formation of free-charge
carriers (i.e., holes or electron vacancies)”1 The film’s electrical conductivity is
directly related to the carrier concentration and the carrier mobility. For both
types of diamond, typical carrier concentrations are in the range of 1018 to 1020
cm'3 with carrier mobilities (holes) of 10-500 cmzN-s. Of course, the mobility is
limited by the defect density within the film. Activation energy of acceptors is
observed to decrease with increasing boron concentration. This decrease of the
energy can be explained by acceptor band width increase, impurity band

formation and conduction as a result of it. For highly doped films, activation

16

energies as low as 2 meV have been reported.132 For these highly doped films,
the impurity band merges with the valence band, and metallic conduction is
observed. The abundance of boron during diamond deposition can degrade the
film quality which is manifested by increased point defects and, in extreme
instances, by graphitization of the diamond film. ‘29'133 Figure 1.1 shows the

bandgap diagram for boron-doped diamond electrode.

 

Conduction Band

 

 

 

Band Gap
6 > ~ 5.4 eV

___________ I ' Band
~0.35 eV I \ mpumy

Valence Band

Energy

 

 

 

 

 

 

Figure 1.1 A schematic for the band structure of boron-doped diamond.

Highly boron-doped polycrystalline diamond electrodes are more or less

degenerate semi-metals. They deviate from ideal p-type semiconductor behavior

17

because of a high density of mid-gap electronic states. The mid gap density of
states results from at least four factors: (i) boron doping level, (ii) lattice hydrogen
content, (iii) inherent grain boundaries and other defects in the polycrystalline
films, and (N) non-diamond carbon impurity at the surface.59'109 Among these
factors, the surface non-diamond carbon phases exert little influence on the
response of doped diamond films.134 The role of morphological defects and grain
boundaries on the electrode kinetics, specifically the electrical conduction, at
polycrystalline film is still not clearly understood. Diamond is host for a variety of
extended defects such as stacking faults, microtwins, dislocation, grain
boundaries, and mixed habit growth features.59 and mamas “W“ These defects
could serve as discrete sites for electron transfer or could simply affect the
electronic properties of the material by increasing the density of states. The mid-
gap density of states arising from grain boundaries and defects, is relatively low
compared to the number of states arising from the boron doping and lattice

hydrogen.

1.4. Outline of the Dissertation

Details of experimental approaches used are given in Chapter 2. A
discussion of the basic electrochemical and physical properties of boron-doped
microcrystalline and nanocrystalline diamond thin films is given in Chapter 3. In
Chapter 4, the voltammetric studies of several different chlorinated phenols are
reported for both microcrystalline and nanocrystalline diamond. The analysis of

the chlorinated phenols in standard solutions by flow injection analysis (FIA) and

18

liquid chromatography (LC), both with amperometric detection, is discussed in
Chapter 5. The analysis of the chlorinated phenols in standard solutions by
capillary electrophoresis (CE) with amperometric detection is described in
Chapter 6. Finally, in Chapter 7, results for the analysis of chlorinated phenol
using off-line solid phase extraction (SPE) and capillary electrophoresis with
amperometric detection are presented. A brief conclusion follows, summarizing

the results and the main contribution of the work.

19

1.5. References

1.

10.

11.

12.

13.

14.

15.

16.

17.

Masque, N.; Pocurull, E.; Marce, R.; Borull, F.; Chromatographia, 1998,
47,176.

Galceran, M.; Jauregui, 0.; Anal. Chim. Acta 1995, 304, 75.

Veningerova, M.; Prachar, V.; Uhnak, J.; Lucacsova, M.; Tmovec, T.; J.
Chromatogr. B 1994, 657, 103.

Lindstrom, K.; Nordin, J.; J. Chromatogr. 1976, 128, 13.
Vuorinen, P.J.; Chemosphere 1985, 14, 1729.

Sarkka, J.; Chemosphere 1985, 14,469.

Realini, P.A.; J. Chromatogr. Sci. 1981, 19, 124.

WHO, Environmental health criteria for chlorophenols other than
pentachlorophenol, Supplement Draft, July 31, 1986.

WHO, Environmental health criteria for pentachlorophenol. Draft, March
19, 1986.

Braun, W.H.; Blau, G.E.; Chenowith, M.B. Toxicol. Appl. Pharrnaool. 1978,
45, 278.

Kishino, T.; Kobayshi, K.; Wat. Res. 1996 30, 393.

WHO. Guidelines for drinking-water quality. Vol. 2. Health criteria and
other supporting information. Geneva 1984.

Tanjore, S.; Viraraghaghavan, Int. J. Environ. Stud. 1994, 45, 155.

Noguera, P.; Maquieira, A.; Puchades, R.; Brunet, E.; Carramolino, M.;
Rodriguez-Ubis, J.C.; J. Environ. Monit. 2002, 4, 442.

National Primary Drinking Water Regulations, Technical Factsheet on
Pentachlorophenol, Office of Ground and Drinking Water, US. EPA, 1998.

Drinking Water Standards and Health Advisories, Summer 200, US. EPA,
2000.

httfizll www.epa.govlwater science/criterialwqcriteria.html accessed June
27

20

 

18.

19.
20.
21.
22.
23.

24.

25.

26.
27.

28.

29.

30.

31.

32.
33.

34.

35.

Andres, M.P.; Leon-Gonzalez, M.E.; Perez-Anibas, L.V.; Polo-Diez, L.M.;
J. High Resol. Chromatogr. 2000, 23, 367.

Coman, V.; Moldovan, 2.; J. High Resol. Chromatogr. 2000, 23, 699.
Xueliang, L.; Frank, H.; J. High Resol. Chromatogr. 1998, 21, 309.
Frebortova, J.; Tatarkovicova, V.; Analyst, 1994, 19, 1519.

Angelino, S.; Gennaro, M.C.; Anal. Chim. Acta. 1997, 346, 61.

Tumes, l.; Rodriguez, l.; Gracia, C.M.; Cela, R.; J. Chromatogr. A 1996,
743, 283.

Hoogerbrugge, R.; Gort, S.M. Van der Velde, E.G.; Zoonen, P.; Anal.
Chim. Acta. 1999, 388, 119.

Rodriguez, l.; Tumes, M.l.; Mejuto, M.C.; Cela, R.; J. Chromatogr. A 1996,
721, 297.

Hedges, J.l.; Ertel, J.R.; Anal. Chem; 1982, 54, 174.
Ruana, J.; Urbe, l.; Borrull, F.; J. Chromatogr. A, 1993, 665, 217.

Pocurull, E.; Sanchez, G.; Borrull, F.; Marce, R.M.; J. Chromatogr. A 1995,
696,31.

EPA Method 604, Phenols, in Federal Register, Environmental
Protection Agency, Part Vlll, 40 CFR Part 136, Washington DC, 1984, 58.

EPA Method 625, Base/Neutrals and Acids, in Federal Register
Environmental Protection Agency, Part Vlll, 40 CFR Part 136, Washington
DC, 1984, 153.

EPA Method 8041, Phenols by Gas Chromatography: Capillary Column
Technique, Environmental Protection Agency, Washington DC, 1995, 1.

Bruno, F.; Pham, M.C.; Dubois, J.E.; Electrochim. Acta. 1977, 22, 451.
Saby, C.; Luong, J.H.T.; Electroanalysis, 1998, 10, 7.

Agui, L.; Serra, B.; Yanez-Sedeno, P.; Reviejo, A.J.; Pingarron, J.M.;
Electroanalysis, 2001, 13, 1231.

Lu, W.; Wallace, G.G.; lmisides, M.D.; Electroanalysis, 2002, 14, 325

21

36.

37.
38.

39.

40.
41.
42.

43.

45.

46.

47.

48.

49.

50

51.
52.

53.

55.

Saterlay, A. J.; Foord, J.S.; Compton, R.G.; Electroanalysis, 2001, 13,
1065.

Gattrell, M.; Kirk, D.W.; J. Electrochem. Soc. 1993, 140, 1534.
Belhadj Tahar, N.; Savall, A.; J. Electrochem. Soc. 1998, 145, 3427.

Rodgers, J.D.; Jedral, W.; Bunce, N.; J. Environ. Sci. Techno/.1999, 33,
1453.

Gattrell, M.; Kirk, D.W.; J. Electrochem. Soc. 1993, 140, 903.

Mengoli, G.; Daolio, S.; Musiani, M.M.; J. Electrochem. Soc.1980,10, 459.
Gattrell, M.; Kirk, D.W.; Can. J. Chem. Eng. 1990, 68, 997.

Xu, J.; Chen, 0.; Swain, G.M.; Anal. Chem. 1998, 70, 3146.

Muna, G.W.; Tasheva, N.; Swain, G.M.; Environ. Sci. Technol. 2004, 38,
3674.

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

Hupert, M.; Muck, A.; Wang, J.; Stotter, J.; Cvackova, Z.; Haymond, S.;
Show, Y.; Swain, G.M.; Diem. Rel. Mater. 2003, 12, 1940.

Kinoshita, K.; Carbon: Electrochemical and Physicochemical Properties;
John wiley & Sons: New York, 1988.

Arihara, K.; Kitamura, F.; Nukanobu, K.; Ohsaka, T.; Tokuda, K.; J.
Electroanal. Chem. 1999, 473, 138.

Edstrom, K.; Herranen, M.; J. Electrochem. Soc.,2000,147, 3628.
Chen, P.; McCreery, R.L.; Anal. Chem. 1996, 68,3958.
Duebgen, R.; Can. Chem. News, 1987, 39, 13.

Engelman, E. E.; Evans, D.H.; Langmuir, 1992, 8, 1637.

Kamau, G.N.; Anal. Chim. Acta. 1988, 207, 1.

Angus, J.C.; Argoitia, A.; Gat, R.; Li, Z.; Sunkara, M.; Wang, L.; Wang, Y.;
Philos. Trans. R. Soc. Lond. Ser. A-Math. Phy. Eng. Sci. 1993, 342, 195.

Angus, J.C.; Collins, A.T.; Nature, 1994, 370, 601.

22

56.

57.

58.

59

60.

61.

62.
63.

64.

65.

66.

67

68.

69

70.

71.

Angus, J.C. Martin, H.B.; Landau, U.; Evstefeeva, Y.E.; Miller, B.; Vinokur,
N.; New Diamond and Frontier Carbon Technology 1999, 9, 175.

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

Katsuki, N.; Wakita, S.; Nishiki, Y.; Shimamune, T.; Akiba, Y.; lida, M.;
Jpn. J. Appl. Phys. Lett. 1997, 36, L260.

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

Anderson, A.; Kang, D.; J. Phys. Chem; 1998, 102, 5993.

Vinokur, M; Miller, B.; Avyigal, Y.; Kalish, R.; J. Electrochem. Soc.1996,
143, L238.

Ramesham, R.; Rose, M.; Thin Solid Films, 1997, 300, 144.
Swain, G.M.; Ramesham, R.; Anal. Chem. 1993, 65, 345.

Xu, J.; Granger, M.; Chen, 0.; Strojek, J.; Lister, T.; Swain, G.M.; Anal.
Chem. 1997, 69, 591A.

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

Rao, T.N.; Loo, B.H.; Sarada, B.V.; Terashima, C.; Fujishima, A.; Anal.
Chem. 2002, 74, 1578.

Fujishima, A.; Popa, E.; Wu, Z.; Rao, T.N.; Novel trends in Electrooorganic
Synthesis, [Papers presented at the Intemational Symposium on
Electroorganic Synthesis], 3’”, Kurashiki, Japan, Sept. 24-27, 1998, 421.
Ohta, N.; Nakabayashi, S.; Fujishima, A.; Chem. Lett. 1999, 1147.

Rao, T.N.; Yagi, l.; Miwa, T.; Tryk, D.; Fujishima, A.; Anal. Chem. 1999,
71, 2605.

Rao, T.N.; Sarada, B.V.; Tyrk, D.; Fujishima, A.; J. Electroanal. Chem.
2000, 491, 175.

Sarada, B.V.; Rao, T.N.; Tyrk, D.; Fujishima, A.; Proceedings
Electrochemical Society 2000, 99-32, 502.

23

72.

73.

74.

75.

76.

77.

78.

79.

80.

81

82

83.

84.

85.

86.

87.

Chailapakul, O.; Aksharandana, P.; Frelink, T.; Einaga, Y.; Fujishima, A.;
Sens. Actuators B 2001, 880, 193.

Notsu, H.; Fukazawa, T.; Tatsuma, T.; Tyrk, D.A.; Fujishima, A.;
Electrochemical and Solid-State Letters 2001, 4, H1.

Tsunozaki, K.; Einaga, Y.; Rao, T.N.; Fujishima, A.; Chem. Lett. 2002,
502.

Pleskov, Y.V.; J. Chem. Vap. Deposition, 1997, 5, 126.

Pleskov, Y.V.; Mishuk, V.Y.;Abaturov, M.A.; Elkin, V.V.; Krotova, M.D.;
Vamin, V.P.; Teremetskaya, I.G.; Russ. J. Electrochem. 1997, 33, 61.

Modestov, A.D.; Evstefeeva, Y.E.; Pleskov, Y.V.; Mazin, V.M.; Varnin,
V.P.; Teremetskaya, I.G.; J. Electroanal. Chem. 1997, 431, 211.

Pleskov, Y.V.; Sakharova, A.Y.; Churikov, A.V.; Vamin, V.P.;
Teremetskaya, l.G.; Russ. J. Electrochem. 1996, 32, 1075.

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

Sunkara, M.; Angus, J.C.; l-layman, C.C.; Buck, P.A.; Carbon 1990, 28,
745.

Lambrecht, W.R.L; Lee, C.H.; Segall, B.; Angus, J.C.; Li, Z.D.; Sunkara,
M.; Nature 1993, 364, 607.

Mearini, G.T.; Krainsky, l.L.; Dayton, J.A.; Wang, Y.X.; Zonnan, C.A.;
Angus, J.C.; Hoffman, R.W.; Appl. Phys. Lett. 1994, 65, 2702.

Hyde, M.; Saterlay, A..J; Wilkins, S..J; Foord, J. S.; Compton, R..;G
Marken, F, J. Solid State Electrochem.2002,6, 183.

Saterlay, A.J.; Foord, J.S.; Compton, R.G.; Electroanalysis, 2001, 13,
1065.

Saterlay, A.J.; Wilkins, S.J.; Holt, K.B.; Foord, J.S.; Compton, R.G.;
Marken, F.; J. Electrochem. Soc. 2001, 148, E66.

Holt, K.B.; Delcampo, J.; Foord, J.S.; Compton, R.G.; Marken, F.; J.
Electroanal. Chem. 2001, 513, 94.

Qui, F.; Compton, R.G.; Marken, F.; Wilkins, S.J.; Goeting, C.H.; Foord,
S.J.; Anal. Chem. 2000, 72, 2362.

24

88.

89.

90.

91.

92.

93.

94.

95.

96.

97.

98.

99.

100.
101.

102.

103.

Goeting, C.H.; Marken, F.; Gutierrez-Sosa, A.; Compton, R.G.; Foord,
S.J.; New Diamond and Frontier Technology. 1999, 9, 207.

Prado, C.; Flechsig, G.-U.; Gruendler, P.; Foord, S.J.; Marken, F.;
Compton, R.G.; Analyst, 2002, 127, 329.

Gruen, D.M.; MRS Bull. 1998, 23, 32,

Krauss, A.R.; Gruen, D.M.; Auciello, O.H.; Carlisle, J.A.; Swain, G.M.;
Ding, M. In PCT Int. Appl.; (The University of Chicago, USA) WO 2002 pp
48.

Chen, 0.; Gruen, D.M.; Krauss, A.R.; Corrigan, T.D.; Witek, M.; Swain,
G.M.; J. Electrochem. Soc. 2001, 148, E44.

Fausett, B.; Granger, M.C.; Hupert, M.L; Wang, J.; Swain, G.M. Gruen,
D.M.; Electroanalysis, 2000, 12, 7.

Zhou, D.; Gruen, D.M.; Qin, L.C.; McCauley, T.G.; Krauss, A.R.; J. Appl.
Phys. 1998, 84, 1981.

Zhou, D.; McCauley, T.G.; Qin, L.C.; Krauss, A.R.; Gruen, D.M.; J. Appl.
Phys. 1998, 83, 540.

Qin, L.C.; Zhou, D.; Krauss, A.R.; Gruen, D.M.; Nanostructured Materials,
1998, 10, 649.

Yoo, K.; Miller, B.; Kalish, R.; Shi, X.; Proceedings-Electrochemical
Society, 2000, 99—17, 440.

Vinokur, N.; Miller, B.; Avyigal, Y.; Kalish, R.; Electrochemical and Solid
State letters, 1998, 1, 265.

Vinokur, N.; Miller, B.; Avyigal, Y.; Kalish, R.; J. Electrochem. Soc.,1999,
146, 125.

Vinokur, M; Miller, B.; Proc-Electrochem. Soc. 1997, 91-17, 384.

Sopchak, D.; Miller, B.; Kalish, R.; Avyigal Y.; Shi, X.; Electroanalysis,
2002, 14, 473.

Muna, G.W.; Quaiserova-Mocko, V.; Swain, G.M.; Electroanalysis, 2005,
1 7, 1 160.

Muna, G.W.; Quaiserova-Mocko, V.; Swain, G.M.; Anal. Chem. Submitted.

25

104.

105.

106.

107.

108.

109.

110.

111.

112.

113.

114.

115.

116

117.

118.

Stotter, J.M.; Haymond, 3.; Butler, J.E.; Swain, G.M.; Zak, J.K.; Abstracts
of papers, 222"" A08 National Meeting, Chicago, IL, USA, August 26-30
2001, ANYL-117.

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

Witek, M.; Wang, J.; Stotter, J.; Hupert, M.; Haymond, S.; Sonthalia, P.;
Swain, G.M.; Zak, J.K.; Chen, 0.; Gruen, D.M.; Butler, J.E.; Kobashi, K.;
Tachibana, T.; Journal of Wide Bandgap Materials, 2002, 8, 171.

Xu, J.; Swain, G.M.; Anal. Chem. 1998, 70, 1502.

Reuben, C.; Galun, E.; Tenne, R.; Kalish, R.; Muraki, Y.; Hashimoto, K.;
Fujishima, A.; Butler, J.M.; Levy-Clement, C.; NATO ASI Ser. 1995, 1,
137.

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

Gherardini, L.; Comninellis, C.; Vatistas, N.; Annalli di Chim. 2001 Mar-
Apr, 91, 161.

Terashima, C.; Rao, T.N.; Sarada, B.V.; Fujishima, A.; Chemical Sensors,
2001, 17, 124.

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

Fujishima, A. Terashima, 0.; Honda, K.; Sarada, B.V.; Rao, T.N.; New
Diamond and Frontier Carbon Technology, 2002, 12, 73.

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

Martin, H.B.; Morrison, P.W.; Jr.; Electrochemical and Solid-State letters,
2001, 4, E17.

Argoitia, A.; Martin, H.B.; Rozak, E.J.; Landau, U.; Angus, J.C.; Mater.
Res. Soc. Symp. Proc. 1996, 416, 349.

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

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

26

119.

120.

121.

122.

123.
124.
125.
126.

127

128.
129.

130.

131.

132.

133.

134.

Sonthalia, P.; McGaw, E.; Show, Y.; Swain, G.M.; Anal. Chim. Acta. 2004,
522, 35.

May, P.W.; Philosophical Transactions of the Royal society of London,
Series A: Mathematical, Physical and Engineering Sciences, 2000, 358,
473.

Angus, J.C.; MRS Bull. 1998, 21,4

Angus, J.C.; Li, Z.; Sunkara, M.; Gat, R.; Anderson, A.B.; Mehandru, S.P.;
Gels, M.W.; Proc-Electrochem. Soc. 1991, 91-8, 125.

Bachmann, P.K.; EMIS Datareviews Series, 1994, 9, 354.
Bachmann, P.K.; EMIS Datareviews Series, 1994, 9, 349.
Bachmann, P.K.; EMIS Datareviews Series, 1994, 9, 364

Gruen, D. M.; MRS Bull. 1998, 23, 32.

Gruen, D. M.; Annu. Rev. Mater. Sci. 1999, 29,211.

Gruen, D.M.; Annual Review of Materials Science, 1999, 29, 211.

Spitsyn, B.V.; Aleksenko, A.E.; Sokorina, G.A.; Laptev, V.A.; New
Diamond Sci. Technol.; Proc. Int. Conf., 2"“ 1991, 909.

Looi, H.J.; Pang, L.Y.S.; Molly, A.B.; Jones, F.; Foord, J.S.; Jackman,
R.B.; Diam. Rel. Mater. 1998, 7, 550.

R. F. Mamin, T. Inushima, Phys. Rev. B 2001, 63, 033201.

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

Spear, K.E.; Synthetic Diamond Emerging CVD Science and Technology.
John Wiley & Sons, Inc. New York, 1994.

Granger, M.C.; Swain, G.M.; J. Electrochem. 800.1999, 146, 4551

27

CHAPTER 2

Experimental Procedures and Instrumentation

2.1. Diamond Thin-Film Deposition

Boron-doped diamond thin films were deposited using a commercial
chemical vapor deposition (CVD) system (1.5 kW ASTEX Corp., Lowell, MA).
Microwave energy originating from a magnetron was directed into a quartz
reaction chamber to generate a plasma above the sample stage. Deposition
occurred on the surface of the substrate, which was placed on a molybdenum
stage positioned in the middle of the reaction chamber. The substrate was in
contact with the lower region of the plasma discharge. Figure 2.1 shows a
schematic of the CVD reactor. Details of the deposition and substrate preparation

are presented in the following paragraphs.

2.1.1. Planar Film Electrodes
2.1.1.1. Microcrystalline Diamond Thin-Film Electrodes

Microcrystalline diamond thin film was deposited on p-Si (100) (Virginia
Semiconductor Inc., Fredricksburg, VA) substrate. The substrate (ca. 1 x 1 cm2)

was prepared for deposition by, first, rinsing with ultrapure water, methanol, and

28

 

Microwave
Generator

 

 

 

 

  

Mass Flow , I

 

 

  
   

 

 

 

 

 

 

 

 

 

 

 

 

Controllers —Antenna
M.F.c.-[><|—
M'F‘C' Plasma
M.F.C. .. . .
Substrate
M.l=.c.-j><j- ~

 

 

 

 

 

Substrate
Holder

 

 

 

 

 

 

 

 

 

 

 

  

Gas Inlet .
Pressure

Gauge
Exhaust
Valve

 

 

 

 

H2 cylinder >'I><}

CH4 cylinder

 

 

Rotary Pump bro vent

 

B2He cylindelD-IX}
Ar cylinder )IDC

 

 

 

 

 

 

 

 

 

Figure 2.1 Microwave-assisted CVD apparatus for diamond growth.

acetone. It was then polished with 0.1 um diameter diamond powder (GE
Superabrasives, Worthington, OH) for 5 min on a felt pad. The scratched
substrate was then ultrasonically cleaned in acetone for several minutes until the
polishing debris was removed. The clean scratches, as well as residual diamond

particles, serve as the initial nucleation sites for film growth.

The microcrystalline film was deposited from a 0.5% CH4/H2 (C/H) source

gas mixture at a total gas flow of 200 sccm, 1 kW of microwave power, a system

29

pressure of 45 torr, an estimated substrate temperature of 800°C, and a growth
time ca. 10 h. Ultrahigh purity (99.999%) CH4 and H2 were used as the source
gases for the growth. Boron doping was accomplished using either a solid state
boron diffusion source (BoronPlus GS 126, Techneglas, Inc. Perrysburg, OH) or
by adding B2H6 diluted in H2 to the source gas mixture. The diffusion source was
placed underneath the substrate with B203 diffusing from the disc and serving as
the source for the dopant atoms. In the case of gas phase doping, B2H6
concentration of 10 ppm was used. After deposition, the CH4 flow was stopped
and the film remained exposed to the H2 plasma for an additional 15 min at the
growth conditions. After this annealing period, the plasma power and system
pressure were gradually reduced over a 15 - 30 min period, to 400 W and 20 torr,
repeatedly in order to cool the samples (<400 °C) in the presence of atomic
hydrogen. Post growth annealing in atomic hydrogen is critical for etching
adventitious non—diamond phases, minimizing dangling bonds, and ensuring full
hydrogen termination. The plasma was then extinguished and the films further
cooled to room temperature under a flow of hydrogen. Table 2.1 lists the typical
deposition conditions for microcrystalline diamond thin film growth. The film
thickness was nominally 5 um and the boron dopant concentration was in the low
to mid 1020 cm“3 range. The doping level was determined from boron nuclear
reaction analysis measurements made at the surface characterization facility at
Case Western Reserve University. The film resistivity was ~0.01 Q—cm, or less,

as measured with a tungsten four-point probe. Therefore, the electrical

30

conductivity of the diamond is comparable to that along the C-axis of the HOPG

(between the large planes).

Table 2.1 Microcrystalline Diamond Thin-Film Deposition Parameters

 

 

Deposition Parameters
Substrate Si
CH4/H2 (v/v) ratio 0.5 %
Plasma fonlvard power 1 kW
Deposition pressure 45 torr
Deposition time 10 h
Substrate temperature 800 °C
Boron dopant source B203 or B2H6
Doping level ~1x1029 B/cm3
Resistivity ~ 0.01 Q-cm or less

 

 

 

2.1.1.2. Nanocrystalline Diamond Thin-Film Electrodes

The boron-doped nanocrystalline diamond thin film was deposited on a p-
Si (100) substrate (~ 10'3 Q-cm, Virginia SemiConductor lnc., Fredricksburg, VA).
The substrate was prepared for diamond growth by, first, mechanically scratching
the surface on a felt polishing pad for 5 min with 0.1 pm diameter diamond
powder (GE Superabrasives, Worthington, OH). The scratched substrate was
then washed with ultrapure water, isopropyl alcohol (IPA), acetone, IPA, and

ultrapure water before being ultrasonically cleaned in acetone to remove the

31

polishing debris. Ultrahigh purity, CH4, Ar, and H2 (99.999%) were used as the

source gases.

The film was grown using flow rates of 1, 94, and 5 sccm, respectively, for
CH4, Ar, and H2. The microwave power and system pressure were maintained at
800 W and 140 Torr, respectively. Table 2.2 lists the typical deposition
conditions. B2H5 was used for doping and was added to the gas source mixture
at a concentration of 10 ppm. The substrate temperature was estimated, by
optical pyrometry, to be about 800°C. The deposition time was 2 h giving

approximately a 4 um-thick film, as estimated from sample weight change.

