..11.1.' wwAEoJII‘li) . r1...

2.1%....

 

This is to certify that the
dissertation entitled

THE APPLICATION OF BORON-DOPED DIAMOND THIN-
FILM ELECTRODES FOR WATER TREATMENT AND
WATER QUALITY MONITORING

presented by

YANG SONG

has been accepted towards fulfillment
of the requirements for the

PhD degree in Chemistry - Environmental
Toxicology

 

 

.41.. M. aim.

V Major Professor’s Signature

Cégp/au/a, /‘é.§bdé

Date

MSU is an Affirmative Action/Equal Opportunity Institution

-r..—-.-—-_——._,_.

LIBRARY
Michigan State
University

 

 

 

 

-‘—-.- ..-.-.-.-.-.-.-7-.-.-.-._.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.—

av-.-.-.-4-

 

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

 

DATE DUE

DATE DUE

DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2/05 p:/ClRC/DateDue.indd-p.1

 

 

THE APPLICATION OF BORON-DOPED DIAMOND THIN-
FILM ELECTRODES FOR WATER TREATMENT AND
WATER QUALITY MONITORING

By

Yang Song

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

2006

ABSTRACT

THE APPLICATION OF BORON-DOPED DIAMOND THIN-FILM ELECTRODES
FOR WATER TREATMENT AND WATER QUALITY MONITORING

By

Yang Song

In this Dissertation, the application of boron-doped diamond electrodes for
water treatment (degradation of an organic compound (atrazine)) and water
quality monitoring (detection of inorganic ions (arsenic)) is described.

In one project, boron-doped diamond thin-film coated on Ti (diamond/T i)
was used for the anodic degradation of atrazine. The specific aim was to learn
how efficiently atrazine can be electro-oxidized at diamond to produce C02. The
diamond/T i anode was first evaluated during anodic polarization in 0.1 M H2804
and was found to exhibit a stable electrochemical response over time, good
activity for the two redox systems, Ru(NH3)63‘"2+ and Fe(CN)53"4', and excellent
morphological and microstructural stability. The electrooxidative degradation of
atrazine was investigated by bulk electrolysis at constant current density (11 or
22 mA/cmz). The intermediate degradation products were detected by HPLC-
UVNis. The three common degradation products for atrazine, hydroxyatrazine
deethylatrazine and deisopropylatrazine, were not detected during the
electrolysis. However, radioisotope labeling studies proved that 81% of the 14C-
Iabeled atrazine was oxidized to “C02 during a 2-h electrolysis period confirming
that diamond is a suitable anode for complete mineralization.

In a second project, a Au-coated boron-doped diamond thin-film electrode

was used to detect total inorganic arsenic in water supplies using differential
pulse anodic stripping voltammetry (DPASV). The specific aim of the work was to
learn how the diamond support influences (i) the formation of Au metal phases
and, in term, the formation of the Au-As complex during the deposition step and
(ii) the resistance of the nearby Au particles toward adsorption of contaminants in
solution (i.e., fouling). In the method, As is codeposited (preconcentrated) with
Au on the diamond surface and detected oxidatively by DPASV. Au deposition
was found to be uniform over the surface with a nominal particle size of 23 i 5
nm under the potentiostatic conditions employed. The electrode provided a
sensitive, reproducible and stable response for total inorganic arsenic detection
and exhibited better performance than did Au-coated glassy carbon or Au foil
electrodes. Sharp and symmetric stripping peaks were generally observed for the
Au-coated diamond electrode. As(V) was chemically reduced to As(lll) by
Na2303 before analysis. Limits of detection (LOD) were 0.005 ppb (S/N = 3) for
As(lll) and 0.08 ppb (S/N = 3) for As(V) in standard solutions. An As(lll)
concentration of 0.6 ppb was found in local river water with a stable electrode
response. lnten‘erences, such as metal ions (e.g., Cu(ll)) and organic matter (e.g.,
humic acid), that commonly cause decreased electrode response sensitivity for
arsenic detection in real samples, were minimized through the use of solid phase
extraction in the sample preparation. The method provided accurate and
reproducible results for total inorganic arsenic detection in two contaminated
water samples. The detected total inorganic arsenic concentration differed from

the specified concentrations by less than 4%.

Copyright by
YANG SONG
2006

ACKNOWLEDGEMENTS

Words can not express the appreciation to my advisor, Dr. Greg Swain.
That is your guidance, patience and support walked me through the four years.
Your hard work, quest for excellence and vast knowledge in analytical chemistry
greatly influenced me along the way.

Thanks also go to my guidance committee: Dr. Blanchard, Dr. Voice and
Dr. Jackson for the helpful comments about my research. Your input and support
has been valuable for my development as a scientist.

Many thanks to the former and present Swain group members: Show,
Prema, Stotter, Grace, Anne, Bennett, Jinwoo, Gloria, Veronika, Elizabeth, Doug,
Karolina, Pushy, Shihua, Audrey, Yingrui, Hua, Luther, Michael, Ayten, Matt,
Martin, Isao, Aihua, Bhavik, Dasa and Jason for your friendship, support, help
and encouragement. Special thanks to Show and Bennett who fabricated the
diamond planer films.

Many “non-Swain-group” friends have also seen my growth in MSU.
Thanks to Ningqing, Xiaoping, Ziyang, Dan, Qing, Guanghui, Yi, Xiaoyun, Joy
and the Detjen family. Your friendship means a lotto me. I would like to extend a
very special thank to Ningqing for his support and encouragement during the four
years. I truly could not have done this without you.

Finally and most importantly, I dedicate this dissertation to my parents,
Mingchao Song and Yanyan Hou, thank you for your unending love,
encouragement and support! I would not be where I am today without you who

have supported and encouraged me to achieve my dreams.

Funding from NASA (NAG9-1481; Advanced Environmental Monitoring
and Control) for the water treatment project and the Office of Naval Research, as
part of the Expeditionary Unit Water Purification Program (N000140310995) and

NASA (NNJO4HI1OG) for the water monitoring project is gratefully acknowledged.

vi

TABLE OF CONTENTS

List of Tables ...................................................................................................... xi
List of Figures .................................................................................................. xiii
CHAPTER 1 - INTRODUCTION .......................................................................... 1
1.1 Importance of Water Treatment ..................................................................... 2
1.1.1 Methods of Water Treatment ................................................................ 2
1.1.2 Electrode Materials and Their Limitations ............................................. 8
1.1.3 Boron-Doped Diamond and Substrate Materials .................................. 9
1.1.4 The Herbicide — Atrazine .................................................................... 11
1.2 Importance of Water Quality Monitoring ....................................................... 12
1.2.1 Metal Ion Contaminants and Arsenic .................................................. 12
1.2.2 EPA Approved Methods for As Detection ........................................... 14
1.2.3 Differential Pulse Anodic Stripping Voltammetry (DPASV) ................. 15
1.2.4 Electrode Materials and Their Limitations for Arsenic Detection ......... 18
1.3 Dissertation Outline ..................................................................................... 20
1.4 References .................................................................................................. 22
CHAPTER 2 - EXPERIMENTAL METHODS ..................................................... 27
2.1 Diamond Thin Film Growth .......................................................................... 27
22 Material Characterization ............................................................................. 30

vii

2.3

2.4

2.5

2.6

2.2.1 Scanning Electron Microscopy (SEM) ................................................. 30

2.2.2 Raman Spectroscopy .......................................................................... 31
2.2.3 X-Ray Diffraction Spectroscopy (XRD) ............................................... 31
2.2.4 Electrical Measurements ..................................................................... 32
2.2.5 Electrochemical Characterization ........................................................ 32
Water Treatment .......................................................................................... 33
2.3.1 Electrolysis of Atrazine ........................................................................ 34
2.3.2 Separation and Detection of Atrazine and Metabolites by HPLC-

UVNis ................................................................................................. 35
2.3.3 Electrolysis of “C-Labeled Atrazine ................................................... 36
Water Quality Monitoring ............................................................................. 37
2.4.1 Gold and Arsenic Electrodeposition .................................................... 37
2.4.2 Preparation of Au-Coated Diamond and Planar Au Electrode ............ 38
2.4.3 Differential Pulse Anodic Stripping Voltammetry ................................. 39
2.4.4 Stability Test in River Water ................................................................ 40
2.4.5 Cation Exchange Method to Remove Interfering Metal Ions ............... 40
2.4.6 Detection of As (V) ................................................................... 41
2.4.7.Inductively Coupled Plasma Mass Spectrometry (lCP-MS) and

Graphite Furnace Atomic Absorption Spectroscopy (GFAAS) ............ 42
2.4.8 Total Inorganic Arsenic Detection in Real Water Samples .................. 43
Reagents ..................................................................................................... 45
References .................................................................................................. 47

viii

CHAPTER 3 — ELECTROCHEMICAL AND PHYSICAL PROPERTIES OF

DIAMOND/1' i ....................................................................................................... 48
3.1 Introduction .................................................................................................. 48
3.2 Results and Discussion ................................................................................ 51

3.2.1 Electrochemical Properties of Diamond/Ti .......................................... 51

3.2.2 Morphological and Microstructural Properties of Diamond/Ti .............. 59
3.3 Conclusions ................................................................................................. 69
3.4 References .................................................................................................. 71

CHAPTER 4 — ELECTROCHEMICAL REMEDIATION OF ATRAZINE USING A

DIAMOND/T i ELECTRODE ............................................................................... 74
4.1 Introduction .................................................................................................. 74
4.2 Results and Discussion ................................................................................ 78
4.2.1 Separation and Detection of Atrazine and Metabolites ....................... 78
4.2.2 Electrochemical Degradation of Atrazine ............................................ 84
4.2.3 Determination of the 002 Conversion Efficiency ................................. 89
4.3 Conclusions ................................................................................................. 92
4.4 References .................................................................................................. 94

CHAPTER 5 - A SENSITIVE METHOD FOR INORGANIC As DETECTION
USING ANODIC STRIPPING VOLTAMMETRY AND A Au-COATED DIAMOND

THIN-FILM ELECTRODE ................................................................................... 96
5.1 Introduction .................................................................................................. 96

5.2 Results and Discussion ................................................................................ 99

ix

5.2.1 Particle Analysis of As and Au Deposition on Diamond Electrode ...... 99

5.2.2 Effect of Au Coverage on the As Stripping Peak Current .................. 102
5.2.3 Electrochemical Measurements ........................................................ 103
5.2.4 Au-Coated Diamond vs. Au Foil for As(lll) Detection ........................ 108
5.2.5 As(V) Detection ................................................................................. 113
5.2.6 Real Sample Analysis ....................................................................... 114
5.3 Conclusions ............................................................................................... 118
5.4 References ................................................................................................ 119

CHAPTER 6 — TOTAL INORGANIC As DETECTION IN REAL WATER
SAMPLES USING DPASV AND A Au-COATED DIAMOND THIN-FILM

ELECTRODE. ................................................................................................... 123
6.1 Introduction ................................................................................................ 123
6.2 Results and Discussion .............................................................................. 126

6.2.1 Effect of Metal Ions on As Detection ................................................. 126

6.2.2 Effect of Humic Substances on As Detection .................................... 131

6.2.3 Analysis of Inorganic As in Real Water Samples .............................. 132
6.3 Conclusions ............................................................................................... 133
6.4 References ................................................................................................ 135
CHAPTER 7 - KEY FINDINGS AND FUTURE WORK .................................... 137
7.1 Water Treatment ........................................................................................ 137
7.2 Water Quality Monitoring ........................................................................... 138

Table 1.1
Table 2.1
Table 2.2

Table 3.1

Table 3.2

Table 3.3
Table 3.4
Table 3.5
Table 4.1

Table 4.2

Table 4.3

Table 4.4

Table 4.5

Table 5.1

Table 5.2

LIST OF TABLES

Chemical processes generating free radicals .................................... 5
Typical microcrystalline boron-doped diamond thin film properties..30
Detection conditions for arsenic analysis using GFAAS ................... 43

Summary of cyclic voltammetric data for diamond/Ti electrodes at a
scan rate of 100 mV/s ...................................................................... 55

Cyclic voltammetric and the apparent heterogeneous electron
transfer rate constant data for diamond/T i electrodes for Fe(CN)53""
and Ru(NH3)(53+’2+ in 1 M KCI before and after electrolysis in 0.1 M

H2SO4 at 22 mA/cm2 ........................................................................ 57
List of standard XRD pattern data for Ti ........................................... 66
List of standard XRD pattern data for TiC ........................................ 67
List of standard XRD pattern data for cubic diamond ....................... 68
Molecular structure and pKa value for ATZ, HA, DEA and DIA ........ 80

Summary of the retention times, capacity factors and plate numbers
for the four analytes: ATZ, HA, DEA and DIA ................................... 83

Atrazine concentration remaining in the anolyte compartment after 20
min of oxidative electrolysis of 5 ppm atrazine in 0.1 M H2804 at a
diamond/T i anode ............................................................................ 85

Reproducibility test of the oxidation of 14C-labeled atrazine in 0.1 M
H2SO4 at a diamond/T i anode using a constant current of 100 mA (22
mA/cm2) for 30 min .......................................................................... 90

Total 14C recovery and 14002 production percentage after 30, 60, 90
and 120 min of electrolysis with convection ..................................... 91

Particle analysis results for Au and As/Au metal phase formation on
diamond ......................................................................................... 101

Comparison of the ASV detection figures of merit for As at Au-coated
diamond and Au foil electrodes ...................................................... 110

xi

Table 5.3

Table 5.4

Table 6.1

Comparison of the ASV detection figures of merit for As at Au-coated
carbon electrodes ........................................................................... 112

Comparison of As Stripping Peak Reproducibility and Stability at Au-
Coated Diamond and Glassy Carbon Electrodes in River Water...1 17

Total Inorganic As Concentration in Real Water Samples ............. 133

xii

Figure 1.1

Figure 1.2

Figure 1.3

Figure 2.1

Figure 2.2

Figure 2.3

Figure 2.4

Figure 2.5

Figure 2.6

Figure 3.1

List of Figures

Scheme of the oxidation of organic compounds with simultaneous
02 evolution (a) interaction of the electrode surface with OHo, (b)
02 evolution by electrochemical oxidation of OH-, (c) formation of
the higher oxide on active anode, (d) 02 evolution by chemical
decomposition of the higher oxide, (e) oxidation of the organics (R)
via the higher oxide on active anode and (f) complete oxidation of
the organics via hydroxyl radicals on inactive anode ....................... 7

The valence electron configuration of arsenic in the ground state.14

The differential pulse anodic stripping voltammetric (DPASV)
potential waveform and resulting current response. (1) The
deposition step involves a cathodic potential-step of certain
magnitude and duration and (2) the stripping step involves a
potential scan in the positive direction to oxidize the predeposited
metal phase formed in step (1). An enlarged view of the applied
pulse program depicting the various parameters is also shown. The
resulting voltammetric i-E curve is constructed by plotting the
current difference of the two sampling points for each pulse as a
function of the potential ................................................................. 15

Diagram of a typical microwave CVD reactor set-up ..................... 27

Design of the single compartment three-electrode electrochemical
cell ................................................................................................. 33

Design of the two-compartment H-cell for atrazine electrolysis ..... 34

Design of the two-compartment H-cell for the 14C-labeled atrazine
analysis ......................................................................................... 37

The differential pulse anodic stripping voltammetric potential
waveform and parameters ............................................................. 39

Total inorganic arsenic detection in real water samples ................ 44
Background cyclic voltammetric i-E curves for diamond/Ti and
diamond/Si electrodes in 0.1 M H2804 over a (A) wide and (B)

narrow potential range. Scan rate = 100 mV/s. Geometric area =
0.2 cm2 .......................................................................................... 52

xiii

Figure 3.2

Figure 3.3

Figure 3.4

Figure 3.5

Figure 3.6

Figure 3.7

Figure 3.8
Figure 3.9
Figure 3.10

Figure 4.1

Figure 4.2

Figure 4.3

Figure 4.4

Figure 4.5

Cyclic voltammetric i-E curves for 1 mM Fe(CN)63"4‘ and
Ru(NH3,)63*’2+ in 1 M KCI at a diamondfl'i electrode. Scan rate = 100

mV/s .............................................................................................. 55
Plots of the anodic eak current versus the square root of scan rate
for 1 mM Fe(CN)53"' and Ru(NH3)53*’2* in 1 M KCI ....................... 56
SEM images of a diamond/T i electrode (A) before and (B) after
electrolysis in 0.1 M H2804 at 22 mA/cm2 for 50 h ........................ 60
Cross sectional SEM image of a diamond/T i electrode ................. 61
Raman spectra of diamond/T i electrode (A) before and (B) after
electrolysis in 0.1 M H2S04 at 22 mA/cm2 for 50 h ........................ 62
XRD pattern for a diamond/T i electrode (A) before and (B) after
electrolysis in 0.1 M H2804 at 22 mA/cm2 for 50 h ........................ 65
Standard XRD pattern for Ti .......................................................... 66
Standard XRD pattern for TiC ....................................................... 67
Standard XRD pattern for cubic diamond ...................................... 68

Proposed atrazine degradation pathway for electrochemical Fenton
reaction with bold arrow shown the major pathway and dashed
arrow shown the minor pathway .................................................... 75

UVNis spectra for atrazine, hydroxyatrazine, deethylatrazine,
deisopropylatrazine (5 ppm each) all dissolved in 0.1 M H2804. A
background spectrum (blank), 0.1 M H2S04, is also shown for
comparison .................................................................................... 79

Isocratic separation of 1.25 ppm (each) ATZ, HA, DEA and DIA
using a reversed-phase HPLC separation and UVNis detection at
216 nm. The mobile phase was 0.05 M phosphate buffer (pH 3.5)
phase flow rate was 1.0 mL/min and the sample injection volume
was 20 pL ...................................................................................... 81

Response curves for standard solutions of ATZ, HA, DEA and DIA
...................................................................................................... 84

Percentage of atrazine remaining in solution as a function of the
electrolysis time at two currents: 100 mA (22 mA/cm2) and 50 mA
(11 mA/cmz) ................................................................................... 86

xiv

Figure 4.6

Figure 4.7

Figure 4.8

Figure 5.1

Figure 5.2

Figure 5.3

Figure 5.4

Figure 5.5

Semilogarithmic plots of atrazine concentration versus electrolysis
time for a diamond/T i anode at two constant currents: 100 mA (22
mA/cmz) and 50 mA (11 mA/cmz) .................................................. 87

Percentage of atrazine remaining in solution as a function of the
electrolysis time at a constant current of 50 mA (11 mA/cmz) with
and without convection .................................................................. 88

Comparison of percentage of atrazine oxidized to 002 over a 120
min electrolysis period at a constant current of 100 mA (22 mA/cmz)
with and without convection ........................................................... 91

SEM images of (a) a bare diamond electrode surface without any
metal deposition, (b) a diamond surface after 90 s of deposition at -
0.45 V in 10 ppm Au(lll) + 1 M HCI solution, (c) a diamond surface
after 90 s of deposition at -0.45 V in 10 ppm As(lll) + 1 M HCI
solution and (d) a diamond surface after 90 s of deposition at -0.45
V in 10 ppm As(lll) and 10 ppm Au(lll) +1 M HCI ....................... 101

DPASV i-E curves for 100 ppb As(lll) + 1 M HCI at a diamond
electrode as a function of the Au(lll) concentration in solution. A
plot of the As stripping peak current versus the Au(lll) solution
concentration is provided as an inset. DPASV parameters: Edep = -
0.45 V. tdep = 120 3, pulse width = 50 ms, pulse amplitude = 50 mV,
step height = 4 mV, cycle period 200 ms, and sampling time = 8.33
ms ................................................................................................ 102

Cyclic voltammetric i-E curves for a diamond electrode in contact
with a 100 ppb Au(ll|) + 1 M HCI solution as a function of the As(lll)
solution concentration, which ranged from 500 to 1000 ppb. Scan
rate = 100 mV/s ........................................................................... 104

A comparison of the effect of the deposition potential on the As
stripping current density at a Au-coated diamond and Au foil
electrode in a solution of 200 ppb As(lll) and 50 ppb Au(lll) + 1 M
HCI. DPASV parameters: taep = 110 s, pulse width = 50 ms, pulse
amplitude = 50 mV, step height = 4 mV, cycle period 200 ms and
sampling time = 8.33 ms .............................................................. 105

Effect of deposition time on the As stripping peak current density at
a Au-coated diamond electrode in a solution of 200 ppb As(lll) and
50 ppb Au(lll) + 1 M HCI. DPASV parameters: Edep = -0.45 V, pulse
width = 50 ms, pulse amplitude = 50 mV, step height = 4 mV, cycle
period 200 ms and sampling time = 8.33 ms ............................... 106

XV

Figure 5.6 Effect of the pulse voltage amplitude on the As stripping peak
current density at a Au-coated diamond electrode in a solution of
200 ppb As(Ill) and 50 ppb Au(Il|) + 1 M HCI. DPASV parameters:
Edep = -0.45 V, tdep = 120 3, pulse width = 50 ms, step height = 4
mV, cycle period 200 ms and sampling time = 8.33 ms .............. 107

Figure 5.7 DPASV i-E stripping curves for As using standard additions of As(lll)
at (A) a Au-coated diamond electrode in a solution of 100 ppb Au(lll)
+ 1 M HCI and (B) a Au foil electrode in a solution of 1 M HCI. The
As(lll) concentration ranged from 1 to 50 ppb. A response curve in
the inset shows the proportionality between the As stripping peak
current and the As(lll) solution concentration. DPASV parameters:
Edep = -0.45 V, tdep = 120 5, pulse width = 50 ms, pulse amplitude =
50 mV, step height = 4 mV, cycle period 200 ms and sampling time
= 8.33 ms ..................................................................................... 109

Figure 5.8 A response curve showing the proportionality between the As
stripping peak current and the As(V) solution concentration at a Au-
coated diamond electrode. The As(V) concentration ranged from 1
to 50 ppb. DPASV parameters: Edep = -0.45 V, tdep = 120 s, pulse
width = 50 ms, pulse amplitude = 50 mV, step height = 4 mV, cycle
period 200 ms and sampling time = 8.33 ms ............................... 113

Figure 5.9 DPASV i-E curves and the corresponding response curve for the
addition of 2.5 - 40 ppb As(lll) to Red Cedar River water acidified
with 100 ppb Au(lll) + 1 M HCI on a Au-coated diamond electrode.
DPASV parameters: Edep = -0.45 V. tdep = 120 5, pulse width = 50
ms, pulse amplitude = 50 mV, step height = 4 mV, cycle period 200
ms and sampling time = 8.33 ms ................................................. 115

Figure 6.1 Effect of Cu(II) on the As stripping peak current at a Au-coated
diamond electrode in a solution of 50 ppb As(lll) and 100 ppb Au(lll)
+ 1 M HCI. (A) DPASV stripping voltammetric i-E curves. (B) Plot of
the normalized As stripping peak current as a function of the
concentration ratio of As(lll) and Cu(ll) in solution. DPASV
parameters: Edep = -0.45 V, tdep = 120 8, pulse width = 50 ms, pulse
amplitude = 50 mV, step height = 4 mV, cycle period 200 ms and
sampling time at the end of the pulse = 8 ms .............................. 127

Figure 6.2 Comparison of the As stripping voltammetric i-E curves in solutions
with and without interfering metal ions after passage through the
cation exchange resin. (A) Curves for 25 ppb As(lll) + 75 ppb Cu(ll)
and (B) Curves for 25 ppb As(lll) + 25 ppb Cu(ll) and 25 ppb As(lll)
+ 25 ppb Cu(ll), Hg(|l), Pb(ll) each. DPASV parameters: Eaep = ‘
0.45 V, tdep = 120 5, pulse width = 50 ms, pulse amplitude = 50 mV,

xvi

Figure 6.3

step height = 4 mV, cycle period 200 ms and sampling time at the
end of the pulse = 8 ms ............................................................... 130

Comparison of the effect of added humic substances (0 to 5 ppm)
on the As stripping peak charge in a solution of 50 ppb As(lll) and
100 ppb Au(lll) + 1 M HCI at a Au-coated diamond and 3 Au foil
electrode. DPASV parameters: Edep = -0.45 V, tdep = 120 5, pulse
width = 50 ms, pulse amplitude = 50 mV, step height = 4 mV, cycle
period 200 ms and sampling time at the end of the pulse = 8 ms
................................................................................................... .131

xvii

CHAPTER 1

INTRODUCTION

Ground water is one of the most important natural resources of the world.
For example, in the United States, ground water accounts for about 40 percent of
the water used for public utility and 97 percent of the drinking water for the rural
population who do not have access to public water-supply systems.1 However,
water pollution has been a serious problem all over the world. It is suggested to
be the leading cause of death and disease worldwide.2 More than 14,000 people
die daily due to water pollution.3 Water pollution has also expedited the depletion
of ground water. Reduction of ground water supplies are expected in many areas
of the world in the coming century as the population increases.2 Thus,
technologies to economically, reproducibly and quickly remediate and monitor
ground water contamination are of great need.

Electrochemical methods, being one of these technologies, have garnered
increasing attention because of their versatility in application, low energy
requirement, environmental compatibility (electrons as the main reagent),
portability and cost effectiveness. The electrode material plays an important role
in the application of the technology. The microstructure and morphological
stability of the electrode and its electrochemical properties will affect the
efficiency and outcome of the electrochemical method.

Electrically conducting diamond thin films provide scientists and engineers
with an entirely new type of electrode material that meets the requirements for a

wide range of electrochemical applications. Boron-doped diamond (BDD) is a

new sp3 carbon material that possesses attractive properties compared with sp2
carbon (e.g., glassy carbon) and metal electrodes (e.g., Hg). Its beneficial
electrochemical properties include: (i) low and stable background current, (ii)
wide working potential window, (iii) good activity for many redox analytes without
conventional electrode pretreatment, (iv) stable surface microstructure and
morphology over a wide potential range and (v) weak adsorption of polar
molecules (i.e., fouling resistance)“ The last two properties are key for long-
term treatment and monitoring of real water samples.

The research in this Dissertation consists of two major parts: (1)
application of a boron—doped diamond electrode on Ti substrate (diamond/T i) for
water treatment to study the physical and electrochemical properties of the
electrode before and after electrolysis using atrazine as a sample pollutant and
(2) developing a sensitive, stable and reproducible method for on-site inorganic

arsenic monitoring using a Au-coated boron-doped diamond electrode.

1.1 Importance of Water Treatment

1.1.1 Methods of Water Treatment

With the increase in public knowledge about the consequences of
environmental pollution and the development of modern industry, the demand for
better water quality has also increased. The US. Environmental Protection
Agency (USEPA) has set more strict regulations to control the effluent of
industrial wastewater. Thus, the role of wastewater treatment has become more

and more important to meet these strict regulations. Interest in the development

of advanced remediation and water purification technologies for the treatment of
hazardous chemical wastes has grown enormously in the past 50 years. The
methods used for highly contaminated wastewater treatment can be divided into
three major categories: physical methods, biodegradation and chemical
processes.8 Currently, a typical treatment plant will use a combination of the
three treatment methods. Physical methods are usually applied first to remove
suspended, undissolved pollutants by sedimentation, flocculation, filtration or
adsorption. Then biological and chemical methods are used to further remove the
remaining pollutants. Biological treatment techniques are well established and
relatively cheap. However, biodegradation depends on a variety of factors such
as the concentration of pollutants, their chemical structure and biodegradability of
the contaminants. The process is slow and not ubiquitous.9' 1° Chemical
processes are very important to efficiently remove persistent organic pollutants.
Oxidation is the most commonly used chemical process and it produces
degradation products that are non-toxic or easily bio-degradable intermediates.8
A special class of chemical oxidation techniques defined as advanced
oxidation processes (AOP) was developed to oxidize pollutants that are difficult
to deal with. The AOP methods are fairly new technologies developed since 1975
and are characterized by the capability to generate highly reactive free radicals,

such as hydroxyl radicals (0H1) by means of energy or catalyst.10

For example,
hydrogen peroxide can generate hydroxyl radicals when irradiated with UV light
(< 254 nm). These hydroxyl radicals are important for degradation of the

pollutants because they are able to add to a double bond, abstract hydrogen or

transfer an electron from a halogenated compound as shown in the reactions:10

- Addition to double bond 0H0 + X2C=CX2 —-> X2C(0H)CX2o
- Hydrogen abstraction 0H1 + RH -—> H20 + R-
- Electron transfer 0H0 + RX —> OH + XR+o

where R represents a typical hydrocarbon and X represents a halogenated group.
The oxidation products could be intermediates (e.g., H003', cr and N05)”
however, complete mineralization or ultimate degradation of the organics is ideal
because the final products are 002 and H20.

Processes of the AOP methods that are capable of generating the free
radicals include photochemical processes, ozonation processes and in situ
generation of free hydroxyl radicals (Fenton or electrochemical methods).8 The
oxidants and methods used in the different processes and the generated free
radicals are listed in Table 1.1.” ‘1 Many of the methods use a combination of
the different processes shown in Table 1.1. The photochemical processes utilize
artificial light source or natural sunlight to generate hydroxyl radicals. For
example, when hydrogen peroxide is added, the reaction is:

H202 + hv —+ 2 OH-
To effectively generate hydroxyl radicals, the photons have to be absorbed by
the chemical partners (03 or H202).11 Therefore, the intensity and wavelength of
the light source has to be carefully selected and the processes are generally slow
and leading to incomplete degradation.8 Ozonation processes utilize dissolved
ozone to attack C=C double bonds. The products of this reaction are compounds

containing carbonyl groups (aldehydes and ketones) and hydrogen peroxide as

well as a group of free radicals as shown in Table 1.1. When used in combination
with H202, the reaction is:11

203+H202—+302+20H0

 

 

 

 

Table 1.1 Chemical processes generating free radicals.
Oxidants and Generated Free
Methods Radicals
UV OH.
Ti02/UV 0H.
Photochemical
Processes 03/ UV 0H-
H202/UV OH.’ H02.
03/UV/H202 0“” 02"- H02" 03".
H030
0 0H°. 02". H02'. 03".
Ozonation Processes 3 H03'
Os/H 0 OH'- 02": H02" 03".
2 2 H030
H o /Fe2* 0H
In Situ Generation of 2 2 '
Free Radicals 2+ .
H202/Fe IUV OH.

 

 

 

Although the ozonation processes are very efficient at producing hydroxyl
radicals, they usually require in situ generation of ozone and are relatively
expensive.8 The Fenton reaction involves the reaction of H202 with Fe2+ to
generate hydroxyl radicals. Under acidic conditions, the reaction is:

H202 + Fe2+ —+ Fe?” + OH- + OH'

This method is efficient and capable of completely degrading organic compounds

to CO2. However, a potential environmental hazardous Fe2+ has to be added to
the water.

Electrochemical processes are good alternatives due to their versatility,
environmental compatibility and cost effectiveness. The electrochemical
oxidation of organic pollutants in water occurs at high anodic potentials with
concomitant generation of oxygen (02 evolution). The nature of the electrode
material is an important factor affecting the selectivity and efficiency of the
process. An anode material that possesses a wide working potential window,
long service-life in harsh chemical environments and resistance to surface fouling
and passivation are desired for good remediation efficiency.

Comninellis et al. proposed a model to comprehensively interpret the
anodic oxidation of organics in acid medium on different electrode materials.
They divided the electrodes into two limiting cases: active and nonactive anodes
(Figure 1.1).12 M represents the reaction sites on electrode surface and R
represents the organic pollutants. Hydroxyl radicals are produced by the
dissociation of water. The first step of reaction is the interaction of the hydroxyl

radicals with the electrode surface (Figure 1.1, reaction a).12

H++C'

 

 

 

 

 

Activcanodc + +R M+RO
c aMO e
d M+'/202
M OH. M(0H)0 02 6v: mm? M + '/2 02 + H+ + c'
a
+R
f = M+mC02+nH20+H++€

Nonactive anode

Figure 1.1 Scheme of the oxidation of organic compounds with simultaneous
02 evolution: (3) interaction of the electrode surface with 0H1, (b) 02 evolution by
electrochemical oxidation of 0H1, (c) formation of the higher oxide on an active
anode, (d) 02 evolution by chemical decomposition of the higher oxide, (e)
oxidation of the organics (R) via the higher oxide on an active anode and (f)

complete oxidation of the organics via hydroxyl radicals on a nonactive anode.

An active anode is defined as one with a strong interaction with the
hydroxyl radicals. Through the interaction, hydroxyl radicals transfer oxygen to
the electrode surface and form a higher oxide M0 (Figure 1.1, reaction c). The
higher oxide M0 can mediate the oxidation of organic pollutants R (Figure 1.1,
reaction e) or dissociate back to M and 02 (Figure 1.1, reaction d). The oxidation
of organics mediated by the higher oxide M0 tends to be much more selective
than the oxidation reaction mediated by hydroxyl radicals on a nonactive anode.
Nonactive anode has weak interaction with hydroxyl radicals. Instead of forming

higher oxide on the electrode surface, hydroxyl radicals on an active anode

mediate the oxidation of organics and result in fully oxidized product, 002 (Figure
1,1, reaction f). In this case, the electrode serves as an inert substrate and a sink
for the removal of electrons without participating in anodic reactions.

The hydroxyl radicals can also result in 02 formation, which is a
competitive process with the oxidation of organics (Figure 1.1, reaction b).12 The
nature of the electrode material determines the strength of its interaction with
hydroxyl radicals, and thus affects the reactivity of 02 evolution and the oxidation
efficiency of organic pollutants. In general, nonactive anodes are desired
because they have weak interaction with hydroxyl radicals, and thus have low

activity toward the formation of 02 and high efficiency for pollutant oxidation.12

1.1.2 Electrode Materials and Their Limitations

Electrochemical remediation treatments are often limited by the availability
of a suitable anode material that is not active toward 02 evolution and electrode
oxidation. Properties such as a wide working potential window, long service-life in
harsh chemical environments, and resistance to surface fouling and passivation
are desired for anode material used for electrochemical treatment of wastewater.
Several anode materials have been used in the past, such as Pt, glassy carbon,
Ir02/T i, Sn02-Sb205fTi, Pb02, and Sn02.13"16 These electrodes all have
limitations in terms of their effectiveness for electrochemical remediation. Some,
like glassy carbon, undergo rapid loss of activity due to surface fouling.13 Toxic
ions are liberated from others due to corrosion, such as Pb2+ from Pb02.14 The

service-life of anodes that undergo corrosion at significant rates is another

limiting factor. For example, the service life of Ti/Sn02-Sb205 is only about 12 h in

1 M H2SO4 due to surface oxidation and corrosion.15

1.1.3 Boron-Doped Diamond and Substrate Materials

A dimensionally stable anode is necessary for the electrochemical
remediation of chemical waste, which can operate in chemically harsh
environments with unchanging performance, resist fouling by matrix components
and efficiently support anodic oxygen transfer reactions. Diamond electrodes
appear to meet all of these requirements. Diamond electrodes have been
successfully used for electrochemical oxidation of chemical compounds in
aqueous media.”21 lniesta et al. used a boron-doped diamond electrode to
oxidize 3-methylpyridine (3-MP). They found that at low current density, 3—MP
was oxidized to nicotinic acid, while at high current density it was directly
converted to CO2.17 Canizares et al. used a diamond eleCtrode to oxidize
carboxylic acid wastes in water by galvanostatic electrolysis. 10 mM of formic
acid, oxalic acid and maleic acid was fully or partially oxidized to 002 within 210
min. No intermediates were found.18 Codognoto et al. studied the selective
oxidation of pentachlorophenol (PCP) on diamond electrode, they found that
under fairly mild conditions (lower applied potential), a dimerization process leads
to the formation of an insoluable species; at more positive potentials, PCP was
initially converted to hydroxyl-quinone and finally to CO2 and water.19 Real
wastewater containing organic pollutants has also been studied by Panizza on a

p-Si substrate diamond electrode.21 All their work showed that diamond

electrodes can effectively degrade the organics in water without losing reactivity.
Electrically conducting (< 0.05 Q-cm) boron—doped diamond thin films can
be deposited on different substrate materials using chemical vapor deposition
(CVD). Currently, the most popular substrate used for diamond growth is Si due
to the simplicity of growth. However, boron-doped diamond electrodes grown on
a Si substrate (diamond/Si) have some limitations for wastewater treatment
applications due to their mechanical weakness and low electrical conductivity.22
To overcome these limitations, diamond films grown on metal substrates can be
used. Despite the good performance of Nb, Ta, and W as substrates for diamond
growth, the high cost of these metals is a disadvantage for large-scale industrial
application. Ti is much cheaper and is a technologically-important metal that
possesses good electrical conductivity and high mechanical strength.23 It is much
more attractive than the others listed above for wastewater treatment
applications. Diamond/T i has been used as an electrochemical electrode for the
remediation of chemical contaminants in water. For example, Chen and
coworkers applied a diamond/T i anode for the oxidation of carboxylic acids,
phenol, orange II, and reactive red HE-3B.23 Fryda and coworkers deposited
large area diamond films on different substrate materials, including Ti, and used
them for the oxidation of alcohols, organic acids and halogenated aromatic
molecules.24 They found that the organic molecules could be oxidized to 002
without major amounts of other detectable by-products even at very low
concentrations (<3 ppm). Even with these reported results, there remains a need

to comprehensively evaluate the electrochemical properties and remediation

10

performance of diamond films coated on Ti substrates.

1.1.4 The Herbicide — Atrazine

In this work, atrazine was the organic pollutant used due to its resistance
to oxidation. Atrazine [2-chloro-4-(ethylamino)-6-(isopropyl-amino)—s-triazine] is a
compound of particular interest due to its wide use as a herbicide”: 26 It is a
chlorinated triazine compound that is difficult to oxidatively degrade using
conventional electrode materials.27 Atrazine undergoes slow hydrolysis, has low
vapor pressure (3><10'7 mm Hg at 20°C), is moderately soluble in water (30 ppm)
and shows moderate adsorption to organic matter and clay. These properties
make atrazine relatively persistent in soils and aquatic systems.28 From 1991 to
1995, thousands of sampled wells and springs across the US. were studied by
the United States Geological Survey (USGS). Among the seven herbicides of
interest (atrazine, cyanazine, simazine, alachlor, metolachlor, prometon and
acetochlor), atrazine was the only one found to exceed the allowable
concentration level in drinking water (3 ppb), as mandated by the USEPA.
Studies have shown that atrazine is harmful to both animals and plants.

Several methods have been employed to treat atrazine-contaminated

29' 3° and photochemical degradation.“ 32 There are

water including ozonation
only a couple of reports describing the electrochemical treatment of such
contaminated water.27' 33 Ventura and coworkers used the electrochemical

generation of Fenton’s reagent and a mercury electrode to degrade atrazine and

observed efficient atrazine degradation with the formation of deisopropylatrazine,

11

deethylatrazine and deethyI-deisopropylatrazine. They claimed no trace of
hydroxyatrazine could be found.” Saltmiras and Lemley also performed atrazine
degradation by anodic Fenton treatment using iron electrodes as both a
sacrificial anode and an inert cathode. They reported that 70% of the atrazine
was degraded in 3 min with the formation of dechlorinated ammeline as the
primary product after 10 min.27 Electrodes that can completely degrade atrazine
to 002 are needed as well as fundamental studies of the reaction kinetics and
mechanisms.
The research on water treatment in this dissertation sought to answer
several key questions:
(1) What are the morphological and electrochemical properties of
diamondfl' i?
(2) What is the efficiency of diamond/Ti for electrochemical oxidation of
atrazine and if the electrode is stable?
(3) What are the oxidation reaction products?

(4) What is the conversion efficiency of the final oxidation product 002?

1.2 Importance of Water Quality Monitoring

1.2.1 Metal Ion Contaminants and Arsenic

Water quality monitoring is very important to determine the level of
contaminants in water. Trace elements, especially heavy metals, are one of the
major groups of contaminants in ground water.34 Heavy metals such as copper,

lead, and mercury are discharged from mining, smelting and industrial

12

manufacturing and can cause long-term effects on the environment and human
health due to their extremely toxic nature even at trace concentrations, and their
non-biodegradability.35 The effects of hazardous metal ions are caused by
mimicking of the essential elements.36 Effects such as birth defects, cancer, skin
lesions, mental and physical retardation, learning disabilities and liver and kidney
damage are linked to many metal contaminations.35

Among the trace element contamination, arsenic contamination has been
one of the most common and harmful types.37 Arsenic naturally occurs in the
ground and can be introduced into water through the erosion of rocks, minerals
and ores. Industrial effluents are another source of arsenic in the environment.
Ground water arsenic contamination has been reported in many countries of the
world, such as Argentina, Bangladesh, Cambodia, Chile, China, Ghana, Hungary,
Mexico, Mongolia, Nepal, New Zealand, Philippines, Taiwan, United States and
Vietnam.38 Chronic exposure to As-contaminated water can cause several health
problems including pigmentation changes, hyperkeratosis and various cancers
(e.g., skin, lung, urinary bladder, and kidney).39 Symptoms can begin with
exposure to concentrations as low as 50 ppb.37

Arsenic has been put into the heavy metal category because of its
semimetal electronic properties. However, being in group five in the periodic
table, arsenic also has many nonmetallic properties that are similar to nitrogen
and phosphorus. Due to its valence electron configuration shown in Figure 1.2,

arsenic can have all oxidation states between - 3 and + 5.