All gases flowed into the reactor simultaneously to initiate the diamond
deposition. At the end of the deposition period, the CH4 flow was stopped and the
Ar and H2 flows continued. The film remained exposed to an H2/Ar plasma for
approximately 10 min. The substrate was then cooled in presence of atomic
hydrogen to an estimated temperature of less than 400°C by slowly reducing the
plasma power and pressure over a period of 4 min. This leads to the formation of
a thin and continuous nanocrystalline diamond thin film in a relatively short period

of time.

32

Table 2.2 Nanocrystalline Diamond Thin-Film Deposition Parameters

 

 

Deposition Parameters
Substrate p-Si (100)
CH4/Ar/H2 (v/v) ratio 1/94/5
Plasma forward power 0.8 kW
Deposition pressure 140 torr
Deposition time 2 h
Substrate temperature 800 °C
Boron dopant source B2H6
Doping level 10 ppm
Resistivity ~ 0.01 O-cm or less

 

 

 

2.1.2. Boron-Doped Diamond Thin-Film Microelectrodes

Boron-doped diamond thin film was deposited on sharpened, 76 pm
diameter, Pt wires (St. Louis, MO, 99.99% pure).1 Prior to deposition, the
sharpened wire was prepared by, first, ultrasonic cleaning in acetone for 10 min.
The wire was then ultrasonically “seeded” for 30 min in a diamond powder
suspension (5 nm particles, ca. 20 mg in 100 mL of ethanol, Tomai Diamond Co.,
Tokyo, Japan). During the seeding process, the metal surface is scratched by the
diamond particles with some getting embedded. Both the scratches and the

embedded particles likely serve as the initial nucleation sites for film growth.

33

The pretreated Pt wire was mounted horizontally in the reactor on top of a
quartz plate (10 x 10 x 1 mm). Four or five wires were coated during a deposition
run. The quartz plate was placed in the center of the molybdenum substrate
stage and served to thermally isolate the wires. Boron-doped diamond thin film
was deposited from a 0.5 % CH4/H2 source gas mixture with doping
accomplished by adding 10 ppm of diborane, B2H6, diluted in H2. All source
gases were ultrahigh purity (99.999 %). For growth, the system pressure was 45
torr, the substrate temperature ~ 600 °C (as estimated with an optical pyrometer),
the microwave power was 400 W, and the total gas flow was 200 sccm. The
deposition time was 10 h. The nominal growth rate under these conditions was

estimated to be about 0.3 - 0.5 um/h, leading to a final film thickness of 3-5 pm.

The Pt wire was sharpened electrochemically using 1 M KOH prepared
with ultrapure water (Bamstead E-Pure system). The etching apparatus
consisted of a small Teflon vessel equipped with four equidistantly-spaced
carbon rod auxiliary electrodes electrically connected to one another. A 1.3-cm
long piece of 76 um diameter Pt wire was attached to a holder fabricated using a
small metal clip soldered to a copper wire, immersed into the etching solution to
a depth of 1 to 2 mm in the center of the four carbon rod electrodes. An AC
polarization of 12 V (60 Hz) was applied using a variable autotransforrner (Staco
Energy Products, Dayton, OH,). The sharpening procedure was considered

complete after approximately 25 s; a point at which visible gas evolution ceased.

34

After deposition, the diamond-coated Pt wire was attached to a copper
wire (~7 cm in length, 0.2 mm in diameter, Goodfellow Cambridge Ltd.,
Huntingdon, England) using conductive silver epoxy (CW2400 Chemtronics,
Kennesaw, GA), which was allowed to cure overnight at room temperature. The
microelectrode was then sealed in a polypropylene pipet tip (Ultramicro pipet tip,
Daigger, Vernon Hills, IL). After inserting the diamond-coated Pt wire into a tip, a
heat gun was used to first melt the polypropylene at the top of the microelectrode
assembly (securing the copper wire). The tip and was then carefully heated,
which softened the polypropylene and formed a conformal coating around the
rough microelectrode. The resulting microelectrode had both a conical and

cylindrical geometry, as shown in Figure 2.2.

insulated in pipet tip

exposed area copper wire

A
r \

.—

Figure 2.2 Schematic of a boron-doped diamond microelectrode sealed in a
polypropylene pipet tip.

2.2. Glassy Carbon Electrode Preparation

The glassy carbon (GC-30, Tokai, Japan) was polished with successively
smaller sizes of alumina powder (Buehler, Lake Bluff, IL) slurried in ultrapure
water (~18 MQ-cm, Barnstead E-Pure) on a felt polishing cloth (Buehler). Grit

sizes of 1.0, 0.3 and 0.05 pm were used. The electrode was sonicated in

35

ultrapure water for 10 min after each polishing step to remove polishing debris.
At the end of the polishing sequence, the electrode was thoroughly rinsed with
water and allowed to soak in distilled IPA for 10 - 20 min prior to use. An
electrode was used in electrochemical measurements immediately after

completion of polishing steps.

2.3. Characterization of the Diamond Thin-Film Electrodes

2.3.1. Scanning Electron Microscopy (SEM)

SEM was performed with JOEL-6400 V electron microscope. Images were
collected at a 2048 X 1536 pixel resolution with an ADDA ll digital image
acquisition system and AnalySlS Pro 3.2 image acquisition software (Soft

Imaging system GmbH, Munster, Germany).

2.3.2. Atomic Force Microscopy (AFM)

AFM was performed with a NanoScope Ill instrument (Veeco Metrology
Group. Inc., Santa Barbara, CA) operated in the contact mode. Pyramidal Si3N4
probe tips, mounted on Au-coated cantilevers (100 pm legs, 0.38 N/m spring

constant), were used in air to acquire topographical images of the films.

2.3.3. Raman Spectroscopy
Raman spectra were obtained at room temperature. The instrument
consisted of a diode-pumped, frequency-doubled CW Nd: yttrium aluminum

garnet (YAG) excitation laser of 523 nm and objective lens of x 20 (Raman 2000,

36

Chromex, lnc., Albuquerque, NM). Spectra were collected with an incident power
density of ~ 500 kW/cm2 (100 mW at the sample and 5 pm diameter spot size), a
10 5 integration time and spectra resolution of about 2 cm". A white light
spectrum collected under the same conditions was used for background

correction.

2.3.4. X-ray Diffraction (XRD)

XRD analysis was using a Rigaku Rotaflex diffractometer equipped with a
rotating anode and a Cu K-alpha radiation source. In the measurements, the thin
film rotates by an angle, 9, with respect to a fixed-positioned x-ray source, and
the detector rotates by 20 to measure the diffracted radiation. Diffraction profiles
were collected using incident angles of 30° to 100° with respect to the sample

surface.

2.3.5. Electrical Resistivity

Electrical resistance measurements were made using a tungsten four-
point probe having a 1 mm probe spacing. Currents (uA) were applied between
the outer two electrodes and the resulting voltage drop between the inner
electrodes was measured. Five measurements were taken from different regions

on each film.

37

2.3.6. Contact Angle Measurements

An automated contact angle apparatus (First Ten Angstroms) was used to
deliver a 10 uL drop of deionized water from a 50-mL syringe onto 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. The angle between a
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 ranging 69.2 :I:

5.5°, 64.0 :I: 3.0°, 72.8 :I: 2.3°, and 75.3 :I: 5.1°.

2.4. Electrochemical Measurements

Cyclic voltammetric measurements were made with a CYSY- 2000
computerized potentiostat (Cypress System Inc. Lawrence, KS) using a single
compartment glass cell. A diagram of the cell is shown in Figure 2.3. The
diamond film or glassy carbon electrode was pressed against a viton o-ring and
clamped to the bottom of the glass cell. Ohmic contact was made by pressing a
copper plate against the backside of the scratched and cleaned Si substrate,
which contained a bead of Ga/ln alloy or silver paste. A graphite rod was used as

the counter electrode and a commercial Ag/AgCl electrode (4 M KCI) served as

38

the reference. The geometric area of the working electrode was ca. 0.2 cm2. All
measurements were made at room temperature. All solutions were
deoxygenated with nitrogen for 5 min prior to any measurement and remained

blanketed with the gas during a measurement.

    
 
    

 

   

 

   

_ _Auxiliary
electrode
Reference _ _ _
electrode ‘ " ’ N2 purge
Clamp
Working --- g! ' -— O-ring
electrode -- Cu Foil

 

Figure 2.3 Design of the three-electrode, single compartment electrochemical
cell.

Electrochemical measurements using the diamond microelectrode were
performed in a 10 mL glass vial with a cover that had openings for the auxiliary,
reference, and working electrodes. A three-electrode system consisting of a
diamond microelectrode, a Ag/AgCl reference electrode, and a carbon rod

auxiliary electrode was connected to the potentiostat. The cell was housed in an

39

electrically-grounded Faraday cage to reduce the ambient electrical noise. All
measurements were performed at room temperature in solutions deoxygenated

with nitrogen for 5 min.

2.5. Flow Injection Analysis

Flow injection analysis (FIA) studies were performed using a home-made
thin-layer electrochemical flow cell with a cross-flow geometry. The experimental
set-up has been described elsewhere.2 A block diagram of the FIA system and a

schematic of the thin-layer flow cell are shown in Figures 2.4 and 2.5.

 

 

 

 

 

 

 

Figure 2.4 A schematic of the FIA system.

40

reference electrode

auxiliary / inlet
electrode

rubber gasket

working
electrode

   
 
  
  
  

Kel-f body, top
piece

current collector

Kel-f body, bottom
piece

screw clamp

Figure 2.5 Design of the thin-layer flow cell.

The flow rate of the carrier solution was regulated with a consta-Metric lll

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

0.6 mUmin, was constantly sparged with nitrogen.

with a Model LP-21 pulse dampener (Scientific Systems) placed between the
pump outlet and the injector valve. 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-

41

Pump noise was reduced

FTM. 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 rubber
(neoprene) gasket separated the surface of the working electrode from the top
piece of the cell. A rectangular groove (7 mm x 1.5 mm x 1 mm) 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 ~8 uL. An Omni-90 analog potentiostat
(Cypress Systems, Inc. Lawrence, KS) was employed to control the potential
applied to the working electrode and to measure the current flowing during
detection. The potential was applied with respect to a commercial Ag/AgCI mini-

electrode (3 M NaCl, Cypress Systems, Inc. Lawrence, KS).

2.6. High Performance Liquid Chromatography

Reversed-phase high performance liquid chromatography (HPLC) was
performed using a commercial system (LC-10A, Shimadzu Corp., Columbia,
MD), which consisted of a controller (SCL-10AVP), a solvent delivery module
(LC-10ADVP), a degasser (DGU-14A), and an auto-injector (SlL-10ADVP). The
same thin-layer flow cell described above was used for electrochemical
detection. Elution was carried out by changing the 50 mM phosphate buffer (pH

3.5)lacetonitrile mobile phase composition from 65:35 to 20:80 (v/v). The flow

42

rate was maintained at 0.6 ml/min. The mobile phase was constantly sparged
with nitrogen (99%). The separation of phenol and the chlorophenols was
carried out using a C18 reversed-phase column (X-Terra, 5 um particle size, 3.9 x
150 mm, Waters). The stationary phase was 3-(chlorodimethylsilyl)propyl-N-
dodecyl carbamate. This column has a modified silica support in which some of
the silicon atoms are replaced with carbon. This substitution reduces the
supports solubility at high pH, yet leaves sufficient silanol activity to enable the
use of traditional bonding chemistry to covalently attach the stationary phase. An
Omni-90 analog potentiostat (Cypress Systems, Inc. Lawrence, KS) controlled
the potential applied to the working electrode and measured the current flowing
during detection. Data acquisition and processing were performed with Class VP

7.0 chromatography software package (Shimadzu).

2.7. Capillary Electrophoresis

Capillary electrophoresis was performed using a home-made system
consisting of CZE1000R 30 kV variable power supply (Spellman, Hauppauge,
NY), a home made electronic timer for the electrokinetic injection, and a plexiglas
box with a safety interlock that housed the 76-cm long, 30 pm i.d. fused-silica
capillary (Polymicro Technologies, Phoenix, AZ), buffer reservoir, sample vessel,

and detection cell. Figure 2.6 show a block diagram of the system.

43

The separations were made by applying the desired positive voltage
between the run buffer reservoir and the electrically-grounded detection

reservoir. Electrokinetic injection was used to introduce the sample onto the

 

| Potentlostat Q
Plexiglas box

Separation

Power supply“ Timer relay “pm”

 

 

 

   
 

 

 

 

 

 

 

 

Figure 2.6 Capillary electrophoresis set up.

column by applying 8 - 10 W for 3s between the sample reservoir and the
electrically-grounded detection reservoir. The plexiglas box was covered with
aluminum foil and was electrically grounded in order to reduce the pick up of

electrical noise.

The end-column detection cell was fabricated from glass and consisted of
(i) four Teflon rods with Kel-F bolts for adjusting the capillary position in the x and
y directions, (ii) two Pt wires with one serving as a ground for the high voltage

and the other as an auxiliary electrode for the electrochemical cell, (iii) a bolt for

adjusting the separation capillary-electrode distance in the z direction, and (iv)
two openings for mounting the working and the reference electrodes.‘ A three-
electrode system was employed. Figure 2.7 shows a diagram of the detection

cell.

separation capillary

  

auxiliary electrode reference electrode

working electrode

Figure 2.7 Design of the end-column electrochemical detection cell.1

All the electrodes were connected to a digital potentiostat (El-400
Bipotentiostat, Cypress Systems, Inc., Lawrence, KS) and the data were
acquired digitally using Powerlab Chart 5 (ADlnstruments, Colorado Springs,
CO). Each new capillary was activated by sequentially flowing (pressure driven)
a solution of 0.5 M NaOH, deionized water, and the run buffer for 20 min each.
The capillary was then equilibrated in the run buffer under an electric field of 263

Vlcm (20 W) for approximately 20 min prior to sample injection.

45

2.7. Solid Phase Extraction

Local river water (Red Cedar River, MSU Campus) was used as the
sample from which to extract the chlorinated phenols. The river water was
collected in a 500 mL Nalgene bottle. The bottle was flushed three times with the
water before collecting a sample. The sample was stored in a capped bottle and
transported back to the laboratory. The river water was first filtered through a
nylon membrane with a 0.4 pm pore size (Sigma-Aldrich, St. Louis, MO). After
filtration, the pH of the water was adjusted to 2.5 using concentrated HCI and
then spiked with chlorinated phenols. An Alltech Vacuum manifold was used for
the SPE procedure. It consisted of a glass chamber, a vacuum Gauge and valve,
manifold lid and stop cocks. The ENVl-Chrom P SPE tubes (80 - 160 um particle
size, 800 — 950 m2/g surface area, 110 — 175 A mean pore size, 3 mL tube
volume and 0.25 g sorbent loading) were used for the extraction (Supelco,
Bellefonte, PA). Prior to extraction, the sorbent were conditioned sequentially by
passing 3 mL of ethyl acetate, 3 mL of methanol and 3 mL of ultrapure water
adjusted to pH 2.5 using concentrated HCI. Solutions were passed through the
sorbent by applying a vacuum (vacuum assisted). The filtered river water sample
was passed through the moist sorbent at a low flow rate of < 5 mL/min. After
sample passage, the solid phase was dried under vacuum for 5 min prior to
analyte elution. Elution was performed by adding 2 mL of ethyl acetate and
allowing it to interact with the solid phase for 5 min before elution. The elution
was then performed at a flow rate of about 2 mL/min. The extract had two

phases, aqueous and non-aqueous, which were separated out. Chlorinated

46

phenols were found in the non-aqueous phase and none in the other phase. The
volume of the non-aqueous phase was adjusted to 1 mL using ethyl acetate. If
the elution volume was greater than 1 mL, the volume was reduced to less than 1

mL by passing a stream of N2, and then readjusted to 1 mL using ethyl acetate.

Figure 2.8 summarizes the steps involved in the extraction.

1

1
I

.1- .l i

 

 

 

 

 

 

 

 

 

1. Sample preparation - filtration
2. Column conditioning
3. Sample application
4. Sample elution

O. O i

O fl

0 0
LL

 

Figure 2.8 SPE extraction steps.

2.8. Materials and Chemicals
Reagent-grade quality chemicals, or better, and ultrapure distilled and
deionized water (~18 MQ-cm, Barnstead E-Pure) were used to prepare all
solutions. Potassium phosphate monobasic (Mallinckrodt, Germany) and

acetonitrile (EM Science, Germany) were used to prepare the mobile phase for

the HPLC measurements. Sodium phosphate monobasic, sodium tetraborate

47

(Na2B4Oy.10H2O), sodium phosphate dibasic, and sodium hydroxide (Spectrum
Chemical, Gardena, CA) were used to prepare run buffer for capillary
electrophoresis measurements. The run buffer was filtered through a nylon
membrane with a 0.4 pm pore size (Sigma-Aldrich, St. Louis, MO) before use.
Ethyl acetate (Mallinckrodt, Germany) and methanol (Sigma-Aldrich, St. Louis,

M0) were used in the SPE procedure.

The analytes of interest, phenol, 2-chlorophenol, 3-chlorophenol, 4-
chlorophenol, 2,4-dichlorophenol, 2,4,6-trichlorophenol, and pentachlorophenol
(Sigma-Aldrich, St. Louis, M0) were used without any additional purification.
Table 2.3 shows the molecular structure, the pKa and the formula weight for each
analyte. The diamond film electrochemical response was characterized by cyclic
voltammetry using the following inorganic and organic redox systems: potassium
ferrocynide (Fe(CN)6'3’4), hexaamine ruthenium (Ill) chloride (Ru(NH3)*3’*2),
iridium (IV) hexachloride (erle'M), methyl viologen dichloride hydrate (Mv*2’*"°),
4-tert-butylcatechol, and ferric sulfate (Fe‘z’fi’) and ferrocene carboxylic acid
(Sigma-Aldrich, St. Louis, MO). All of these chemicals were reagent-grade
quality, or better, and used without additional purification. The supporting
electrolyte for Fe(CN),~;3’4, Ru(NH3)*2’*3, erl6'2"3, Mv*2’*"° was 1 M KCI. The
supporting electrolyte for 4-tert-butylcatechol and l=e*2"‘3 was 0.1 M HCIO4. The
potassium chloride (Fisher) and perchloric acid (Sigma-Aldrich, St. Louis, M0)

were ultrahigh purity (99.999%).

48

Table 2.3 Properties of Phenol and the Chlorinated Phenols

 

 

 

 

 

 

 

 

 

Compound Molecular structure pK. Formula
values Mass
(glmol)
OH

Phenol i! 10.00 94
2—chlorophenol 8.52 128
3-chlorophenol 8.97 128
4-chlorophenol 9.37 128
2,4-dichlorophenol 7.90 163
2,4,6-trichlorophenol 6.10 197
Pentachlorophenol 4.74 266

 

 

 

 

 

49

 

2.10. References

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

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

50

CHAPTER 3

The Physical and Electrochemical
Characterization of Boron-Doped Diamond Thin-

Film Electrodes

3.1. Introduction

Boron-doped diamond thin film can be grown with two different
morphologies and microstructures: microcrystalline and nanocrystalline. The
distinctions between the two are the surface roughness and the nominal grain
size. For example, microcrystalline films have a nominal grain size of ~ 2 pm, or

greater, and nanocrystalline films have a grain size on the order of 15 nm. 1 °"°

references therein

It is important to have a good knowledge of the physical and chemical
properties of the film if one is going to understand the electrochemical behavior.
Therefore, detailed structural and electrochemical characterization of the films
was carried out. The motivation for studying these two types of diamond stems

from the fact that the different morphology and microstructure of each could

51

influence the electrode reaction kinetics, particularly for the electro-oxidation of

chlorinated phenols.

Table 3.1 lists the instrumental techniques used to characterize the
physical, chemical and electronic properties of diamond thin film along with the
information gained from each.2'3 Results from the characterization by atomic
force microscopy, visible Raman spectroscopy, XRD and cyclic voltammetry
using Fe(CN)6’3"‘, Ru(NH3)*2’*3, erls'm, Mv*2’*"°, 4-tert-butylcatechol, l=e“"”*3 as

redox test probes, are presented in this Chapter.

Table 3.1 Analytical Techniques Used for Film Characterization

 

 

 

Technique Information
Scanning Electron Microscopy Morphology
Atomic Force Microscopy Morphology
Raman Spectroscopy Microstructure
X-ray Diffraction Crystallinity
4-pt Probe Electrical Measurements Electrical Resistivity
Contact Angle Measurements Hydrophobicity

 

 

52

 

3.2. Results and Discussion
3.2.1. Morphology and Microstructure

Atomic force microscopy (AFM) was used to probe the morphology of the
diamond thin film. Figure 3.1 shows typical images of a microcrystalline and

nanocrystalline thin film.

   

Figure 3.1 AFM images showing a (A) microcrystalline and (B) nanocrystalline
diamond thin film deposited on Si.

Microcrystalline films are polycrystalline and well-faceted with a
predominance of octahedral and cubo—octahedral crystallites. The nominal
crystallite size (lateral dimension at the base) is 2 pm or greater."4 and ”mm“
"‘W" The nanocrystalline film is also polycrystalline but with a smoother surface,
smaller feature sizes, and a greater fraction of grain boundary, as compared to
the microcrystalline film. The nanocrystalline film consists of nodular features ~50

— 100 nm in diameter. The nodular features are actually clusters of individual

53

diamond crystallites that are 10 — 15 nm in diameter, formed as a result of the

high renucleation rate.5

The different morphologies result from the growth conditions employed. As
mentioned in Chapter 2, the conventional microcrystalline film is grown using 0.5
— 1.5 % CH4/H2 source gas mixtures (i.e., hydrogen rich). It is known that, under
such growth conditions, molecular H2 must be dissociated into atomic hydrogen
in order to accomplish a sequence of hydrogen abstraction reactions on the
surface and in the gas phase. Methyl radicals (CH3') and or (C2H2) are the
primary growth species in hydrogen rich environments. Heterogenous nucleation
of the diamond phase must compete with regasification of the embryonic nuclei
by atomic hydrogen.°8 The end result is the survival of crystallites that become
ever larger as a result of grain coarsening. The surface roughness of such films
tends to track the grain size and film thicknessf’8 A nanocrystalline film can be
grown from CH4/Ar source gas mixture with little or no added H2. The
nanocrystalline films, on the other hand, result from a growth and nucleation
mechanism involving the insertion of the primary growth species, carbon dimer-
C2, into surface C-H bonds. The C2 addition is believed to occur by a two-step
growth mechanism.9 A C2 molecule approaches the unreconstructed
monohydride surface and inserts into a C-H bond. The C2 then rotates to insert
its other carbon into a neighboring C-H on the surface. A C2 molecule then
inserts into an adjacent C-H bond, parallel to the newly inserted C2 dimer. The

original state of the surface is recovered by the formation of a bond between

54

carbon atoms in the adjacent dimers. Very high renucleation rates are observed,
on the order of 101°cm'2, and the resulting film is morphologically smooth.° The
smooth nanocrystalline films possess interesting mechanical, tribological, and
electrical properties owing to their grain size. For example, the film transitions
from an electrically insulating to a conducting material with a reduction in

'10

crystallite size from the micrometer to the nanometer leve This is largely due

to the presence of high-energy. high-angle grain boundaries that contain 11:-
bonded carbon atoms. The grain boundaries are conducting because of the It-
states, and since their number vastly increases with decreasing crystallite size,

the entire film becomes electrically conducting and functions as an electrode

material.11

Figure 3.2 shows visible-Raman spectra that reflect the microstructures for
the microcrystalline and nanocrystalline films. Raman spectroscopy has been
used extensively to characterize carbonaceous materials, including diamond.12'15
Raman spectroscopy can be used to discriminate between sp2 and sp3-bonded
carbon as well as to probe the local bonding environment of the carbon. A
spectrum with a single sharp peak at 1332 crn‘1 is seen for microcrystalline
diamond. This is the first-order diamond phonon mode that originates from the
spa-bonded carbon microstructure. An intense and narrow peak is characteristic
of high quality diamond.1°'17 The spectrum shows negligible scattering in the
region between 1500 and 1600 cm", where intensity is normally observed when

significant amounts of spz-bonded carbon are present in several dispersed forms

55

within the CVD-grown polycrystalline diamond films. These include point defects
within the diamond lattice and extended defects, such as twin planes and

stacking faults at grain boundaries.18

 

 

15,0 16.0
. 1470 cm-1
1332 cm-1 I B 1333 \m" 1 1550 cm-1

/ 12.0

8.0

in:
E"
0

it 7 I >

Intensity (a.u.)

 

 

 

 

 

 

0.9 1 1 l 1 l 1 1 1 1 1 l 1 0.0 t 1 1 1 1 L1 1 1 1 1 1
00 900 1100 1300 1500 1700 1900 700 900 1100 1300 1500 1700 1900

 

 

Raman Shift cm'1

Figure 3.2 Visible Raman spectra for a (A) microcrystalline and (B)
nanocrystalline diamond thin film.

The spectrum for nanocrystalline diamond spectrum is quite different.
Broad peaks are seen at 1150, 1333, 1470, and 1550 cm". The peak at 1333
cm", sitting atop a large background signal, is probably associated with the first-
order phonon mode. Due to resonance effects, the Raman cross-section
scattering coefficient for spz-bonded carbon (e.g., graphite) is 50 times larger
than that for spa-bonded diamond.13 The peak width (fvvhm) for diamond line is
much broader than that for the microcrystalline diamond. This is because of the
small grain size.5 There are two possible explanations for the line broadening.

The first is the surface confinement model which states that the smaller the

56

domain size, the larger the range of phonon modes that are allowed to participate
in the Raman process. The broad Iinewidth is due to the spread in phonon
energy.19 Another explanation is phonon scattering by impurities and defects
(i.e., grain boundaries) as, to a first approximation, the phonon lifetime is
inversely related to the defect density. The high grain boundary density reduces
the phonon lifetime resulting in the increased Iinewidth.19 The peak at 1150 cm”1
is often used as 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, specifically transpolyacetylene segments at the grain
boundaries.20 Their assignment of the 1150 crn'1 peak to spz-rather than spa-
bonded carbon, as is often proposed, is based on the observations that the peak
position changes with excitation energy, the peak intensity decreases with
increasing excitation energy, and the peak is always accompanied by another
peak at ~ 1450 cm“, which behaves similarly with excitation energy. The 1550
cm’1 peak is assigned to disordered sp2 bonded carbon primarily 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 spa-
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.21

XRD was employed to characterize the bulk crystallinity of the diamond

thin films. Figure 3.3 shows the XRD spectra for the (A) microcrystalline and (B)

57

nanocrystalline. 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.5 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)

(220)
(311)

 

 

mount. Units)
I
I

 

 

..JL ”.1 It

30 40 50 60 70 80 90 100
2-Theta (degrees)

 

 

 

Figure 3.3 XRD patterns for a (A) microcrystalline and (B) nanocrystalline

diamond thin film.