13

4s 4p 4d
II I I T

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 1.2 The valence electron configuration of arsenic in the ground state.

In ground water, the As(lll) and As(V) inorganic forms are the most abundant
forms and are much more toxic than the organic forms, with As(lll) being the
most toxic”: 4° The research in this dissertation is therefore focused on the

detection of the toxic inorganic forms: As(lll) and As(V) in water.

1.2.2 EPA Approved Methods for As Detection

In an effort to better protect society, the USEPA recently lowered the
maximum contaminant level (MCL) for total arsenic in drinking water from 50 to
10 ppb.41 This lowering creates a need for more sensitive, reliable and stable
detection methods. The USEPA-recommended methods for arsenic analysis
presently include inductively coupled plasma-mass spectrometry (ICP-MS)
(Method 200.8), inductively coupled plasma-atomic emission spectrometry (ICP-
AES) (Method 200.7) and graphite furnace atomic adsorption spectrometry
(GFAAS) (Method 200.9). Each of these methods provide detection limits at or
below the MCL (1.4, 8 and 0.5 ppb, respectively);37 however, they are expensive
in terms of the instrumentation cost, time consuming, labor intensive and the
equipment is not easily deployed in the field. There is also a USEPA approved
anodic stripping voltammetric method that makes use of a Au-coated glassy

carbon electrode (Method 7063).42

14

1.2.3 Differential Pulse Anodic Stripping Voltammetry (DPASV)

Anodic stripping voltammetry (ASV) is generally used for metal ion
analysis in water samples due to its high sensitivity and good resolution. ASV is
capable of measuring four to six metal ions simultaneously at concentrations as
low as the sub part-per-billion (ppb) level. The instrumentation is capable of
being used on-site and is relatively low-cost.“'3

ASV measurements involve two steps: (1) deposition of a metal phase
onto a solid electrode surface or into Hg (liquid) at cathodic potentials and (2)
selective electrooxidation of predeposited metal on the electrode surface as

shown in Figure 1.3.

Step Praise
(1) Deposition step .. W1 rim 3"" W
(2)Stnpp1ng step m

—u

PIloePeflod

 

Potentlal (V)

 

 

 

Potential (V)—>

 

Current (A)

 

Figure1.3 The differential pulse anodic stripping voltammetric (DPASV)
potential waveform and resulting current response. (1) The deposition step

involves a cathodic potential-step of certain magnitude and duration and (2) the

15

stripping step involves a potential scan in the positive direction to oxidize the
predeposited metal phase formed in step (1). An enlarged view of the applied
pulse program depicting the various parameters is also shown. The resulting
voltammetric i-E curve is constructed by plotting the current difference of the two

sampling points for each pulse as a function of the potential.

Due to the two-step measurement scheme, ASV offers a 2 to 3 order of
magnitude lower limit of detection as compared with solution-phase voltammetric
measurements.43 In the deposition step, a controlled potential is applied for a
definite time under reproducible hydrodynamic (mass-transport) conditions in the
solution. The trace metal ions of interest are reduced and preconcentrated on the
electrode surface at the controlled cathodic potential:

Mrl+ + n e' —+ M
The preconcentration greatly enhances the amount of metal at the electrode
surface and it can be improved by increasing the deposition time or optimizing
the deposition potential.43 In the stripping step, pulses of equal amplitude are
superimposed on a positive potential sweep and the predeposited metal is
oxidized and stripped from the electrode surface:

M - n e' —> M'”
The pulses usually have a 10-100 mV pulse amplitude, a 5-100 ms pulse width
and a 0.5-4 s pulse period. During each pulse period, the current is sampled
twice: once right before the application of the pulse and once right before the

termination of the pulse. The current difference between these two sampling

16

points is plotted against the scanned potential (Figure 1.3). Using this method,
the charging current is subtracted from the total current and the signal to
background ratio is greatly enhanced.43 This is because when applying the pulse,
the faradaic current and charging current both increase. However, the charging
current (ic ~ e") decays much faster than mass transfer limited faradaic current (ii
~ t'1’2). At the sample point right before the termination of the pulse, the current is
mostly faradaic current. At both sampling points, the charging current is
essentially the same, so the current difference between these two sampling
points is only attributed to faradaic current thus increasing the signal to
background ratio and sensitivity of the method.“ A further gain of the signal to
background ratio is obtained because the stripped metal ions could redeposit
onto the electrode surface during the waiting period between pulses, thus the
same material can be repetitively stripped and deposited and contribute to the
total current at potentials negative of the standard redox potential.43

The equation describing the differential pulse stripping peak current was
derived by Lund and Onshus for hanging mercury drop electrode:44

Aip = knerEUI’ztdCb

where k is the heterogeneous electron transfer rate constant, n is the number of
the electrons transferred per equivalent, r is the radius of the mercury drop, AE is
the pulse amplitude, u is the rotation speed of the stirrer, td is the deposition
period and Cb is the concentration of metal ion in the sample solution. Even

though it might not correctly describe the current response for a planar electrode,

the equation shows the linear relationship between the stripping peak current and

17

the deposition time, pulse amplitude and metal ion concentration. These are

important parameters to study when optimizing the stripping peak current.

1.2.4 Electrode Materials and Their Limitations for Arsenic
Detection

For arsenic monitoring using DPASV method, an important factor that can
strongly influence the detection is the electrode material. Arsenic detection has
been accomplished using different electrodes: Hg,‘°’9"“"5“‘8 PL”!51 Au‘w' 52'” and

Au-coated carbon electrodes (graphite‘o' 54' 57' 58

and glassy carbon‘m“). The
limited solubility of As in Hg, as well as its toxicity make Hg less desirable for this
application. Platinum electrodes are of limited use because of a narrow potential
window and large background current.53 It turns out that Au is the electrode of
choice for measuring this analyte. This is because of the Au-As interrnetallic
compound that forms during the preconcentration step with the general formula
AuxAsy, where x=1-3 and y=2-6.62 In this form, all of the deposited As can be
oxidatively removed at lower potentials than the Au, thus enabling its detection.
Tikhomirova and Akhmedzhanova analyzed the Au-As interrnetallic compound
stability in HCI under different conditions using X-ray diffraction and found that
the compounds decompose to arsenic oxides of the arsenolite type (A8203) and
fine-grained gold when exposed to oxidizing potentials.62' 63 This facilitates the
release of the As(lll) during stripping step. Limits of detection (LOD) in the low
ppb to sub-ppb range for As(lll) are typical for Au and Au-coated carbon

electrodes. Generally, Au particles supported on carbon materials provide lower

18

LODs and are more cost effective than solid Au electrodes due to the lower mass
of metal needed. To cite a few examples in the literature, Kopanica et al.
detected As(lll) by ASV using a Au disc electrode and achieved an LOD of 0.15
ppb.64 Feeney and Kounaves used a Au ultramicroelectrode array with 80 s
deposition time and reported an LOD of 0.013 ppb and a sensitivity of 0.044
ttA/ppb.40 Dai et al. reported a low LOD of 0.0096 ppb for As(lll) in 1 M HCI using
linear sweep voltammetry with a Au-coated glassy carbon electrode,59 and Simm
et al. investigated the sonochemical detection of As with a Au electrode,
obtaining an LOD of 0.28 ppb.52 A general limitation with Au and Au-coated sp2
carbon electrodes is the poor response precision and stability. Examples of this
include the work by Davis et al. and Viltchinskaia et al. who reported passivation
of the Au-coated graphite,“ 6° Feeney and Kounaves who observed a 30%
attenuation of the As stripping peak current with a Au ultramicroelectrode over 30
successive runs,40 and Prakash et al. who reported a 20% variation in the results
for 5 measurements using a Au disk rotating electrode.56

Diamond is a suitable alternate to Hg for ASV and has been successfully
applied for the detection of several metal ions: Ag(l),4' 65' 66 Cu(ll),“ 67 Pb(ll),4' 6759
Cd(ll),4' 69' 7° Zn(ll),4 Mn(ll),71 Sn(l|),65 Hg(ll)72 and Au(lll).73 However, there are
only a few reports of arsenic detection using diamond electrodes. Simm et al.
detected As(lll) using a Au plated BDD microelectrode array and reported an
LOD of 2.85 ppb and a sensitivity of 17.6 th/ppm.74 Hignett et al. detected As(lll)
using linear sweep voltammetry by iodine-mediated oxidation of As(lll) to As(V).

They compared diamond with Au, Pt and glassy carbon and found that the LOD

19

to be lowest for BDD.50 Salimi et al. modified a BDD electrode by
electrodepositing hydrous iridium oxide, which enabled the oxidative detection of
As(lll) to As(V) at 0.5 V. They reported an LOD of 0.15 ppb and a sensitivity of 56
llA/ppm.75
The research on water monitoring in this Dissertation sought to answer
several key questions:
(1) What are the morphological and electrochemical properties of Au-coated
diamond electrode and can it be used for As analysis in water?
(2) What is the nature of Au deposition on diamond in terms of particle
nucleation and growth, size and dispersion?
(3) How do the figures of merit for a Au-coated diamond electrode compare
with planar Au and Au-coated sp2 carbon electrodes for arsenic monitoring?
(3) How can common interferences (e.g. Cu(ll) and humic material) in real
water be minimized?
(4) How do the electrode and method function for total inorganic As analysis

in real samples?

1.3 Dissertation Outline

This dissertation describes studies using boron-doped diamond electrode
for the application in both water treatment and water quality monitoring.
Specifically, boron-doped diamond electrode grown on Ti substrate (diamond/T i)
was applied for atrazine remediation and Au-coated diamond electrode was

applied for total inorganic arsenic monitoring in water supplies. In Chapter 2, the

20

experimental parameters of the work including materials, instrumentation and
methods are presented. The water treatment studies are presented in Chapters 3
and 4. In Chapter 3, evaluation of the physical and electrochemical properties as
well as the stability of diamond/T i electrode is presented. Application of the
electrode for the electrochemical remediation of atrazine is described in Chapter
4. The water monitoring studies are presented in Chapters 5 and 6. In Chapter 5,
the Au-coated diamond electrode is evaluated in terms of the morphology,
response stability and detection figures of merit for inorganic arsenic detection.
Its performance for As detection is compared with planar Au electrode and Au-
coated glassy carbon electrode. Chapter 6 presents the application of the
established method for real water analysis with demonstrated mitigation of
common interferences such as trace metal cations and humic material. The
method is demonstrated for two different contaminated water samples provided
by the US. Bureau of Reclamation. Finally, Chapter 7 provides a summary of the

conclusions as well as future directions for this research.

21

1.4

(1)
(2)

(3)

(4)

(5)

(7)

(8)

(9)

(10)

(11)
(12)

(13)

(14)

References

U. 8. Geological Survey 1995.

Pink, D. H. http://finance.yahoo.com/columnist/article/trenddesk/3748?
p=_1, Investing in Tomorrow's Liquid Gold, Accessed July 23, 2006.

West, L. http://environment.about.com/ch/environmentqlevents/a/water
da mahtm, World Water Day: A Billion People Worldwide Lack Safe
Drinking Water, Accessed July 23, 2006.

Sonthalia, P.; McGaw, E.; Show, Y.; Swain, G. M. Analytica Chimica Acta
2004, 522, 35-44.

Muna, G. W.; Tasheva, N.; Swain, G. M. Environmental Science and
Technology 2004, 38, 3674-3682.

Granger, M. 0.; 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. Analytical Chemistry 2000, 72, 3793-3804.

Strojek, J. W.; Granger, M. C.; Swain, G. M.; Dallas, T.; Holtz, M. W.
Analytical Chemistry 1996, 68, 2031 -2037.

Merli, 0.; Petrucci, E.; Pozzo, A. D.; Pemetti, M. In Wastewater Treatment
Technologies and Environment; Kumar, A., Ed.; Daya Publishing House:
Tri Nagar, 2004; Vol. 1, pp 48-62.

Troster, l.; Fryda, M.; Herrmann, D.; Schafer, L.; Hanni, W.; Perret, A.;
Blaschke, M.; Kraft, A.; Stadelmann, M. Diamond and Related Materials
2002, 11, 640-645.

Tabrizi, G. B.; Mehrvar, M. Journal of Environmental Science and Health
2004, A39, 3029-3081.

Gulyas, H. Water Science and Technology 1997, 36, 9-16.

Marselli, B.; Garcia-Gomez, J.; Michaud, P.-A.; Rodrigo, M. A.;
Comninellis, C. Journal of the Electrochemical Society 2003, 150, D79-
D83.

Gattrell, M.; Kirk, D. W. Canadian Journal of Chemical Engineering 1990,
68, 997-1003.

Tahar, N. B.; Savall, A. Journal of the Electrochemical Society 1998, 145,
3427-3434.

22

(15)

(16)

(17)

(18)

(19)

(20)

(21)

(22)

(23)

(24)

(25)

(26)

(27)

(28)

(29)

Correa-Lozano, B.; Comninellis, C.; Battisti, A. D. Journal of Applied
Electrochemistry1997, 27, 970-974.

lniesta, J.; Michaud, P. A.; Panizza, M.; Cerisola, G.; Aldaz, A.;
Comninellis, C. Electrochimica Acta 2001, 46, 3573-3578.

lniesta, J.; Michaud, P. A.; Panizza, M.; Comninellis, C. Electrochemistry
Communications 2001, 3, 346-351.

Canizares, P.; Garcia-Gomez, J.; Lobato, J.; Rodrigo, M. A. Industrial and
Engineering Chemistry Research 2003, 42, 956-962.

Codognoto, L.; Machado, S. A. S.; Avaca, L. A. Journal of Applied
Electrochemistry 2003, 33, 951 -957.

Panizza, M.; Michaud, P. A.; Cerisola, G.; Comninellis, C. Journal of
Electroanalytical Chemistry 2001, 507, 206-214.

Panizza, M.; Michaud, P. A.; Cerisola, G.; Comninellis, C.
Electrochemistry Communications 2001, 3, 336-339.

Chen, X.; Chen, G. Journal of the Electrochemical Society 2004, 151,
8214-B219.

Chen, X.; Chen, G.; Gao, F.; Yue, P. L. Environmental Science and
Technology 2003, 37, 5021 -5026.

Fryda, M.; Dietz, A.; Herrmann, D.; Hampel, A.; Schafer, L.; Klages, C.-P.;
Perret, A.; Haenni, W.; Comninellis, C.; Gandini, D. Proceedings -
Electrochemical Society 2000, 99-32 (Diamond Materials), 473-483.
National Agriculture Statistics Service 1993.

Environmental Fact Sheet, Atrazine Label Amendment, January 23, 1990,
US EPA Office of Pesticide Programs.

Saltmiras, D. A.; Lemley, A. T. Water Research 2002, 36, 51 13-51 19.

Erikson, L. E.; Lee, K. H. Critical Reviews in Environmental Control 1989,
19, 1-14.

Benitez, F. J.; Beltran-Heredia, J.; Gonzalez, T.; Acero, J. L. Ozone:
Science & Engineering 1995, 17, 237-258.

23

(30)

(31)

(32)
(33)
(34)
(35)

(36)
(37)
(38)
(39)
(40)
(41 )
(42)
(43)
(44)
(45)
(45)
(47)

(48)

Beltran, F. J.; Garcia-Araya, J. F.; Alvarez, P. M.; Rivas, J. Journal of
Chemical Technology & Biotechnology 1998, 71, 345-355.

Bourgine, F. P.; Chapman, J. l.; Kerai, H.; Duval, J. L.; Green, J. 6.;
Hamilton, D. Journal of the Chartered Institution of Water and
Environmental Management 1995, 9, 417-423.

Hequet, V.; Gonzalez, 0.; Le Cloirec, P. Water Research 2001, 35, 4253-
4260.

Ventura, A.; Jacquet, G.; Berrnond, A.; Camel, V. Water Research 2002,
36, 3517-3522.

Schwartz, F. W.; Zhang, H. Fundamentals of Ground Water, John Wiley &
Sons, Inc.: New York, 2003.

Singh, P.; Cameotra, S. S. Biochemical and Biophysical Research
Communications 2004, 319, 291-297.

Anon, E. Principles of Biochemical Toxicology, John Timbrell ed., 2000.
World Health Organization, May 2001, pp Fact sheet N. 210.

Dal, X.; Compton, R. G. Electroanalysis 2005, 17, 1325-1330.

Li, H.; Smart, R. B. Analytica Chimica Acta 1996, 325, 25-32.

Feeney, R.; Kounaves, S. P. Analytical Chemistry 2000, 72, 2222-2228.
Ferreira, M. A.; Barros, A. A. Analytica Chimica Acta 2002, 459, 151-159.
Pyle, S.; Miller, E. L. USEPA 1996.

Wang, J. Stripping Analysis; VCH publishers, Inc.: Deerfield Beach, 1985.
Lund, W.; Onshus, D. Analytica Chimica Acta 1976, 86, 109-122.
Sadana, R. S. Analytical Chemistry 1983, 55, 304-307.

Holak, W. Analytical Chemistry 1980, 52, 2189-2192.

Barra, C. M.; Correia dos Santos, M. M. Electroanalysis 2001, 13, 1098-
1104.

Adeloju, S. B.; Young, T. M.; Jagner, D.; Batley, G. E. Analytica Chimica
Acta 1999, 381, 207-213.

24

(49)
(50)

(51)
(52)

(53)

(55)

(55)
(57)

(58)

(59)
(50)

(51)

(52)
(53)

(64)

Williams, D. G.; Johnson, D. 0. Analytical Chemistry 1992, 64, 1785-1789.

Hignett, G.; Wadhawan, J. D.; Lawrence, N. S.; Hung, D. 0.; Prado, 0.;
Marken, F.; Compton, R. G. Electroanalysis 2004, 16, 897-903.

Wei, 2.; Somasundaran, P. Journal of Applied Electrochemistry 2004, 34,
241-244.

Simm, A. 0.; Banks, 0. E.; Compton, R. G. Analytical Chemistry 2004, 76,
5051-5055.

Forsberg, G.; O'Laughlin, J. M.; Megargle, R. G. Analytical Chemistry
1975, 47, 1586-1592.

Davis, P. H.; Dulude, G. R.; Griffin, R. M.; Matson, W. R.; Zink, E. W.
Analytical Chemistry 1978, 50, 137-143.

Feeney, R.; Kounaves, S. P. Talanta 2002, 58, 23-31.

Prakash, R.; Srivastava, R. 0.; Seth, P. K. Electroanalysis 2003, 15, 1410-
1414.

Kamenev, A. l.; Orlov, S. E.; Lyakhov, A. B. Jouranal of Applied
Electrochemistry 2001, 56, 850-854.

Simm, A. 0.; Banks, G. E.; Wilkins, S. J.; Karousos, N. G.; Davis, J.;
Compton, R. G. Analytical and Bioanalytical Chemistry 2005, 381, 979-
985.

Dai, X.; Nekrassova, 0.; Hyde, M. E.; Compton, R. G. Analytical
Chemistry 2004, 76, 5924-5929.

Viltchinskaia, E. A.; Zeigman, L. L.; Garcia, D. M.; Santos, P. F.
Electroanalysis 1997, 9, 633-640.

Munoz, E.; Palmero, S. Electroanalysis 2004, 16, 1528-1535.

Tikhomirova, V. l.; Akhmedzhanova, G. M.; Nekrasov, I. Y. Experiment in
Geosciences 1994, 3, 1-9.

Akhmedzhanova, G. M.; Tikhomirova, V. l. Experiment in Geosciences
1995, 4, 26-28.

Kopanica, M.; Novotny, L. Analytica Chimica Acta 1998, 368, 211-218.

25

(55)

(55)

(57)

(58)

(59)

(70)

(71)

(72)

(73)

(74)

(75)

Chatterjee, A.; Wiltshire, R.; Holt, K., B.; Compton, R. G.; Foord, J. S.;
Merken, F. Diamond and Related Materials 2002, 11, 646-650.

Saterlay, A. J.; Marken, F.; Foord, J. S.; Compton, R. G. Talanta 2000, 53,
403-415.

Prado, 0.; Wilkins, S. J.; Marken, F.; Compton, R. G. Electroanalysis 2004,
14. 262-272.

Manivannan, A.; Tryk, D. A.; Fujishima, A. Electochemical and Solid-State
Letters 1999, 2, 455-456.

Babyak, 0.; Smart, R. B. Electroanalysis 2004, 16, 175-182.

Banks, 0. E.; Hyde, M. E.; Tomcik, P.; Jacobs, R.; Compton, R. G.
Talanta 2004, 62, 279-286.

Saterlay, A. J.; Foord, J. S.; Compton, R. G. Analyst 1999, 124, 1791-
1796.

Manivannan, A.; Seehra, M. S.; Fujishima, A. Fuel Processing Technology
2004, 85, 513-519.

Hian, L. 0.; Grehan, K. J.; Goeting, C. H.; Compton, R. G.; Foord, J. 8.;
Marken, F. Electroanalysis 2003, 15, 169-174.

Simm, A. 0.; Banks, 0. E.; Ward-Jones, 8.; Davies, T. J.; Lawrence, N. S.;
Jones, T. G. J.; Jiang, L.; Compton, R. G. Analyst 2005, 130, 1303-1311.

Salimi, A.; Hyde, M. E.; Banks, 0. E.; Compton, R. G. Analyst 2004, 129,
9-14.

26

CHAPTER 2
EXPERIMENTAL METHODS

2.1 Diamond Thin Film Growth

Microcrystalline boron-doped diamond thin film was grown on a Ti or Si
substrate using microwave plasma-assisted chemical vapor deposition (1.5 KW,

2.54 GHz, ASTeX lnc., Lowell, MA). A diagram of a typical reactor system is

shown in Figure 2.1.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

I ROWY Pump IZTo vent

I BzHo cylindeID-IXI
I CH.cylinder)'IX'r

I H2 cylinder

 

 

 

 

 

 

Figure 2.1 Diagram of a typical microwave CVD reactor set-up. (Copyright ©
2003 From Electroanalytical Chemistry by Swain, G. M./Rubinstein, |..

Reproduced by permission of Routledge/Taylor & Francis Group, LLC)

Each component of the source gas mixture (0H4, H2 and B2H6) was

introduced into the reaction chamber at a flow rate determined by a mass flow

controller. Microwave energy is generated and focused into the quartz chamber
(1 L bell jar) to ignite the plasma. The plasma, which is in close proximity to the
substrate, contains the reactive species important for growth. The chamber
pressure is maintained by a throttle exhaust valve and rotary pump that is
operated continuously during film growth.

Microcrystalline boron-doped diamond thin-film on Ti substrate
(diamond/T i) was grown on a polycrystalline Ti plate (0.15 x 2 x 2 cm3, Sigma-
Aldrich Chemical or Goodfellow Metals). The Ti substrate was first degreased
sequentially with acetone, isopropanol and water, and then cleaned and
roughened in a solution of 1% HF for 10 min. This was followed by ultrasonic
cleaning for 3 min each in acetone, isopropanol and ultrapure water. The metal
substrate was then dried in air before being mechanically polished on a felt pad
for 5 min using 0.1 pm diamond powder (GE Superabrasives, Worthington, OH)
slurried with ultrapure water. The polished substrate was then once again
ultrasonically cleaned in acetone, isopropanol and ultrapure water for 3 min each
to remove polishing debris. After cleaning, the substrate was examined under an
Olympus BX60M optical microscope (Olympus America Inc.) to assess the
surface morphology and cleanliness (removal of polishing debris). Finally, the
substrate was placed in the CVD reactor for diamond growth (base pressure ~10
mtorr). The microwave power used for film growth was 500 W. The source gas
mixture was 0.5% 0H4 in 99.5% H2 (v/v) with 10 ppm B2H6 added for boron
doping. Ultrahigh purity (99.999%) CH4 (AGA Specialty Gas, Cleveland, OH) and

(H2 (B00 Group, Inc., Murray Hill, NJ) were used as the source gases. The gas

28

mixture was supplied to the reaction chamber at a flow rate of 197 sccm for H2,
1.00 sccm for CH4, and 2.00 sccm for the 0.1% B2H6 (Matheson) in H2 to achieve
a 10 ppm boron gas phase concentration and a 200 sccm total flow rate.
Diamond was deposited for 10 h. The growth temperature was estimated by
optical pyrometry to be between 700 and 800 °C. After deposition, the CH4 and
B2H6 gas flows were stopped (plasma still ignited) and the films remained
exposed to an H2 plasma (atomic hydrogen) for 10 min under the deposition
conditions. The samples were cooled in the H2 plasma to a temperature less than
400 °C by slowly reducing the power and pressure over a 15 min period.
Microcrystalline boron-doped diamond thin-film on Si substrate
(diamond/Si) was grown on a p-type Si (100) substrate (~10'3 Q-cm, Virginia
Semiconductor Inc., Fredricksburg, VA). The substrate (0.1 x 1 x 1 cm3) was first
cleaned sequentially with ultrapure water, methanol, and acetone, and then hand
polished on a felt pad for 5 min with 0.1 pm diam. diamond powder (GE
Superabrasives, Worthington, OH) mixed with ultrapure water. The substrate was
then ultrasonically cleaned in acetone, isopropanol and ultrapure water (5 min
each) to remove the polishing debris. The clean scratches and residual diamond
particles serve as the initial nucleation sites for diamond growth. The growth
conditions for diamond/Si were the same as those for diamond/Tl described
above except that the microwave power used for diamond/Si growth was 1000 W.
The resulting film properties (thickness, resistivity, carrier concentration and hole
mobility) are listed in Table 2.1. The boron dopant concentration was determined

from boron nuclear reaction analysis measurements of other films grown in an

29

identical manner. The film resistivity was measured with a tungsten four-point

probe.

Table 2.1 Typical microcrystalline boron-doped diamond thin film propertieS" 2.

 

Microcrystalline Diamond Electrode

 

Film Thickness 5-7 pm
Resistivity s 0.05 Q-cm
Carrier Concentration ~ 1 x 1020 cm'3
Carrier Mobility 1-10 cmZN-s

 

2.2 Material Characterization

The physical, chemical and electrochemical properties of the diamond/T i
electrode were probed by scanning electron microscopy, Raman spectroscopy,
X-ray diffraction analysis, electrical measurements and electrochemical

measurements.

2.2.1 Scanning Electron Microscopy (SEM)

The film morphology was probed by SEM using a S-4500 Hitachi electron
microscope equipped with LaB6 emitter at an accelerating voltage of 15 W or a
JSM-6300 F electron microscope (JEOL, Ltd., Tokyo, Japan) at an accelerating
voltage of 10 W. A 15 mm working distance was used for imaging the secondary
and backscattered electrons. The electrodes were affixed to an aluminum

specimen stub using double-sided carbon tape.

30

2.2.2 Raman Spectroscopy

The microstructure of diamond/T i electrodes was evaluated by Raman
spectroscopy. Spectra were obtained at room temperature using a Chromex
RAMAN 2000 spectrograph (Chromex, lnc., Albuquerque, NM). The instrument
consisted of a diode-pumped, frequency-doubled 0W Nd:YAG laser (50 mW at
532 nm, COHERENT), a Chromex 500is spectrometer (f/4, 1200 grooves/mm
holographic grating), and a thermoelectrically-cooled, 1024 x 256 element
charge-coupled device (00D) detector (ANDOR Tech., Ltd.). Spectra were
collected with an incident power density of ca. 50 kW/cm2 (10 mW at the sample
and 5 pm diameter spot size) and a 10-s integration time. A white-light spectrum
was collected under the same conditions and used as the background to which to
reference the film spectra. The Raman instrument was calibrated by the
spectrum of 4-acetamidophenol and the diamond scattering was calibrated using
a high-pressure, high temperature (HPHT), single crystal diamond sample (first

order phonon position = 1332 cm").

2.2.3 X-Ray Diffraction Spectroscopy (XRD)
The crystallinity of diamond/Ti electrode was studied by XRD. Spectra
were obtained by a Rigaku Rotaflex RTP300 R0 diffractometer equipped with a

rotating anode and a Cu Kat radiation source (1.540 A) scanning 29 from 30 to

90°.

31

2.2.4 Electrical Measurements

The electrical resistivity of diamond electrodes was measured using a
tungsten tip, four-point probe connected to an analog multimeter (HP 3478A,
Hewlett-Pachard, Palo Alto, CA), which was operated in a four-wire resistance
measurement mode. The probe spacing was 0.1 cm. A constant current was
applied between the outer two probes, the voltage drop between the inner two
probes was measured. The measured resistance, R, from the l-V data was
converted to a resistivity, p, according to the equation:

p (Q-cm) = 4.5321 R (Q)
where l is the film thickness (cm). Multiple measurements were usually made at

different locations on film.

2.2.5 Electrochemical Characterization

Cyclic voltammetry was used to access the electrochemical properties of
the diamond/Ti electrode. Two redox systems, Fe(0N)63"4‘ and Ru(NH3)63+’2*,
were used. Prior to a measurement, the diamond electrode was soaked in
distilled isopropanol for at least 20 min and dried with N2. The electrode was
pressed against a Viton® o-ring (0.22 cm2) and clamped to the bottom of the cell.
Ohmic contact was made to the backside of the conducting Si substrate with a
Cu plate. Before contact, the Si substrate surface was scratched, cleaned and
coated with graphite or colloidal Ag paint. The copper plate was secured against
the back of the electrode using a clamp to assure a good electrical connection to

the electrode. The cell is shown in Figure 2.2. The cell was housed in a Faraday

32

cage for electrical shielding. All of the electrochemical measurements were
performed using a CH instruments Model 650A electrochemical workstation (CH
Instruments, Inc., Austin, TX) with a Ag/AgCl reference electrode (4 M NaCl, E°’
= -45 mV vs. SCE) and a graphite rod counter electrode. All measurements were
made in solutions purged with N2 for at least 15 min and at room temperature (25

:t 1 °C).2 The solution remained blanketed with N2 at all times.

 

 

 

 

 

 

Figure 2.2 Design of the single compartment three-electrode electrochemical
cell. (Reproduced with permission from Anal. Chem. 2000, 72, 3793-3804.

Copyright 2000 American Chemical Society.)

2.3 Water Treatment

The degradation of atrazine was used to evaluate the effectiveness of the
diamond/Ti electrode for oxidative decontamination. Oxidation reaction products

‘were monitored by HPLC. The final degradation product, 002, was determined

33

by measuring the radioactivity of a 1‘i'C-labeled atrazine solution.

2.3.1 Electrolysis of Atrazine

Atrazine (ATZ) electrolysis was performed in an H-cell using a diamond/T i
anode at constant current as shown in Figure 2.3. The glass cell consisted of two
compartments separated by a porous glass frit, which allowed for ion transport
but inhibited bulk solution mixing. The cell volume in the anolyte and catholyte
compartments was ca. 6 mL each. A commercial Ag/AgCI (4 M KCI, Cypress
Systems, Inc.) electrode was used as the reference and was positioned in the
anolyte compartment. A diamond/T i electrode was used as both the working and
counter electrodes. The electrodes were pressed against a Viton® o-ring (i.d. 4.5
cm2) at the opening of each compartment using a metal clamp. A constant

current was applied by a Model 273 potentialstat/galvanostat (EG&G).

AglAgc
Ref.

    
 
  

Diamonlei

 

 

 

A / DiamondrTi

 

 

 

O-ring

Figure 2.3 Design of the two-compartment H-cell for atrazine electrolysis.

Prior to electrolysis, the catholyte compartment was first filled with 6 mL of

34

0.1 M H2804 and then the anolyte compartment was filled with 6 mL of 0.1 M
H2804 solution containing 5 ppm of ATZ. This is about the highest concentration
that can be solubilized in the solution. The electrolysis was conducted for 120
min using a constant current of 50 or 100 mA. The concentrations of ATZ and its
three major metabolites, hydroxyatrazine (HA), deethylatrazine (DEA) and
deisopropylatrazine (DIA), were monitored by HPLC every 20 min during the
electrolysis. After each electrolysis period, the solution in the anolyte
compartment was quantitively removed and 1.0 mL of this solution was mixed
with 1.0 mL of 0.05 M phosphate buffer (pH 3.5) in acetonitrile (60/40 WV). 20 pL
of the resulting solution was then injected into an HPLC system for separation
and detection. A fresh ATZ solution was used for each electrolysis period. The

experimental details of the HPLC system and analysis are given below.

2.3.2 Separation and Detection of Atrazine and Metabolites by
HPLC-UVNis

A Shimadzu UV-2401 PC UVNis spectrophotometer was used to
measure the optical absorption of ATZ, hydroxyatrazine (HA), deethylatrazine
(DEA) and deisopropylatrazine (DIA). The maximum absorbance for all four was
found to be around 216 nm so this wavelength was used for analysis. Reversed-
phase HPLC with UVNis detection at 216 nm was performed using a home-
made system consisting of a Milton Roy ConstaMetric Ill metering pump, an
Adsorbosphere HS 013 column, (150 mmX4.6 mm, 5 pm particle size), a

Shimadzu SPD-6AV UV-Vis spectrophotometric detector and an XYT strip chart

35

recorder. Separation was performed isocratically in a solution of 0.05 M
phosphate buffer (pH 3.5) mixed with acetonitrile at a volumetric ratio of 60/40.
This ratio was carefully selected to achieve the best separation of the

compounds. The mobile phase flow rate was 1.0 mL/min.

2.3.3 Electrolysis of 14C-Labeled Atrazine

To determine the oxidative conversion efficiency to 002, 14C-ring-labeled
ATZ was electrolyzed and the 14CO2 produced was trapped in a 0.1 M KOH
solution. For these experiments, all openings of the cell were sealed so as to
eliminate gas escape. A syringe needle directed the gas generated from both
compartments to a 002 trap (0.1 M KOH) through rubber tubing, as shown in
Figure 2.4. The 002 trap was sealed with a balloon. 30 pL of 14C-labled ATZ
(stock solution in methanol, 2.2 g/L) was added to a solution containing 5 ppm
ATZ and 0.1 M H2S04. 6 mL of this mixed solution was then added to the anolyte
compartment and 6 mL of 0.1 M H2804 was added to the catholyte compartment
of the H-cell. Bulk electrolysis was conducted as described above. A constant
current (50 or 100 mA) was delivered with a DC power supply (Model WP 711 A,
VIZ MFG Co.). The conversion efficiency to CO2 was monitored by dividing the
radioactivity of the KOH trap solution by the total radioactivity of the ATZ solution
before electrolysis. The radioactivity of the solution was determined in a ready-to-
use high flash point safety-solve counting cocktail (Research Products
International Corp.) using a 1500 Tri-Carb liquid scintillation Analyzer (LSA)

(Packard Instrument Company).

36

Ag/AgCl
Ref.

 

 
      
 
  

 

"C-Iabeled

Diamond/Ti Atrazine

 

\
Diamond/Ti

 

 

 

 

  

0.1 M KO
(002 trap)

Figure 2.4 Design of the two-compartment H-cell for the 14C-Iabeled atrazine

analysis.

2.4 Water Quality Monitoring

2.4.1 Gold and Arsenic Electrodeposition

A Au-coated diamond/Si electrode was used to monitor the inorganic
arsenic concentration in water supplies. The deposition of Au and As was
performed by an amperometric method. Four electrodes, all grown in one batch,
were used for these studies. One electrode was bare, another was coated with
As for 90 s at -0.45 V from a solution containing 10 ppm As(lll) in 1 M HCI,
another was coated with Au for 90 s at -0.45 V from a solution containing 10 ppm
Au(lll) in 1 M HCI and another was coated with Au and As for 90 s at -0.45 V
from a solution containing 10 ppm As(lll) and 10 ppm Au(lll) in 1 M HCI. After
deposition, the electrodes were carefully rinsed with ultrapure water and dried

with N2. They were then attached to a specimen stub for SEM analysis using

37

double-sided carbon adhesive tape. A JSM-6300F scanning electron microscope
(JEOL, Ltd., Tokyo, Japan) was used with an accelerating voltage of 10 W to
obtain the surface images of these electrodes. Particle size and density were
determined using the AnalySlS imaging software (Software Imaging System
Corp., Lakewood, 00). All particle analysis data are reported as a mean i

standard deviation.

2.4.2 Preparation of Au-Coated Diamond and Planar Au

Electrode

Au-coated diamond electrode was used for arsenic detection. No
pretreatment was applied to the diamond electrode prior to use other than a 15
min soak in distilled isopropanol. 100 ppb of AuCl3 was added to the solution
containing 1 M HCI and arsenic before detection. Au was co-deposited with As at
a deposition potential of -0.45 V vs. Ag/AgCl for 120 s.

The performance of a planar Au electrode was compared with that of a
Au-coated diamond electrode for arsenic analysis. The Au foil was supported on
a 1x1 cm2 piece of glass and polished under clean conditions using 0.05 um
diameter (deagglomerated) alumina powder slurried in ultrapure water for 15 min
prior to use. The electrode was then rinsed and ultrasonically cleaned in
ultrapure water and 0.1 M H2804 for 3 min each to remove polishing debris. The

planar Au electrode was not repolished between measurements.

38

2.4.3 Differential Pulse Anodic Stripping Voltammetry (DPASV)

Differential pulse anodic stripping voltammetry (DPASV) method was used
for arsenic analysis. The same electrochemical instrumental set-up was used as
described above. The typical measurement with the Au-coated diamond or
glassy carbon electrode consisted of 3 steps: (i) a 120 s deposition at -0.45 V (vs.
Ag/AgCl) to preconcentrate the As, (ii) a differential pulse anodic scan (stripping
step) from -0.45 to 0.5 V using a 4 mV step height, a 50 mV pulse amplitude, a
50 ms pulse width and a 200 ms cycle period and (iii) a constant potential of 0.6
V for 180 s after completion of the anodic sweep to fully oxidize all metal deposits
prior to the next measurement. The typical analysis with the Au foil was the same
except that the optimum deposition potential was -0.15 V. For codeposition on
diamond and glassy carbon, a solution containing both As(lll) and Au(lll) was
added to the cell. Standard addition was employed to determine the detection
figures of merit and the As(lll) concentration in the unknown solutions. All data
are reported as mean 1: standard deviation. All the DPASV i-E stripping

voltammograms were background corrected.

 

T

50mV

_I_

Figure 2.5 The differential pulse anodic stripping voltammetric potential

 

 

 

 

 

w—ZOOms—H

39

waveform and parameters.

2.4.4 Stability Test in River Water

The response reproducibility and stability of a Au-coated diamond and a
planar Au electrode were evaluated in spiked river water samples. For real water
analysis, samples were collected from the Red Cedar River (Lansing, MI, March
2005) using a 500 mL Nalgene bottle. Each bottle was rinsed with the river water
three times before collecting a sample. It was then sealed, labeled and stored in
a plastic bag until time for analysis. Prior to measurement, the river water was
filtered by passage through a sterilized 0.22 pm PES (polyethersulfone), low
protein binding membrane (Corning Incorporated Filter System). 250 mL of the
filtrate was then acidified to pH 0 by adding 20.5 mL of 37% HCI. The
reproducibility and long-term response stability of the Au-coated diamond was
assessed using acidified Red Cedar River water (1 M HCI) spiked with 20 ppb
As(lll). Au-coated diamond and glassy carbon electrodes were compared under
the same conditions. A run-to-run reproducibility test was performed over 10
consecutive measurements. Longer term stability studies were carried out by
leaving the electrode exposed continuously to the As(lll) solution for 10 h and
periodically measuring the As stripping peak current (each hour). The relative
standard deviation of peak current and peak charge was then calculated and

reported.

2.4.5 Cation Exchange Method to Remove Interfering Metal Ions

4O

A cation exchange method was developed for the removal of metal ions
that interfere with arsenic detection. Water samples containing As(lll) or As(V)
and other metal ions were passed through the sorbent. This was performed using
an Alltech vacuum manifold consisting of a glass chamber, a vacuum gauge and
valve, manifold lid, and stopcocks. 1.0 g of a Chelex 100 cation exchange resin
(50-100 dry mesh, Sigma) was packed into a 3 mL cartridge (Supelco). The
cartridge was then conditioned by 10 bed volumes (~ 30 mL) of ultrapure water
before use at a flow rate of ~ 5 mL/min. After conditioning, the solid phase was
dried under vacuum for 5 min prior to introducing the analyte solution. The
analyte solution (5 mL) was then added and immediately pulled through the
cartridge by vacuum at a constant flow rate of ~ 5 mL/min. The solution passing

through the solid phase contained the inorganic arsenic for analysis.