3.2.2. Electrochemical Responsiveness

Figure 3.4 shows a background voltammetric i-E curve in 0.1 M HCIO4 for
a microcrystalline diamond film deposited on Si. The nanocrystalline diamond
film exhibits a similar curve shape. Background voltammetric curves are useful

for assessing diamond film quality.4 The magnitude of the background current

58

and working potential window are highly sensitive to the presence of non-
diamond spz-bonded carbon impurity.22 This impurity, when present at significant
levels, is located at the grain boundaries and defects. The i-E curve is flat and
featureless with no peaks present in the 0 — 0.5 V range that are indicative of
redox-active surface carbon—oxygen functionalities.23 The background current is
low over the entire potential range between the solvent electrolysis limits. For
example, the anodic current at 0.5 V is 1.5 uA (8 uA/cmz), whereas for glassy
carbon the current is variable from 7 -10 uA (35 - 50 uA/cmz). The current at
positive potentials begins to increase at 1.8 V due to the onset of oxygen
evolution. The current at negative potentials begins to increase at -1.5 V due to
hydrogen evolution. The scans were unchanged with multiple cycles reflecting
the stability of the surface morphology and microstructure. The working potential
window, defined as the potential range between current limits of :I: 50 uA, is ~ 3.2
V. The wide working potential window permits the electrochemistry of redox
systems with large positive or negative standard reduction potentials to be
studied, something that would otherwise be impossible due to interference from

solvent electrolysis.

The background current for both microcrystalline and nanocrystalline is
less than that of a comparably sized glassy carbon electrode (a factor of 5 - 7
higher). A low background current and wide working potential window (3 - 4 V)

are characteristic features of high quality boron-doped diamond electrodes.24

59

This leads to improved signal-to-background (SBR) and signal-to-noise (SNR)

ratios in electroanalytical measurements.25

 

N A at
G G 6

(M)

G

 

Current
IL)
G

.L
c

-60'

 

 

 

-80 . . . . . . .
-1500 -1000 -500 0 500 1000 1500 2000
Potential (mV vs. Ag/AgCl)

Figure 3.4 Background cyclic voltammetric i—E curve in 0.1 M HCIO4 for a
microcrystalline diamond film deposited on Si. Scan rate = 0.1 We
Electrochemical geometric area = 0.2 cm2.

The electrochemical response of both boron-doped microcrystalline and
nanocrystalline diamond thin-film electrodes towards different redox systems was
also investigated using cyclic voltammetry. The physical, chemical and electronic
properties of the diamond influence the electrode kinetics and reaction
mechanism for these systems to differing extents.5'25'26 Ru(NH3)6*2’*3. methyl

viologen (MV'2’*"°), |rC|6‘2"3, Fetm, Fe(CN)6°’4, and tert-butylcatechol were

60

used to probe the reactivity of the electrodes. These redox systems were chosen,

in part, based on the fact that their E°' values span a wide potential range.

Redox reactions are generally of two types. One type is reactions that
proceed by simple diffusion of the analyte to the electrode-solution interface, with
the electrode serving solely as a source or sink for electrons. In this case, the
electrode-reactions are relatively insensitive to factors such as surface chemistry
and microstructure, but are very sensitive to the density of electronic states near
the formal potential (so-called outer-sphere reaction). The other type includes
reactions that occur via some specific interaction with the electrode surface; for
example, reactions that proceed through an adsorbed state. In this case, the
electrode reaction kinetics are sensitive to surface chemistry and microstructure
as well as the density of electronic states near the formal potential (so called
inner-sphere reaction). The redox systems Ru(NH:4)r;“2”'3 , erls'Z”3 , and methyl

viologen (MV‘Z’Hm

) are all outer-sphere in nature with the electrode kinetics
being relatively insensitive to the physicochemical properties of diamond.
Apparent heterogeneous electron-transfer rate constants, k°app, between 0.01 and
0.1 cm/s are commonly observed for both microcrystalline and nanocrystalline
films without any conventional pretreatment.“'27'32 Normally, carbon and metal
electrodes must be pretreated to achieve such a large rate constant. Fe(CN)6’°"‘
proceeds through more of an inner-sphere electron-transfer pathway, as
electrode kinetics are highly sensitive to the surface chemistry of diamond.31

Presumably, the reaction proceeds through a specific surface interaction

61

 

available on the hydrogen-terminated surface. Such surface interactions seem to
be blocked on the oxygen-terminated surface. k°4.pp values ranging from 0.01 to
0.1 cmls are commonly observed for clean, hydrogen-terminated films, but the
rate constants decrease by over two orders of magnitude for oxygen-terminated
films.” For spz-bonded carbon electrodes, the electrode kinetics are strongly

influenced by the fraction of exposed, clean edge plane?4

Figure 3.5 and 3.6, respectively, shows cyclic voltammetric i-E curves on
boron-doped diamond microcrystalline and nanocrystalline in the presence of six
redox systems (A) Fe(CN)6‘3"‘, (B) Ru(Nl-l3)6*2’*3, (C) erIe'z"3, (0) methyl
viologen in 1 M KCI, and (E) Fe ‘2’”, (F) tert-butyl catechol in 0.1 M HCIO4. The
potential scan rate was 0.1 V/s. The voltammograms were obtained with
hydrogen-terminated electrodes. A summary of the voltammetric data is
presented in Table 3.2 and 3.3. The CV data presented are the oxidation
potential, Ep°", oxidation peak current, ip°", oxidation and reduction peak current
ratio, ip°"flp'°°, and the peak potential separation, AEp. The AEp (AEp = B,” —
Ep'°°) is of particular importance because it is related to the heterogenous

electron transfer rate constant.

The E°' values for these redox systems range from ~ +0.8 to -1.1 V, so
they are useful for probing the film's electronic properties over a wide potential
range. A nearly reversible response is observed for Fe(CN)(~,'3"4 with a AEp of 61

mV for both diamond film types. The i,4°"/ip'°d was 1.1 and 1.0 for microcrystalline

62

LL

and
sens
borc
sou;
bon
surl

inIIl

IIaI

SUI

65
An
als
nal
nar
ISdl
Sp .
mall

eacl

and nanocrystalline diamond thin film, respectively. This redox couple is quite
sensitive to the physical and chemical properties of both glassy carbon and
boron—doped diamond electrodes. 5'28'32'34 The electrode reaction kinetics for this
couple are strongly influenced by the amount of exposed edge plane on spz-
bonded carbon, as well as the surface cleanliness.34 It has been shown that
surface carbon-oxygen functionalities on microcrystalline diamond significantly
influence AEp, with increasing oxygen content causing an increase in the peak
potential separation.33 Such an effect was also observed by Fujishima and co-
workers.35 The effect of surface oxygen is reversed if the film is rehydrogenated
in a hydrogen plasma.25 The small AEp seen for both microcrystalline and
nanocrystalline boron-doped diamond electrodes is indicative of a high level of

surface cleanliness, low oxygen coverage and high electrical conductivity.

A nearly reversible response is seen for Ru(NH3)6"2’*3 with of AEp 63 and
65 mV, respectively, for microcrystalline and nanocrystalline diamond thin film.
An i,,°"/i,,'°d of 1.0 was observed for both films. A nearly reversible response is

also seen for erl6‘2"3

with a AEp of 63 mV on both microcrystalline and
nanocrystalline diamond. The ip°"/i,,red was 1.0 and 1.1 for microcrystalline and
nanocrystalline diamond, respectively. The electrode kinetics for both of these
redox systems are much less sensitive to the physicochemical properties of both
spa-bonded carbon and sp3-bonded diamond electrodes.28'32'34 The kinetics are

mainly influenced by the density of electronic states near the formal potential of

each redox system. The relatively low AEp (or relatively rapid kinetics) for both

63

A /_z T""""“"““""1—

   

 

 

 

 

 

 

 

 

 

 

60- 60 B
40_ A 40"
20_ 20»
0 ' 0 ‘—
.20
.201-
.40
40’
so-
‘6903 L 0.4 A ofo A 014 A ofa -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4
15 40
10.. C 20,. D
A 5
i o _. °' <—
v
‘E -5 .20.
g-10-
=_15_ .40.
U

 

 

 

 

 

 

0.4 A 0.5 A 0.6 A 0.7 A 0.8 A 0.9 A 1.0 A 1.1 -1.4 A -1.2 A -1.0 A -0.8 A -0.6 A -0.4 A -0.2

 

 

 

 

 

 

 

60 150
40" E 100- F
20-
50~
Ol-
—>
-20- 0’ r,
40- -50- \f
-60-
1 ’100 A

 

 

0.4 A 010 A 0.4 A 018 A 1:2 '04 A 0-0 A 0-4 A 0.8 A 1-2

Potential (v vs. AgIAgCI)

Figure 3.5 Cyclic voltammetric i-E curves for (A) Fe(CN)5'3"‘ (B) Ru(NH3)4*2’*3
(C) "013°"3 (D) methyl viologen in 1 M KCI, and (E) Fe‘z’“, (F) tert-butyl catechol
in 0.1 M HCIO4 at a boron-doped microcrystalline diamond thin-film electrode.
Scan rate = 0.1 Vls. Electrode geometric area = 0.2 cm2.

64

80-

40~

.40-

(11A) .,

Current

-20 1 1 1 1 1 1 1 1 . 1 .
(14 (15 (16 (17 (18 (19

60

4o-

20

 

 

 

1111111111111

 

 

 

_L
o
I

O
l

I
A
o

r

 

 

 

10

 

E

 

 

 

80

40-

~80

41.8

 

 

¢

 

 

A416

A-0.4

-O.2A 0.00.2 OA

 

40-

20-

40

 

[8F

 

 

-1.4A

-1.2A

-1.0 A

-0.8 A

-0.6 A -0.4 A -0.2

 

150-

100»

50-

0..

-50 ..

 

-100

F

 

 

414

0.0 A

Potential (V vs. AglAgCl)

Figure 3.6 Cyclic voltammetric i-E curves for (A) Fe(CN).~;3"4 (B) Ru(NH3),,*2’+3
(C) erl5'2"3 (0) methyl viologen in 1 M KCI, and (E) Few”, (F) tert-butyl catechol
in 0.1 M HCIO4 at a boron-doped nanocrystalline diamond thin-film electrode.

Scan rate = 0.1 Vls. Electrode geometric area = 0.2 cm2.

65

Table 3.2 Cyclic Voltammetric Data for a Microcrystalline Boron-Doped

Diamond Electrode

 

 

 

Analyte A5,, (mV) E,“ l,,°" var—i?—
1 mM Fe(CN)4'3"‘ in 1 M KCl 611: 3 314 60 1.1
1 mM Ru(NH;,)4"2’+3 in 1 M KCl 63 1 4 -140 63 1.0
0.25 mM erl,,'°"3 in 1 M KCl 63 1 4 807 15 1.0
0.5 mM wt”:1 in 1 M KCl 60 1 5 -601 32 1.2
1 mM t-butylcatechol in 0.1 M HCIO4 460 1 7 624 134 2.0
1 mM Few“ in 0.1 M HCIO4 574112 857 43 0.9

 

 

Scan rate = 100 mV/s. Electrode area = 0.2 cm2

Table 3.3 Cyclic Voltammetric Data for a Nanocrystalline Boron-Doped

Diamond Electrode

 

 

 

Analyte A5,, (mV) F.,ox l,°'* i971,“
1 mM Fe(CN)6'3” in 1 M KCl 61 1 2 311 73 1.0
1 mM Ru(l\lH3)4*2’+3 in 1 M KCl 65 1 4 -137 65 1.0
0.25 mM erl,~,'2"3 in 1 M KCl 63 1 3 809 14 1.1
0.5 mM Mvt2l+1 in 1 M KCl 60 1 4 -611 31 1.1
1 mM t-butylcatechol in 0.1 M HCIO4 483 1 9 635 134 2.1
1 mM ref”+3 in 0.1 M HCIO4 685 1 10 921 45 0.9

 

Scan rate = 100 mV/s. Electrode area = 0.2 cm2

66

 

microcrystalline and nanocrystalline diamond indicates that both diamond types
possess a high density of electronic states at both -0.2 V over a wide potential

range.3°'37

A reversible response is also seen for MV‘Z’“ with an AED of 60 mV for
both microcrystalline and nanocrystalline thin film. An i,,°"/ip'°d of 1.2 and 1.1 was
observed for microcrystalline and nanocrystalline, respectively. Like
Ru(NH3)(3"2”'3 and erl6'2"3, the electrode reaction kinetics for MV“2’+1 are
relatively insensitive to the physicochemical properties of both spz-bonded and
diamond electrodes.2°'3A‘"38 The kinetics are mainly influenced by the density of
electronic states at the formal potentials. MV undergoes two redox reactions
(MV‘M and MV*’°). Good responsiveness was also observed for the other redox
couple, MV‘1’°. The cathodic peak at ~-1.0 V is associated with the reduction of
MV+1 to MV°. The peak shape is consisted with diffusion-controlled reaction. The
data for these four redox systems indicate that both diamond film types possess
a high density of electronic states over a wide potential range.

The cyclic voltammetric i-E curves for tert-butylcatechol (t-BC) and Few+3
have much larger AEp values and more asymmetric peak shapes than do the
curves of the other redox couples. AEp values of 460 and 574 mV are observed
for t-BC and Fem”, respectively, for microcrystalline diamond. The large AEp
values reflect more electrochemical irreversibility (i.e., more sluggish electrode

reaction kinetics).""28 The i,,°"/ip'°d ratios are 2.0 and 0.9 for t-BC and Few”,

67

respectively, for microcrystalline diamond due to the reaction irreversibility. The
AEp values of 483 and 685 mV are for t-BC and Fem”, for nanocrystalline
diamond. The i.=,°"/ip'°d ratios are 2.1 and 0.90 for t-BC and Few”, respectively.
The catechol redox reaction involves a transfer of both electrons and protons,
and the electrode kinetics are highly sensitive to the surface microstructure, the
presence of surface carbon-oxygen functionalities, and the surface cleanliness of
spz-bonded carbon electrodes.34 Surface adsorption catalyzes the reaction rate.39
It is postulated that the slow kinetics for diamond result from a lack of adsorption
on the spa-bonded hydrogen-terminated diamond surface.40 The electrode

kinetics of Fe+2I+3

are strongly influenced by the presence of carbon-oxygen
functionalities, specifically carbonyl groups, which catalyze the reaction rate at
spz-bonded carbon electrodes.40 The hydrogen-terminated diamond surface is
void of such functionalities, so this is the postulated reason for the sluggish
kinetics. Increasing the surface oxygen content on diamond has been observed

to slightly decrease the AEp.‘°’5 More research is needed to fully understand why

the kinetics of this redox system are so sluggish on diamond.

Plots of ip°" versus scan rate"2 (vy‘) were constructed for Fe(CN)6‘3"‘,
Ru(NH3)5+2"’3, erl6'2"3, methyl viologen, Few”, tert-butyl catechol redox systems
at various scan rates (100 — 500 mV/s). Figure 3.7A presents plots of ip°" versus
v” for the different redox analytes at a boron-doped microcrystalline diamond
electrode. It was observed that the oxidation peak current for each redox system

varied with the scan rate” with a near- zero y-axis intercept, indicative of

68

 

reactions limited by semi-infinite linear diffusion of the reactant to the electrode
surface. A similar trend was observed on the nanocrystalline diamond. Figure
3.7B shows a comparison for the experimental plot and theoretical prediction for
the Fe(CN)3'°” for a microcrystalline diamond. The two curves are more or less
comparable. The quasi-reversible to reversible voltammetry for these couples
indicates that the boron-doped microcrystalline and nanocrystalline films have

sufficient density of electronic states over a wide potential range to support rapid

rates of electron-transfer.

 

 

 

 

 

 

 

 

350
A
300.
El t-BC

v O +ZI+3
E 200A ‘ RU(NH3)° 11:1 Experimental B
2 Few“ 120 _¢ Theoretical
I- 150 ~
: .XMV‘Z;
o 100‘ erl, 80 g

50- 40 ~

0 “i’— r T I I 0 I T— _'7
0 5 10 15 20 25 0 10 20
Scan Rate (mV/s)"2

Figure 3.7 (A) Plots of 1,0" vs. scan rate"2 for Fe(CN)6’3"4, Ru(NH3)4*3’*°, IrCl5'2"3,
methyl viologen, Few”, tert-butyl catechol at a boron-doped microcrystalline
diamond thin-film electrode. (B) Experimental and theoretical curves for Fe(CN)5'

3". Scan rate = 0.1 Vls. Electrode geometric area = 0.2 cm2. The scan rate was
varied from 100 to 500 mV/s.

69

3.3. Conclusions

The physical characterization and electrochemical properties of boron-
doped diamond thin film were discussed. Two types of conducting diamond,
microcrystalline and nanocrystalline, with different morphologies and
microstructures were investigated and compared. Both film types exhibited a
wide working potential window, a low voltammetric background current and good
responsiveness for Fe(CN)6'3’4, Ru(NH3)5+2’*3, methyl viologen (MV*2’”’°), and
IrCls’z"3 without any conventional pretreatment. The nearly reversible
voltammetry (at 100 mV/s)for all four of these redox systems indicates that
boron-doped microcrystalline and nanocrystalline diamond have a sufficient
charge carrier concentration and carrier mobility over a wide potential range to
support relatively rapid ratios of electron-transfer for these analytes. More
sluggish kinetics were found for tert-butylcatechol and Fe‘z’+3 at both diamond
types. The results were reproducible from electrode-to-electrode. The sluggish
kinetics have been attributed to weak adsorption for the former, and to the

absence of catalyzing surface carbonyl groups for the latter.

7O

3.4. References

10.

11.

12.
13.
14.
15.
16.

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

Xu, J.; Granger, M.C.; Chen, 0.; Strojek, J.W.; Lister, T.E.; Swain, G.M.
Anal. Chem; 1997, 69, 591 A.

Hupert, M.; Muck, A.; Wang, J.; Stotter, J.; Cvackova, Z.; Haymond, S.;
Show, Y.; Swain, G.M.; Diamond and Rel. Mater. 2003, 12, 1940.

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

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

Butler, J.E.; Windischmann, H.; MRS Bull. 1998, 23, 22.

Gwen, D.M.; MRS Bull. 1998, 23,32.

Gruen, D.M. Annu. Rev. Mater. Sci. 1999, 29, 211.

McCauley, T.G.; Gruen, D.M.; Krauss, A.R. Appl. Phys. Lett. 1998, 73,
1646.

Bhattacharyya, S.; Auciello, 0.; Birrello, J.; Carlisle, J.A.; Curtiss, L.A.;
Goyette, A.N.; Gruen, D.M.; Krauss, A.R.; Schlueter, J.; Sumant, A.;
Zapol, P. Appl. Phys. Lett. 2001, 79, 1.

Chen, 0.; Gruen, D.M.; Krauss, A.R.; Corrigan, T.D.; Witek, M.; Swain,
G.M. J. Electrochem. Soc. 2001, 148,44.

Dennison, J.R.; Holtz, M.W.; Swain, G.M.; Spectroscopy1996, 10, 1.
Knight, 0.8.; White, W.B.; J. Mater. Res. 1989, 4, 385.

Nemanich, R.J.; Solin, S.A.; Phys. Rev B 1979, 20, 392.

McCreey, R.L.; Packard, R.T.; Anal. Chem. 1989, 61, 775A.

Keblinski, P.; Wolf, 0.; Phillpot, S.R.; Gleiter, H.; J. Mater. Res. 1998, 13,
2077.

71

 

 

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

Nemanich, R.J.; Glass, J.T.; Lucovsky, G.; Shroder, R.E.; J. Vac. Sci.
Technol. 1988, A6, 1783.

Robins, L.H.; Farabaugh, E.N.; Feldman, A.J.; J. Mater. Res. 1990, 5,
2456.

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

Ferrari, A.C.; Robertson, J.; Phys. Rev. B 2001, 63, 121405.

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

Bennett, J.A.; Wang, J.; Show, Y.; Swain, G.M.; J. Electrochem. Soc.
2004, 151, E306.

Chen, 0.; Swain, G.M.; Langmuir 1998, 14, 7017.

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

Strojek, J.W.; Granger, M.C.; Swain, G.M.; Dallas, T.; Holtz, M.W.; Anal.
Chem. 1996, 68, 2031.

Witek, M.; Wang, J.; Stotter, J.; Hupert, M.; Haymond, S.; Sonthalia, P.;

Swain, G.M.; Zak, J.K.; Chen, 0.; Gruen, D.M.; Butler, J.E.; Kobashi, K.;
Tachibana, T.; J. WrdeGap Mater. 2002, 8, 171.

Chen, 0.; Gruen, D. M. ; Krauss, A. R.; Corrigan, T. D.; Witek, M.; Swain,
G.M.; J. Electrochem. Soc. 2001, 148, E44.

Wang, J.; Swain, G.M.; Mermoux, M.; Lucazeau, G.; Zak, J.; Strojek, J.W.;
New Diamond and Frontier technology 1999, 9, 317.

Haymond, 3.; Babcock, G. T.; Swain, G. M.; Electroanalysis, 2003, 15,
249.

Xu, J.; Granger, M.C.; Chen, 0.; Strojek, J.W.; Lister, T.E.; Swain, G.M.;
Anal. Chem. 1997, 69, 591A.

72

31.

32.

33.

35.

36.

37.

38.

39.

40.

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

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

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

Chen, P.; McCreeey, R.L.; Anal. Chem. 1996, 68, 3958.

Yagi, |.N.H.; Kondo, T.; Tryk, D.A.; Fujishima, A.; J. Electroanal. Chem.
1999, 173,473.

Alehashem, 3.; Chambers, F.; Strojek. J.W.; Swain, G.M.; Ramesham, R.;
Anal. Chem. 1995, 67, 2812.

Pleskov, Y.V.; Mazin, V.M.; Evstefeeva, Y.E.; Vamin, V.P.; Teremeskaya,
l.G.; Laptev, V.A.; Electrochem. Solid, State Left. 2000, 3, 141.

Qui, F.; Compton, R.G.; Marken, F.; Wilkins, S.J.; Goeting, C.H.; Foord,
J.S.; Anal. Chem. 2000, 72, 2362.

Vail, S.D.; Yang, H.-H.; McCreery, R.L.; Proceedings-Electrochemical
Society 1999, 99-5, 33.

McDermott, C. A.; Kneten, K. R.; McCreery, R. L. J. Electrochem. Soc.
1993, 140, 2593.

73

CHAPTER 4

Cyclic Voltammetric Studies of the Electro-
oxidation of Chlorinated Phenols at Boron-Doped
Diamond Thin-Film Electrodes

4.1. Introduction

The phenol oxidation reaction mechanism is complex with multiple
pathways possible. The first is the formation of a phenoxy radical due to the loss

of 1 proton and 1 electron (Figure 4.1).

OH

M+H++CA

Figure 4.1 Phenoxy radical formation during the initial stage of the electro-

oxidation of phenol.

This species then undergoes further oxidation to form soluble products
(e.g., quinones and hydroquinones) and/or anion-radical or radical-radical

coupling

74

Phenol
Reactive Polymer on

   
 

OH
R R
R -e , -H+
e',-H+ Polymer Radicag o
o- o o o-
(“\g R R R I R R I c/R
| I <3— 4)— | I «a» I «a»
R R
Oxidation to Soluble olymerizationw

Products Oxidation to Quinone Structures

Et er ¢-e,+H20H 51']

Formation
61°11 H19 fij °"
Catechol

Hydroquinone idea-ZIP

refit-1120, -H+

R R
O O
O .. ‘1
O

p-Benzoquinone O-Benzoquinone

 

Figure 4.2 The different possible reaction pathways for phenol oxidation (where
R = hydrogen or an adjacent ring in the polymer structure).5

to form oligomeric and polymeric products."7 Some coupling reactions may

produce ether type linkages, resulting in structures less reactive than phenol. The

75

oligomeric and polymeric radical, like the phenol radical, can be further oxidized
to form hydroquinone and catechol species.7 Further oxidation to benzoquinone
moieties are also possible. Figure 4.2 depicts the possible electro-oxidation
pathways for a phenol molecule. The polymeric products are characterized by
low permeability, low solubility in aqueous media, and a tendency to strongly
adsorb on electrode surfaces.1 The tendency of the polymeric products (e.g.,
polyphenol) to deposit on the electrode surface is the main reason for electrode
deactivation during the electro-oxidation of phenols. The adsorption passivates

the electrode for further electron transfer.

Most electrodes (e.g., Fe, Cu, Ni, Cr, Ti, Zn, Au, Pt, and A9) are rapidly
deactivated and fouled during the early stages of phenol oxidation.“"’“2 The
extent of passivation/fouling, however, depends on several parameters including
the phenol concentration, surface chemistry of the electrode, solution pH, solvent
additives, electrode potential, and current densityz'5 For example, in alkaline
solution, the formation of the C-O-C bonds prevails over the C-C coupling, while
in acidic solution, the C-C coupling is more favored compared to formation of
ether linkage.°'12 The anodic oxidation of phenols bearing halogen substituents
proceeds with some halogenlde elimination in addition to the carbon-oxygen
coupling mechanisms.° In general, elimination of the halogen during oxidation of

halophenols proceeds following a substitution radical-nucleophilic mechanism.13

76

Various approaches have been used to minimize the deactivation and
fouling problem. One is to optimize the electro-oxidation conditions as fouling can
be minimized at low phenol concentrations, elevated temperatures, and large
values of applied overpotential. The latter increases the rate of anodic discharge
of H20 to generate OH-.“'7 The OH- oxidatively degrades the adsorbed
oligomeric and/or polymeric reaction products, thereby removing them from the
surface. Another approach is to chemically modify the electrode surface. For
example, electrode surfaces have been modified with enzymes to enhance the
oxidative reactivity. 1‘ However, this approach has the drawbacks of complexity

and maintenance of the enzyme activity overtime.