2.4.6 Detection of As(V)

The detection of As(V) consisted of two steps: (i) a chemical reduction
from As(V) to As(lll) and (ii) DPASV detection of reduced As(V). Na2803 was
used as the reducing agent because it did not interfere with the electrochemical
detection and had high conversion efficiency in reasonable reaction time. 4 mL of
an As(V) solution in 1 M HCI was added to 0.01 g N32803 (~ 20 mM) in a sealed
plastic test tube. The solution was then heated in an oven (VWR Scientific, VWR
1410) at 50 °C for 30 min. N32803 was used as the reducing agent, converting
As(V) to As(lll) by the following reactions:

N32803 + 2 HCI —-> 2 NaCl + H2803

41

H3ASO4 ‘1' H2803 --> H3A803 ‘1' H2304
After cooling the solution to room temperature, ASV was performed as described

above.

2.4.7 Inductively Coupled Plasma Mass Spectrometry (ICP-MS)
and Graphite Furnace Atomic Absorption Spectroscopy
(GFAAS)

The results obtained by DPASV method were compared with those
obtained by lCP-MS and GFAAS. An inductively-coupled plasma mass
spectrometer (Micromass Platform lCP-MS), fitted with a hexapole collision cell,
was used for the measurements. A Meinhard concentric nebulizer was used for
sample introduction. The As(lll) concentration was determined from a response
curve generated using a series of standard solutions of known As(lll)
concentration. Each standard was prepared from a 1000 ppm commercial stock
solution of As(lll) (Spex Certiprep, Metuchen, NJ) diluted with 2% HN03. For the
analysis, 2 mL of the standard solution was transferred to a sample tube
containing 2 mL of a 10 ppb ln(lIl) internal standard. A measurement time of 1
min was used for each sample. The torch was thoroughly rinsed with 2% HN03
between measurements. The response curve for calibration was generated using
peak intensities. For the GFAAS measurements, a Hitachi Z-9000
spectrophotometer with a graphite tube furnace attachment was used. The
instrument possessed a diffraction grating monochromator, which was set to 195

'nm for monitoring arsenic. Hitachi arsenic hollow cathode lamp was used with

42

emission wavelengths of 195 nm. The lamp current was set at 10 mA. The
sheath gas flowing through the tube furnace was argon at 200 mL/min. A 20 (IL
sample volume was introduced into the furnace via an auto injector. The sample
was dried and ashed using manufacturer-specified times and temperatures. The
experimental parameters are summarized in Table 2.2. Calibration was
accomplished by recording the responses for standard solutions containing

known concentrations of arsenic.

Table 2.2 Detection conditions for arsenic analysis using GFAAS.

 

Lamp Current 10 mA
Wavelength 195 nm
Slit 1.2 nm
Carrier Gas 200 mL/min
1:332:22? 3......
Sample Volume 20 pL

 

2.4.8 Total Inorganic Arsenic Detection in Real Water Samples

Well water samples (Well 119 water and UV plant influent) from two
contaminated sites were provided by the US. Bureau of Reclamation. The
samples were received after being acidified with HN03 to a pH of ~2. The
samples contained no visible sediment so filtration was not performed. The
samples were first neutralized by adding solid NaOH to adjust the pH to > 4.

'Then, 4 mL of the solution was passed through the cation exchange cartridge, as

43

described above, to remove interfering metal ions. 2 mL of this solution passing
through was collected and mixed with 2 mL of 2 M HCI. After adding 0.01 g
Na2803, the solution was then mixed and heated at 50 °C for 30 min to convert
the As(V) to As(lll). Minimal volume change occurred during this period. After
cooling to room temperature, the solution was ready for analysis by DPASV, as
described above. Quantification of the-total inorganic arsenic was accomplished
by the standard addition method. The procedure used for analyzing the water

samples is illustrated in step-wise fashion in Figure 2.6.

 

Water sample
acidified with HNO3
mpH=2

 

 

1)
Adjusted to pH > 4.0 l
b addi solid NaOH

 

 

 

Passed through a
cation exchange
resin

Collected solution and acidified with
2 M HCI (1:1 v/v)

As(lll) W
analysis analysis

 

 

 

 

 

 

 

 

 

 

Detected As(lll) by DPASV Reduced As(V) to As(lll) by
with a Au—coa’ted diamond adding Na.‘,SO3 and
electrode heating at 50 °C for 30 min

Standard addition used

to determine the arsenic

concentration

 

 

 

Figure 2.6 Total inorganic arsenic detection in real water samples.

44

2.5 Reagents

All chemicals were analytical-grade quality, or better, and were used
without additional purification. All solutions were prepared with 18 Mfltcm
ultrapure water (Bamstead E-pure).

For the water treatment experiments, the chemicals and suppliers were as
follows: atrazine [2-chloro-4-ethylamino-6-isopropylamino-s-triazine] (97.2%,
Crop Protection, Inc.); hydroxyatrazine [2-hydroxy-4-isopropylamino-6-
ethylamino-s-triazine] (97%, Crop Protection, Inc.); deethylatrazine [2-amino-4-
isopropylamino-6-chloro-s-triazine] (94%, Crop Protection, Inc.);
deisopropylatrazine [amino-2-chloro-6-ethylamino-s-triazine] (96%, Crop
Protection, Inc.); “C-ring-labeled atrazine (100 uci, Sigma); HPLC grade
acetonitrile (EMD Chemicals Inc); sulfuric acid (98%, Aldrich); phosphoric acid
(Aldrich); potassium phosphate monobasic (Spectrum Chemical Mfg. Corp.) and
isopropanol (Columbus Chemical Industry Inc.). All glassware was cleaned in a
KOH-ethanol bath followed by thorough rinsing with ultrapure water.

For the water quality monitoring experiments, all glassware was acid
washed sequentially in 1 M HCI (reagent grade) and 1 M HN03 (reagent grade),
followed by rinsing thoroughly with purified water before use. The chemicals and
materials used were arsenic trioxide (A8203, Aldrich, 99.95%-100.0%), arsenic
pentaoxide (A8205, MV Laboratories, Inc., 99.999%), gold foil (Au, Aldrich, 0.1
mm thick, 99.99%), sodium hydroxide (NaOH, Spectrum Quality Products, Inc.

~ Reagent A.C.S.), hydrochloric acid (HCI, Aldrich, 37%, 99.999% purity), gold (Ill)

45

chloride (AuCla, Aldrich, 99.99+%), cupric nitrate (Cu(N03)2, 99.999%, Aldrich),
lead nitrate (Pb(N03)2, 99.999%, Aldrich), mercuric acetate (Hg(CH3000)2,
99.999%, Aldrich), sodium sulfite (Na2803, Columbus Chemical Industries, Inc.,
ACS grade), nitric acid (HN03, EM Science, Reagent A.C.S.), Chelex 100
sodium form ion exchange resin (50-100 dry mesh, Sigma), 3 mL cartridge

(Supelco). and humic acid, sodium form (Aldrich).

46

2.6 References

(1)

(2)

Swain, G. M. In Electroanalytical Chemistry, Rubinstein, L, Ed.; Marcel
Dekker, Inc.: New York, 2003; Vol. 22, pp 181-277.

Granger, M. 0.; 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. Analytical Chemistry 2000, 72, 3793-3804.

47

CHAPTER 3
ELECTROCHEMICAL AND PHYSICAL PROPERTIES OF
DIAMOND/Ti

3.1 Introduction

Ti is an attractive substrate material for diamond growth due to its low cost,
good electrical conductivity and high mechanical strength.1 However, compared
with Si, Ti has received less attention as a substrate material due to the difficulty
of depositing a high quality, crack-free and well adhering diamond film on Ti. In
this work, we opted to use Ti as a substrate for the growth of conductive diamond
film and to employ this electrode for wastewater treatment. In order to optimally
remediate organic contaminants with a boron-doped diamond electrode on a Ti
substrate (diamond/Ti), it is necessary to first characterize the electrochemical,
morphological and microstructural properties of the diamond/Ti electrodes.

There are two challenges to depositing a well adhering, conductive
diamond overlayer on Ti or Ti alloy.2' 3 One is the severe embrittlement of Ti that
can occur from the diffusion of hydrogen and carbon into the lattice to form
hydrides and carbides during the early stages of deposition. The hydrides can
erode the diamond surface and inhibit diamond growth.4 Carbide formation
competes with diamond nucleation and forms a layer on Ti preferentially to
diamond nucleation. Although this carbide layer ensures good conductivity
between the diamond and Ti, it can result in poor adhesion of the diamond layer.5

,The other challenge is the residual stress, primarily compressive in nature, that

48

develops from the difference in thermal expansion coefficients, as the coefficient
for Ti (8.5OX10’6 C") is about 10 times larger than the coefficient for diamond
(0.84X10'6 c"). Additionally, the thermal expansion coefficient of the TiC
interfacial layer (7.42x10'6 0") is significantly larger than the coefficient for
diamond.5 The compressive stress that is introduced into the diamond is
detrimental to good film adhesion and leads to the formation of microcracks and
even debonding if severe enough. The presence of any microcracks is critical for
diamond electrochemical electrodes because the electrolyte solution can
permeate them and reach the underlying metal substrate. Contact with solution
combined with anodic electrode potentials can cause corrosion of the Ti through
the formation of soluble Ti+2 and Ti+4 species and this leads to the debonding of
the diamond overlayer.“ 6

Several factors are known to affect the adhesion of diamond to Ti during
hot filament CVD growth, such as substrate temperature, filament temperature,
filament substrate distance, the deposition temperature,6 substrate surface
roughness7' 8 and boron concentration.5 The substrate temperature determines
the thickness of TiC layer. A high substrate temperature will cause the formation
of a thick TiC layer and poor diamond film adhesion, while a low substrate
temperature can cause the formation of sp2 carbon and poor diamond quality and
durability. An optimum substrate temperature was found to be about 850 °C.6
High filament temperature, short filament-substrate distance and long deposition
time tend to cause film debonding.6 To prevent this effect, polishing the substrate

before growth to increase the surface roughness can improve the adhesion

49

strength and decrease the internal stress within the film.7' 8 Boron can positively
influence diamond deposition on Ti as it was found to increase the diamond
adhesion with Ti, facilitate diamond nucleation and inhibit the formation of a TIC
layer by forming borides or carbo-borides in the interlayer.5 Judicious choice of
the growth temperature and control of the substrate morphology and chemistry
through appropriate pretreatment are, therefore, of paramount importance.

An ideal anode for wastewater treatment should exhibit (i) a large
overpotential for oxygen evolution, (ii) high electron-transfer rates for general
redox reactions and (iii) morphological and microstructural stability as discussed
in Chapter 1. Up to now, few studies have been performed comparing the
diamond/Ti electrode properties before and after an electrolysis event.

It has been reported that diamond/T i electrodes have a large working
potential window of 2.5 to 4.7 V in acidic, neutral or alkaline solution.5' 6' 9‘”
However, in terms of the electrochemical activity, the reported cyclic

voltammetric response for Fe(CN)63"4'

, a common redox test system, is
quasireversible to have AEp values between 96 and 188 mV.6' ‘2 The
heterogeneous electron-transfer rate constant, k°app, was calculated to be (1.0
~1.5) x 10'3 cm/s,"' ‘2 lower than a typical k°app value for this system at high
quality diamond (0.050 cm/s).13 Moreover, a 3-4 fold decrease in response
sensitivity was reported for diamond/Ti after 1 month of use14 and a passivation
layer was found after 10 cycles in 1 M H2804.11 The passivation was probably

caused by poor film adhesion, microcrack formation and debonding, and the

subsequent formation of Ti02 because of solution reaching the Ti substrate (i.e.,

50

the passivation layer).

In this Chapter, the physicochemical and electrochemical properties of the
diamond/Ti electrode are compared before and after exposure to an anodic
current of 22 mA/cm2 for 50 h in 0.1 M H2804. Cyclic voltammetry was used to
study the electrochemical activity of the electrode using two different redox
systems, Fe(CN)63"4‘ and Ru(NH3)53+’2+. The electrode morphology and
microstructure were characterized by scanning electron microscopy (SEM),

Raman spectroscopy and X-ray diffraction spectroscopy (XRD).

3.2 Results and Discussion

3.2.1 Electrochemical Properties of Diamond/Ti

Background cyclic voltammetric i-E curves can be very informative about
the boron-doped diamond film quality. The technique provides information about
the presence of nondiamond sp2 carbon impurity at the surface, the electrode
working potential window and the electrical conductivity. Figure 3.1 shows
background cyclic voltammetric i-E curves for diamond/Ti and diamond/8i
electrodes (same geometric area in each case) in 0.1 M H2804 at the scan rate
of 100 mV/s. The curves for both electrodes are flat and featureless with a wide
potential window and low background current. The flat and featureless curve
shape is typical of high quality, hydrogen-terminated boron-doped diamond as
the electrode is devoid of redox-active impurities, reactive sp2 carbon and redox-
active carbon-oxygen functional groups (e.g., quinone/hydroquinone) that will

L15

‘ result in faradaic current and cause an increase of the background curren For

51

wastewater treatment, a large overpotential for oxygen evolution is a useful
property for an anode in terms of enabling high current efficiency during pollutant

oxidation.

 

 

 

 

   

 

 

 

 

 

 

A
3 J
i=2 s
t . .. .. t
a a . .
— Diamond/Ti -1.0 - ——- D'amond/T'
, ..... Diamond/8i . ----- Diamond/Si
-2 -1 0 1 2 3 ”1'5 -0.4 0.0 0.4 0.8 1.2
Potential (V) vs. Ag/AgCl Potential (V) vs. AgIAgCI

Figure 3.1 Background cyclic voltammetric i-E curves for diamond/T i and
diamond/Si electrodes in 0.1 M H2804 over a (A) wide and (B) narrow potential

range. Scan rate = 100 mV/s. Geometric area = 0.2 cm2.

Figure 3.1A shows that the working potential window for both electrodes is about
3.2 V. The onset of hydrogen evolution for diamond/Ti occurs slightly more
positive of that for diamond/8i while the onset of oxygen evolution for both
electrodes is greater than 2 V vs. Ag/AgCl. The potential window seen for this
diamond/Ti electrode is comparable to what has been reported for diamond/1' i
(2.5 to 4.7 V)5' 6' 011 and is characteristic of good quality diamond. From Figure
3.1B, it is clear that the background current for both electrodes is low and
featureless. The anodic current at 0.5 V was 0.25 pA for diamond/T i and 0.34 pA

for diamond/8i. In contrast, spz-bonded carbon and metal electrodes usually

52

have about 5-10 times higher background current. The low background current
for diamond is due to a reduced pseudocapacitance because of the absence of
redox-active and/or ionizable surface carbon-oxygen functional groups and a
reduced capacitance due to a slightly lewer internal charge carrier
concentration.15 The lower background current leads to improved signal to
background ratio in electroanalytical measurements.

The diamond/Ti electrode response for two different redox systems,
Fe(0N)63"4’ and Ru(NH3)53*’2+, was studied by cyclic voltammetry method. Redox
reactions of Fe(CN)63"4' and Ru(NH3)63+’2* are redox systems that represent two
major categories of heterogeneous electron-transfer reactions: inner-sphere (or
surface sensitive) and outer-sphere (or relatively surface insensitive) reactions.
Inner sphere redox reactions involve specific interaction with the electrode and
heterogeneous electron-transfer rate constants are sensitive to surface

cleanness and surface termination. Fe(CN)63"*

proceeds through a more inner-
sphere electron-transfer pathway. At diamond electrodes, the peak separation,
AEp, for Fe(CN)63"4' is particularly sensitive to the surface chemistry.16 The
lowest AEp is generally observed for the hydrogen-terminated surface,
introducing surface oxygen can block the reactive surface sites, slow the
electrode reaction kinetics and cause an increase in AEp.16 The apparent

heterogeneous electron-transfer rate, k°app, for Fe(CN)63"4'

at a high quality
diamond/Si is typically 0.01-0.1 cm/s.1621 Outer-sphere reactions do not involve
interaction with the electrode and proceed by simple mass transport of the

electrolyte to the electrode-solution interface, with the electrode only serving as a

53

source or sink of the electrons. Ru(NH3)63’”2+

proceeds through an outer-sphere
electron-transfer pathway. AEp for this redox system at carbon electrodes (e.g.,
diamond, glassy carbon and HOPG) is most sensitive to the electronic properties
of the electrode, that is, the density of electronic states (electrons and holes)
near the standard redox potential. AEp of Ru(NH3)63“’2+ is less sensitive to factors
such as the surface microstructure, surface chemistry and surface cleaness.16
The most important factor controlling the density of electronic states in
polycrystalline diamond is probably the boron-doping level, but other factors such
as non-diamond sp2 carbon impurity, defects, and lattice hydrogen may also be
influential.16 The lowest AEp is generally seen for the high electron density
surface with high electron-transfer rate constant. The apparent heterogeneous
electron-transfer rate constant, k°app, for Ru(NH3)63*’2+ at a high quality
diamond/8i is typically 0.01-0.2 cm/s."‘*21

The electrochemical responses for Fe(CN)63"4' and Ru(NH3)53+’2+ were
evaluated at five different diamond/Ti electrodes in order to test the electrode
activity and response reproducibility. Representative cyclic voltammetric i-E
curves for diamond/Ti in 1 mM Fe(CN)63"4‘ and Ru(NH3)63*’2* in 1 M KCI are
shown in Figure 3.2. A summary of the data is presented in Table 3.1. The
voltammetric responses for these five electrodes are reproducible and stable
from film-to-film. All five electrodes exhibited active responses for Fe(CN)63”“

and Ru(NH3)63+’2+ without any conventional pretreatment (e.g., polishing);

something that is normally required for spZ-bonded carbon and metal electrodes.

54

 

Current (pA)

 

-501

_80 1 g . 1 . 1 1 1 . 1 ,

-O.6 -O.4 -O.2 0.0 0.2 0.4 0.6
Potential (V) vs. Ag/AgCI

 

 

 

Figure 3.2 Cyclic voltammetric i-E curves for 1 mM Fe(CN)63"4' and

Ru(NH3)63+’2+ in 1 M KCl at a diamond/Ti electrode. Scan rate = 100 mV/s.

Table 3.1 Summary of cyclic voltammetric data for diamond/Ti electrodes at a

scan rate of 100 mV/s.

 

 

Redox System AEp (mV) 15,,2 (mV) i,” (M) i,,°"rl,,"ml
litu(Nl-l4).,3:"2+ 76 1 8 -188 1 5 58 1 5 1.0 1 0.1
Fe(CN)33"" 76 1 5 251 1 2 50 1 16 1.0 1 0.1

 

AEp for Fe(CN)53"4' is 76 t 5 mV (RSD = 7%) at 100 mV/s. The value is
comparable with others reported for a diamond thin film on Si (65-70 mV at 100
mV/s)20 and is smaller than values reported for diamond on Ti by other
researchers (96-188 mV at scan rate of 20-200 mV/).6' 1" 12' ‘4 The results for

-Fe(CN)63"" indicate that the diamond/Ti has a clean hydrogen-terminated

55

surface. AEp for Ru(NH3)53+’2+ is 76 1 8 mv (RSD = 11%) at 100 mV/s. This is
comparable to values reported by Granger et al. for diamond thin films on Si (74
mV).16' 22 The results for Ru(NH3)53+’2+ demonstrate that the diamond/Ti electrode
possesses a density of electronic states sufficiently high to support rapid
electrode reaction kinetics. AEp values for both redox systems increased with
increasing scan rate, which is consistent with quasireversible reaction kinetics.
ip°" varied linearly with the scan rate"2 between 10 and 500 mV/s with a near
zero y-axis intercept, as shown in Figure 3.3, indicating the current is limited by
semi-infinite linear diffusion. The i,,°"/ipred and Qp°"/Qp"’d ratios were near 1.0 as

expected for reversible systems and the values were stable with cycling.

 

 

 
 
  

     

 

 

 

 

 

 

.. 120 Fe(CN)43"" .. 120 .’ Ru(Nl-l3).’*””

31001 3100:

E 80- 5 80:

g 60» S 60*

O ' y=1.610E-04x O 40; y=1.643E-04x

x 40‘ x

3 20. +2.199E-6 3 20: 4.215555

0. l R2=0.996 o. i 2 . ,R2,= @999 ‘

8.0 0.2 0.4 0.6 0.8 8.0 0.2 0.4 0.6 0.8

Scan Rate"2 Scan Rate"2

Figure 3.3 Plots of the anodic peak current versus the square root of scan rate

for 1 mM Fe(CN)e3"4‘ and Ru(l~lH3).;°"’2+ in 1 M KCI.
The electrode apparent heterogeneous electron-transfer rate constant,

1839,, for Fe(CN)53"4‘ and Ru(NH3)63*’2* before and after a 50-h electrolysis at 22

, mA/cm2 in 0.1 M H2804 were compared and the results are listed in Table 3.2.

56

k°,,.pp was determined from cyclic voltammetric data (AEp- v dependence) using a
method developed by Nicholson.17 According to the theory, AEp for an
electrochemical reaction is dependent upon the scan rate (v, V/s), transfer
coefficient (a) and apparent heterogeneous electron-transfer rate constant (kOapp.

cm/s). He derived an equation for quasireversible reactions showing this relation:

(Dox /Dred )a/Zkoapp

W = (”Dex )1/2 (3-1)

 

where Dox and Dred (cmzls) are the diffusion coefficients of the oxidized and

reduced forms of analyte, respectively, f is the Faraday constant divided by the

ideal gas constant and temperature (F/RT) and ‘1’ is a dimensionless parameter.
Nicholson derived an experimental relationship between \P and AE,,, so that k0,,pp

can be calculated from the above equation at different scan rates.17

Table 3.2 Cyclic voltammetric and the apparent heterogeneous electron

3-l4-

transfer rate constant data for diamond/Ti electrodes for Fe(CN)5 and

Ru(NH3)63*’2* in 1 M KCI before and after electrolysis in 0.1 M H2804 at 22

 

 

mA/cmz.
Fe(CN)63-/4- RU(NH3)53+,2+
Before After Before After
AEp (mV) 83 99 71 75
‘P 1.08 0.62 2.25 1.75
k°app (cm/s) 0.010 0.006 0.019 0.014

 

Before electrolysis, for both Fe(CN)63"4' and Ru(NH3)53*’2*, the electrode

57

k"app was near 0.01 cm/s. This value is consistent with others reported for
diamond/8i (0.01-0.2 cm/s)6' 11' 12' 14 and is higher than the values previously
reported for diamond/T i electrodes (1.94 x 104 to 1.5 x 10'3 cm/s).7'8' "' ‘2’” The
results indicate that prior to electrolysis, the diamond/Ti electrode surface is
clean and hydrogen-terminated and possesses a sufficiently high density of
electronic states to support rapid rates of electron-transfer. However, after a 50-h
electrolysis at 22 mA/cm2 in 0.1 M H2804, the k°,—,.pp value for Fe(CN)63""
decreased to 0.006 cm/s while k°app for Ru(NH3)53"’2+ decreased only from 0.019

to 0.014 cm/s. Because the k°app value for Fe(CN)63"“'

is very sensitive to the
surface carbon-oxygen functional groups, the introduction of surface oxygen can
inhibit this redox reaction presumably by blocking active sites. Surface oxygen
typically causes an increase in AEp, thus, a decrease of k°a.pp for this redox
system.” During the electrolysis, hydroxyl radicals as well as oxygen are formed
which leads to the incorporation of oxygen functional groups. The functional
groups cause an increase in the Fe(CN)63”* AEp and a decrease in the k°app.
Apparently, the density of electronic states on the surface does not change

significantly because AEp and k°app for Ru(NH3)63“’2+

is fairly constant before and
after electrolysis.

The cyclic voltammetry results indicate that the diamond/T i electrode
possesses the requisite surface structure, chemical composition and electronic
properties to support relatively rapid electron transfer for these two redox

systems. It is noteworthy that low AEp values were observed even though the

electrodes were exposed to the laboratory atmosphere for days to weeks prior to

58

use with no pretreatment applied for activation. The diamond electrode, due to
the non-polar nature of the sp3-bonded surface (absence of rr electrons), is
resistant to deactivation by contaminant adsorption and typically does not require

pretreatment for activation.

3.2.2 Morphological and Microstructural Properties of

Diamond/T i

The morphological and microstructural stability of the diamond/Ti
electrode during the 50-h electrolysis in 0.1 M H2804 (at 22 mA/cmz) was
evaluated. Characterization was performed with scanning electron microscopy
(SEM), Raman spectroscopy and X-ray diffraction spectroscopy (XRD).

Figure 3.4A shows a typical SEM image of a diamond/Ti electrode prior to
electrolysis. A well-faceted, polycrystalline diamond film completely covers the
substrate with no sign of any cracks or pinholes. The film possesses a wide
variety of crystallite sizes ranging from several micrometers down to several
hundred nanometers. There is also considerable secondary nucleation observed
as evidenced by the large number of smaller crystallites formed on the larger
crystallites or in the grain boundaries between the larger crystallites. Clearty, the
film grows via a more progressive nucleation mechanism across the metal
surface. This probably is due to the spatial variability in the substrate morphology
and chemistry as well as non-uniform seeding of diamond particles during the
preparation of the electrode substrate for deposition. The diamond film has a

, rolling, undulating character to it, which tracks the topography of the underlying Ti

59

substrate. For comparison, an SEM image of a diamondfl'i electrode after 50 h of
polarization at 22 mA/cm2 in 0.1 M H2804 is shown in Figure 3.4B. The diamond
film morphology was unaltered with no sign of cracks, pinholes or debonding.
The observed dimensional stability is consistent with previous observations for
diamond/Si electrodes.” 2325 The SEM images confirm that the diamond/Ti

electrode is morphologically stable during these electrolysis conditions.

 

Figure 3.4 SEM images of a diamond/T i electrode (A) before and (B) after

electrolysis in 0.1 M H2804 at 22 mA/cm2 for 50 h.

Figure 3.5 shows the cross sectional SEM image of' a diamond/Ti

60

electrode including three phases: diamond, a TiC layer and the Ti substrate. The
thickness of the diamond film is about 10 pm which is typical for microcrystalline
diamond films. The film is dense and homogeneous. The TiC layer is clearly
observed between the diamond film and the Ti substrate. It is porous and rough
with uneven thickness ranging from 5 to 20 pm. The TIC layer is formed by the
diffusion of carbon atoms into the Ti substrate during the early stages of film
growth. The formation of the TIC layer competes with diamond nucleation during
the growth. The diamond nucleation can only occur after the saturation of the Ti
surface with carbon and the carbide layer has certain thickness.5' 26 Even though
the TiC layer can cause the poor adhesion of the diamond layer, it is observed in
most diamond/Ti films and its conducting property is important to ensure good

electrical connection between the Ti substrate and the diamond coating.12

Diamond film
TiC layer

Ti substrate

 

Figure 3.5 Cross sectional SEM image of a diamond/Ti electrode.

The microstructural quality and stability of the diamond/T i film was

61

assessed by Raman spectroscopy. A typical Raman spectrum for the electrode

prior to electrolysis is shown in Figure 3.6A.

 

 

 

 

 

 

 

 

1-0 A 1.0 B
‘7 0-3' ’T 0.8-
=’: . 3. .
3 0.6. 3 0.6-
:E‘ ’ 1327 cm-1 ,1} *
in 0.4 - at 0.4) 1327 cm"
5 . 5 t
'5 0.2» E 0.2)
0.0 - 1 A 1 1 1 . 1 L 4 0.0 1111111111
0 400800120016002000 0 400800120016002000
Raman Shift (cm-1) Raman Shift (cm-1)

Figure 3.6 Raman spectra of diamond/Ti electrode (A) before and (B) after

electrolysis in 0.1 M H2804 at 22 mA/cm2 for 50 h.

The spectral features reflect a heavily boron-doped film and are quite
different from those typical of a single crystal diamond reference sample. The
apparent doping level in this film is 2 1020 cm'3 based on results from other
groups who compared Raman spectra with actual doping levels.3' ”' 27' 28 A weak
first-order diamond phonon mode is seen at 1327 cm". This peak is asymmetric
and down-shifted from the expected position for a single crystal diamond at 1332
cm". The weak intensity is attributed to the high optical density of the film due to
the high doping level. The asymmetry is attributed to Fano-type interference,
which is caused by a quantum interference between the phonon and a continuum
of electronic transitions around the same energy involving the boron impurity

band.3 There is very weak scattering intensity between 1500 and 1600 cm‘1,

62

which is associated with amorphous or nondiamond sp2 carbon?"2 The sp2
carbon (e.g., graphite) has a larger Raman scattering cross section than diamond
so the weak intensity indicates that there is very little sp2 carbon impurity in this
film. A broad continuum exists toward lower wavenumbers with peaks at ca. 500
and 1200 cm". These two maxima are often observed for heavily boron-doped
films and their intensity tracks the boron-doping level.3 The origin of the peak at
500 cm‘1 was assigned to local vibrational modes of boron pairs by Bernard et
al..”31 Disordered growth can result from a modification of the growth
mechanism at the surface related to the high boron concentration.3 The
theoretical phonon density of state maximum for diamond occurs near 1200 cm'
1,3' ”therefore, the peak iS attributed to the disordered spa bonding by the
incorporation of the lighter boron.3' ””31 Gonon and coworkers proposed that
these bands appear whenever a high concentration of impurities or defects is
present.” This Spectrum is characteristic of a heavily doped film, with the
disorder in the diamond crystal structure resulting from the incorporation of
boron.32 Quite Similar observations were reported by Gonon and coworkers27 for
boron-doped diamond on Si with a high doping level of 1.1x1021 B cm'3 and by
Fryda and coworkers32 for heavily doped diamond on metal substrates. Figure
3.6B shows 3 Raman spectrum of the diamond/T i anode after a 50-h electrolysis
(22 mA/cmz) in 0.1 M H2804. No significant difference is observed in the spectral
features suggesting that no microstructural changes occurred during the
electrolysis.

The crystallinity of the diamond/Ti electrode before and after electrolysis

63

was examined by XRD. Figure 3.7A shows a typical diffraction pattern that
reveals reflections for Ti, TIC and diamond. This is consistent with the three
phases evidenced in the cross sectional SEM image shown in Figure 3.5. Peak
assignments were made by comparing this diffraction pattern with patterns for
Ti,” no34 and cubic diamond standards35 (shown in Figure 3.8, 3.9 and 3.10,
respectively). The reference data for Ti, TiC and diamond are also summarized in
Table 3.3, 3.4 and 3.5, respectively.

There are two primary diamond peaks present at 43.9 and 75.3 degrees
(29) and are assigned to (111) and (220) reflections of cubic diamond. The
relative intensity ratio of the (111) and (220) peaks in Figure 3.7A is 100/26 and
is in good agreement with the expected ratio of 100/25 for cubic diamond.35 This
XRD pattern is in good agreement with that reported by Chen.1 Figure 3.7B
Shows the XRD pattern for a diamond/Ti electrode after the 50-h electrolysis (22
mA/cmz). Again, in terms of shape, no Significant pattern changes are observed.
The diamond, TiC and Ti peaks are still present, which indicates that the
diamond film remained attached to the surface and that all these phases
underwent no crystallographic changes during the electrolysis. The difference in
signal intensity between the two XRD patterns is attributed to the fact that two
different films were used. In summary, the characterization results indicate that
the diamond/Ti electrodes are morphologically and microstructurally stable during

these electrolysis conditions.

64

700

 

 

 

 

 

  
 

 

 

 

   

 

 

 

 

 

 

 

, "(151)
600 _ A
500 -
3 mung TIC(111) TIC(311)
c 400 - /
3 : T|C(220) Dlam(220)
8 300: / TiC(200) “(110)
200 - /Dia;_'llt((: (1) 21)) TIC(222)
100 _ / 11(103) Aim”)
0 I ' I r 1 . y . . M
30 40 50 60 70 80 90
29 (degree)
2500 —'l'lT‘l'01)
. B
2000 -
m TIC(111)
E 1500 Til100) / TIC(311)
3 .
81000 TIC(200) TiC(220) Dlam(220)
fiamlm) “(110) 110(222)
500 "(102) /Tl(201)
- 1(103) IM’
0‘ v ' ' ' ' ' ' I ‘ I v
30 40 50 60 70 80 90

Figure 3.7 XRD pattern for a diamond/T i electrode (A) before and (B) after

29 (degree)

electrolysis in 0.1 M H2804 at 22 mA/cm2 for 50 h.

65

100

 

904

.--_--..--_-

...................................................................................................

..............................................................................................

Intensity (Relative Units)

 

 

 

 

 

 

 

 

 

50-, .., ..
30..) ......... .-
0 ' I it 1 Ijfi 1 1 I 1 I 1 I I it
40 5‘0 60 7'0 80 90 100 110 120
2Theta (Degrees)
Figure 3.8 Standard XRD pattern for Ti.
Table 3.3 List of standard XRD pattern data for Ti.

20 d (A) Int-f h k I 20 d (A) Int-f h k I
35.093 2.5550 25 1 0 0 82.290 1.1707 1 0 0 4
38.421 2.3410 30 0 0 2 86.759 1.1215 1 2 0 2
40.170 2.2430 100 1 0 1 92.729 1.0643 1 1 0 4
53.004 1.7262 13 1 0 2 102.361 0.9887 2 2 0 3
62.949 1.4753 11 1 1 0 105.798 0.9658 1 2 1 0
70.660 1.3320 11 1 0 3 109.042 0.9459 4 2 1 1
74.157 1.2776 1 2 0 0 114.278 0.9170 3 1 1 4
76.218 1.2481 9 1 1 2 119.256 0.8928 1 2 1 2
77.368 1.2324 6 2 0 1

 

66

 

100

80-..-.- _ . ..............................................................
25
60* ...................
,9
a
E 50.. ..... ... _
:3.
g 40.- _ , ___________________
c
.9
E 30- ............
20- ..... -.__-..., _. . .....................

10.-. ..-... -.... ..-

 

 

 

 

 

 

 

 

 

 

 

 

4'0 6‘0 820 9:0 100 110 120 180 140
2Theta (Degrees)
Figure 3.9 Standard XRD pattern for TiC.
Table 3.4 List of standard XRD pattern data for TiC.
29 d (A) Int-f h k I
35.906 2.4990 80 1 1 1
41.710 2.1637 100 2 0 0
60.448 1 .5302 60 2 2 0
72.369 1 .3047 30 3 1 1
76.139 1.2492 17 2 2 2
90.802 1.0818 10 4 0 0
101.781 0.9927 13 3 3 1
105.498 0.9677 25 4 2 0
121.372 0.8834 25 4 2 2
135.348 0.8327 16 5 1 1

 

67

 

100-

 

 

 

 

 

90-“... ..... , ........ . ............................

80-.---. ........ . ....... . ....................
g
0) 60. ......... . ... .. - ... ..........
>
'5
8‘:
C
95.
.E 30-. . .

20- ..... - .. .

4O 50 60 70 80 90 100 110 120 130 140
2Theta (Degrees)

Figure 3.10 Standard XRD pattern for cubic diamond.

Table 3.5 List of standard XRD pattern data for cubic diamond.

 

 

20 d (A) Int-f h k I
43.915 2.0600 100 1 1 1
75.302 1.2610 25 2 2 0
91.495 1.0754 16 3 1 1
119.521 0.8916 8 4 0 0
140.587 0.8182 16 3 3 1

 

68

3.3 Conclusions

The electrochemical and physical properties of the diamond/T i electrode
were investigated before and after 50 h of anodic polarization in 0.1 M H2804 at
22 mA/cmz. Background cyclic voltammetric i-E curves Showed that the electrode
was characterized by a featureless and low background current (e.g., 0.25 11A at
0.5 V vs. Ag/AgCI). This indicated the electrode surface was devoid of redox-
active impurities, sp2 carbon and carbon-oxygen functional groups. The
diamond/T i electrode also possessed a wide working potential window of 3.2 V, a
property that is characteristic of good quality diamond. This value is comparable
to that for diamond/8i electrodes and assures high current efficiency during
pollutant oxidation. The electrode response for two redox probes, Fe(CN)6M'
and Ru(NH3)53+’”, for five electrodes was active and reproducible from film-to-
film before and after electrolysis. AEp for Fe(CN)63"* and Ru(NH3)63+’” were low
with nominal value of 76 1: 5 mV and 76 :I: 8 mV, respectively. These values are
comparable with those reported for diamond/8i electrodes and lower than what
have been reported for diamond/Ti. The results indicate that the diamond/Ti
electrode is clean and hydrogen-terminated with density of high electronic states
to support fast electron-transfer kinetics. No conventional pretreatment was
required for activation. Apparent heterogeneous electron-transfer rate constants,
18..., for Fe(CN)63"4‘ and Ru(NH3)63*’2* before electrolysis were 0.010 and 0.019
cm/s, respectively. After electrolysis, k°app for Fe(0N)63"4' decreased to 0.006
cm/s because of surface oxide incorporation on the electrode. k°,,,pp for

' Ru(NH3)63*’”, though, was almost unchanged at 0.0144 cm/s after electrolysis.

69

While the electrolysis changed the electrode surface chemistry, it did not cause a
major change in the density of electronic states. In other words, the electrodes
remain highly conductive after the electrolysis. The SEM, Raman spectroscopy
and XRD all confirmed that the diamond/T i electrode was morphologically and

microstructurally stable under the reported polarization conditions.

70

3.4 References

(1)

(2)

(3)

(4)

(5)

(7)

(8)

(9)

(10)

(11)

(12)

(13)

' (14)

Chen, X.; Chen, G.; Gao, F.; Yue, P. L. Environmental Science and
Technology 2003, 37, 5021-5026.

Fryda, M.; Herrmann, D.; Schafer, L.; Klages, C.-P.; Perret, A.; Haenni, W.;
Comninellis, 0.; Gandini, D. New Diamond and Frontier Carbon
Technology 1999, 9, 229-240.

Agerlll, J. W.; Walukiewicz, W.; McCluskey, M.; Plano, M. A.; Landstrass,
M. I. Applied Physics Letters 1995, 66, 616-618.

Baranauskas, V.; Ceragioli, H. J.; Peterlevitz, A. 0.; Fontana, M. Surface
and Coatings Technology 2005, 200, 2343-2347.

Gerger, l.; Haubner, R.; Kronberger, H.; Fafilek, G. Diamond and Related
Materials 2004, 13, 1062-1069.

Chen, X.; Chen, G. Jouma/ of the Electrochemical Society 2004, 151,
B214-B219.

Pleskov, Y. V.; Evstefeeva, Y. E.; Krotova, M. D.; Lim, P. Y.; Shih, H. 0.;
Varnin, V. P.; Teremetskaya, l. G.; Vlasov, l. l.; Ralchenko, V. G. Joumal
of Applied Electrochemistry 2005, 35, 857-864.

Pleskov, Y. V.; Evstefeeva, Y. E.; Krotova, M. D.; Lim, P. Y.; Chu, 8. 8.;
Ral'chenko, V. G.; Vlasov, l. l.; Kononenko, V. V.; Varnin, V. P.;
Teremetskaya, I. G.; Shi, H. 0. Russian Journal of Electrochemistry 2005,
41, 337-345.

Beck, F.; Krohn, H.; Kaiser, W.; Fryda, M.; Klages, C. P.; Schafer, L.
Electrochimica Acta 1998, 44, 525-532.

Beck, F.; Kaiser, W.; Krohn, H. Electrochimica Acta 2000, 45, 4691-4695.

Hian, L. 0.; Grehan, K. J.; Compton, R. G.; Foord, J. 8.; Marken, F.
Journal of The Electrochemical Society 2003, 150, E59-E65.

Fisher, V.; Gandini, D.; Laufer, 8.; Blank, E.; Comninellis, C.
Electrochimica Acta 1998, 44, 521 -524.

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

Terranova, M. L.; Piccirillo, 8.; Sessa, V.; Compagnone, D.; Sbomicchia,
P.; Rossi, M. Diamond and Related Materials 2001, 10, 627-630.

71

(15)

(15)

(17)

(18)

(19)

(20)

(21)

(22)

(23)

(24)

(25)

(25)

(27)

(28)

(29)

Fagan, D. T.; Hu, l.-F.; Kuwana, T. Analytical Chemistry 1985, 57, 2759-
2763.

Granger, M. 0.; Witek, M.; Xu, J.; Wang, J.; Hupert, M.; Hanks, A.;
Koppang, M. 0.; Butler, J. E.; Lucazeau, G.; Mermoux, M.; Strojek, J. W.;
Swain, G. M. Analytical Chemistry 2000, 72, 3793-3804.

Nicholson, R. 8. Analytical Chemistry 1965, 37, 1351 -1 355.

Show, Y.; Matsukawa, T.; Ito, H.; Iwase, M.; lzumi, T. Diamond and
Related Materials 2000, 9, 337-340.

Swain, G. M.; Rajeshuni, R. Analytical Chemistry 1993, 65, 345-351.
Hupert, M.; Muck, A.; Wang, J.; Stotter, J.; Cvackova, Z.; Haymond, 8.;
Show, Y.; Swain, G. M. Diamond and Related Materials 2003, 12, 1940-
1949.