Boron-doped diamond thin film is an advanced electrode material that
shows less tendency to deactivate and foul during phenol oxidation. There have
been several reports on the phenol and chlorinated phenols at conductive
diamond.15"°'"'""19 Fujishima and co-workers discussed the electrochemical
oxidation and detection of chlorinated phenols at anodically pretreated diamond
(i.e., oxidized). The oxidized electrodes exhibited good response stability and
reproducibility with low detection limits for several chlorinated phenols. They
reported that some fouling occurred at high phenolic compound concentrations (5
mM), but the electrodes could effectively be reactivated by anodic treatment at
2.6 V vs. SCE for 4 min.15 The authors postulated that the treatment generates
OHo, which oxidatively degrades the passivating polymer layer. In another report,

Comninellis and co-workers investigated the electrochemical oxidation of 4-

77

chlorophenol at a boron-doped diamond thin-film electrode.16 Under their
conditions, the formation of phenoxy radicals, 1,4-benzoquinone, and polymeric
products were observed. O’Grady and co-workers demonstrated that phenol can
be oxidatively degraded to CO2 at boron-doped diamond stably over time without
electrode fouling.17 Several other workers have addressed the electro-oxidation
reaction mechanism in acidic media at diamond electrodes.‘°'2°'22 For example,
Zhi et al. reported evidence for the direct electrochemical oxidation pathway for

phenol at diamond electrodes.18

In this Chapter, a comparison of the electrochemical performance of
boron—doped, hydrogen-terminated microcrystalline and nanocrystalline diamond
thin-film electrodes for the electro—oxidation of chlorinated phenols is described.
Cyclic voltammetry (CV) was used to study the oxidation of several chlorinated
phenols. The technique involves linearly scanning the potential applied to an
electrode from an initial value to a switching potential 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 plot of current versus
potential called a cyclic voltammogram. CV is a versatile technique useful for
studying the thermodynamics of redox processes, evaluating the effects of
various parameters and conditions on the redox reaction kinetics and
mechanisms, and investigating coupled chemical reactions and adsorption

processes. The technique is also useful for determining the redox potentials of

78

the electroactive species.23'2‘"25 The goal for this work was to learn how the
physicochemical properties of the diamond electrode affect the oxidation reaction
kinetics and mechanism of the chlorinated phenols. The analytes studied were
phenol, 2-chlorophenol, 3-chlorophenol, 4-chlorophenol, and pentachlorophenol

in acidic medium.

4.2. Results and Discussion

Cyclic voltammetric measurements using both microcrystalline and
nanocrystalline diamond thin film were conducted. The first task was to compare
the background voltammetric responses of both film types with that of glassy
carbon (GC), a commonly used carbon electrode material. Background cyclic
voltammetric i-E curves for GC, and microcrystalline and nanocrystalline
diamond thin film in 0.05 M phosphate buffer, pH 3.5, are presented in Figure
4.3. The cyclic voltammetric i-E curves for both diamond types are featureless
with no evidence for any surface redox chemistry. It can also be seen that the
two types of diamond have a larger overpotential for oxygen evolution than does
GC. The background current for G0 is larger than for either diamond type. For
example, the anodic background current at 0.45 V for GC is a factor of 5 larger
than that for microcrystalline diamond and a factor of 3 larger than that for
nanocrystalline diamond. The magnitude of the background current showed
some variability from GC sample-to-sample but, in all cases, was larger than

either diamond type. As a result, the oxidation current for the chlorophenols at

79

25.0

 

 
   

 

 

A - Microcrystalline diamond
20,0 _ B — Nanocrystalline diamond
C - GC
A 15.0 _
é
1a 10.0 —
i‘...’
g 5.0 _
U
0.0 _
-5 0

 

 

 

. . l . l . l . l
-500 0 500 1000 1500
Potential [mV vs. AgIAgCI]

Figure 4.3 Cyclic voltammetric i-E curves for a (A) microcrystalline, (B)
nanocrystalline diamond thin-film electrodes, and (C) GC in 0.05 M phosphate
buffer, pH 3.5. Scan rate = 100 mV/s. Electrode geometric area = 0.2 cm2.

GC is superimposed on a large and sometimes changing background signal,
while the oxidation current at either type of diamond is recorded against a low
and unchanging background. The broad oxidation and reduction peaks present
near 0.2 V, are' possibly associated with redox-active surface carbon-oxygen
functionalities (e.g., quinone and hydroquinone couple) formed at the exposed
edge planes.26 During the surface pretreatment of GC, carbon-carbon bonds are
broken, leaving dangling bonds on the surface. Upon air exposure, atmospheric
oxygen and/or water will react with the dangling bonds to form carbon-oxygen

functional groups.26 Some of the resulting functional groups are electroactive.

80

The low and stable background current, as well as the large overpotential
for oxygen evolution are characteristic features of the diamond electrodes.
Improved S/B and signal-to-noise (S/N) ratios are often observed in
measurements with this electrode material because of the low background
current. There are three factors responsible for the low background current.
First, a reduced pseudocapacitance due to the absence of redox-active and/or
ionizable carbon-oxygen functional groups on the surface. The increased
background current of GC is, in part, due to surface carboxylic acid functional
groups that deprotonate at pH values above the pKa of ca. 4.5.27 The excess
surface charge increases when these functional groups are deprotonated and
this leads to an increased double layer charging current. Second, the semimetal-
semiconductor nature of the diamond electrode leads to a slightly lower density
of surface electronic states near the Fermi level. The lower surface charge
density results in a reduced accumulation of counterbalancing ionic charge on
the solution side of the interface during double-layer charging. This results in a
lower background current. Third, the diamond film appears to consist of regions
of high electrical conductivity separated by zones of less conductivity, as recently
been observed by conductivity-probe AFM. This means that the effective
electrode area is less than the geometric area; the latter is the value used to

normalize currents.

The background current for nanocrystalline diamond is slightly larger than

that for microcrystalline diamond presumably due to the increased fraction of

81

grain boundaries and rt-bonded carbon that resides there. The tit-bonded carbon
is a source of charge carriers, which can increase the capacitive component of
the background current. Some of these carbons may also contain
electrochemically-active surface carbon-oxygen functionalities that would give

rise to a faradaic component in the background current.2°

Cyclic voltammetric i-E curves for phenol, 2-chlorophenol (2-CP), 3-
chlorophenol (3-CP), 4-chlorophenol (4-CP), and pentachlorophenol (PCP), all in
50 mM phosphate buffer, pH 3.5, were recorded for both diamond types. The
cyclic voltammetric responses for phenol, 2-CP, and 4-CP are similar in shape
for both diamond types (Figure 4.4). The electrochemical response of 2-CP is
used herein to exemplify the voltammetric behavior for all the three. The first
forward scan exhibits a single, well-defined oxidation peak at ca. 1.1 V (peak a).
The relatively stable peak charge after five scans ranged from 23 to 30 170 for all
three analytes. This current is due to the oxidation (via loss of 1 e‘ and 1 H*) to
form 2-chlorophenoxy radical. This oxidation peak occurs near the onset of
oxygen evolution so that oxidant, OH-, is also being generated. This oxidant can
attack the 2-chlorophenoxy radical and oxidize it via an anodic oxygen transfer
reaction. Of particular importance is the addition of OH groups to the aromatic
ring.1°' ‘7 Oxidation products, such as hydroquinone and catechol, are initially
formed.15 However, at these anodic potentials, the hydroxylated products can

undergo further oxidation (via electron and proton loss) to form quinone-like

species. It is these quinone-like oxidation products that are reduced during the

82

Current (11A) Current (11A)

Current (pA)

Figure 4.4 Cyclic voltammetric

20.0

 

15.0
10.0 _

 

55
A1.

20.0

 

 

20.0

 

15.0 s
10.0 e
5.0 _
0.0 _
-5.0 _

 

-10.0

 

 

.5007

l
0

‘ 500 ‘ 10100715100

Potential [mV vs. AglAgCl]

 

15.0 _
10.0 -
5.0 a
0.0 _

 

 

 

'5300 ‘

20.0

0

. l
500

L 10100 71500

20.0
1 5.0
10.0
5.0
0.0

'5300 ‘ 0

 

.—

h

 

l—

 

 

‘ 500 ‘ 1000 51500

Potential [mV vs. AglAgCl]

 

15.0

10.0 _

5.0
0.0

r—

L.

 

'53

Phenol

WSW—“1W

A

 

20.0
15.0
10.0
5.0
0.0

 

.—.

r—

.—

 

-5.0

.500 ‘

 

 

l
0

Potential [mV vs. AglAgCl]

‘ 500 ‘ 1000 ‘ 15'00

i-E curves for 2-CP, 4-CP, and phenol at a (A)

microcrystalline and, (B) nanocrystalline diamond thin-film electrode in 0.05 M
phosphate buffer, pH 3.5. Scan rate = 100 mV/s. Electrode geometric area =

0.2 cm2.

first negative-going sweep, as evidenced by the two reduction peaks at 0.55 and
0.48 V (peaks b and c). There is also a broad reduction wave at 0.05 V (peak d).
The reduced species at 0.55 and 0.48 V can be reoxidized at 0.55 and 0.60 V
(peaks e and f) during the positive-going scan. The peak currents for these
reactions grew with cycle number, stabilizing after the fifth scan. Reduction peaks
b and c are associated with the oxidation products of peaks e and f, as all were
present when the forward scan was reversed at 0.8 V. The peak currents for a,
ip°", and d, ip'°°, at both electrode types varied linearly with v"2 between 0.05 to
0.25 V/s (r2 > 0.990) indicating that the reaction rates for the initial oxidation of 2-
CP and the reduction of the product(s) formed are controlled by semi-infinite
linear diffusion of the reactant to the electrode surface. This means that soluble
rather than surface-confined reactants and products are involved in the redox
reactions. Rodrigo et al. observed similar voltammetric features for 4-CP and
reported that the peak currents, a and (I, increased with the addition of
hydroquinone to the solution.16 Therefore, we suppose that peak d is associated

with the reduction of a dissolved quinone-like product formed at 1.1 V.

The currents for peaks b and c, and e and f, on the other hand, increased
linearly with scan rate (r2 > 0.990), consistent with surface-confined redox
processes. Further evidence that these redox reactions involve adsorbed species
is the fact that the peak currents were, for most part, unaffected by electrolyte
stirring. In these measurements, five cyclic voltammetric scans were recorded

between -0.5 and 1.5 V to form the adsorbed oxidation reaction products. The

84

20.0

 

 

 
   
 

 

 

 

 

  
 
 
 

 

 

 

 

   

 

 

 

 

 

 

 

 

 

 

 

  

 

 

 

 

 

 

 

 

Ph I 20.0
000
15.0 _ A 15.0 _
"‘ I“ scan "'
gt 10'0 7 After agitation 3 10 0 A
A5 5.0 . ‘5 . _
i t“
0 0.0 _ ,7 o 0.0 -
W '5'$00 ‘ 0 ‘ 500 ‘ 1000 ‘ 1500 ‘
Potential [mV vs. AglAgCl] Potential [mV vs. AglAgCl]
20.0 20.0
2-CP
15.0. A 15.0,
g 10.0. 1"sca g 10.0.
After agitation -’ ‘E
E 5.0. g 5.0 .
3
3 0.0. .._ A . 0 0.0.
ngwfiislluw '5-3,‘ .....:
- 1 oo o 500 1000 1500
Potential [mV vs. AglAgCl] Potential [mV vs. AglAgCl]
20.0 20.0
15.0 _ TOP A 15.0 _
10.0 _ 10.0 _
3 After its g
5.0 - 5.0 -
E A: é
o 0.0 F 1 1. 0 0.0 _
$10 0 500 1000 W "53:00 0 500 1000 1500

Potential [mV vs. AglAgCl]

Potential [mV vs AglAgCl]

 

Figure 4.5 Cyclic voltammetric i-E curves for phenol, 2-CP, and 4-CP at (A)
microcrystalline and (B) nanocrystalline diamond thin-film electrode in 0.05 M

phosphate buffer, pH 3.5. Scan rate = 100 mV/s. Electrode geometric area =
0.2 cm2.

85

analyte containing solution was then vigorously agitated for several minutes by
bubbling with nitrogen gas. Figure 4.5 shows voltammetric responses for 0.1 mM
solutions of phenol, 2-CP, and 4-CP for (A) microcrystalline and (B)
nanocrystalline diamond electrodes before and after agitation. Oxidation peaks, e
and f, and reduction peaks, b and c, were still present after the bubbling,
although with minor current attenuation. However, if the analyte-containing
solution was removed from the cell, the electrode copiously rinsed with ultrapure
water, and the cell filled with fresh electrolyte, then the background voltammetric
i-E curves were devoid of these redox peaks. This indicates that these oxidation
reaction products are weakly adsorbed to the diamond surface. We suppose that
these surface confined oxidation reaction products are either individual
molecules adsorbed on the diamond surface, perhaps at localized sites where
some oligomeric quinone structures has formed. For 2-CP and 4-CP, the
corresponding chlorinated or non-chlorinated hydroquinone/benzoquinone is
expected.22 The fact that the primary oxidation peak current at 1.1 V is not
significantly attenuated with prolonged potential cycling indicates that a blocking
polyphenol layer does not form on the surface. In other words, some polymer
may form on the electrode surface, but large uncoated areas of the electrode

exist for the primary oxidation reaction to occur.
We suppose that the oxidation current at 1.1 V is due to a direct one-

electron, one- proton oxidation of the phenol to form 2-chlorophenoxy radical.2'5 It

has been reported that the Ep°" for the oxidation of 2,4—dichlorophenol at “as

86

deposited” (i.e., presumably hydrogen-terminated) diamond shifts less positive
potentials with increasing pH up to the pKa value of ca. 8. A change of -59
mV/dec was found.16 This trend is consistent with an equal number of protons
and electrons being transferred during the redox reaction. Rodrigo et al. have
reported that this oxidation peak results from the one-electron/one-proton
oxidation to phenoxy radical or from an additional one-electron oxidation to form
phenoxy cation.17 The seminal mechanistic work by Gattrell and Kirk indicates
that the phenol oxidation reaction (and presumably for chlorinated phenols as
well) proceeds initially through the formation of the phenoxy radical."2'5'7 Once
formed the radicals can react via two pathways: (i) radical-radical coupling to
form oligomeric and polymeric species or (ii) subsequent direct or indirect
oxidation to produce soluble or surface confined oxidation products, like

benzoquinone and catechol.

The voltammetric responses for 3—CP and PCP were similar in shape at
both electrode types, but quite different from the responses for the other three
phenols. Figure 4.6 shows five consecutively recorded cyclic voltammograms for
0.1 mM PCP at a (A) microcrystalline and (B) nanocrystalline diamond thin-film
electrode. The first forward scan for both electrodes shows a well-defined
oxidation peak at ca. 0.90 V with a corresponding reduction peak at ca. —0.20 V
during the reverse scan. The oxidation peak occurs at a potential less positive
than the peak for phenol, 2-CP and 4-CP by approximately 200 mV. This is

because the pKa for PCP is 4.74, as compared to values of 8 to 10 for phenol

87

 

 

 

 

 

Potential [mV vs. AglAgCl]

20.0

15.0L B
10.0
1

5.0 _

 

 

0.0- V

-5.0_ l

“filo ‘ 0 ‘W

Potential [mV vs. AglAgCl]

Current (11A)

 

 

 

Figure 4.6 Cyclic voltammetric i-E curves for 0.1 mM PCP at a (A)
microcrystalline and (B) nanocrystalline diamond thin-film electrode in 0.05 M

phosphate buffer, pH 3.5. Scan rate = 100 mVls. Electrode geometric area = 0.2

cm2.

and the other chlorinated phenols. The lower value reflects a greater ease of
deprotonating the phenolic hydrogen. The oxidation peak charge, after

stabilization, was nominally 18 110. The 0.90 V peak is presumed to be

88

associated with the formation of the dimer 2,3,4,6-pentachlorophenoxy—2,5-
cyclohexadienone.11 Gattrell and MacDougall studied the electrochemical
oxidation of PCP and concluded that the reaction involves the loss of one-
electron and one proton to form pentachlorophenoxy radical. The radical then
undergoes radical-radical coupling to form a dimer.11 A slight decrease in the
oxidation current and a slight increase in the reduction current are observed for
the second and all subsequent scans. It is unclear at this point what reaction is
associated with the reduction current. The pentachlorophenoxy radical most
likely dimerizes to form a water insoluble species through C-O rather than C-C
coupling due to steric limitations. The dimers formed could be 2,3,4,5,6-
pentachloro-4-pentachlorophenoxy-2,5-cyclohexadienone and/or 2,3,4,5,6-
pentachloro-2-pentachIorophenoxy-3,5-cyclohexadienone.11 Figure 4.7 shows
preliminary AFM images of a microcrystalline diamond electrode after 25
voltammetric scans in 0.1 mM PCP. The image reveals the presence of deposits
on the surface, as indicated by the arrows. The deposits are formed all over the
surface, but large clusters are found to be localized along grain boundaries. The

deposits are water insoluble species mentioned above.

89

deposits

 

Figure 4.7 AFM images of a microcrystalline diamond thin-film electrode after 25
cycles in 0.1 mM PCP in 0.05 M phosphate buffer, pH 3.5. Experimental
conditions as in Figure 4.6.

Figure 4.8 shows five consecutive cyclic voltammograms for 0.1 mM 3-CP
recorded at a microcrystalline (A) and nanocrystalline (B) diamond thin-film
electrode. An oxidation peak is observed during the forward scan at ca. 1.2 V,
with only a weak corresponding reduction wave at ca. —0.20 V. The anodic peak
charge, after stabilization, was nominally 26 pC. This oxidation peak is believed
to be due to the 3-chlorophenoxy radical. Unlike the other four solutes, the
oxidation peak currents decreases more rapidly with cycle number moreso for
the nanocrystalline film, before reaching a pseudo-steady—state value. Contrary

to the behavior of phenol, 2-CP, and 4-CP, no new oxidation or reduction peaks

9O

develop at lower potentials. Apparently, the surface-confined electroactive
quinones are not by-products of the oxidation reaction. For both 3-CP and PCP
at microcrystalline and nanocrystalline diamond, the oxidation peak current (1.2
and 0.9 V, respectively) varied linearly with v from 0.05 to 0.25 V/s (r2>0.990)

consistent with a surface-confined redox reaction.

20.0

15.0-A

10.0 .

 

5.0 .

Current (uA)

 

 

0.0. ,1

'513110‘ 0 ‘5011‘10‘00—‘1500'

Potental [mV vs. Ag.AgCl]

 

 

 

 

20.0

 

15.0
1 0.0

5.0

Current (pA)

0.0

'5'51110 1m) an urine—etc

Potential [mV vs. AglAgCl]

 

 

 

 

 

Figure 4.8 Cyclic voltammetric i-E curves for 0.1 mM 3-CP at (A) microcrystalline
and (B) nanocrystalline diamond electrodes in 0.05 M phosphate buffer, pH 3.5.
Scan rate = 100 mVls. Electrode geometric area = 0.2 cm2.

91

The 3-CP and PCP and/or their products adsorb on the electrode strongly
than do other phenolic compounds studied as featureless background
voltammogram was only observed if the passivated electrode was first cycled
(ca. 10 scans) well into the oxygen evolution regime and then exposed to a fresh
electrolyte solution. This suggests that the adsorbed products need to be

oxidatively degraded by OH' in order to be removed from the surface.

The same cannot be said for GC, as similar agitation did not restore the
original current. The trends in the shapes of the cyclic voltammograms for each
of the phenolic compound with cycle number were the same for both GC and
diamond. Figure 4.9 A shows five consecutive cyclic voltammograms for 0.1 mM
2-CP at freshly polished GC. A difference from diamond was the ca. 100 mV less
positive oxidation peak for GC. The oxidation reaction products adsorb very
strongly to the GC surface resulting in significant response deactivation. For
example, GC was irreversibly fouled after five consecutive cycles in 2-CP and 3-
CP. The oxidation peak current at 1.0 V decreased with cycle number and the
original value could not be regained even with vigorous solution agitation or
replacement with fresh electrolyte (see Figure 4.9 B). The current decreased by
36 %, and the oxidation peak was significantly shifted to the positive potentials.

Polishing was the only means to remove all traces of the phenols.

92

20.0 20.0

 

 

 

 

 

 

 

 

 

15.0 . A 15.0 _
“10.0 _ -1” _
3 5
a 5.0 - 1 .., 5.0 _
g c
a o o _ g 0.0 _
0

3
-5.0 _ 0 -5.0 _
-10.0 1 t . t . t 1 - . . . 1
-500 o 500 1000 1500 10300 0 5110 1000—Woo
Potential [mV vs. AglAgCl] Potential [mV vs- AglAgCl]

Figure 4.9 Cyclic voltammetric i-E curves for 0.1 mM 2-CP at GO in 0.05 M
phosphate buffer, pH 3.5, (A) and after solution agitation (B). Scan rate = 100
mVls. Electrode geometric area = 0.2 cm2.

4.3. Conclusions

Cyclic voltammetric studies of phenol and chlorinated phenol oxidation
were conducted using two types of diamond. The surface morphology and
microstructure of each were different. The results revealed that (i) all five
phenolic compounds can be electro-oxidized at both diamond types without
electrode deactivation and fouling, as is the case for other electrodes like GC, (ii)
the electro-oxidation reaction mechanism proceeds in a similar manner at both
diamond types regardless of the differences in film morphology and
microstructure, with phenol, 2-CP, and 4-CP yielding similar responses, (iii) the
oxidation of phenol, 2—CP, and 4-CP leads to the formation of surface confined
quinone/hydroquinone moieties, whereas the oxidation of 3-CP and PCP do not,
and (iv) adsorbed reaction products were easily removed from the diamond

surface by copiously rinsing with ultrapure water and exposure to fresh

93

electrolyte. In some cases, such as that for PCP, limited potential cycling was

necessary to remove all traces of the reaction products.

94

4.4. References

10.

11.
12.
13.
14.

15.

16.

17.

18.

19.

Gattrell, M.; Kirk, D.W.; J. Electrochem. Soc., 1992, 139, 2736.
Gattrell, M.; Kirk, D.W.; J. Electrochem. Soc., 1993, 140, 1534.
Belhadj Tahar, N.; Savall, A.; J. Electrochem. Soc., 1998, 145, 3427.

Rodgers, J.D.; Jedral, W.; Bunce, N.; J. Environ. Sci. Technol., 1999, 33,
1453.

Gattrell, M.; Kirk, D.W.; J. Electrochem. Soc., 1993, 140, 903.

Mengoli, G.; Daolio, S.; Musiani, M.M.; J. Appl. Electrochem 1980,10, 459.
Gattrell, M.; Kirk, D.W.; Can. J. Chem. Eng., 1990, 68, 997.

Mengoli, G.; Musian, M.; J. Electrochem. Soc.1987,134, 643C.

Bruno, F.; Pham, M.C.; Dubois, J.E.; Electrochim. Acta., 1977, 22,451.

Fleischmann, M.; Hill, |.R. Mengoli, G.; Musian, M.M. Electrochim. Acta.,
1983, 28, 1545.

Gattrell, M.; MacDougall, B.; J. Electrochem. Soc. 1999, 146, 3335.
lotov, P.l.; Kalcheva, S.V.; J. Electrochem. Soc. 1998, 442, 19.
Ezerskis, Z.; Jusys, Z.; J. Appl. Electrochem. 2002, 32, 543.

Lu, W.; Wallace, G.G.; lmisides, M.D.; Electroanalysis, 2002, 14, 325.

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

Rondrigo, M.A.; Michaud, P.A.; Duo, l.; Panizza, M.; Cerisola, G.; Ch.
Comninellis, G.; J. Electrochem. Soc. 2001, 148, D60.

Hagans, P.L.; Natishan, P.M.; Stoner, B.R.; O’Grady, W.E. J.
Electrochem. Soc. 2001, 148, E298.

Zhi, J.-F.; Wang, H.-B.; Nakashima, T.; Rao, T.N.; Fujishima, A.; J. Phys.
ChemB 2003, 107, 13389.

Prado, C.; Murcott, G.G.; Marken, F.; Foord, J.S.; Compton, R.G.;
Electroanaysis, 2002, 14, 975.

95

20.

21.

22.

23.

24.

25.
26.

27.
28.

Polcaro, A.M.; Vacca, A.; Palmas, S.; Mascia, M.J.; J. Appl. Electrochem.
2003, 33, 885.

Codognoto, L.; Machado, S.A.S.; Avaca, L.A.; J. Appl. Electrochem. 2003,
33, 951.

Canizares, P.; Garcia-Gomez, J.; Saez, C.; Rodrigo, M.A.; J. Appl.
Electrochem. 2003, 33, 917.

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

Kissinger, P.T.; Heineman, W.R.; J. Chem. Ed. 1983, 60, 702.

Wang, J.; Analytical Electrochemistry, 2nd ed.; Wiley: New York, 2000.

Runnels, P.L.; Joseph, 10.; Logman, M.J.; Wightman, R.M. Anal. Chem.
1999, 71, 2782.

Duvall, S.H.; McCreery, R.L.; J. Am. Chem. Soc. 2000, 122, 6759.

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

96

CHAPTER 5

Detection of Chlorinated Phenols by Flow
Injection Analysis and High Performance Liquid
Chromatography with Amperometric Detection

using a Boron-Doped Diamond Thin-Film

Electrode

5.1. Introduction

Results presented in the previous Chapter showed that chlorinated
phenols can be oxidized at boron-doped microcrystalline and nanocrystalline
diamond thin film electrodes without significant deactivation or fouling."5 The
results presented vide intra further demonstrate the capability of this new
electrode for the sensitive, reproducible and stable amperometric detection of
these priority pollutants, when coupled with either flow injection analysis (FIA)

and high performance liquid chromatography (HPLC).

In FIA, a liquid analyte sample is injected into the nonsegmented flowing

carrier solution, propelled by a constant flow pump. The injected sample forms a

97

well-defined zone that is transported to the detector. The analyte concentration
profile entering the detection cell depends upon the mode of the sample
introduction, longitudinal diffusion, the flow parameters, and the geometry of the
FIA channel situated between the sampling point and the detection site.° 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 an analyte zone. The output is a detector signal that
varies with time and is used for quantitative analysis (peak height or area).7'8 In
FIA, the primary cause of zone (i.e., peak) broadening is longitudinal diffusion.
The benefits of FIA include automated sample processing, high repeatability,
short analysis time, adaptability to miniaturization, containment of chemicals, and
waste reductions"1o Liquid chromatography (LC) is a separation technique in
which the solute is injected into a mobile phase and is transported through a
column and to a detector. In HPLC (with a packed column), zone broadening
arises from (i) eddy diffusion, (ii) longitudinal diffusion, and (iii) mass transfer

within the stationary phase.