Swain, G. M. In E/ectroana/ytical Chemistry, Rubinstein, l., Ed.; Marcel
Dekker, Inc.: New York, 2003; Vol. 22, pp 181-277.

Granger, M. 0.; Xu, J.; Strojek, J. W.; Swain, G. M. Analytica Chimica
Acta 1999, 397, 145-161.

Swain, G. M. Journal of the Electrochemical Society 1994, 141, 3382-
3393.

Ramesham, R.; Rose, M. F. Diamond and Related Materials 1997, 6, 17-
27.

Stotter, J.; Zak, J.; Behler, Z.; Show, Y.; Swain, G. M. Analytical Chemistry
2002, 74, 5924-5930.

Fu, Y.; Yan, B.; Loh, N. L. Journal of Materials Science 1999, 24, 2269-
2283.

Gonon, P.; Gheeraert, E.; Deneuville, A.; Fontaine, F.; Abello, L.;
Lucazeau, G. Jouma/ of Applied Physics 1995, 78, 7059-7062.

Ushizawa, K.; Watanabe, K.; Ando, T.; Sakaguchi, |.; Nishitani-Gamo, M.;
Sato, Y.; Kanda, H. Diamond and Related Materials 1998, 7, 1719-1722.

Pruvost, F.; Bustarret, E.; Deneuville, A. Diamond and Related Materials
2000, 9, 295-299.

72

(30)

(31)

(32)

(33)

(34)
(35)

Bernard, M.; Baron, 0.; Deneuville, A. Diamond and Related Materials
2004, 13. 896-899.

Bernard, M.; Deneuville, A.; Muret, P. Diamond and Related Materials
2004, 13, 282-286.

Fryda, M.; Dietz, A.; Herrmann, D.; Hampel, A.; Schafer, L.; Klages, C.-P.;
Perret, A.; Haenni, W.; Comninellis, 0.; Gandini, D. Proceedings -
Electrochemical Society 2000, 99-32 (Diamond Materials), 473-483.

Sailer, R.; McCarthy, G. [000 Grant-in-Aid; North Dakota State University:
Fargo, North Dakota, USA., 1993.

National Bureau of Standards (U. 8.), Monogr. 1981, 25, 73.

National Bureau of Standards (U. 8.), Circ. 1953, 539, 5.

73

CHAPTER4
ELECTROCHEMICAL REMEDIATION OF ATRAZINE
USING A DIAMOND/Ti ANODE

4.1 Introduction

In Chapter 3, the physical and electrochemical properties of the boron-
doped diamond/Ti electrode were reported on. The electrode structure and
electrochemical properties were stable during a 50-h electrolysis at 22 mA/cm2 in
0.1 M H2804. These results indicate that diamond can function as a
dimensionally stable electrode during anodic polarization; a requisite property for
the remediation of organic contaminants. In this Chapter, the oxidative
remediation of atrazine in aqueous media is reported on.

Various methods have been used to remediate atrazine (ATZ) in soil and
water, such as ozonation”, photochemical degradation""7 and bioaugmentation.8'
9 These methods are able to degrade atrazine; however, by-products, such as
hydroxyatrazine (HA), deethylatrazine (DEA) and deisopropylatrazine (DIA), are
commonly formed. The effect of these by-products on the environment is still not
fully understood, so they may be secondary contaminates. Therefore, a method
that can fully oxidize ATZ to 002 would be desirable.

Besides the advantages of environmental compatibility and low energy
consumption, electrochemical methods afford the possibility of completely
oxidizing organic contaminants, either directly or indirectly, to 002. However,

- there are only two reports describing the electrochemical remediation of ATZ and

74

both used an indirect approach involving Fenton reaction.” 11

Q

27)-

Q
o«--
I
)— )—
\ / E
)
Q

2

)

IZ
>__.

Z
>\__

IZ

tit 1 it”; 1*:
H2N N u ”n N n u N NH2
DEA l-lA DIA
" f1
j , H2N N NH2 ‘ 1
0H 0H
NJSN NJSN
HZN’tN’ANk I ”NANANW
”2“. OH "(H
.......... 2 14:1 1
H2N N NH2
ammeline

Figure 4.1 Proposed atrazine degradation pathway for electrochemical Fenton
reaction with bold arrow shown the major pathway and dashed arrow shown the
minor pathway. (Reproduced with permission from Water Research 2002, 36,

5113-5119. Copyright 2002 Elsevier Science Ltd.)

75

Degradation via the Fenton reaction is based on the chemical generation of 0H-
as Shown in the following reaction:5
Fe” + H202 —> Fe3+ + OH' + H0'
One of the Fenton reagents, Fe”, is generated electrochemically using a
sacrificial iron anode to deliver Fe” into solution and another iron electrode
serves as the inert cathode for the reduction of water. The other Fenton reagent,
H202, is added to the contaminated solution. Using this method, Saltmiras et al.
were able to degrade 70% of ATZ in 3 min with the formation of seven
degradation products, as shown in the proposed degradation pathway in Figure
41 .11, 12
Anode: Fe _- Fe” + 2e’
Cathode: 2 H20 + 2e’—-> H2 + 2 OH'

They proposed that dealkylation was the preferred pathway for atrazine
degradation, as shown by the bold arrows. Dechlorination was also observed but
was the minor pathway, as shown by the dashed arrows. Ammeline was the final
degradation product after 10 min. Interestingly, they did not observe any ring
cleavage.11 Using a Similar approach, Ventura et al. generated Fe” and H202
simultaneously in solution and reported that the reaction is more efficient at
generating 0H0 than the method Saltmiras et al. used.”

For both works, the major ATZ degradation product was not 002.
Saltmiras et al. detected only 12.5% 002 after ATZ oxidation using the Fenton
treatment. The factor limiting the complete mineralization may be the electrode

. material. To effectively degrade organic contaminants to 002, the electrode

76

needs to have a wide working potential window, long service-life in harsh
chemical environments, and good activity for OH- generation. Diamond/Ti
electrodes possess these properties (as described in Chapter 3) and have been
used electrochemically treat several types of chemical contaminants in water. For
example, Chen et al. used a diamond/T i anode for the oxidation of carboxylic
acids, phenol, orange II and reactive red HE-3B.” Compared with a Ti/Sb205-
Sn02 electrode, diamond/T i had 1.6-4.3 times higher current efficiency and was
stable for over 300 h.” Fryda et al. deposited diamond films on different large
area substrate materials, including Ti, and used them for the oxidation of alcohols,
organic acids and halogenated aromatic molecules.14 They found that the organic
molecules could be oxidized to 002 without major amounts of other detectable
by-products, even at very low concentrations (<3 ppm).14

The oxidation reaction mechanism for organic pollutants in acidic solution
at boron-doped diamond electrodes has been studied. Marselli15 and Michuaud”
et al. proposed that the electrochemically generated hydroxyl radicals, 0H0, are
the important oxidant generated at diamond during the electrolysis. The
generation of this oxidant is complex with a series of possible parallel reactionsz”
(i) discharge of water on electrode surface with the formation of hydroxyl radicals,

H2O -- OH- + H“ + e'
(ii) oxidation of organic pollutants by the interfacial hydroxyl radicals,
Organic pollutants + OH- —-> 002 + H2O

(iii) formation of H202 by two hydroxyl radicals,

2 OH. —-> H202

77

(iv) oxygen evolution or regeneration of OH' by the dissociation of H202.
H202 ——> 02 + 2H“ + 2e'
H202 _- 2 OH-

Diamond electrodes possess a highly stable microstructure and morphology
during anodic polarization and a large overpotential for oxygen evolution. This
gives the diamond anode high activity for organic pollutant oxidation to 002.15

In this Chapter, the electrochemical degradation of atrazine using a
diamond/Ti anode was studied and the CO2 conversion efficiency was
determined. The atrazine concentration was monitored as a function of
electrolysis time over 120 min. The effect of the current on the remediation was
investigated by performing the electrolysis at 50 and 100 mA (11 mA/cm2 and 22
mA/cmz). The effect of convection on the conversion of atrazine was also studied.
The main atrazine degradation intermediates were separated and identified by
high performance liquid chromatography (HPLC) with UVNis detection. The
atrazine degradation product, 002, was determined by the electrolysis of “0-
labeled atrazine. The main goal of this work was to determine if ATZ could be

completely mineralized to 002 at a diamond anode.

4.2 Results and Discussion

4.2.1 Separation and Detection of Atrazine and Metabolites

Atrazine (ATZ) and its three major intermediate degradation products,
hydroxyatrazine (HA), deethylatrazine (DEA) and deisopropylatrazine (DIA), are

shown in Figure 4.1. These compounds were detected by UVNis spectroscopy,

78

as they are all chromophoric. Figure 4.2 shows UVNis spectra for ATZ, HA, DEA
and DIA (5 ppm each in 0.1 M H2804). A 0.1 M H2804 background spectrum was

also recorded and is shown for comparison (i.e., blank).

 

 

0.8
A ' ,1 ——Atrazlne(222.4 nm)
=3 05 . ,o’r - - - Deethylatrazlne(214.2 nm)
19, :I Deisopropylatrazlne (213.6 nm)
0 :3 ----- Hydroxyatrazlne (215.0 nm)
e .i
c 0.4l-
m e
g .
g 0.2
<
0.0 '-

 

 

 

200 250 ' 300 ' 350 i 400
Wavelength (nm)
Figure 4.2 UVNis spectra for atrazine, hydroxyatrazine, deethylatrazine,

deisopropylatrazine (5 ppm each) all dissolved in 0.1 M H2804. A background

spectrum (blank), 0.1 M H2804, is also shown for comparison.

The blank spectrum is flat and has near zero absorbance over most of the
wavelength range. All four compounds have high absorbance at wavelengths
between 215 and 225 nm. It can be seen that the distinction coefficient, 8, is
similar for ATZ, DEA and DIA within this wavelength range, but smaller for HA.
The absorbance is attributed to the nto 11* transition of electrons in the triazine

ring. For ATZ, the maximum absorbance occurs at 222.4 nm and for HA, DEA

and DIA, the maximum occurs at 215.0, 214.2 and 213.6 nm, respectively. Based

79

on these results, 216 nm was for use in the HPLC analysis.

Reversed-phase HPLC separation was used to determine if the
electrolysis decreased the ATZ concentration and formed any of the intermediate
degradation products (e.g., HA, DEA and DIA). The isocratic separation was
optimal with a mixed mobile phase consisting of 0.05 M phosphate buffer (pH 3.5)
and acetonitrile at a volumetric ratio of 60/40. The mobile phase flow rate was 1.0
mUmin. Table 4.1 lists the molecular structure and pKa values for ATZ and three

possible intermediates: HA, DEA and DIA.

Table 4.1 Molecular structure and pKa value for ATZ, HA, DEA and DIA.

 

 

 

Name ATZ HA DEA DIA
st 2” 2 E
Structure N \N "f \N "f \N N. \N
”NAN/”ML /‘N’I‘N”kN’k HZNJ‘NANk /\NJ\N’)\NH2
H H H H H H
PKa 1.71 5.15 4.57 4.65

 

 

A successful reversed-phase HPLC method requires the proper mobile
phase to achieve optimum peak shape, low limit of detection and reproducible
retention time. The concentrations of acid and conjugate base required to
prepare the pH 3.5 buffer were calculated using the Henderson-Hasselback

equafion:

[A—I
PH = PK. + logifl] (4.1)

80

At pH 3.5, about 90% of all four compounds exist in a single form because the
solution pH is about one unit above or below the analytes’ pKa. A single form of
each analyte assures good peak shape and resolution. The chromatogram for a
solution of ATZ, HA, DEA and DIA is shown in Figure 4.3. The peaks were

identified by peak retention time matching.

DIA

ATZ

 

 

 

0 2 4 6 B 1 0
Minutes (mln)

Figure 4.3 Isocratic separation of 1.25 ppm (each) ATZ, HA, DEA and DIA
using a reversed-phase HPLC separation and UVNis detection at 216 nm. The
mobile phase was 0.05 M phosphate buffer (pH 3.5) mixed with acetonitrile at a
volumetric ratio of 60/40. The mobile phase flow rate was 1.0 mL/min and the

sample injection volume was 20 pL.

Key parameters indicative of the efficiency of the separation are the

retention time (tR), retention factor or capacity factor (k’) and plate number (N).

81

Retention time is the time during which the analyte moves from injection port to
the detector. The longer the retention time, the more strongly the analyte is
retained on the column.17 Capacity factor is an important parameter that

describes the migration rate of a solute. It can be calculated using the equation:

 

k'- (4.2)

where k’ is the capacity factor, tp is the retention time of analyte and try. is the
retention time of unretained species, also called the dead time. An unretained
analyte will have a capacity factor close to unity while a strongly retained analyte
will have a capacity factor >> 1. A capacity factor in the range of 2 to 10 is
desired to avoid excessive zone broadening.17 Plate number is used to
quantitatively evaluate the efficiency of the separation. The efficiency of a
separation increases with the plate number. The plate number can be calculated

using the equation:

2
t
N =16(WR] (4.3)

where N is the plate number, tR is the retention time and W is the peak width at
the base.17

Table 4.2 summarizes the retention time, capacity factor and plate number
of the four analytes: ATZ, HA, DEA and DIA. The elution order is consistent with
the polarity of the compounds. For a reversed-phase separation, a more polar
analyte will elute easier than a less polar one. HA with the hydroxyl group on the

triazine ring is the most polar of the four solutes. DEA and DIA contain primary

82

amine groups and, at low pH, with the buffer protonates the acidic silanols on the
silica surface, the amine retention times shorten and peak Shape sharpens. The
ATZ retention time is considerably longer than that for the degradation by-
products. The retention time for ATZ under these conditions is 8.2 min, giving a k’
of 11 and an N of 2165 plates. The retention times and capacity factors for the
more polar degradation by-products are 1.7 min and 1.5 for HA, 2.4 min and 2.6

for DIA, and 3.1 min and 3.6 for DEA.

Table 4.2 Summary of the retention times, capacity factors and plate numbers

for the four analytes: ATZ, HA, DEA and DIA.

 

 

Analyte t” (min) tn (min) k' W (min) N (plates)
ATZ 0.67 8.2 11 0.71 2165
DEA 0.67 3.1 3.6 0.39 1037
DIA 0.67 2.4 2.6 0.32 894

HA 0.67 1.7 1.5 0.29 558

 

For each of the solutes, the peak height increased linearly with increasing
injected concentration between 10 ppb and 10 ppm (r2 > 0.998), as shown in
Figure 4.4. The concentration of each compound in the electrolysis solution was

then determined from the response curves generated from a set of standards.

83

 

0.14_
0.12-
0.10-
0.081
0.061

0.04 3
0.02 -

 

Absorbance (a. u.)

0.00

. ATZ
,.
. ’ y = 0.0131x
/ R2 = 0.9993

A l A l 4“ 11111

 

 

0.30

g 0.25:
53, 0.20-
8 0.15I
0.101

Absorban

Figure 4.4 Response curves for standard solutions of ATZ, HA, DEA and DIA.

 

0.05 1

. DEA /
F ;///
y = 0.0278x
R2 = 0.9986
/'

 

0.00 ’

 

 

Concentration (ppm)

Absorbance (a. u.)

Absorbance (a. u.)

0.35

0.30 '-
0.25 L
0.20 '-
0.15 ’
0.10 _
0.05 '

0.00

0.5

 

 

HA /)-’
l' ////
* y = 0.0307x
y/ R2 = 0.9994

 

 

Concentration (ppm)

 

0.4 -
0.3 i
0.2 l
0.1 l

 

0.0

. DIA /./
/, y = 0.04441:

. / R2 = 0.9981

,4“ .........

 

Concentratlon (ppm)

4.2.2 Electrochemical Degradation of Atrazine

The ATZ electrolysis tests were conducted in the H-cell shown in Figure
2.3. The electrolysis of 5 ppm ATZ in 0.1 M H2804 was investigated as a function
of the electrolysis time and current density. The first task was to evaluate the
anode response reproducibility and stability for ATZ oxidation. This was done

using a constant current of 100 mA (22 mA/cmz) for 20 min. A summary of the

results are presented in Table 4.3.

84

 

Table 4.3 Atrazine concentration remaining in the anolyte compartment after

20 min of oxidative electrolysis of 5 ppm atrazine in 0.1 M H2804 at a diamond/1' i

 

 

anode.
Atrazine Concentration (ppm)
1 2 3 4 5 Average RSD (%)
Electrode 1 1.7 1.5 1.6 1.5 1.5 1.6 2.7
Electrode 2 2.4 2.3 2.2 2.1 1.9 2.2 4.1

 

Electrolysis was performed at a constant current of 100 mA (22 mA/cmz). The
data presented are the remaining ATZ concentrations in the anolyte
compartment for five different runs, as determined by HPLC analysis. Results for
five measurements using two different anodes are presented. The results
indicate that the ATZ electrolysis is quite reproducible at the diamond/Ti anode.
The relative standard deviation in the concentration remaining after electrolysis
was 2.7% and 4.1%, respectively, for the two anodes. After electrolysis, the most
intense peak in the chromatogram was for ATZ as its concentration decreased
from 5 down to 1 or 2 ppm with three new peaks emerging. The peaks were for
unknown solutes more polar than ATZ as their elution times were from 1 to 3 min.
All three peaks were weak, as compared to the signal for ATZ, with the intensity
being greatest at Short electrolysis times. Their retention times did not match
those for HA, DEA or DIA. No additional work was performed to identify these
three apparent products.

In order to verify that the decreased ATZ concentration was not due to the

85

diffusion from the anolyte to the catholyte compartment, 3 test was performed to
measure the ATZ concentration in the anolyte compartment every 20 min over a
120 min period without any electrolysis. 6 mL of 5 ppm ATZ in 0.1 M H2804 was
placed into the anolyte compartment and 6 mL of 0.1 M H2804 solution was
placed into the catholyte compartment. For the 120 min period and 6
measurements, the ATZ concentration in the anolyte compartment remained
constant with a variance of only 5.5%. It was, therefore, concluded that the
reduced ATZ concentration after electrolysis resulted from the electrochemical
oxidation.

The effectiveness of the diamond/1' i anode for the oxidative degradation of
ATZ, as a function of electrolysis time and current density, was then studied. The
initial ATZ concentration was 5 ppm in 0.1 M H2804. Plots of the percentage of
ATZ remaining in solution, as measured by HPLC, as a function of the

electrolysis time at two different current densities, are shown in Figure 4.5.

 

 

 

2 100 -

'7: n 050m
5 c 30 * I100mA
<9

~53 60 .

9’ O

9’11) 40 t

9 e

:0-

9 l
8 0 -

 

 

 

 

 

 

 

0 5 10 20 40 60 80 100120
Electrolysis Time (min)

Figure 4.5 Percentage of atrazine remaining in solution as a function of the

electrolysis time at two currents: 100 mA (22 mA/cmz) and 50 mA (11 mA/cmz).

86

At both currents, the ATZ concentration decreased over time, with the
largest decrease seen during the first 60 min of electrolysis. After 120 min,
greater than 97% of the ATZ was oxidatively removed at a current of 100 mA (22
mA/cmz) and 83% of the ATZ at a current of 50 mA (11 mA/cmz). As expected,
ATZ degradation is more complete at the higher current density (i.e., more
charge passed). At both current densities, the ATZ concentration decreased
exponentially with time as semilogarithmic plots of the concentration and

electrolysis time were linear (r2 > 0.99), as shown in Figure 4.6.

 

 

 

 

 

 

 

 

g 1.0 \ .5 1.0 \

g . ‘\ 50 mA ‘5 051 >,\ 100 mA

H l H . > >‘\.

g 0.5- \\ § 0.0.. ‘\

g ’ \\‘\ g 0.5: \\\‘

u 0.0T f\\ o .1.o: \\\\g

0 l !\ 0 > \\.

'fi -0.5- y=-0.0145x+0.9134 ‘~'\ '5 1'5. giggszgoean K.\

g * R2: -o.9932 “x g '2- 0f I:

15"”) o ‘20‘40‘60A80A100L120 £45 0 ‘20‘40‘60‘80‘100‘120
Electrolysis Time (min) Electrolysis Time (min)

Figure 4.6 Semilogarithmic plots of atrazine concentration versus electrolysis
time for a diamond/T i anode at two constant currents: 100 mA (22 mA/cmz) and

50 mA (11 mA/cmz).

The linear relationship indicates that the reaction is first-order and reaction rates
for each current density are the negative of the slopes.18 So at the current of 100

mA (22 mA/cmz), the reaction rate is 0.0145 min'1 and at the current of 50 mA

87

(11 mA/cmz), the reaction rate is 0.0245 min".