Several detection methods are normally coupled to FIA and HPLC, with
electrochemical detection being one of the most common.7 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 in which the potential is

98

scanned with time. Several types of electrochemical cell designs have been
used in FIA and HPLC with the most common being a thin-layer or cross flow
type. In this design, the working electrode is positioned in a thin channel through
which the carrier solution or mobile phase flows.11 A schematic of the cell used in
these studies was presented in Chapter 2. The cell design must give rise to a
high S/N ratio, have low dead volume and well-defined hydrodynamic flow,
minimal contribution to ohmic resistance, and be easy to construct and
maintain.” With electrochemical detection, the mobile phase or carrier solution
contain a supporting electrolyte. Oxidation or reduction reactions occur when a
small volume of the solute flows past an electrode in the detection cell. Mass-
transfer of the analyte to the electrode occurs by diffusion and also by convection
due to the fluid flow. Therefore electrochemistry, in a flowing stream is referred to
as a hydrodynamic technique. The enhanced rate of mass-transfer yields higher
currents and higher S/N ratios."'” The conversion efficiency, which is defined
as the fraction of solute molecules entering the cell that are electrolyzed, is
typically less than 5% in a thin-layer cell with a crossflow geometry.” The
limiting current for a thin-layer flow cell is described by the equation,”
in... = 1.47nFC(pA/p)°’3u"3

where ium is the limiting current (A), n is electron stoichiometry, F is the Faraday
constant, C is concentration (mol/cma), D is diffusion coefficient (cmzls), A is
electrode area (cmz), b is channel height (cm), and U is average volume flow rate

(cm3/s).

99

FIA and HPLC coupled with amperometric detection (FIA-EC and HPLC-
EC) were used to analyze for chlorinated phenols. Both microcrystalline and
nanocrystalline electrodes were used for the oxidative detection of these
pollutants. The objectives of the work were (i) to compare the performance of
both diamond thin film types, (ii) to comprehensively evaluate the analytical
detection figures of merit for both electrode types, and (iii) to demonstrate the
capability of the method for the analysis of chlorinated phenols in a real sample.
The electrode performance was evaluated in terms of the linear dynamic range,

sensitivity, limit of quantitation, and response precision.

5.2. Results and Discussion

5.2.1. Flow Injection Analysis with Electrochemical Detection

Hydrodynamic voltammetric background i-E curves for microcrystalline
diamond, nanocrystalline diamond, and freshly polished glassy carbon (G0) in
0.05 M phosphate buffer, pH 3.5, are shown in Figure 5.1. The mobile phase
flow rate was 0.6 mL/min. The current was measured after 20 min at each
potential and recorded. The curves were obtained by changing the applied
potential in 50 mV increments beginning at 600 mV. The current magnitude and
the peak-to-peak noise were recorded at the end of this period. The current for
both diamond types is low and about the same magnitude over the entire
potential range. The GC response, on the other hand, is much larger (factor of 5—

10) at all potentials, especially positive of 1.2 V. As previously discussed in

100

400

 

 

 

 

I so

a; 300 A microcrystalline
I: . nanocrystalline
:
g 200 _
3
U

100 _

L 11"1’ l AAA-l
200 800 1000 1200

Potential [mV vs. AglAgCl]

Figure 5.1 FIA-EC hydrodynamic voltammetric background i-E curves for
microcrystalline and nanocrystalline diamond, and GO in 0.05 M phosphate
buffer, pH 3.5. Flow rate = 0.6 mUmin. Electrode geometric area = 0.11 cmz.

Chapter 4, the larger background for CC is due to the microstructural changes
associated with the oxidation (i.e., formation of surface carbon-oxygen
functionalities) at the edge plane and defect sites, which occurs over a wide
potential range, as well as the onset of oxygen evolution.“ One of the
advantages of using diamond in amperometric detection is the low and stable
background current and the large overpotential for oxygen evolution. Diamond is
much more resistant to morphological and microstructural degradation than is
GC. Therefore, the background current is much more stable over time at a given
potential, particularly positive values. The background current for GC is also

dynamic, progressively increasing with time due to the dynamic morphological

101

changes that occur at positive potentials. The oxidation causes an increase in
the true electrode surface area, which consequently leads to an increase in
current. On the other hand, the background current for diamond was very stable
over time and reached a constant value soon after detector tum-on (Eapp. = 1.1 V)
within 5-10 min. The diamond surface is morphologically and microstructurally
stable at anodic potentials so no dynamic oxidation processes or changes in
surface structure. Surface oxidation does occur on diamond but is confined to a
simple functional group exchange on the surface.15 In contrast, the background
current for GC often took 45—60 min to stabilize. A similar observation was
reported by Terashima et al.2 The rapid stabilization time is an advantageous

property of diamond because it leads to a shorter overall analysis time.

Figure 5.2 shows hydrodynamic i-E curves for microcrystalline and
nanocrystalline diamond thin-film electrodes during 20-uL injections of 40 (1M
phenolic solutions in 50 mM phosphate buffer, pH 3.5. The mobile phase flow
rate was 0.6 mUmin. Each datum corresponds to the average response for 5
injections. The error bars were within the size of the marker and are not shown.
The responses for phenol, 2-CP, 3-CP, and 4-CP do not exhibit a well-defined
sigmoidal shape (i.e., limiting current) - a shape expected if the reaction rates is
limited exclusively by mass transfer at an electrode with well-defined area. The
only sigmoidally-shaped curve is the one for PCP. These observations are
different from those of Terashima et al. as they reported sigmoidalIy-shaped

hydrodynamic voltammograms for several di- and trichlorophenols.2 It is

102

supposed that the maximum oxidation current for phenol, 2-CP, 3-CP, and 4-CP
in the potential region positive of 800 mV is under mixed control by direct and
indirect electro-oxidation reaction rates as well as mass transfer. As mentioned in
Chapter 4, the electro-oxidation of these phenols involves the initial formation of
a phenoxy radical species. This is followed by either radical-radical coupling to
form polymeric products, or direct and indirect (OH- generation) oxidation
reactions to form quinone-like products.‘“'”*21 The maximum current in the
hydrodynamic curves varied depending on the phenolic compound with the

largest magnitude observed for 2-CP at both electrodes.

Figure 5.3 show plots of the signal-to-background (S/B)(I(o. — Ibkgllbkg) ratio
as a function of the applied potential for the phenol, 2-chlorophenol, 3-
chlorophenol and 4-chlorophenol. Such plots are useful for determining the
optimum detection potential when sigmoidal hydrodynamic voltammograms are
not observed. The optimum detection potential for the analytes is in the 1100-
1200 mV range for both diamond types. A sharp decrease in SIB for the
nanocrystalline diamond is observed at potentials positive of 1200 mV. This is
due to the onset of oxygen evolution, which occurs at a slightly less positive
potential than for microcrystalline diamond. The most negative onset potential for
the oxidation and the largest current response is observed for 2-CP. Irrespective
of the order of injection, the largest SIB is seen for 2-CP at both electrode types.
SIB plot for the PCP is not presented because It exhibited a well-defined

sigmoidal current response.

103

1200

 

 

 

 

   
  

 

  

 

   

 

900 _ + .213 /
-D— Bk:

3t: 600 - / /
g ,
g 300 r (l {f ,
o I?

g I I .—I —-l 4.1.1.411." ' ' '.-£-.'.-:-'

01811 1000 1200
Potential [mV vs. AglAgCl]
1200
B Phenol
900 ._ i
+

A -D— Bk:
:5. 600 _
E
t
5’ 300 _

0 l _ I I '

600 800 1000 1200

Potential [mV vs. AglAgCl]

Figure 5.2 Hydrodynamic voltammetric i-E curves for 40 (N concentrations of
each phenolic compound in 50 mM phosphate buffer, pH 3.5, at a (A)
microcrystalline and (B) nanocrystalline diamond thin-film electrode. Flow rate =
0.6 mUmin. Each datum represents the average of 5 injections.

In order to evaluate the effectiveness of diamond for phenol and
chlorinated phenols analysis, the analytical detection figures of merit (e.g., linear

dynamic range, sensitivity, limit of quantitation (LCD), and response precision)

104

30.0

 

+ Phenol
A _._

+ 4-CP
200 'th'“”

I

SIB

10.0

      

 

 

L

I
1000 1200
Potential [mV vs. AglAgCl]

30.0

 

  
 
 

+ Phenol
B +
4—CP
+ 2-CP
Q 2000

S

 

10.0 ._

  

 

 

 

0.0 1 1 1 1
600 800 1000 1200
Potential [mV vs. AglAgCl]

Figure 5.3 SIB vs potential plots for 40 (N concentrations of each phenolic
compound in 50 mM phosphate buffer, pH 3.5, at a (A) microcrystalline and (B)
nanocrystalline diamond thin-film electrode. Flow rate = 0.6 mUmin. Injected
volume = 20 uL. Each datum represents the average of 5 injections.

105

were determined for both diamond thin film types. 60 was also used for
comparison. A calibration curve for each analyte was obtained in 50 mM
phosphate buffer, pH 3.5, as a function of the injected concentration. Plots of the
peak height versus the concentration are shown in Figure 5.4 for microcrystalline
diamond. Similar calibration curves were obtained for nanocrystalline diamond
(not shown). The injected volume was 20 (1L and the flow rate was 0.6 mLImin.
The data presented are average for at least three electrodes. Detection was
made at 1.1 V for microcrystalline and 1.2 V for nanocrystalline diamond. The
linear dynamic range, LOQ, and the sensitivity for the phenol and chlorinated
phenols both electrode types are summarized in Tables 5.1 and 5.2. The linear
dynamic range for all the analytes at both diamond types is 4 - 5 orders of
magnitude (r2 > 0.990) from the nanomolar to micromolar range. Overall, slightly
higher sensitivities were observed for nanocrystalline diamond by a factor of 1.3
— 1.5 at least for some analytes. The sensitivities for all analytes, except for PCP,
are approximately the same for a given diamond type with the value for PCP
being about a factor of 3 lower. The lower sensitivity for PCP, as compared to
other analytes, could be due to a lower number of electrons transferred in the
overall electrochemical reaction. The data for only one electrode type, shown in
Figure 5.3, reveal a sizable difference in the maximum response for each of the
compounds. However, we observed this to be the exception rather than the rule.
The more common observation is reflected in the sensitivity data in Tables 5.1
and 5.2. The minimum concentration detected (LOQ) for each analyte is also

similar for both diamond types. The mass limit of quantitation is in the high pg to

106

a
.3

Phenol

   
   

3

Current(nA)
3 s 3 §

 

 

 

g g
1 1

Current (nA)
§

141

as

 

 

 

 

Concentration (pM)

 

 

f

0 2(1) 400 6(1) 300 1111) 1200

Concentration (11M)

Figure 5.4 Calibration curves for the phenol and chlorinated phenols obtained for
microcrystalline diamond. The carrier solution was 50 mM phosphate buffer, pH

3.5. Flow rate = 0.6 mLImin. Injected volume = 20 uL. Electrode geometric area
= 0.11 cmz. Detection potential = 1.1 V.

107

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table 5.1 FIA-EC Data for Phenol and Chlorinated Phenols at
Microcrystalline Diamond Thin-Film Electrode
Analyte Linear Dynamic Sensitivity LOQ‘ LOQ
Range 1le (anIM) (11M) (n9)
phenol 0.3 — 100 9.9 1: 1.2 0.3 (SIN =7) 0.77
r2 = 0.9903 (28 ppb)
2-CP 0.05 - 200 14.2 1 0.5 0.05 (SIN = 6) 0.13
r2 = 0.9923 (6 ppb)
3-CP 0.1 — 100 10.0 1: 2.2 0.1 (SIN =7) 0.26
r2 = 0.9902 (13 ppb)
4-CP 0.5 — 100 11.0 :I: 1.3 0.5 (SIN = 4) 1.28
r2 = 0.9928 (64 ppb)
PCP 0.6 — 1200 3.4 1 0.1 0.6 (SIN = 6) 3.19
r2 = 0.9980 (160 ppb)
Table 5.2 FIA-EC Data for Phenol and Chlorinated Phenols at
Nanocrystalline Diamond Thin-Film Electrode
Analyte Linear Dynamic Sensitivity L00“ L00
Range (11M) (nAIlIM) (an ("9L
phenol 0.3 - 100 18.4 1 1.3 0.3 (SIN =8) 0.77
r2 = 0.9958 (28 ppb)
2-CP 0.05 - 200 17.9 1: 1.8 0.05 (SIN = 7) 0.13
r2 = 0.9946 (6 ppb)
3-CP 0.1 — 100 18.8 :I: 1.4 0.1 (SIN =7) 0.26
:2 = 0.9977 (13 ppb)
4-CP 0.5 - 100 16.4 :I: 1.5 0.5 (SIN = 6) 1.28
:2 = 0.9964 (64 ppb)
PCP 0.6 — 1200 6.9 :t 1.9 0.6 (SIN = 6) 3.19
r2 = 0.9979 (160 ppb)

 

 

 

 

 

 

a limit of quantitation = lowest concentration injected and detected.

Note: The data are reported for 3 electrodes for each diamond type. Carrier
solution = 50 mM phosphate buffer, pH 3.5. Injection volume = 20 uL. Flow rate
= 0.6 mUmin. Electrode geometric area = 0.11 cmz. The detection potential was

1.1 and 1.2 for microcrystalline and nanocrystalline diamond, respectively.

108

low ng range (SIN 2 3). By way of comparison, a reduced linear dynamic range,

a higher limit of detection, and greater response variability were observed for GC.

The response precision was evaluated over twenty consecutive 20-uL
injections of 40 11M analyte at 1-min intervals. Both diamond types exhibit good
response precision for these analytes without significant electrode deactivation.
For example, Figure 5.5 shows the response for multiple injections of 2-CP at a
microcrystalline film. A summary of the response precision for all the analytes at
both diamond types and GC is presented in Table 5.3. The data represents an
average of three runs on each electrode type. The RSD values range from 1.9 to
5.7% for microcrystalline diamond, 1.4 to 3.9% for nanocrystalline diamond, and
4.6 to 20.7% for GC. The good response precision for diamond is due to the
excellent morphological and microstnlctural stability of the material. In contrast,
the response precision was poor for GC due to the morphological and
microstructural instability and deactivation by reaction intermediates and
products. Figure 5.6 shows multiple injections of phenol, 2-CP, 3-CP, 4-CP and
PCP on GC. The current response for the analytes decreases with time moreso
for phenol and 3-CP. On GC, the edge plane carbon atoms are terminated by
oxygen functionalities, making the surface polar and hydrophilic. Strong
adsorption occurs near these sites probably due to 11-11, dipole-dipole and/or ion-

dipole interactions between the polar surface and the solute.

109

 

[100M

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

JLhusc-LLLLLLLnnLnLLLt
Time

 

 

Figure 5.5 FIA-EC response for multiple injections of 40 pM 2-CP. The carrier
solution was 50 mM phosphate buffer, pH 3.5. Detection potential = 1.2 V.
Injection volume = 20 uL. Flow rate = 0.6 mUmin.

Table 5.3 Response Precision Data for Phenol and Chlorinated Phenols for
Microcrystalline and Nanocrystalline diamond, and GC.

 

 

 

Analyte Microcrystalline Nanocrystalline GC
RSD% RSD% RSD%
Phenol 4.1 3.9 7.7
2-chlorophenol 1 .9 1.4 4.6
3-chlorophenol 5.6 2.2 20.7
4-chlorophenol 1.7 2.2 6.1
Pentachlorophenol 1 .6 2.6 6.5

 

 

 

 

Note: The response precision was evaluated from 20 consecutive injections at 1-
min intervals. The data are reported as a mean for 3 measurements. Carrier
solution = 50 mM phosphate buffer, pH 3.5. Analyte concentration = 40 11M.
Injection volume = 20 uL. Flow rate = 0.6 mUmin. Electrode geometric area =
0.11 cmz. The detection potential was 1.1, 1.2 and 1.0 V for microcrystalline
diamond, nanocrystalline diamond, and GC, respectively.

110

 

 

1000 _..
. w Wag-'3'”-

800 ; ' " ' "WWW A 21:1»
. "l'I... DPhenol

Q 3-CP
O 4-ce
A PCP

O
O
o
o

 

Current (nA)
N is a
o
c

O
O

 

 

 

Q
GT“

10.0 ‘ 20.0 ‘ 30.0
No. of Injection

Figure 5.6 FIA—EC responses for multiple injections of 40 pM phenol, 2-CP, 3-
CP, 4-CP, and PCP on GC. The carrier solution was 50 mM phosphate buffer,
pH 3.5. Detection potential = 1.0 V. Injection volume = 20 pL. Flow rate = 0.6
mUmin.

5.2.2. High Performance Liquid Chromatography with Electrochemical

Detection

Reversed-phase HPLC separation with electrochemical detection was
used to analyze solutions containing phenol and chlorinated phenols. Significant
optimization was performed initially to develop a satisfactory isocratic separation
of these solutes with maximum efficiency. In the isocratic elution, the retention
time of the least polar solute (PCP) was quite long (> 30 min), therefore a step
was employed in the elution procedure. A 50 mM phosphate buffer/acetonitrile,

pH 3.5, 65:35 (vIv) mobile phase was used to separate phenol, 2—CP, 3-CP, and

111

4-CP. The mobile phase was then switched to a 20:80 (v/v) composition in order
to elute the last analyte (PCP). Peak assignments were made based on retention

time matching of the individually injected solutes.

lsocratic elution is usually the separation method of choice with
electrochemical detection because gradient elution or often causes drifting
baseline currents and long stabilization times for many electrode materials. Both
diamond types exhibited low and stable background currents in the mixed mobile
phase system, even at high positive potentials. The background current for
diamond at 1.1 or 1.2 V was low at about 5 - 8 nA and increased only slightly with
an increase in the organic modifier content. The peak-to-peak noise was ca. 200
pA. Rapid stabilization was observed for both diamond types after detector turn

on in this mobile phase.

Figure 5.7 shows a typical chromatogram for the five phenols detected
amperometically using a microcrystalline diamond electrode. A similar
chromatogram was observed for a nanocrystalline diamond. The expected
reversed-phase elution order, based on the solute polarity, was observed.
Phenol, being the most polar solute elutes first, whereas PCP, being the least
polar molecule, elutes last. The separation is complete in about 16 min and the
peaks are fronting slightly. The fronting worsened with prolonged column use.

Typically, no fronting was seen with a new column. This is presumably because

112

of stationary phase degradation which creates exposed silanol sites that can

interact strongly with the phenols leading to a secondary retention mechanism.

A comprehensive set of detection figures of merit for phenol and the chlorinated
phenols was obtained for both nanocrystalline and microcrystalline diamond.
Calibration curves for all the solutes are shown in Figure 5.8 for (A)
microcrystalline and (B) nanocrystalline thin film. The detection figures of merit
are summarized in Table 5.5 and 5.6. The linear dynamic range for all the solutes
is at least 2-3 orders of magnitude (r2 > 0.99). It should be emphasized that the
maximum concentration of 80 (M does not represent the upper limit of the
dynamic range but just simply the maximum solute concentration used in this
work. The response sensitivity is approximately the same for both electrode
types, but lower in the mixed aqueous/organic solution than they are in the pure
aqueous medium. Both diamond types exhibited good run-to-run response
reproducibility with no evidence of electrode deactivation or fouling. The
response precision (2 - 6%) is identical for both electrode types and was
determined from repetitive injections of a 20 (M analyte concentration. The
concentration limit of quantitation was in the low to mid ppb range and
corresponds to a mass limit of quantitation ranging from 0.1 to 0.5 ng (SIN 2 3).
In general, the detection figures of merit for the microcrystalline and
nanocrystalline diamond films compare favorably with those previously reported
for an oxygen-terminated diamond electrodez'22 and are somewhat improved over

those reported for other electrodes.”25 In addition, the limits of quantitation for

113

the phenolic solutes in our HPLC-EC method are comparable to those reported

for EPA Method 1625.26

 

 

 

 

 

 

 

a b
c e
d
20nA
0 4 8 12 16 18
Minutes

Figure 5.7 Reversed-phase chromatogram for phenol and the chlorinated
phenols detected with amperometric detection using a microcrystalline diamond
thin-film electrode. The concentration of each solute was 20 pM. Injection volume
= 20 uL. Detection potential = 1150 mV vs. AglAgCl. Flow rate = 0.6 mUmin. C13
column (X-Terra, 5 pm particle size, 4.6 mm x 150 mm). The compounds are (a)
phenol, (b) 2-CP, (c) 3-CP, (d) 4-CP, and (e) PCP. A two-step gradient elution
was used. Initially, a 50 mM phosphate buffer, pH 3.5/acetonitrile (65:35, v/v)
was used and was switched to a 20:80 mixture after 10 min. Electrode geometric
area = 0.11 cmz.

114

 

 

 

160 Q phenol A
O 4-CP .,
A El 24:9
<t120 ‘ A 3.0,,
c O
I: A PCP ,
C 80
0 A
h
'5 40 r A
-< / 6
U ,/
A? ‘
0 / I

 

o 20 Jo 6'0 go
Concentration (pM)

 

160< <> phenol
0 24:? B

E] 4.09
I 3-CP o
A PCP

Current (nA)
8 a

«b
o
1

 

\

\

 

o
l

 

T T

20 40 so so
Concentration (pM)

C

Figure 5.8 Calibration curves for phenol and chlorinated phenols at a (A)
microcrystalline and (B) nanocrystalline diamond thin-film electrode. The
separation was performed using the same conditions as in Figure 5.6. The
detection potentials used were 1100 and 1200 mV vs. AglAgCI for A and B,
respectively. Injection volume = 20 uL. Electrode geometric area = 0.11 cmz.

115

Table 5.4 HPLC-EC Detection Figures of Merit for Phenol and the
Chlorinated Phenols at a Microcrystalline Diamond Thin-Film Electrode

 

 

 

 

 

 

 

Analyte Linear Sensitivity Response LOQ‘I LOQ
Dynamic (nAIpM) Precision (pM) (ng)
Range (pM) (n = 7)

phenol 0.1 — 80 1.9 1 0.1 3.8 0.1 (S/N = 4) 0.19
r2 = 0.9983 (9 ppb)

2-CP 0.1 - 60 2.0 1 0.4 3.2 0.1 (SIN = 9) 0.26
12 = 0.9994 (13 ppb)

3-CP 0.1 — 60 1.0 1 0.2 2.9 0.1 (SIN = 4) 0.26
12 = 0.9991 (13 ppb)

4-CP 0.1 — 60 1.4 1 0.3 4.4 0.1 (SIN = 4) 0.26
:2 = 0.9984 (13 ppb)

PCP 0.1 - 80 0.9 1 0.1 3.8 0.1(SIN = 6) 0.53
r2 = 0.9902 (27 ppb)

 

 

 

 

 

 

 

Table 5.5 HPLC-EC Detection Figures of Merit for Phenol and the
Chlorinated Phenols at a Nanocrystalllne Diamond Thin-Film Electrode

 

 

 

 

 

 

 

Analyte Linear Sensitivity Response LOQ‘ LOQ
Dynamic (nAIuM) Precision (pM) (ng)
Range (nM) (n = 7)

phenol 0.1 — 80 2.0 1 0.5 5.2 0.1 (SIN = 4) 0.19
r2 = 0.9974 (9 ppb)

2-CP 0.1 - 60 2.3 1 0.4 4.1 0.1(SIN = 10) 0.26
r2 = 0.9946 (13 ppb)

3-CP 0.1 — 60 1.0 1 0.2 3.2 0.1(SIN = 4) 0.26
r2 = 0.9930 (13 ppb)

4-CP 0.1 — 60 1.4 1 0.3 5.1 0.1(SIN = 5) 0.26
:2 = 0.9990 (13 ppb)

PCP 0.1 — 80 1.0 1 0.1 5.6 0.1 (SIN = 6) 0.53
r2 = 0.9984 (27 ppb)

 

 

 

 

 

 

8limit of quantitation = lowest concentration injected and detected.

Note: The data are reported for 3 electrodes. Experimental conditions as in
Figure 5.6. The detection potential was 1.1 and 1.2 V for microcrystalline and
nanocrystalline films, respectively.

116

 

5.2.3. Soil Sample Analysis

The utility of the HPLC-EC method for the analysis of chlorinated phenols
in a soil sample was demonstrated. A 10 9 soil sample containing at least 60
different organic pollutants was obtained from Absolute Standards, Inc (Hamden,
CT). The sample was prepared according to principles outlined in the “National
Standard for Water Proficiency Testing Studies, Criteria Document” — EPA
12/30/98. The sample came with a list of concentration ranges for each pollutant
present. No information was provided about the mineralogy of the soil or how the
sample was prepared. 2-chlorophenol (2-CP) was one of the electroactive
contaminants present and, therefore, the soil was analyzed for it. The sample
was prepared for analysis in the following manner. First, all 10 g of the soil
sample was added to 40.0 mL of acetonitrile and ultrasonically mixed for 30 min.
This served to extract 2-CP into the liquid phase. Second, the sonicated sample
was then centrifuged for 30 min, and the liquid phase was carefully decanted into
a 250 mL round bottom flask. Third, the solvent was evaporated on a rotatory
flash evaporator down to a volume of 8 - 10 mL. The sample was again
centrifuged and the extract designated as the crude sample. Fourth, solid phase
extraction was carried out using a 3 mL SuperCIean LC-18 cartridges (Supelco,
491 mzlg). The C-18 catridges were conditioned with 6.0 mL of methanol
followed by 6.0 mL of ultrapure water. After conditioning, 2.0 mL of the crude
sample was mixed with 4.0 mL of ultrapure water in a clean vial. The sample was
passed through the solid phase (vacuum-assisted) - a process that took

approximately 10 min. The sample was finally eluted off the solid phase with 6.0

117

mL of a solvent mixture containing dichloromethane, hexane, and acetonitrile
(50:47:13 v/v). The solvent was evaporated by passing N2 down to a final volume
of ca. 1.0 mL. The concentrated sample was then adjusted to 4.0 mL with the
addition of the mobile phase solution (65:35 (WV), 50 mM phosphate buffer, pH
3.5/acetonitrile). This sample was then injected onto the HPLC column for

separation and detection.

Figure 5.9 shows a chromatogram for the sample extract. Amperometric
detection was accomplished at 1.1 V using a microcrystalline diamond electrode.
Elution was performed using the two-step gradient. A 65:35 50 mM phosphate
buffer, pH 3.5/acetonitrile mobile phase was used initially to elute the more polar
analytes, and this composition was adjusted to 20:80 at ca. 10 min mark to elute
the more nonpolar solutes. The chromatogram reveals the presence of 21
different peaks, each of which corresponds presumably to an electroactive
solute. No efforts were made to identify the solutes. Most importantly for this
work, 2-CP elutes at 6.2 min and is labeled on the chromatogram. Peak
identification was performed by retention time matching. The peak represents the
raw signal for the unknown concentration of 2-CP in the soil sample. A standard
calibration curve was generated between 0.1 to 80 (M for quantitation of 2-CP. A
linear response was observed with a regression correlation coefficient of 0.9963.
Standard addition is the method of choice for quantifying 2-CP in this complex
matrix, but we did not have enough of the soil to conduct this type of calibration.

Based on the calibration curves generated from a series of standard solutions,

118

the concentration of 2-CP in the soil was determined to be 2650 pg (2-CP)lkg
(soil). The supplier of the contaminated soil sample specified a low and high
acceptance value of 2643 and 5983 ug/kg, and an assigned value of 4313 ug/kg
for 2-CP. The experimentally determined value is lower than assigned value and
is on the low end of the acceptance range. However, given that specific details
regarding the sample preparation and the length of time the sample was stored
were not readily available and the presumption that 100% recovery was
achieved, we suppose the experimentally determined value reflects the true

value of the 2-CP present.

 

 

 

 

 

 

 

Z-CP
I 20 nA
0 io 90 30 40
Minutes

Figure 5.9 Reversed-phase chromatogram for the separation and amperometric
detection of 2-CP in a contaminated soil sample at a microcrystalline diamond
thin-film electrode. Injection volume = 20 (IL. Detection potential = 1150 mV vs.
AglAgCI. Flow rate = 0.6 mUmin. C13 column (X-Terra, 5 pm particle size, 4.6
mm x 150 mm). A two-step elution was used. Initially, a 65:35 (WV) 50 mM
phosphate buffer, pH 3.5/acetonitrile was used and was switched to a 20:80
mixture at the 10 min mark. Electrode geometric area = 0.11 cmz.

119

5.3. Conclusions

Electrically conducting microcrystalline and nanocrystalline diamond thin-
fllm electrodes Show promise for the amperometric detection of chlorinated
phenols, in either flow injection analysis (FIA) or high performance liquid
chromatography (HPLC). Their low background current, good performance for
phenol oxidation, and low susceptibility to fouling make them useful for phenol
and chlorinated phenol analysis in water samples. Both types of diamond
outperformed GC, which exhibited short-lived activity as a consequence of
electrode deactivation. For example, the relative standard deviation of the peak
height for 20 injections of 3-CP was 2.2% for nanocrystalline diamond and 20.7%
for glassy carbon. Low limits of quantitation were observed for both diamond
types for all the phenols. For example, a limit of quantitation of 50 nM or 6 ppb
(20-uL inj.) was achieved for 2-chlorophenol with both diamond types in FIA-EC.
When coupled to HPLC, both diamond types gave good detection figures of merit
for all the chlorinated phenols. For example, the limit of quantitation for all the
analytes was 0.1 (M. The preliminary result for the soil sample illustrates the
potential for using this method to monitor phenol and chlorinated phenols in real

samples.

120

5.4. References

1.

10.

11.

12.

13.

14.

15.

Muna, G.W.; Tasheva, N.; Swain, G.M.; Environ. Sci. Technol. 2004, 38,
3674.

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

Rondrigo, M.A.; Michaud, P.A.; Duo, |.; Panizza, M.; Cerisola, G.; Ch.
Comninellis, G.; J. Electrochem. Soc. 2001, 148, D60.

Hagans, P.L.; Natishan, P.M.; Stoner, B.R.; O’Grady, W.E. J.
Electrochem. Soc. 2001, 148, E298.

Zhi, J.-F.; Wang, H.-B.; Nakashima, T.; Rao, T.N.; Fujishima, A.; J. Phys.
Chem.B 2003, 107, 13389.

Toth, K.; Stulik, K.; Wlodzimierz, K.; Zsofia, F .; Emo, L.; Pure Appl. Chem.
2004, 76, 1119.

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

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

ln http://www. Iachatinstryments.com/pmdycts/gcfia/FIAPfimerfiQ.,
accessed April 2005.

LaCourse, W.R.; Pulsed Electrochemical Detection in High Performance
Liquid Chromatography, 1997.

Wang, J.; Analytical Electrochemistry, 2nd ed.; Wiley: New York, 2000.
Dryhurst, G.; NATO ASI Sen, Ser. A 1994, 267, 171.

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

Granger, M.C.; Witek, M.; Xu, J.; Wang, J.; Hupert, M.; Hanks, A.;

Koppang, M.D.; Butler, J.E.; Lucazeau, G.; Mermoux, M.; Strojek, J.W.;
Swain, GM. Anal. Chem. 2000, 72, 3793.

121

16.
17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

Gattrell, M.; Kirk, D.W.; J. Electrochem. Soc., 1993, 140, 1534.
Gattrell, M.; Kirk, D.W.; J. Electrochem. Soc., 1993, 140, 903.

Polcaro, A.M.; Vacca, A.; Palmas, S.; Mascia, M.; J. Appl. Electrochem.
2003, 33, 885.

Codognoto, L.; Machado, S.A.S.; Avaca, L.A.; J. Appl. Electrochem. 2003,
33, 951 .

Canizares, P.; Garcia-Gomez, J.; Saez, C.; Rodrigo, M.A. J. Appl.
Electrochem, 2003, 33, 917.

Zhi, J.-F.; Wang, H.-B.; Nakashima, T.; Rao, T.N.; Fujishima, A.; J. Phys.
Chem.B 2003, 107, 13389.

Prado, C.; Murcott, G.G.; Marken, F.; Foord, J.S.; Compton, R.;
Electroanalysis 2002, 14,975.

Agui, L.; Serra, B.; Yanez-Sedeno, P.; Reviejo, A.; Pingarron, J.M.;
Electroanalysis 2001, 13, 1231.
Liu, X.; Frank, H.; J. High Resol. Chromatogr. 1998, 21 , 309.

Dressman, S.F.; Simeone, A.M.; Michael, A.C.; Anal. Chem. 1996, 68,
3121.

Standard Methods for the Examination of Water and Wastewater, 18th
Edition, US Environmental Protection Agency, 1992

122

CHAPTER 6

Chlorinated Phenol Analysis by Capillary Zone
Electrophoresis with Amperometric Detection

using a Diamond Microelectrode

6.1. Introduction

Capillary zone electrophoresis (CZE) is a powerful separation method for
a variety of analytes, offering the advantages of high separation efficiency, short
analysis time, and low reagent consumption.“2 Separation is based on the
differential rates of migration of charged solutes in an applied electric field. Small
inorganic ions and organic species, as well as large biomolecules such as DNA
and proteins, can be separated by this method. During the past two decades,
great attention has been given to different electrophoretic methods in which the
separation occurs in a small bore capillary. ln capillary zone electrophoresis
(CZE), separation is performed in a solvent-filled glass or fused-silica capillary
tube across which a high voltage (e.g., 20 W) is applied. Solutes are transported
across the column by two mechanisms. The first is electroosmosis, which refers

to the movement of solution due to an applied potential gradient.3 The second is

123

electromigration, which involves the movement of a charged species in an
electric field. When charged solutes are introduced into the capillary, they

migrate at rate that depends on their charge-to-size ratio.

Adaptation of a detection method to a small volume separation method,
like CZE, is a challenge. The detection method must not introduce additional
band broadening while providing high sensitivity, high spatial resolution, and
rapid response time.4 Several detection methods have been used in conjunction
with CZE including absorption and fluorescence spectrometry, electrochemistry
with a microelectrode, and mass spectrometry.2"'”‘2 Reviews covering various
detection methods have appeared in the literature."‘10 Electrochemical detection
(EC) is often coupled with CZE and is attractive for a number of reasons. Some
advantages of electrochemical detection are high sensitivity, which is not
compromised by miniaturization, and rather simple and inexpensive
instrumentation?“12 Two technical challenges with electrochemical detection are
(i) decoupling the separation voltage from the electrode response (ii)
reproducible positioning of the working electrode at the end of the separation
capillary in end-column detection. Inadequate decoupling gives rise to noise at
the electrochemical detector, which leads to decreased sensitivity.”12 Another
important issue with electrochemical detection is the working electrode and its
electrochemical properties. Due to the very small sample volume injected, a
microelectrode must be used to achieve high detection sensitivity. Several

different electrode materials and geometries have been used with metal (Pt) and

124

 

carbon fiber microelectrodes being the most popular.4 A major issue, as
discussed previously, is electrode deactivation and fouling due to the irreversible
adsorption of reaction products and intermediates, and the resulting need for
electrode pretreatment to reactivate the surface. Carbon fiber microelectrodes, in
particular, are susceptible to strong adsorption of polar aromatic molecules, such
as chlorophenols, in part, because of the presence of carbon-oxygen
functionalities that make the surface polar and hydrophilic in nature and the

extended n-electron system.

Boron-doped diamond microelectrodes have recently been employed in
CZE, specifically for the amperometric detection of nitroaromatics,
organophosphate agents, phenols and catecholamines."""“'15 Diamond exhibits a
reproducible and stable oxidation response for these analytes without any
surface pretreatment. The application of a diamond microelectrode for the direct
and indirect amperometric detection of chlorinated phenols in CZE is reported
herein. The objectives of this work were (i) to verify that the boron-doped
diamond microelectrode can be used to detect these priority pollutants in CZE,
(ii) to demonstrate that the microelectrode is resistant to fouling and provides a
sensitive, reproducible and stable oxidation response for the chlorinated phenols,
and (iii) to compare the detection figures of merit obtained by direct and indirect

amperometric detection.

125

6.2. Results and Discussion
6.2.1. Surface Morphology and Microstructure of the Diamond
Microelectrodes
Figure 6.1 a and b show SEM images of a sharpened Pt wire coated with
a thin layer of diamond. The SEM image reveals a diamond ovelayer consisting
of randomly oriented, well-faceted diamond crystallites with base diameters from
1 to 3 pm. A wide crystallite size distribution exists.13 Films grown on Pt exhibit
more morphological defects than do films deposited on Si substrates under

similar conditions.13

 

Figure 6.1 SEM images of an electrochemically etched 76-pm sharpened
platinum wire coated with a polycrystalline diamond film. Images a and b show
the film at different magnifications.

126

 

 

 

 

 

 

2.5 _. 1332cm'
33‘ 2.0 -
I:
D
E 1.5 -
E
g 1.0 _
a

 

1000 l 1250 I 1500 1750 l 2000 ' 2250
Raman shift (cm‘)

Figure 6.2 Raman spectrum for a microcrystalline boron-doped diamond thin film
deposited on a 76-pm platinum wire. Integration time = 10 5. A4”, = 532 nm.

Figure 6.2 shows a Raman spectrum for a diamond film grown on a
sharpened 76-pm diameter Pt wire. The spectrum has an intense one-phonon
diamond line at 1332 cm". The full width at half-maxima (fwhm) is ~ 12 cm". By
comparison, the Iinewidth for single-crystal diamond in the same spectrograph is
2 cm". The larger fwhm for the polycrystalline film is indicative of a higher defect
density as, to a first approximation, the peak width is inversely related to the
phonon lifetime. This lifetime is reduced because of phonon scattering at the
grain boundaries and defects.13 There is minimal scattering intensity in the region
near 1500 cm", indicating that the film contains little nondiamond, spz-bonded

carbon impurity.

127

6.2.2. Basic Voltammetrlc Response

The diamond microelectrode was initially characterized by cyclic
voltammetry. A background voltammetric i-E curve in 1 M KCl at 0.1 V/s, is
shown in Figure 6.3A. Voltammetric features characteristic of Pt are not observed
which is exemplified by the large overpotential for hydrogen evolution. The
absence of these features indicates that cracks and pinholes are absent in the
diamond overlayer. The increase in anodic current at ca. 1500 mV is due to the
onset of chlorine evolution, which occurs on diamond with relatively low
overpotential as compared to oxygen evolution. The potential region between -
1000 and 1500 mV is featureless and characterized by a low and stable current.
Figure 6.3 B shows a cyclic voltammetric i-E curve for 1 mM Fe(CN)6"3"4 in 1M
KCl at 10 mVls. A sigmoidally shaped i-E curve is seen with an Em of 285 mV
and a limiting current, inm, of ca. 40 nA. The curve shape remained unchanged
with repeated cycling. The size of the electrode was such that a sigmoidal
response was only observed at low scan rates. The exposed area was
determined from the limiting current, ium, at the conicaIIy-shaped diamond
microelectrode, which is described by a cylindrical and a conical term according

to the equation16'17

-cone

qss +11 (1)

- __ -cylinder
llim —l

For a quasi-steady-state current (long times), it holds that the cylindrical term,

icylinder

gs, , is given by‘6-17

128

icylinder _ anADC
q” _ r In 2' (2)

-cone

and the conical term, 135 .by16

iffne = 4nFDCr(1+ qI-IP) (3)

where n is the number of electrons transferred per equivalent, F is the Faraday
constant, D is the diffusion coefficient (cmzls), C is the bulk concentration of the
analyte (mol/cma), r is the is the cone radius (cm), 1 = 4Dt/r2, q = 0.30661, p =
1.14466 and, H is the aspect ratio h/r where h is the height of the cone (cm). The
measurement period, t (S), is the time of the forward voltammetric scan. The
cone dimensions were determined from the SEM image. The surface area of a
cone is related to its aspect ratio according to the equation,
A = m2 (H+1)"2

For all the calculations, diffusion coefficients of 6.7 x 10'6 cm2/s for Fe(CN)5’3"‘,
aspect ratio of 4.0, and a time t (s) of 30 s were used.13 The theoretical current
calculated from just the cone dimensions (equation 3) was calculated to be 30
nA. This is slightly lower than the experimental value of 40 nA (see Figure (6.33).
This difference is due to the fact that some length of the cylindrical portion is also
exposed to the solution. Due to the manner in which the polypropylene insulation
layer is applied to the microelectrode, it is difficult to reproducibly control the
overall electrode length and, therefore, the amount of cylinder exposed. When
the measured current exceeded the theoretical value, a portion of the cylindrical

part was assumed exposed. In this case, the length of the cylindrical part was

129

 

0.10

 

0.06 _

0.02 _

Current [pA]

 

Ki
-0 02

———————>

1

 

 

. . 1
-1500 -500

Potential [mV vs. AglAgCl]

l
500

1
1 500

 

0.04 — B

0.03 _

0.02 .

Current [pA]

0.01 _

0.00

 

 

 

 

-100 * 160

300

500

Potential [mV vs. AglAgCl]

Figure 6.3 (A) Background cyclic voltammetric i-E curve for a boron-doped

diamond microelectrode in 1 M KCl. Scan rate

100 mV/s. (B) Cyclic

voltammetric i-E curve for 1 mM Fe(CN)5'3’4 in 1 M KCl at a boron-doped
diamond microelectrode. Scan rate = 10 mVls.

130

calculated using equations 1 and 2, in which A = 2an. Nominally, the diamond
microelectrode had a calculated surface area (cone + cylinder) of 2.2 (:t 0.7) x 10'

4 cm2.

A cyclic voltammetric i-E curve for 50 uM 2—chlorophenol in 10/20 mM
borate/phosphate buffer, pH 8.4, at a diamond microelectrode is shown in Figure
6.4. The potential sweep rate was 100 mVls. The faradaic current for the phenol
oxidation commences at ca. 0.6 V and reaches a pseudo-steady state value of
ca. 10 nA at potentials between 0.8 and 1.2 V. Beyond 1.2 V, the current
increases due to the onset of oxygen evolution. The phenol oxidation occurs near
the onset of oxygen evolution; therefore, both direct and indirect (via OH-)
oxidation reactions are believed to occur.18 No other redox peaks develop in the
voltammogram during subsequent scans between -0.4 and 1.6 V. This is different
from what was observed for a planar diamond thin-film electrode when mounted
at the bottom of an electrochemical cell (see Chapter 4).18 Presumably, the
quasi-steady state diffusional flux at the diamond microelectrode transports the
oxidation reaction products away from the surface before adsorption can occur

and they are not detected.

131

 

 

 

 

 

 

 

40~
30~

3?

5 20-

g-I

C

2 10-

h

8
0-___: /
_10 L 1 1 1

-0.4 0.0 I 0.4 ‘ 0.8 ‘ 1.2 ' 1.6
Potential [V vs. AglAgCl]

Figure 6.4 Cyclic voltammetric i-E curve for 50 (1M 2-chlorophenol in a 10/20 mM
mixed-borate/phosphate run buffer, pH 8.4, at a boron-doped diamond
microelectrode. Scan rate = 100 mVls.

6.2.3. Direct Amperometric Detection of Chlorinated Phenols

Decoupling the separation voltage from the detection electrode signal is a
key issue for CE-EC. Decoupling refers to the prevention of the electrophoretic
current from returning to ground via the detection electrode. Increased
background current and noise, greater response instability and potential shifts
are observed when there incomplete decoupling.19 Two approaches are routinely
used to isolate the detection electrode from the separation voltage: end-column
and off-column detection. In end-column detection, the electrode is placed just

outside the separation capillary. In this arrangement, the electrode is positioned

132

 

outside of the separation electric field. In off-column detection the electrode is
isolated from the separation voltage by means of a “decoupler”.‘9'24 We isolated
the diamond microelectrode by positioning it just outside the end of the capillary
column (so called end-column detection). The distance between the capillary and
the microelectrode was adjusted by positioning the capillary, which was mounted
in the detection cell normal to the working electrode. The distance between the
end of the capillary and the microelectrode was kept as short as possible (by
eyesight) in order to minimize dilution effects and band broadening.25 We
estimated that the separation distance was on the order of 20 - 30 pm. If this
distance is too great then the solute zone gets diluted leading to reduced peak
heights, broader peaks, and higher detection limits results. The capillary-
electrode spacing also influences the separation efficiency. By maintaining a low
separation current and an adequate distance between the microelectrode and
capillary, the electrode response was relatively unaffected by the separation
voltage and acceptable signal-to-noise ratios and minimal potential shifts were

obtained.

CE-EC experiments were performed using a mixed 10/20 mM
borate/phosphate run buffer at pH 8.4, a 76-cm-long fused silica capillary (27 um
i.d.), and a separation voltage of 20 W (263 Vlcm). This run buffer was selected
based upon results from numerous optimization runs in which the buffer pH and
composition were varied. Solutes were injected electrokinetically for 3 s at 8 W.

The injected volume of an electroosmotic flow marker, methanol, was calculated

133

 

to be 1.1 nL. Based on the migration time and column length, the EOF was
calculated to be 6.2 x 10‘4 cmst. After application of the detector potential, the
separation voltage across the column was applied. The electrode current was
allowed to stabilize before making an injection. In general, the background
current stabilized within 3 - 4 minutes after detector tum-on. This is a useful

material property.

Hydrodynamic voltammetric i-E curves were recorded for each of the
chlorinated phenols in order to determine the optimum detection potential. Figure
6.4 shows i-E curves for 20 (N concentrations of each pollutant. A potential
increment of 50 mV was used. Each datum shown is the average current value
for at least three injections. The oxidation current for each solute commences
around 800 mV and reaches a near limiting value at ca. 1050 mV. A more or less
sigmoidally-shaped response is observed for each solute. This suggests that the
oxidation current is limited by the flux of analyte to the electrode surface. The
limiting current at the diamond microelectrode in this cross-flow geometry is
given, at least approximately, by the following expression26

i, = nFCKmAU"
in which i, is the mass-transfer limited current, n is the number of electrons
transferred per equivalent, C is the solute concentration, A is the electrode area,
Km is the mass transfer coefficient, and U is the average volume flow rate. a is a
constant that ranges from 0.33 to 1.0. The limiting current for each of the

chlorinated phenols differs by a maximum of a factor of three even though the

134

 

 

4-CP
A _<>_ 3-cp
< 400 " _@_ 2-CP
E; + 2,4-DCP
PCP
‘5 300 _ +9
L-
h
5 200 _
100 _
A/a

 

 

 

O —--—-—= 1 1 . 1 . 1 .
750 800 900 1000 1100 1200
Potential [mV vs. AglAgCI}

Figure 6.5 Hydrodynamic voltammetric i-E curves for the Six chlorinated phenols
at a boron-doped diamond microelectrode recorded in a 10/20 mM
borate/phosphate run buffer, pH 8.4. Separation voltage = 20 kV. Capillary = 27
um id. x 76 cm long. Injection = 8 W for 3s. Solute concentration = 20 (M.

Table 6.1 pK. Values and the Elution Order for Phenol and Different
Chlorinated Phenols.

Compound pK.” Elution order
Phenol 10.0 1
4-chlorophenol 9.43 2
3-chlorophenol 8.97 3
2-chlorophenol 8.52 4
2,4—dichlorophenol 7.89 5
2,4,6-trichlorophenol 6.1 6
Pentachlorophenol 4.74 7

135

solution concentration, the number of electrons transferred (assumed to be 1 for

oxidation to phenoxy radical), and the hydrodynamic flow conditions were the
same for all the solutes. There are two factors that lead to the differences in
limiting current. One is the diffusion coefficient for each, which we did not
measure nor did we find values for in the literature. The other, and the one most
influential, is the different mass of each solute (i.e. concentration) loaded onto the
column during the electrokinetic injection. The mass loading depends on the
electrokinetic injection conditions, the solution concentration, and the
electrophoretic mobility of the solute. In the present case, all the solutes
possessed some degree of negative charge because the run buffer pH was 8.4
and the pKa values for each range from 4.74 to 10. Since the solutes are
negatively charged, the direction of their electrophoretic movement opposes the
electroosmotic flow. Therefore, the solute with the lowest loading should be the
one with the most negative charge/size ratio, that is the one with the lowest pKa,
and the solute with the highest loading should be the one with the least negative
charge/size ratio, that is the one with the highest pKa. The limiting currents do in
fact follow this trend with the magnitude decreasing in order of decreasing pKa: 3-
CP (pKa = 8.97) > 4-CP (PKa = 9.43) > 2-CP (pKa = 8.52) > 2,4-DCP (pKa = 7.89)
> 2,4,6-TCP (pKa = 6.1) > PCP (pKa = 4.74).27 The observance of sigmoidally-
Shaped hydrodynamic i-E curves at the microelectrode is different from our
previously reported work on the electrooxidation of chlorinated phenols at boron-
doped microcrystalline and nanocrystalline diamond thin-film electrodes.18 In that

work, sigmoidally-shaped hydrodynamic i-E curves were not observed for all the

136

pollutants (e.g., phenol, 2-chlorophenol, 3-chlorophenol, and 4-chlorophenol) in
FIA-EC, with exception of PCP, and we concluded that the maximum oxidation
current was under mixed control by direct and indirect (anodic oxygen transfer
reactions) electrooxidation reactions and mass transport. The i-E curves for 2-
chlorophenol, 3-chlorophenol, and 4-chlorophenol in Figure 6.5 also show a
current increase at 1150 mV. This is due to the onset of oxygen evolution which
begins above 1100 mV (background hydrodynamic i-E curves not Shown). Based
on the hydrodynamic voltammograms, the optimum detection potential for all the
pollutants was determined to be 1050 mV. The normal background current at
this potential was ~ 15 pA and the peak-to-peak noise was 2 to 3 pA. The
background noise was unaffected by the separation voltage provided there was
sufficient distance between the electrode and the end of the capillary (i.e.,

adequate decoupling).

Figure 6.6 shows a typical electropherogram for an equimolar mixture (20
(M) of all seven pollutants. All peaks are well resolved from one another but
each has an asymmetric shape (tailing). Separation of all seven solutes is
accomplished in less than 14 min and all migration times were very reproducible
during an 8-h run period. The an buffer pH influences an electrophoretic
separation by affecting the analyte charge, thus, its electrophoretic mobility and
the electroosmotic flow (EOF). Furthermore, selectivity, resolution, and peak
shape are also all pH-sensitive. Small differences in solute pKa (e.g., 2-

chlorophenol, 3-chlorophenol, and 4-chlorophenol) are enough to enable

137

separation by CE. In the present case, separation was performed at an optimum
pH of 8.4. Below this pH, Significant peak overlap was observed for phenol, 4-
chlorophenol, and 3-chlorophenol. Above this pH, 2,4—dichlorophenol, 2,4,6-
trichlorophenol, and pentachlorophenol were poorly resolved from one another.
The electrophoretic mobility of a solute is governed by its charge/size ratio as
well the concentration and viscosity of the buffer. At pH 8.4, all the chlorophenols
possess some negative charge (deprotonated) and the elution order exhibits a
correspondence with the pKa (see Table 6.1), as expected. Phenol, having the
least negative charge (highest pKa), elutes first and pentachlorophenol, having
most negative charge (lowest pKa), elutes last. The direction of electrophoretic
movement of each solute is against the electroosmotic flow direction. The
differences in peak height are partly attributed to differences in the solute loading
during the electrokinetic injection. The peak tailing appears to be related to the
solute flow across the diamond microelectrode during migration, and not to on-
column effects. The migrating solute first encounters the end of the electrode as
it migrates off the column. This is followed by flow and diffusional spreading
across the length of the exposed electrode. Dilution of the solute occurs as it
travels across the fiber length so the current goes down. This influence was
confirmed by using diamond microelectrodes of differing lengths. The extent of
tailing increased with increasing electrode length. Shin et al. also observed some
peak tailing when using a diamond microelectrode for the detection of
catecholamines.14 Table 6.2 presents a summary of the detection figures of merit

for the chlorinated phenols. The data are reported at the 95% confidence level.

138

1200

 

 

 

 

 

 

 

P 3

900 _ 4
2 2
Q
:: 600 _
E»
I- T 1
8 300 _

o L 0L1

 

o—‘Too ‘ 400 ‘ Mo 800 1 00

Time (s)

Figure 6.6 A typical electropherogram for phenol and six chlorinated phenols
using direct amperometry recorded in a 10I20 mM borate/phosphate run buffer,
pH 8.4. Separation voltage = 20 kV. Capillary = 27 pm id. x 76 cm long.
Detection potential = +1050 mV (vs. AglAgCI). Injection = 8 W for 3S. Solute
concentration = 20 (M.

The reproducibility of the electrode response was determined from 9 injections of
a solution containing all seven pollutants (20 uM each) prepared in the run buffer.

The separation efficiency ranged from 85,000 to 302,000 plates/m

_ 41.7(tm/wm)2
b/a+1.25

 

(N

where wo,1 is the width of the peak at 10% from the base,

b/a is the asymmetry factor). The linear dynamic range for each was at least 3
orders of magnitude with correlation coefficients greater than 0.993. The

sensitivity for all the pollutants ranged from 6 to 16 pA/uM. The minimum

139

 

Table 6.2 Detection Figures of Merit for Phenol and the Chlorinated Phenols
ln Direct Amperometry Coupled with Capillary Electrophoresis.'

 

 

 

 

 

 

 

 

 

Analyte LDR Sensitivity LOQ Estimated Response Efficiency
(nM) (pAIuM) (nmolIL) mass precision (plates/m)
LOQ (fg) (n = 9)
RSD%
Phenol 0.5 — 100 10.7 1 7.3 500
r2 = 0.994 (SIN =3) 10 5.6 1 0.2 302,000
(47 ppb)
2-CP 0.1 - 100 100
r2 = 0.999 14.7 1 4.2 (SIN = 3) 14 4.7 1 0.5 133,000
(13 ppb)
3-CP 0.1-100 16.3141 100
r2 = 0.998 (SIN = 4) 14 5.2 1 0.2 121,000
113 ppb)
4-CP 0.1 - 100 15.0 1 4.8 100
r2 = 0.993 (SIN = 4) 14 4.5 1 1.2 116,000
(13 ppb)
2,4- 0 5 - 100 9.9 1 3.1 500
DCP r2 = 0.999 (S/N = 3) 90 5.0 1 0.2 129,000
(82 ppb)
2,4,6- 0 5 - 100 5.9 1 3.9 500
TCP r2 = 0.998 (SIN = 3) 107 5.4 10.7 85,000
(99 ppb)
PCP 0 5 — 100 6.2 1 2.4 500
r2 = 0.997 (SIN = 3) 146 5.6 1 0.5 187,000
(133 ppr

 

 

 

 

 

 

 

aCE-EC parameters: E5") = 8 kV; tn] = 3 s; capillary dimensions = 27 um id. x 76

cm long; E53p = 20 kV; and Edet = 1.05 V. The data are based on the response of

3 microelectrodes and are reported at the 95% confidence level. LDR = linear

dynamic range. LOQ = limit of quantitation (the minimum concentration of analyte

injected and measured). The mass limit of quantitation was estimated from the

calculated injection volume for methanol, 1.1 nL. This was calculated using the

expression:

Vin] = (Ein) X tin) X Vcapillary)/Esep X tmig

concentration measured (SIN 2 3) was 100 nM (13 ppb) for 2-chlorophenol, 3-

chlorophenol and 4- chlorophenol, and 500 nM (47-133 ppb) for phenol, 2,4-

dichlorophenol, 2,4,6-trichlorophenol, and pentachlorophenol. The estimated

140

 

 

mass limits of quantitation were in the femtogram (fg) range, calculated using the
injected volume for methanol, 1.1 nL. Good peak height reproducibility was
observed for all analytes with a nominal value of 5.6 1 0.5%, or less. In summary,
the diamond microelectrode exhibited good electroanalytical performance with a
low peak-to-peak noise (2 to 3 pA); low and stable background current under the
electrophoretic conditions imposed; and sensitive, reproducible and stable
responses for all the chlorinated phenols without electrode deactivation (fouling)
during extended use. Additionally, no pretreatment was required for electrode
activation. The low and stable background current and noise allowed for
relatively low mass limits of detection to be achieved. These results are
consistent with two previous reports that diamond thin-film electrodes are useful
for the detection of chlorinated phenols with relatively high resistance to surface

fouling.""28

The detection figures of merit obtained for the diamond microelectrode
appear superior to other data reported in the literature, such as for UV detection.
2932 As far as electrochemical detection is concerned, there have been relatively
few reports of chlorinated phenol analysis presumably because of the electrode
deactivation and fouling. Several approaches have been employed to try and
minimize electrode fouling. Hilmi and co-workers used a Pt electrode modified
with electrodeposited Sn to detect chlorinated phenols. The modified Pt electrode
exhibited enhanced sensitivity as well as reduced electrode fouling by phenol

oxidation reaction products, as compared to bare Pt.33 The limits of detection

141

 

achieved were in the low micromolar range and the separation efficiencies
ranged from 88,000 to 145,000 plates/m. The response precision was 4%, or
less. In another report, the authors reported improved detection figures of merit
for chlorinated phenols using nonaqueous capillary electrophoresis with end-
column detection.34 The nonaqueous medium offered high sensitivity and
alleviated the electrode fouling problem. Concentration limits of detection were in
the nanomolar range with a response precision of 4%, or less. Wang and co-
workers utilized electrochemical detection with a thick screen-printed carbon
electrode in combination with microchip electrophoresis to analyze chlorinated
phenol solutions. The concentration limits of detection were in the 1 to 2 pM
range. The response precision was 6.6%, or less, for the chlorinated phenols.”
The group also reported on the use of a boron-doped diamond electrode with
microchip electrophoresis for the detection of nitroaromatics, organophosphate
agents as well as chlorinated phenols.15 No detection figures of merit were
reported for the chlorinated phenols, but diamond did provide higher sensitivity,
lower noise and better resistance to fouling for the nitroaromatics than did a
screen-printed carbon electrode. Gaitonde and oo-workers used an on-column
electrochemical detection with a modified carbon fiber for the analysis of
chlorinated phenols extracted from industrial effluents.36 The limits of detection

reported were in the low ppb range.

As mentioned before, amperometric detection of phenols is commonly

plagued by surface deactivation and fouling due to the formation of a polymeric

142

 

 

film on the electrode surface.“38 The mechanistic work by Gattrell and Kirk
reveals that the phenol oxidation reaction proceeds initially through the formation
of a phenoxy radical species, which can subsequently undergo radical-radical
coupling to form polymeric and oligomeric species. Follow-up direct and indirect
oxidation reactions, producing soluble products such as hydroquinone and
catechol, are also possible."3'2""°’7'38 No passivation of the diamond
microelectrode was observed, as it could be used in multiple measurements for
several weeks with little deactivation. This is a major attribute of diamond for

monitoring this important class of pollutants.

6.2.4. Indirect Amperometric Detection of Chlorinated Phenols

The general applicability of CE-EC can be expanded if indirect detection
methods are employed. Therefore, we also conducted a study of the indirect
amperometric detection of the same chlorinated phenols. Although direct
amperometry is a sensitive and stable means for solute monitoring, many solutes
are not electroactive and, therefore, cannot be detected. Indirect amperometric
detection circumvents this limitation and is based on the displacement of an
electroactive molecule added to the run buffer (electrophore) by the
electroinactive solute.39 Due to the need to preserve charge neutrality, the
electrophore in the buffer is displaced by the analytes zones. For analytes with
the same charge as the electrophore, displacement of the detectable ions results
in a lower level of the electrophore and a negative peak for the analyte results.39

Quantitation is made by measuring a negative peak current or charge. The extent

143

 

 

 

of electrophore displacement by the solute is known as the transfer ratio (TR). It
is defined as the number of moles of electrophore displaced by one mole of the
analyte.39 Indirect UVNis detection has often been utilized with CE, however,
there are only a few reports of indirect electrochemical detection in aqueous or
non-aqueous media.4042 Olefirowicz and Ewing were one of the first to
demonstrate the practicality of coupling indirect amperometric detection with CE
as they applied this method for the analysis of electroinactive cationic amino
acids and dipeptides. The authors used dihydroxybenzylamine (DHBA) as the
cationic electrophore. A stable background signal during its continuous oxidation
at the detection potential of 700 mV vs. SCE was observed. Zones of
electroinactive cations displaced the DHBA during the electrophoretic separation
and were detected as negative peaks. Mass detection limits were in the

femtogram (fg) range, or lower.40

In the present work, we tested the diamond microelectrode for the indirect
amperometric detection of chlorinated phenols and compared the detection
figures of merit with those obtained by the direct method. Several criteria must be
met by an ion or molecule to function as a suitable electrophore. For example,
the electrophore must have well—defined electrochemical behavior at the detector
electrode, the electrode must exhibit good electrochemical response stability for
the electrophore's redox reaction, and the electrophore must have a charge and
electrophoretic mobility similar to that of the solutes being separated. In our

studies, ferrocene carboxylic acid was chosen as the anionic electrophore.

144

Figure 6.7 shows a cyclic voltammetric i-E curve for 0.8 mM ferrocene carboxylic
acid in 0.01 M phosphate buffer, pH 8.1, at a diamond microelectrode. A well-
defined sigmoidal curve is observed at this slow scan rate with an E112 of 340 mV
(vs. AglAgCI) and a steady state oxidation current of 35 nA. Based on the curve

shape, a detection potential of 450 mV was selected.

 

0.029 __

P
o
A
to
1

0.009 _

Current (pA)

 

 

 

 

l l

 

'0'00-11 500

l
300
Potential [mV vs. AglAgCl]

00 T 100

Figure 6.7 Cyclic voltammetric i-E curve for 0.8 mM ferrocene carboxylic acid in
10 mM phosphate buffer, pH 8.1, at a boron-doped diamond microelectrode.
Scan rate = 5 mVls.

A series of experiments was first conducted to determine the appropriate
run buffer composition, ionic strength, and pH for the separation and indirect

detection of the chlorinated phenols. After much optimization, a run buffer

145

 

 

composition of 0.01 M phosphate buffer at pH 8.1 was found to be appropriate. A
typical electropherogram for three of these pollutants is Shown in Figure 6.8. The
electropherogram shows the separation of 2-chlorophenol, 3-chlorophenol, and
2,4-dichlorophenol obtained using a 30 um id. x 76 cm fused silica capillary. The
large background current, due to the oxidation of ferrocene carboxylic acid in the
run buffer, decreases when each chlorinated phenol migrates through the column
and reaches the electrode. The analytes were identified by injecting the individual
solutes and matching the migration times. It should be noted that all seven
solutes can be separated and detected, even though the electropherogram
Shows for only three. The first peak, A, in the electropherogram is the system
peak observed at a migration time corresponding to the rate of electroosmotic
flow.40 The origin of this system peak is unclear, but it is related in some manner
to the electrophore as the peak magnitude (anodic current) increased with
increasing electrophore concentration. The peak marked B represents a zone in
which the ferrocene carboxylic acid is virtually absent (i.e., vacancy zone). If a
species present in the run buffer is also contained in the sample at a different
concentration, then a vacancy will be created. The signal magnitude for this
vacancy will depend on the ratio of its concentration in the sample and the run
bufl‘er.“3'44 Thus, if a component is absent in the injected sample but present in
the run buffer, then the sample plug represents a vacancy and during application
of a separation voltage, this vacancy migrates as an individual zone and is
detected.43 This vacancy (negative peak) is present because all solute samples

were prepared in the run buffer without adding the electrophore.4o Proof that this

146

 

 

 

 

1 .24 [A

 

 

 

 

1- 3-CP
2- 2-CP

A 1,20 - 3-2,4-DCP
.1
5
'5 l

g 1.16-- F("VF*""41’
3
O 3 a

1.12-
1 ‘2

 

 

 

 

0 I 150 I 300 l 450 600 . 750 A
Time (s)

Figure 6.8 An electropherogram for 3-chlorophenol, 2-chlorophenol, and 2,4-
dichlorophenol using indirect amperometry recorded in a 10 mM phosphate run
buffer, pH 8.1. The electrophore was 0.8 mM ferrocene carboxylic acid. Detection
was accomplished with a boron-doped diamond microelectrode. Separation
voltage = 20 kV. Capillary 30 pm id. x 76 cm long. Detection potential = 450 mV
(vs. AglAgCI). injection = 8 W for 3 s. Solute concentrations = 200 (M.

peak is associated with the electrophore vacancy is the fact that the peak
magnitude increased with increased concentration of ferrocene carboxylic acid in
the run buffer. The peak shapes for the solutes are sharp and narrow with no
evidence of tailing or fronting. This is consistent with the electrophore having a
mobility similar to those of the solute.“"6 Peak fronting or tailing can occur when
the electrophoretic mobilities of the analyte and electrophore are significantly

different.47 Ferrocene carboxylic acid was found to have an electrophoretic

147

mobility that is similar to the mobilities of the chlorinated phenols under the
electrophoretic conditions used. This is important in order to maximize sensitivity
and minimize dispersion effects.“47 Electromigration dispersion effects are
related to the relative mobility differences between a solute and the added
electrophore. Tailing results when the conductivity within the zone is lower than
in the electrolyte, and peak fronting results when the reverse happens.46 Proper
choice of the electrophore provides the best peak shapes, maximizes the transfer
ratio, and minimizes competitive displacement.45 The lack of tailing due to the
Shape of the microelectrode as observed in direct measurements, could be
possibly due to the fact that detection here is not due to the direct oxidation of the

analyte but rather a displacement reaction.

A key parameter in indirect amperometric detection is the dynamic reserve
(DR), which is a dimensionless parameter defined as the ratio of the background
signal magnitude to the noise in that signal. The DR is one factor controlling the

minimum concentration (Cl ) theoretically detectable for a given set of

im

conditions. C is defined as

lim

C

m

Clim =
(DR)(TR)
in which Cm is the concentration of the electrophore and TR is the displacement

).39

ratio (transfer ratio The DR limits the fractional change in the background

signal that can be recognized. The greater the DR, the lower theCfim, For the

chlorinated phenols, the DR was determined by measuring the steady-state

148

current for the electrophore at 450 mV and dividing it by the corresponding peak-
peak noise (~ 3 - 4 pA) at concentrations from 50 pM — 1 mM. There was a clear
dependence of DR on the electrophore concentration up to 0.8 mM. As the
electrophore concentration increased, the DR increased from 150 at 50 (M to
1530 at 0.8 mM. In addition, as the concentration of the electrophore increased,
the magnitude of displacement by the 2-chlorophenol, 3-chlorophenol, and 2,4-
dichlorophenol also increased. For example, with 0.5 mM ferrocene carboxylic
acid, the magnitudes of the 2-chlorophenol, 3-chlorophenol, and 2,4-
dichlorophenol were 37, 49, and 14 pA, respectively. When the concentration of
the electrophore was increased to 0.8 mM the negative peak magnitudes
increased to 102, 161, and 40 pA, respectively. Above this electrophore
concentration, no addition increase in the chlorinated phenol signal was seen.
This could be due to the high background signal resulting from the electrophore.
This makes it more difficult to measure changes in the background signal due to
displacement by a solute. Thus, 0.8 mM was determined to be the optimum
electrophore concentration for the indirect detection. It’s worth noting that DR is
related to Cm so decreasing the electrophore concentration will not necessarily

improve the detection limit because DR is simultaneously reduced.47

The separation efficiency and the detection figures of merit for 2-
chlorophenol, 3-chlorophenol, and 2,4-dichlorophenol were evaluated. Table 6.3
contains a summary of the data. The peaks for these three pollutants were well-

resolved with separation efficiencies of 101,000, 224,000, and 69,000 plates/m

149

for 2-chlorophenol, 3-chlorophenol, and 2,4-dichlorophenol, respectively. A linear
dynamic range for all three is seen to range between 30 and 600 (M with
correlation coefficients greater than 0.996. The sensitivities for all three solutes
are between 300 and 900 meM and are lower than those observed in direct
amperometry. The concentration limits of quantitation were 30 (M (4 ppm) for 2-
and 3-chlorophenol, and 50 (1M (8 ppm) for 2,4—dichlorophenol at a SIN 2 3.
These values are higher than those observed in direct amperometry. The mass
limits of quantitation, based on a 1.1 nL injection volume, are from 5 to 10 pg.
The electrode response reproducibility was excellent for 10 injections with values
less than 3.2 1 0.7% for all three solutes. The diamond microelectrode exhibited
low peak-to-peak noise (3 - 4 pA) and stable background current under the
imposed electrophoretic conditions. The background current at the detection
potential remained stable over periods of days. In fact, the electrode could be
used for several weeks without any response attenuation. In summary, the
results indicate that the diamond microelectrode functions well for the indirect
amperometric detection of the chlorinated phenols. The linear dynamic range is
lower, the sensitivity is lower and the lowest concentration measured and
detected is higher for the three pollutants as compared to direct amperometry.
However, the response reproducibility and response stability are equally as good.
The figures of merit reported herein compare favorably with those reported for

other in indirect amperometric detection studies."'°"‘2

150

 

Table 6.3 Detection Figures of Merit for Chlorinated Phenols in CZE with
Indirect Amperometric Detection.b

 

 

 

 

 

 

 

 

 

Compound 2-CP 3-CP 2,4-DCP
Linear Dynamic Range
(1.1M) 30 - 600 30 - 600 50 - 600
Correlation Coefficient,
r2 0.999 0.997 0.996
Sensitivity (meM) 774 1 143 906 1 108 339 1 29
Concentration LOQ 30 3O 50
(pmolIL) (SIN = 6) (SIN = 6) (SIN = 6)
Estimated Mass
LOQ (pg) 4.7 4.7 9.8
Peak Height RSD%
(n =10) 3.2 1 0.7 1.6 1 0.8 3.0 1 0.8
Efficiency (plates/m) 101,000 224,000 69,000

 

 

 

 

 

 

b CE-EC parameters: Em) = 8 kV; tin] = 3 s; capillary dimensions = 30 um id. x 76
cm long; Esep = 20 kV; and E391 = 0.45 V. The data is based on the response of 3
microelectrodes and are reported at the 95% confidence level. LOQ = limit of
quantitation (the minimum concentration of analyte injected and measured).The
mass limit of quantitation was estimated from the calculated injection volume for
methanol, 1.1 nL.

6.3. Conclusions

Analysis of aqueous solutions containing chlorinated phenols was
accomplished by capillary electrophoresis with direct and indirect amperometric
detection using a boron-doped diamond microelectrode. The diamond
microelectrode, used in the end-column arrangement,, exhibited a low and stable

background current with low peak-to-peak noise and good electrochemical

151

 

 

activity for the pollutants without any conventional pretreatment. The electrode
performance was evaluated in terms of the linear dynamic range, sensitivity, limit
of quantitation, and response precision for the detection of several priority
pollutants (2-chlorophenol, 3-chlorophenol, 4-chlorophenol, 2,4-dichlorophenol,
2,4,6-trichlorophenol, and pentachlorophenol). The diamond microelectrode
yielded good detection figures of merit for these contaminants in the direct
amperometric mode with no evidence of any deactivation or electrode fouling. As
an example, the lowest concentration measured for 2-chlorophenol was 100 nM
or 13 ppb (SIN = 3) and the relative standard deviation of the peak height for 9
injections was 4.7 1 0.5% (est. 1.1 nL inj.). The separation efficiency ranged

between 85, 000 and 302,000 plates/m for the seven solutes.

The microelectrode was also employed for the indirect detection of the
chlorinated phenols. In this approach, which is useful for detecting electroinactive
solutes, ferrocene carboxylic acid was added to the run buffer as the
electrophore. Good detection figures of merit were also achieved for the
separation and detection of 2-chlorophenol, 3-chlorophenol, and 2,4-
dichlorophenol in this mode, although the lowest concentration measured were
higher than those obtained by direct amperometric detection. For example, the
lowest concentration measured for these pollutants was in the mid micromolar

range (1-10 ppm) with excellent response reproducibility of 3.2 1 0.8%, or less.

152

6.4. References

10.

11.

12.

13.

14.

15.

16.

17.

Holland, L.A.; Leigh, M.A.; Electrophoresis 2002, 23, 3649.

Deterding, L.J.; Parker, C.E.; Perkins, J.R.; Moseley, M.A.; Jorgenson,
J.W.; Tomer, K.B.; J. Chromatogr. 1991, 554, 329.

Foret, F.; Krivankova, L.; Bocek, P.; Capillary Zone Electrophoresis 1993
VCH Verlagsgessellschaft, Weinheim.

You, T.Y.; Yang, X.R.; Wang, E.; Electroanalysis, 1999, 11,459.

He, Y.; Pang, H.M.; Yeung, E.S.; J. Chromatogr. A 2000, 894, 179.

Kang, S.H.; Wei, W.; Yeung, E.S.; J. Chromatogr. B 2000, 744, 149.
Voegel, P.D.; Zhou, W.; Baldwin, R.P.; Anal. Chem. 1997, 69, 951.
Swiney, K.; Bomhop, D.J.; Electrophoresis 2000, 21 , 1239.

Matysik, F.-M.; Electrophoresis 2002, 23,400.

Baldwin, R.P.; Electrophoresis 2000, 21,4017.

Ewing A.G.; Wallingford, R.A.; Olefirowicz, T.M.; Anal. Chem. 1989, 61,
292A.

Lacher, N. A.; Garrison, K. E.; Lunte, S.M.; Electrophoresis 2001, 22,
2526.

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

Shin, D.; Sarada, B.V.; Tryk, D.A.; Fujishima, A.; Anal. Chem. 2003, 75,
530.

Wang, J.; Chen, G.; Chatrathi, M.P.; Fujishima, A.; Tryk, D.A.; Shin, D.;
Anal. Chem. 2003, 75, 935.

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

Zoski, C.G.; Mirkin, M.V.; Anal. Chem. 2002, 74, 1986.

153

 

 

18.

19.

20.

21.

22.

23

24.

25.

26.

27.

28.

29.

30.

31.
32.
33.
34.

35.

Muna, G.W.; Tasheva, N.; Swain, G.M.; Environ. Sci. Technol. 2004, 38,
3674.

Matysik, F-M.; J. Chromatogr. 1996, 742, 229.

Wallingford, R.A.; Ewing A.G.; 1987, 59, 1762.

Qian, J.H.; Wu, Y.Q.; Yang, H.; Michael, A.C.; Anal. Chem. 1999, 71,
4486.

Lacher, N. A.; Lunte, S.M.; Martin, R.S.; Anal. Chem. 2004, 76, 2482.
Huang, X.; Zare, R.N.; Sloss, S.; Ewing, A.G.; Anal. Chem. 1991, 63, 189.
Lu, W.; Cassidy, R.M.; Anal. Chem. 1994, 66, 200.

Gehardt, G.C.; Cassidy, R.M.; Baranski, A.S.; Anal. Chem. 1998, 70,
2167.

Wang, J. Analytical Chemistry, Wiley, New York, 1994, p. 61.

Ugland, K.; Lundanes, E.; Greibrokk, T.; Bjorseth, A.; J. Chromatogr.
1981, 213,83.

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

lVlardones, C.; Rios, A.; Valcarcel, M.; Electrophoresis, 1999, 20, 2922.
Martinez, D.; Pocurull, E.; Marce, R.M.; Borull, F.; Calull, M.; J.
Chromatogr. 1996, 734, 367.

Gonnord, M.F.; Collet, J.; J. Chromatogr. 1993, 645, 327.

Jauregui, 0.; Puignou, L.; Galceran, M.T.; Electrophoresis 2000, 21,611.
Hilmi, A.; Luong, J.H.T.; Nguyen, A.; J. Chromatogr. 1997, 761, 259.
Luong, J.H.T.; Hilmi, A.; Nguyen, A.; J. Chromatogr. 1999, 864, 323.

Wang, J.; Chatrathi, M.P.; Tian, B.; AnaI.Chim. Acta 2000, 416, 9.

154

36.
37.
38.
39.
40.

41.

42.

43.

45.
46.

47.

Gaitonde, C.D.; Pathak, P.V.; J. Chromatogr. 1990, 514, 389.
Gattrell, M.; Kirk, D.W.; J. Electrochem. Soc. 1993, 140,903.
Gattrell, M.; Kirk, D.W.; J. Electrochem. Soc. 1993, 140, 1 534.
Doble, P.; Haddad, P.R.; J. Chromatogr. 1999, 834, 189.
Olefirowicz, T.M., Ewing, A.G.; J. Chromatogr. 1 990, 499, 713.

Woolley, A.T.; Lao, K.; Glazer, A.N; Mathies, R.A.; Anal. Chem. 1998, 70,
684.

Matysik, F.M.; Marggraf, D.; Glaser, P.; Broekaert, J.A.C.; Electrophoresis
2002, 23, 3711.

Desiderio, C.; Fanali, S.; Gebauer, P.; Bocek, P.; J. Chromatogr. 1997,
772, 81.

Macka, M.; Haddad, P.R.; Gebauer, P.; Bocek, P.; Electrophoresis 1997,
18, 1998.

Wang, T., Hartwick, R.A.; J. Chromatogr. 1992, 607, 119.

Li, G.; Locke, D.C.; J. Chromatogr. 1995, 669, 93.

Foret, F.; Fanali, S.; Ossicini, L.; Bocek, P.; J. Chromatogr. 1989, 470,
299.

155

 

 

 

CHAPTER 7

Chlorinated Phenol Analysis Using Solid Phase
Extraction Coupled with Amperometric Detection

7.1. Introduction

The analysis of phenols in water supplies is an important analytical
problem because of their presence in waste water effluents and their potential to
enter to water supply from point sources, such as landfills.1 Capillary
electrophoresis (CE) is an important analytical tool in environmental analysis.2
Many of the detection methods employed with CE for chlorinated phenol analysis
do not achieve the low detection limits (e.g., 1 ppb for PCP) required for water
quality monitoring. Therefore, new approaches for the sensitive and stable
detection of these pollutants is an important area of research. In the previous
Chapter, it was demonstrated that CE-EC, with a diamond microelectrode can be
used to analyze for chlorinated phenols in standard aqueous solutions down to
the mid to low ppb level. The linear dynamic range can be extended and the

detection limit lowered if an extraction and preconcentration step is utilized. In

156

this Chapter, the application of solid phase extraction (SPE) with the CE-EC

method is reported.

Techniques, such as liquid-liquid (LLE), solid phase extraction (SPE) and
solid-phase microextraction (SPME) are routinely used for analyte
preconcentration and sample clean up. Liquid-liquid extraction (LLE), for
instance, has been widely employed for analyte preconcentration and is a step in
several standard methods for phenol analysis (EPA methods 604, 625 and
8041)?”5 LLE, however, suffers from several undesirable features including the
requirement for a large volume of organic solvent, foam formation, a lengthy
analysis time and difficulty with automation. Solid phase extraction (SPE) is a
sample enrichment procedure that does not suffer from the same undesirable
features. SPE requires low eluent volumes, involves short analysis times, is easy
to automate and can even be integrated on-line. Even in the off-line mode, it is a
less time-consuming and more efficient preconcentration method than LLE. SPE
is not only used to extract trace levels of organic compounds from aqueous
samples but also to reduce the effect of interfering components in the sample
matrix. One important parameter to be considered with SPE is the selectivity of
the stationary phase for the solute of interest, especially when compounds are to

be extracted from complex matrices.6

The most widely used solid phases for the extraction and enrichment of

phenolic compounds are Ca and C13 phases chemically bonded to silica, carbon

157

 

black and polymeric resins."10 Retention of phenols from water samples by Cg
and C13 material is the result of a reversed-phase mechanism. The effectiveness
of an extraction is given by the analyte recovery, which is defined as the ratio of
the number of molecules extracted to the number of molecules loaded onto the
stationary phase. Another performance parameter is the breakthrough volume.
This is the maximum sample volume that can be passed through the sorbent
while still maintaining, at least theoretically, 100% retention of the solute
molecules. The breakthrough volume depends on the sorbent surface area
available for solute retention as well of the free energy of the interaction between
solute and the sorbent. In aqueous media, polar solutes have the lowest
breakthrough volumes due to their high solubility in the aqueous phase
particularly with Cg and C13 solid phases. Highly cross-linked packing materials,
such as LiChrolut EN, Styrosorb and Macronet Hypersol, lsolute ENV,
HYSphere-1 and ENVI- Chrom P, have higher breakthrough volumes for polar
analytes.”15 These stationary phases are chemically stable over a broad range
of pH, which extends the applicability of the method, and the cross-linking
increases the specific surface area and allows for greater 1t-1t interaction between

the analytes and the solid phase.6

The ENVl-Chrom P columns used in this work were apparently developed
specifically for the extraction of polar compounds from aqueous media. The solid
phase consists of small, nonionic, styrene-divinylbenzene copolymer beads that

offer a much greater surface area than does a typical silica-based packing.16

158

Phenols are sometimes difficult to retain on C-18 modified silica under reversed
phase conditions, mainly due to their greater solubility in water (at least for the
less chlorinated ones) than in organic matrices. The numerous active aromatic
sites available in ENVl-Chrom P columns enables strong interaction with the
aromatic phenols, thereby improving the recovery rates. Retention of the phenols
occurs by a reversed-phase partitioning mechanism as well as by n—n interaction.
Chlorinated phenol recoveries of greater than 90% have been observed with this

solid phase material.16

This Chapter describes the use of off-line SPE with the CE-EC method to
analyze chlorinated phenols in spiked river water. Electrochemical detection was
performed with a diamond microelectrode - a new electrode material that
provides a highly sensitive, reproducible and stable response for the oxidative

detection of these pollutants.

7.2. Results and Discussion

7.2.1. Separation and Amperometric Detection of Chlorinated Phenols

AS described previously, decoupling the separation voltage from the
electrode signal is a critical issue in CE-EC. Decoupling was accomplished in this
work by using end-column detection. The microelectrode was positioned near the
end of the capillary at a distance estimated to be 20 - 30 um, being controlled by
adjusting the capillary. The microelectrode response was relatively unaffected by

the separation voltage with acceptable signal-to-noise ratios and minimal

159

 

 

 

potential shifts observed. The average peak-to peak noise at the detection

potential ranged from 0.5 -1 pA.

Electrophoretic separation of the chlorinated phenols was accomplished
using a mixed 10/20 mM borate/phosphate run buffer at pH 8.4, a 76-cm-long
fused silica capillary (30 pm id.) and a separation voltage of 20 W.17 The sample
was injected electrokinetically by applying 10 W for 3s. The injected volume of
an electroosmotic flow marker, methanol, was calculated to be 1.4 nL under
these conditions. Based on the migration time, the applied voltage and the
column length, the EOF was calculated to be 6.2 x 10“ cmst. After applying
the detection potential, the separation voltage was then turned on, and the
detection electrode’s background current was allowed to stabilize prior to an
injection. This usually occurred within 3 - 4 minutes after separation voltage tum-
on.

Figure 7.1 displays a typical electropherogram obtained for a standard
solution of 2-chlorophenol (2-CP), 3-chlorophenol (3-CP), 4-chlorophenol (4-CP),
2,4-dichlorophenol (2.4-DCP), 2,4,6-trichlorophenol (2,4,6-TCP) and
pentachlorophenol (PCP). Each analyte concentration was 40 11M and the
detection potential was 1.05 V.17 Separation of all six pollutants is accomplished

in less than 14 min and all migration times were very reproducible over an 8-h

41.7(1,,,Iw,,,)2
b/a+l.25

 

period of use (RSD s 0.5 %). The separation efficiencies (N =

where wo_1 is the width of the peak at 10% from the base, b/a is the asymmetry

160

600

 

A a - 4-CP b
E b — 3—CP a c
v 400 _ c — 2-CP
H
c d — 2.4-DCP
g e - 2,4,6-TCP
:1 f - PCP d
o
200 _

 

 

 

 

IL 11b

0 200 400 600 800
Time (s)

 

0

Figure 7.1 An electropherogram for a standard solution of chlorinated phenols
recorded in a 10I20 mM boratelphosphate run buffer, pH 8.4, by direct
amperometry at 1.05 V (vs. AglAgCI). Separation voltage = 20 kV. Capillary = 30
pm id. x 76 cm long. Injection = 10 W for 3s. Solute concentration = 40 (1M.

factor) were 61,000, 78,000, 116,000, 60,000, 144,000, and 139,000 plates/m for
4-CP, 3-CP, 2-CP, 2.4-DCP, 2,4,6-TCP and PCP, respectively. The low
separation efficiencies for some analytes is due to the asymmetry arising from
peak tailing due to mass transport across the microelectrode during elution, as
discussed previously. Previous work showed that the diamond electrode (both
planar and microelectrode architectures) exhibited good figures of merit for the
oxidative detection of chlorinated phenols in FIA-EC, LC-EC and CE-EC.‘7"° For
example, the lowest concentration injected and detected via LC-EC ranged from

12 to 133 ppb, depending on the pollutant.18

161

 

 

 

7.2.2. Off-line Coupling of SPE with 05-56

The previously reported CE-EC method, using a diamond microelectrode,
achieved detection limits that were below the health advisory levels of 40, 20 and
4000 ppb for 2-chlorophenol, 2,4—dichlorophenol and phenol, respectively.
However, the method did not yield the 1 ppb detection limit required for
pentachlorophenol. Therefore, an SPE procedure was implemented to extract
the chlorinated phenols from the aqueous solution for preconcentration, thereby
extending the linear dynamic range and lowering the limit of detection. Locally
obtained river water was used as the sample matrix in which to test the method
for the detection of chlorinated phenols. Prior to spiking with the chlorinated
phenols, the river water was filtered through a nylon membrane with a 0.4 pm
pore size to remove solid particulates. The extraction process was evaluated
using different applied volumes of spiked river water. By performing the analysis
on a natural water sample, we were able to assess to what extent the matrix
(e.g., humic and fulvic acids) of the river water affects the efficiency of the
chlorinated phenol extraction. Humic matter can bind organic pollutants and
cause a decrease in the breakthrough volume and a lowering of the recovery
because only the soluble fraction will be enriched.“19 The interaction of
pollutants with humic material will be pH dependent; therefore, the extraction of
the Chlorinated phenols from both acidic and neutral solution was evaluated.
Preparation of the basic solution involved simply spiking the river water with the
chlorinated phenols as the solution pH was ~ 7.5. Therefore, no pH adjustment

was made. Preparation of the acidic solution involved spiking the river water with

162

 

 

 

the chlorinated phenols and adjusting the pH to 2.5 with a few drops of
concentrated HCI. Acidification is a normal practice in water analysis in order to
preserve the sample and to avoid partial deprotonation of low-pKa analytes. The
recovery of each chlorinated phenol was determined by applying 50 mL of the
river water spiked with 0.1 — 0.2 ppm of each chlorinated phenol. The recoveries
for 2-chlorophenol, 3-chlorophenol, 4-chlorophenol, 2,4-dichlorophenol and
pentachlorophenol remained fairly constant near 100 % from both low and high
pH solutions, while the recovery for 2,4,6-trichlorophenol decreased from near
100% at low pH to 64 % at high pH (7-8). This decrease was caused by analyte
binding with the humic acids.13 Interaction of phenol having a low pKa value with
fulvic or humic acid material is known to occur at neutral pH.13 Based on these
results, the sample pH was adjusted to 2.5 in order to maximize the recovery of
all the six pollutants. Calibration curves were prepared for standard solutions of
each chlorinated phenol. Plots were made from concentrations ranging from 0.1
— 12.8 ppm for 2-chlorophenol, 3-chlorophenol and 4-chlorophenol; 0.2 — 16.3
ppm for 2,4-dichlorophenol; 0.2 — 19.7 ppm for 2,4.6-trichlorophenol and 0.3 —
26.6 ppm for pentachlorophenol (1 to 100 (M). The linear regression correlation
coefficients for all the plots were greater than 0.99. The spiked concentrations of
the chlorinated phenols (0.1 to 0.2 ppm) fell within the linear dynamic range after
preconcentration. Figure 7.2 shows an example of a calibration curve for the 2-
chlorophenol that was used in calculation of the recovery. The percent recoveries
ranged from 95 - 100% for all six pollutants as shown in Table 7.1. Given the fact

that the phenols are usually present in natural water samples at trace levels,

163

 

 

application of a relatively large volume of a sample is often needed to achieve
low limits of detection. Therefore, the percent recovery was also determined for a
higher applied sample volumes (250 mL) spiked with chlorinated phenol
concentrations ranging from 0.7 — 15 ppb. Good recoveries were observed for all
the pollutants (greater than 95%). The relative standard deviation in the recovery

values from measurement to measurement ranged between 1 and 4 %.

 

160~
R = 0.99577
120-
3‘1
*5 80- '
.41
40-
0-

 

 

 

I I I I I I I

0 2 4 6 8 10 12 14
Concentration (ppm)

Figure 7.2 Calibration: curve for a river water sample spiked with different

concentrations of 2-CP

164

Table 7.1 Recovery of chlorinated phenols.°

 

 

 

 

 

 

 

 

Analyte % recovery RSD% (n = 6)
2-chlorophenol 97 2.4
3-chlorophenol 96 3.0
4-chlorophenol 95 2.1
2,4—dichlorophenol 98 3.9
2,4,6—trichlorophenol 97 3.8
pentachlorophenol 100 1.0

 

 

 

 

c50 mL of the river water, pH 2.5, was spiked with 0.1 ppm of 2-CP, 3-CP and 4-
CP, 0.2 ppm of 2.4-DCP and 2,4,6-TCP, and 0.3 ppm of PCP.

 

 

500

Current (pA)
N u -h
c c
o c
m
o.

O
O
(D

 

 

 

 

 

 

 

2

C
'Q.

F1

O
6.:

 

 

 

_____l_ 1.1
200 400 600 800

Figure 7.3 An electropherogram for a river water SPE extract. 50 mL of the river
water, pH 3.5, was spiked with 0.1 ppm of 2-CP, 3-CP and 4-CP, 0.2 ppm of 2,4-
DCP and 2,4,6-TCP, and 0.3 ppm of PCP (1 11M) and then extracted. Run buffer
= 10/20 mM boratelphosphate, pH 8.4. Separation voltage = 20 kV. Capillary =
30 pm id. x 76 cm long. Detection potential = 1.05 V (vs. AglAgCI). Injection = 10
W for 35.

165

Figure 7.3 shows a typical electropherogram obtained after SPE. The
chlorinated peaks are sharp and well resolved for all six pollutants. Good
separation efficiencies were obtained with plate numbers greater than 100,000
plates/m for all the analytes. A distorted peak of unknown origin appeared with a
migration time of ~ 500 s. This peak was more pronounced in samples containing
lower concentrations of the chlorinated phenols, but did not appear to interfere
with the quantitation of the analytes. The response linearity for the total
procedure was checked using 250 mL of river water spiked with different
concentrations of each chlorinated phenol ranging from 0.02 - 200 ppb. Table
7.2 presents a summary of the separation and detection figures of merit. A wide
linear dynamic range was achieved for each pollutant with correlation coefficients
greater than 0.997. The linear dynamic range was 0.02 — 100 ppb for 2—
chlorophenol, 3-chlorophenol and 4-chlorophenol, 0.04 — 130 ppb for 2,4-
dichlorophenol, 0.2 — 150 ppb for 2,4,6-trichlorophenol and 0.05 - 200 ppb for
pentachlorophenol. The good response sensitivity of the diamond
microelectrode, as well as low peak-to-peak background noise (0.5 - 1 pA),
enabled minimum detectable concentrations in the high ppt to low ppb range for
all the pollutants. Good peak height response reproducibility was observed for all
the pollutants with a nominal value of 5.4 %, or less. In summary, the diamond
microelectrode exhibited good electroanalytical performance with a low peak-to-
peak noise (0.5 to 1 pA), a low and stable background current under the
electrophoretic conditions imposed, and a stable and reproducible response for

the chlorinated phenols with no evidence of any electrode deactivation during

166

Table 7.2 Detection Figures of Merit for Chlorinated Phenols using off-line

SPE with Capillary Electrophoresis and Amperometric Detectioncl

 

 

 

 

 

 

 

 

 

 

 

 

 

Analyte LDR (ppb) Sensitivity Response LOO Separation
(pAInM) Precision (ppb) Efficiency
(n=9) (plates/m)
2-CP 0.02 - 100 3.2 1 1.4 4.6 0.02 185,000
r2 = 0.9984 (SIN = 8)
3—CP 0.02 —100 5.0 1 1.2 5.2 0.02 188,000
r2 = 0.9988 (SIN = 3)
4—CP 0.02 - 100 5.3 1 0.8 4.5 0.02 179,000
12 = 0.9979 (SIN = 4)
2.4-DCP 0.04 — 130 1.9 1 0.5 5.0 0.04 173,000
12 = 0.9982 (SIN = 3)
2,4,6- 0.20 — 150 1.2 :1: 0.3 5.4 0.20 292,000
TCP r2 = 0.9998 (SIN = 3)
PCP 0.05 - 200 1.7 1 0.8 5.2 0.05 172,000
12 = 0.9991 (SIN = 3)

 

 

 

aCE-EC parameters: Eini = 10 kV; tin) = 3 s; capillary dimensions = 30 pm i.d. x 76
cm long; Essp = 20 kV; and Egg, = 1.05 V. LOQ = limit of quantitation (the
minimum concentration of analyte injected and measured). The data is based on

the response of 3 diamond microelectrodes.

extended use. Additionally, no pretreatment was required for electrode activation
making the procedure more reproducible and less time consuming. The low and
stable peak—to-peak variations in the background current allowed for low
detection limits to be achieved. These results are in agreement with previous
reports that diamond thin-film electrodes are useful for the stable detection of
resistance to surface deactivation and

chlorinated phenols with high

fouling.‘7'""20 No deactivation of the diamond microelectrode was observed in

167

 

 

this work at the chlorophenol concentrations studied; therefore, an electrode
could be used in measurements for weeks. The low limit of detection, the wide
linear dynamic range, the excellent response precision and the superb response
stability obtained with diamond shows that the SPE/CE-EC method could be
employed for analysis of these priority pollutants in drinking water or in more
complex environmental samples like river water. Other examples of where this
method could be applied are in monitoring of chlorinated phenols in landfill
leachates where the concentration ranges from the sub-ppb to ppm levels“22 or
in coal tar distillation plant or wood preservation site waste in which the

concentrations in waste ground water can be at ppm levels.”24

Finally, it is important to evaluate how this method compares against
others reported on in the literature. In general, it appears that no analytical
method, without some form of sample enrichment, is capable of achieving the
low detection limit required for PCP (1 ppb), as mandated by EPA. SPE is now a
common sample enrichment method that can be utilized in combination with any
analytical method to increase the linear dynamic range and to lower the detection
limit for an analyte.25 Chlorinated phenol analysis has been carried out using
SPE and gas chromatography (GC), liquid chromatography (LC) and CE.2'3'13'26‘
33 with electrochemical, UVNis and mass spectrometric detection. Chlorinated
phenol limits of detection in the low (~ 2) to mid (~0.05) ppb range (SIN 2 3) are
typical for preconcentration factors of 250 to 500, depending on the analyte. For

example, Cheung et al. reported a detection limit for PCP of 0.05 ppb using GC-

168

MS.26 Galceran et al. reported a detection limit of 0.04 ppb for PCP using LC-
EC.27 Bruijnsvoort et al. reported a detection limit of 0.2 ppb for PCP using CE-
EC (MEKC).33 The detection limits reproducibly obtained with the CE-EC method
are as good or better than those reported in the literature. Moreover, the
response reproducibility for this CE-EC method is superior to that reported for the
other methods. The most noteworthy feature of this method, as compared to the
others using electrochemical detection, is the stable response and the fact that
no reconditioning or self-cleaning is necessary for optimum electrode

performance.

7.3. Conclusions

An analytical method with improved detection figures of merit over
commonly employed methods is described for chlorinated phenol analysis in river
water. SPE was employed for sample enrichment and CE-EC was used for
efficient separation and detection of the electroactive chlorinated phenols. A key
enabling feature of this method is the use of a diamond microelectrode for
electrochemical detection. The method provides detection figures of merit
superior to other methods reported on in the literature and exceeds EPA
requirements for drinking water monitoring in terms of the MCL for
pentachlorophenol (1 ppb) and the health advisory action levels for 2-
chlorophenol, 2,4-dichlorphenol and phenol (40, 20 and 4000 119/L (ppb),
respectively). Diamond outperforms all other bare metal and carbon electrodes

for the amperometric detection of chlorinated phenols, in large part, because of

169

its resistance to deactivation and fouling. Rapid deactivation and fouling plague
EC detection with other electrode materials and limit its application for the
detection of phenols and chlorinated phenols. Diamond exhibited good
electroanalytical performance with a low peak-to—peak noise (0.5 to 1 pA), a low
and stable background current under the electrophoretic conditions imposed, and
a sensitive, reproducible and stable oxidation response for the chlorinated
phenols over many days to weeks of use. Reproducible detection limits for all the
chlorinated phenols were in the mid ppt range with the exception of 2,4,6-
trichlorophenol which was in the high ppt range. The power of CE separations,
the simplicity of electrochemical detection, and the properties of this new
electrode material render this method viable for chlorinated phenol analysis in

real water.

170

7.4. References

10.
11.
12.
13.
14.

15.

16.

Heberer, T.; Stan, H.; Anal. Chim. Acta. 1997, 341, 21.

Martinez, 0.; Cugat, M.J.; Borrull, F.; Calull, M. J. Chromatogr. A, 2000,
902, 65.

EPA Method 604, Phenols, in Federal Register, Environmental
Protection Agency, Part VIII, 40 CFR Part 136, Washington DC, 1984, 58.

EPA Method 625, Base/Neutrals and Acids, in Federal Register
Environmental Protection Agency, Part Vlll, 40 CFR Part 136, Washington
DC, 1984.153.

EPA Method 8041, Phenols by Gas Chromatography. Capillary Column
Technique, Environmental Protection Agency, Washington DC, 1995, 1.

Leon-Gonzalez, M.E.; Prez-Arribas, L.V. J. Chromatogr. A 2000, 902, 3.

Pocurull, E.; Marce, R.M.; Borull, F.; Chromatographia. 1995, 41, 521.

Achilli, G.; Cellerino, G.P.; Melzi d’EriI, G.; Bird, 8.; J. Chromatogr. A
1995, 697, 357.

Somsen, G.W.; Jagt, I.; Gooijer, C.; Velthorst, N.H.; Brinkman, U.A.Th,
Visser, T.; J. Chromatogr. A 1996, 756, 145.

Pichon, V.; Hennion, M.C.; J. Chromatogr. A 1994, 665,269.

Pocurull, E.; Marce, R.M.; Borull, F.; J. Chromatogr. A 1996, 738, 1.
Fiehn, O.; Jekel, M.; Anal. Chem. 1996, 68, 3083.

Puig, D.; Barcelo, D.; J. Chromatogr. A 1996, 733, 371.

Rodriguez, l.; Llompart, M.F.; Cela, R.; J. Chromatogr. A 2000, 885, 291.

Pocurull, E.; Calull, M.; Marce, R.M.; Borull, F.; J. Chromatogr. A 1996,
719, 105.

Supelco, Application Notes, www.sigma-aldrich.com accessed February
2005

171

 

17.

18.

19.

20.

21.

22.

23.

24.
25.
26.
27.

28.

29.

30.
31.

32.

Muna, G.W.; Quaiserova-Mocko, V.; Swain, G.M.; Electroanalysis, 2005,
17, 1 160.

Muna, G.W.; Tasheva, N.; Swain, G.M.; Environ. Sci. Technol. 2004, 38,
3674.

Porchmann, J.; Stottmeister, U.; Chromatographia, 1993, 36, 207
Terashima, C. Rao, T.N.; Sarada, B.V.; Tryk, D.A.; Fujishima, A.; Anal.
Chem. 2002, 74, 895.

Vollmuth, S.; Niessner, R.; Chemosphere 2002, 30, 2317

Gade, B.; Layh, M.; Westerrnann, H.; Amsoneit, N.; Waste Manage. Res.
1996, 14, 553.

Thornton, S.F.; Quigley, S.; Spence, M.J.; Banwart, S.A.; Bottrell, S.;
Lerner, D.N.; J. Contam. Hydrol. 2001, 53, 233.

Galceran, M.; Jauregui, 0.; Anal. Chim.Acta1995, 304, 75.

Cheung, J.; Wells, R.J.; J. Chromatogr. A 1997, 771, 203.

Reitzel, L. A.; Ledin, A.; 2 J. Chromatogr. A 2002, 972, 175.

Galceran, M.; Jauregui, 0., Anal. Chim.Acta1995, 304, 75.

Cardellicchlo, N.; Cavalli, S.; Piangerelli, V.; Giandomenico, S.; Ragone,
P.; Fresenius J. Anal. Chem. 1997, 358, 749.

Pocurull, E.; Sanchez, G.; Borrull, F.; Marce, R.M.; J. Chromatogr. A 1995,
696, 31.

Bangheri, H.; Saraji, M.; J. Chromatogr. A 2001, 910, 87.
Mardones, C.; Rios, A.; Valcarcel, M.; Electrophoresis, 1999, 20 2922.

Bruijnsvoort, M.; Sanghi, S.K.; Poppe, H.; Kok, W. Th.; J. Chromatogr. A
1997, 757,203.

172

CHAPTER 8

Summary

1. Boron-doped microcrystalline and nanocrystalline diamond exhibited a
wide potential window, a low background current and good responsiveness for
Fe(CN)g'3"4, Ru(NHg)g"2"’3, erlg'M, and methyl viologen (Mv*2’*"°) without any
surface pretretreatment. The quasi-reversible voltammetry for all the couples
indicated that the diamond electrodes has sufficient charge carrier density over
the wide potential Window range to support a rapid electron transfer. More
sluggish kinetics were found for 4-tertbutylcatechol and Few“. The sluggish
kinetics are attributed to weak surface adsorption for the former, and to the

absence of catalyzing surface carbonyl groups for the latter.

2. Chlorinated phenols can be quantitavely oxidized at both boron-doped
microcrystalline and nanocrystalline. The electro-oxidation reaction mechanism
appears to proceed in a similar manner at both diamond types, regardless of the
differences in film morphology and microstructure. The oxidation of phenol, 2-CP,

and 4-CP leads to the formation of surface confined hydroquinone/p-

173

benzoquinone and catechol/o-benzoquinone redox couples, whereas the

oxidation of 3-CP and PCP do not.

3. FIA-EC and HPLC-EC results indicate that chlorinated phenols can be
stably and reproducibly detected at constant potential at diamond electrodes.
Both microcrystalline and nanocrystalline are responsive for these analytes
without extensive pretreatment and provide a stable response over time with
minimal electrode fouling. The response precision for all the chlorinated phenols
were 2 — 6% on both diamond types. The limits of quantitation were in low ppb

levels (9 — 27 ppb).

4. The use of boron-doped diamond microelectrode in capillary
electrophoresis with both direct and indirect amperometric detection yielded
excellent detection figures of merit in direct amperometry in terms of the linear
dynamic range, limit of quantitation, response precision, and response stability.
The minimum concentration measured (SIN 2 3) was 100 nM (13 ppb) for 2-
chlorophenol, 3-chlorophenol and 4- chlorophenol, and 500 nM (47-133 ppb) for
phenol, 2,4-dichlorophenol, 2,4,6-trichlorophenol, and pentachlorophenol. Good
peak height reproducibility was observed for all analytes with a nominal value of
5.6 1 0.5%, or less. Good detection figures of merit were also observed in
indirect amperometry. The concentration limits of quantitation were 30 (1M (4
ppm) for 2- and 3-chlorophenol, and 50 11M (8 ppm) for 2,4—dichlorophenol at a

SIN 2 3. These values are higher than those observed in direct amperometry.

174

The electrode response reproducibility was excellent for 10 injections with values

less than 3.2 1 0.7% for 2-chlorophenol, 3-chlorophenol, and 2,4—dichlorophenol.

5. Coupling the CE-EC method to SPE extended the linear dynamic range
and lowered the limit of detection exceeding the EPA requirements for drinking
water monitoring in terms of the MCL for pentachlorophenol (1 ppb) and the
health advisory action levels for 2-chlorophenol, 2,4-dichlorphenol and phenol

(40, 20 and 4000 pglL (ppb), respectively).

Boron-doped diamond thin-film electrodes is a new electrode that provides
electrochemists with a new material for a number of electrochemical technologies
which could not be carried out previously due to limitation by the stability of the
electrode material. Among these applications is the electrochemical detection
and remediation of chlorinated phenols. Thus, with diamond, these areas of

research will greately benefit from its unique properties.

The electroanalytical assays developed in this work provide an
inexpensive, reliable, fast, and sensitive method for monitoring chlorinated
phenols in the aqueous environments. The methods meet the EPA
recommendations for the analysis of chlorinated phenols in drinking water.
Therefore, they can be implemented for electrochemical technologies for water

quality monitoring without being limited by the stability of the electrode material.

175

   
 

lIllillllllllilllll11111111

1