An effort to further increase the reaction rate was accomplished by
introducing convection to the solution. A small stirring bar was placed in the
anolyte compartment and rotated at 4 rps. The effect of convection was studied
at the constant current of 50 mA (11 mA/cmz). Plots of the percentage of ATZ
remaining in solution, as assessed by HPLC, as a function of the electrolysis time

with and without convection are shown in Figure 4.7.

 

e 100 .

.5 I WIth convection

E 80 Dwithout convection
< .5

'05 5 60

0 o

g’ (D 40

c .E

2° I |
w * l
“- o

 

 

 

 

 

 

 

0 5 10 20 40 60 80 100120
Electrolysis Time (min)

Figure 4.7 Percentage of atrazine remaining in solution as a function of the
electrolysis time at a constant current of 50 mA (11 mA/cmz) with and without

convecfion.

With convection, the electrolysis rate of ATZ was increased due to an increased
flux to the electrode. After 120 min, 96% of the ATZ was oxidized at the constant
current of 50 mA (11 mA/cmz) with convection while only 83% of the ATZ was
oxidized without. The reaction rate was not first-order for ATZ degradation when

convection was present. The logarithmic plots of the ATZ concentration and the

88

electrolysis time were not linear.

As mentioned earlier, none of the three major metabolites (HA, DEA or
DIA) were detected in the HPLC analysis of the anolyte solutions, at least down
to the 10 ppb level, although three low intensity peaks associated with the
unidentified oxidation reaction products did develop. These observations are
consistent with the conclusion that ATZ can be oxidized reproducibly and stably
at a diamond/T i anode and that the three major degradation products are not
formed at appreciable levels under these conditions. One possible reaction
product is 002, as shown in the reaction below:

CaH14N5C| + 31 H20 -+ 8 002 + 5 N03- 4' 76 H+ 4' 70 e’ + Cl-

4.2.3 Determination of the C02 Conversion Efficiency

In order to test this supposition and to determine the conversion efficiency
of ATZ oxidation to C02, if in fact oxidation does occur, the electrooxidation of
14C-labeled atrazine was performed in a sealed assembly using a 0.1 M KOHW
trap for the generated 002 (see Figure 2.4). The reproducibility of the experiment
was first tested by conducting the electrolysis three times using the same
diamond/T i anode under the exact conditions. A constant current of 100 mA (22
mA/cmz) was applied to the cell, initially for 30 min, to oxidize 5 ppm of 14C-
labeled ATZ in 0.1 M H2804. Any ”C-labeled ATZ oxidized to 14002 was trapped

in the KOH solution and analyzed by a liquid scintillation counter.

89

Table 4.4 Reproducibility test of the oxidation of 1“C-labeled atrazine in 0.1 M

H2804 at a diamond/T i anode using a constant current of 100 mA (22 mA/cmz)

 

 

for 30 min.
14 14
Time(min) Total C C02
recovery percentage
30 93.6% 38.5%
30 87.5% 31.2%
30 86.2% 33.4%
Average 89.1 1 3.9% 34.4 :1: 3.7%
RSD 4.4% 10.8%

 

The results in Table 4.4 show that the percentage of total 14C recovered as well
as the percentage of 14CO2 trapped in the KOH(aq) were reproducibly observed.
The percentage of total 1"'C recovered after electrolysis is equal to the total
counts for 14C measured from the anolyte and catholyte compartments of the cell
and the KOH trap divided by the total counts recorded for the starting ATZ
solution. The percentage of 14C trapped in the 0.1 M KOH(aq) solution (trap level
count divided by total count) reflects the amount of ATZ ring cleavage to 002. All
of the counts recorded from the trap should be for 14002. The nominal 14C
recovered after 30 min electrolysis was 89.1% and indicating that the H-cell
assembly was largely free of air leaks. Nominally, 34.4% of the atrazine was
converted to 002 via ring cleavage during the period. Good reproducibility in
terms of the CO2 produced was seen as the standard deviation of these tests

was about 11%.

9O

The electrolysis was then conducted for 60, 90 and 120 min periods in
separate tests to examine how completely the ATZ can be mineralized to CO2.
The results are summarized in Table 4.5. The results for the electrolysis, with

and without added convection, are presented in Figure 4.8.

Table 4.5 Total 14C recovery and 14CO2 production percentage after 30, 60,

90 and 120 min of electrolysis with convection.

 

 

T‘mlm‘") 1335;? 29:33:99
30 89.1% 34.4%
so 96.6% 58.2%
90 99.8% 75.8%
120 99.9% 81.2%

 

 

 

 

 

g 100 “I --— Atrazine degradation without convection) 100 3
3;. -0— C02 production without convection 3
T; —A— C02 production with convectlon i g
(n 80 - 80 g
.s , / . ‘8
o
fi 60 - . 60 90"
«I
u. . O
o H
o - \I o/ l g.
m I C
.3 20 - /<- l 20 g,
o

g . e \I\ , 3
0 I\
a 0 -1 . I a r . I - . . . . 1 0

0 20 40 60 80 100 120

Electrolysis Time (min)

Figure 4.8 Comparison of percentage of atrazine oxidized to 002 over a 120
min electrolysis period at a constant current of 100 mA (22 mA/cmz) with and

without convection.

91

The data clearly show that C02 production increases with electrolysis time. From
30 to 120 min, the percentage of 002 produced increased from 34.4 to 81.2%.
When convection was introduced at 120 min, for example, the 002 production
increased from 51.8 to 81.2%. The data unequivocally demonstrate that the
diamond/T i anode is effective at oxidatively degrading ATZ to 002. The 002
conversion percentage at 120 min (81.2%), is greater than the total ATZ
converted to 002 (12.5%) as reported by Saltmiras et al. using anodic Fenton

treatment, even with a longer electrolysis period.11

4.3 Conclusions

The electrolysis of 5 ppm ATZ in 0.1 M H2804 using a diamond/T i anode
was investigated at two currents: 50 mA (11 mA/cmz) and 100 mA (22 mA/cmz).
The diamond/T i anode was shown to function reproducibly and stably for the
mineralization of ATZ. At the current of 100 mA (22 mA/cmz), 97% of the ATZ
was oxidized during a 120 min electrolysis period and the 14C-labeled
experiments showed that 81.2% of the contaminant was converted to 002.
Convection had a significant effect on the degradation rate suggesting that the
oxidation was occurring at a mass transfer limited rate. At a constant current of
50 mA (11 mA/cmz), without convection in the solution, only 83% of ATZ was
degraded in 120 min while with convection, 96% of the ATZ was oxidized. None
of the three common degradation products, HA, DEA and DIA, for ATZ were

detected by HPLC-UVNis during the electrolysis. However, three unknown

92

peaks associated with compounds more polar than ATZ appeared at shorter
degradation times. They might be degradation products containing multiple
hydroxyl groups. These results prove that the major oxidative degradation
product of ATZ during electrolysis on diamond/T i was 002. The reaction was
first-order as the logarithm of the ATZ concentration changed linearly with the
electrolysis time. These results showed that a diamond/1' i anode is effective for
the oxidative degradation of organic pollutants to 002. They portend the
possibility of using the electrode as an industrial anode for efficient wastewater

decontamination.

93

4.4 References

(1)

(2)

(7)

(3)

(10)

(11)
(12)

(13)

(14)

Benitez, F. J.; Beltran-Heredia, J.; Gonzalez, T.; Acero, J. L. Ozone:
Science & Engineering 1995, 17, 237-258.

Beltran, F. J.; Garcia-Araya, J. F.; Alvarez, P. M.; Rivas, J. Journal of
Chemical Technology & Biotechnology 1998, 71, 345-355.

Nelieu, S.; Stobiecki, M.; Einhom, J. Journal of Chromatography A 2000,
866, 195-201.

Bourgine, F. P.; Chapman, J. l.; Kerai, H.; Duval, J. L.; Green, J. 0.;
Hamilton, D. Journal of the Chartered Institution of Water and
Environmental Management 1995, 9, 417-423.

Hequet, V.; Gonzalez, 0.; Le Cloirec, P. Water Research 2001, 35, 4253-
4260.

Arantegui, J.; Prado, J.; Chamarro, E.; Santiago; Esplugas Journal of
Photochemistry and Photobiology, A: Chemistry 1995, 88, 65-74.

Schmitt, P.; Freitag, D.; Sanlaville, Y.; Lintelmann, J.; Kettrup, A. Journal
of Chromatography, A 1995, 709, 215-225.

Runes, H. B.; Jenkins, J. J.; Bottomley, P. J. Applied Microbiology and
Biotechnology 2001, 57, 427-432.

Silva, E.; Fialho, A. M.; Sa-Correia, l.; Burns, R. 6.; Shaw, L. J.
Environmental Science and Technology 2004, 38, 632-637.

Ventura, A.; Jacquet, G.; Berrnond, A.; Camel, V. Water Research 2002,
36, 3517-3522.

Saltmiras, D. A.; Lemley, A. T. Water Research 2002, 36, 5113-5119.
Walling, 0. Accounts of Chemical Research 1975, 8, 125-131.

Chen, X.; Chen, G.; Gao, F.; Yue, P. L. Environmental Science and
Technology 2003, 37, 5021-5026.

Fryda, M.; Herrmann, D.; Schafer, L.; Klages, C.-P.; Perret, A.; Haenni, W.;

Comninellis, 0.; Gandini, D. New Diamond and Frontier Carbon
Technology 1999, 9, 229-240.

94

(15)

(15)

(17)

(18)

Marselli, B.; Garcia-Gomez, J.; Michaud, P.-A.; Rodrigo, M. A.;
Comninellis, C. Journal of the Electrochemical Society 2003, 150, D79-
D83.

Michaud, P.-A.; Panizza, M.; 0uattara, L.; Diaco, T.; Foti, G.; Comninellis,
0. Journal of Applied Electrochemistry 2003, 33, 151 -1 54.

Skoog, D. A.; Holler, F. J.; Nieman, T. A. Principles of Instrumental
Analysis Fifth Edition; Saunders College Pub. : Harcourt Brace College
Publishers: Philadelphia, 1998.

Atkins, P. The Elements of Physical Chemistry Second Edition; W. H.
Freeman and Company: New York, 1996.

95

CHAPTER 5

A SENSITIVE METHOD FOR INORGANIC As DETECTION
USING ANODIC STRIPPING VOLTAMMETRY AND A Au-
COATED DIAMOND THIN-FILM ELECTRODE

5.1 Introduction

Inorganic arsenic contamination in drinking water is a serious worldwide
threat to human health. As(lll) and (V) are inorganic forms found in ground water
and are much more toxic than the organic forms with As(lll) being the most toxic
of them all.“ 2 The EPA required arsenic maximum contaminant level (MCL) in
drinking water is as low as 10 ppb.3 Therefore, a sensitive, reproducible and
inexpensive method is desired to accurately monitor inorganic arsenic in water.
The methods that are used for arsenic monitoring have been discussed in
Chapter 1. Anodic stripping voltammetry (ASV) is the method of choice because
it provides high sensitivity, good resolution, portability and relatively low cost.
Electrode material is one of the most important factors that affect the detection of
inorganic arsenic using ASV. Several electrode materials were used for arsenic

detection such as Hg," 3'7 Ptf"10 Auz' "’15 and Au-coated carbon electrodes

2,13,16,17 18—20

(graphite and glassy carbon ) and their limitations were discussed in
Chapter 1. Au-coated diamond turns out to be the electrode of choice because
the interaction of As with Au facilitating its deposition and stripping“ 22 and the
stability of diamond makes it an ideal support material for reproducible and stable

detection of arsenic.2‘°"36

96

The As stripping peak current and potential for Au-As interrnetallic
compounds are sensitive to both the solution pH and the electrolyte composition.
For example, it has been reported that the As oxidation half-wave potential (E1,2)
is strongly influenced by the activity of not only H+ but also Cl.“ 38 The stripping
peak potential is sensitive to pH as can be seen in the redox reaction:

HAsO2+3H++3e'«-+As+2 H20

0.059

 

E = E0 + 1og[H*]3[HAs0,]

With increasing pH, the standard reduction potential for this reaction shifts
negatively. In addition, the As stripping peak current and potential are affected by
the Cl' activity as Arnold and Johnson reported the anion can facilitate As
formation by acting as a salt bridge between the working electrode and the
arsenic ions.“ 38

As(OH)2+ + Cl' + e' —+ As (0H)Cl + OH' (fast)

2 As(OH)CI + Cl' + e' —-) 2 AsCl + 2 OH' (slow)

AsCl + e‘ —> As° + or (fast)

In terms of the supporting electrolyte, HCI at concentrations greater than 0.5 M
has been reported by many researchers to be the best for As(lll) detection based

on the sharpness and reproducible nature of the stripping peak.
Due to the stability of As(V), an extremely negative potential is required to
reduce it electrochemically.15'19 However, at such potentials, hydrogen evolution
becomes problematic and greatly inhibits As formation. As(V) is usually

determined by first chemically reducing it to As(lll) and then detecting the As(lll)

species by ASV. Several reducing agents have been investigated, such as

97

N2H4/H0l/HBr,39 NaBr/N2H4/H2SO4,4° CU2(3I;2,13'41 Na2SOg,12' 14' ‘2 NaHSOa,43

I.“ ascorbic acid and Kl.6'

gaseous SO2,5 H28,44 Na2S203,3 L-cysteine,7 mannito
4648 Some of these agents are not suitable for ASV and are commonly used for
cathodic stripping voltammetry with a Hg electrode, such as N2H4/HCl/HBr,
NaBr/N2H4/H2SO4, ascorbic acid and Kl, Na2S203, and mannitol. Some of these
agents can also chemically react with gold, such as L-cysteine. The use of
Cu20|2 requires a complex setup and would not facilitate on-site application.
M2803 and NaH803 provide efficient conversion of As(V) to As(lll) and both
have a minimal effect on Au. They are also more easily used and controlled than
gaseous reagents (e.g., $02 and H28).

In this Chapter, we report on a sensitive, reproducible and stable method
for total inorganic arsenic analysis in water samples using DPASV and a Au-
coated diamond electrode. The electrode and method were tested using both
laboratory-prepared and local river water samples. We sought (i) to better
understand Au metal phase formation on diamond surfaces, (ii) to optimize the
DPASV experimental parameters for arsenic determination and (iii) to improve
the sensitivity, reproducibility and stability of the method for total inorganic
arsenic analysis as compared to the figures of merit commonly achieved with Au

and Au-coated glassy carbon electrodes. The electrode and method were tested

using both laboratory-prepared and local river water samples.

98

5.2 Results and Discussion

5.2.1 Particle Analysis of As and Au Deposition on a Boron-

Doped Diamond Electrode

As discussed in the Introduction, Au is the electrode of choice for As(lll)
detectionby ASV with the best supporting electrolyte being HCI.2 The first task
was, therefore, to learn about Au deposition on diamond thin-film surfaces. SEM
was used to investigate the morphology and distribution of the Au and Au + As
deposits on diamond after electrodeposition at a constant potential. (shown in
Figure 5.1). A summary of the particle analysis data is presented in Table 5.1.

Figure 5.1 (a) shows a bare diamond surface without any metal deposition.
The polycrystalline morphology is evident with a crystallite size ranging from a
few micrometers down to several hundred nanometers. The surface is largely
free of secondary growths (i.e., smaller crystallites formed on the primary ones).
There is some surface charging (bright spots) on some of the crystalline facets
indicative of unequal electrical conductivity.“53 For comparison, Figure 5.1(b)
shows a diamond surface after 90 s of deposition at -0.45 V in 10 ppm Au(lll) + 1
M HCI. The Au particles are evenly distributed over the diamond surface and are
located in both the grain boundaries and on the facet surfaces. The particle
shapes are generally round or elliptical. The particle size is relatively uniform with
a nominal diameter of 23 i 5 nm, a particle area of 3.6 (i 1.6) x 10'12 cm2 and a
particle density of 6.4 (i 1.2) x 109 cm'2. The narrow particle size distribution is
consistent with that expected for an instantaneous nucleation mechanism and a

relatively constant rate of particle growth.“ 55 This nominal particle density is

99

comparable to the value Dai et al. reported (~7.5 x 109 cm'2 for a 15 s deposition)
for Au deposited on glassy carbon. However, the Au particle diameter is larger in
the present case because of the longer deposition time (90 s).18

Figure 5.1(c) shows a diamond surface after 90 s of deposition in 10 ppm
As(lll) + 1 M HCI at -0.45 V. No As particles were detected, even at a
magnification of 50,000x. This indicates that little or no As deposits on the bare
diamond surface. At best, only a monolayer of As would be expected because of
the semiconducting electronic properties of the metal.2' 11' 18 This is why bare
diamond can not be used for arsenic analysis. Figure 5.1(d) shows a diamond
surface after 90 s of deposition at -0.45 V in 10 ppm As(lll) and 10 ppm Au(|ll) +
1 M HCI. Au provides a medium onto which As can deposit and form a stable
intermetallic compound. The nominal particle diameter is 22 i 3 nm, the particle
area is 2.4 (:l: 0.7) x 10‘12 cm2 and the particle density is 3.8 (i 0.5) x 1010 cm'z.
The nominal particle size is about the same but the particle density is higher than
that for the Au-only coated surface (Figure 5.1(b)). In both cases, the particles
are dispersed over the entire diamond surface. Presumably because of its
electronic properties, As reduces the Au particle size when co-deposited.56 It has
been reported that As deposits on the surface of the Au rather than within the
bulk metal.56 Therefore, we suppose the reason for the larger particle density
when Au and As are co-deposited is because once As covers a Au particle
surface, further growth of the Au particle is inhibited. Therefore, additional Au

deposition can only occur at new sites on the surface.56

100

 

Figure 5.1 SEM images of (a) a bare diamond electrode surface without any
metal deposition, (b) a diamond surface after 90 s of deposition at -0.45 V in 10
ppm Au(lll) + 1 M HCI solution, (c) a diamond surface after 90 s of deposition at -
0.45 V in 10 ppm As(lll) + 1 M HCI solution and (d) a diamond surface after 90 s

of deposition at -0.45 V in 10 ppm As(lll) and 10 ppm Au(l|l) + 1 M HCI.

Table 5.1 Particle analysis results for Au and As/Au metal phase formation on

 

 

diamond.
10 ppm Au(lll) +
10 ppm Au(lll) 10 ppm As(lll)
Mean particle diameter (nm) 23 i 5 22 i 3
Mean particle area (cmz) 3.6 (i 1.6) x 10'12 2.4 (i 0.7) x 10‘12

Particle density (cm'z) 6.4 (i 1.2) x 109 3.8 (a 0.5) x 101°

 

101

5.2.2 Effect of Au Coverage on the As Stripping Peak Current

The effect of the Au coverage on the As stripping peak current was
investigated. Figure 5.2 shows DPASV i-E curves for 100 ppb As(lll) + 1 M HCI
at a diamond electrode as a function of the Au(|ll) concentration in solution

added by standard addition from 0 to 750 ppb.

 

750 ppb

.5
I
?-f-

N
A l A

w
l

Peak Current (M)

.5
- l

 

 

 

o '260'460'860'860'
Au Concentration (ppb) I

  

A
I

 

As Peak Current (uA)
N

O
l

0 ppb

 

 

 

 

-0.4 -0.2 0.0 0.2 0.4 0.6
Potential (V) vs. AgIAgCI

Figure 5.2 DPASV i-E curves for 100 ppb As(lll) + 1 M HCI at a diamond
electrode as a function of the Au(lll) concentration in solution. A plot of the As
stripping peak current versus the Au(lll) solution concentration is provided as an

inset. DPASV parameters: Edep = -0.45 V, tdep = 120 s, pulse width = 50 ms, pulse

102

amplitude = 50 mV, step height = 4 mV, cycle period 200 ms, and sampling time

= 8.33 ms.

In these measurements, Au was co-deposited along with As. A plot of the As
stripping peak current versus the Au(|ll) concentration in solution is provided as
an inset. Note that no As stripping peak current was recorded for the bare
diamond. The As stripping peak, centered at 0.13 V and observed only with co-
deposited Au, is narrow and sharp for all the Au(lll) concentrations. The peak
widths (AW‘l/Z) range from 30 to 35 mV, depending on the Au surface coverage.
The As stripping current scaled proportionally with the Au(|ll) solution
concentration (i.e., Au coverage), as the response curve has a slope of 5.19
prm and a linear regression correlation coefficient of 0.997. The fact that the
stripping peak current increased proportionally with the Au(l|l) concentration
indicates the As is depositing on and interacting with the Au rather than the
diamond. Based on these results, an Au(lll) concentration of 100 ppb was

selected for use in the method.

5.2.3 Electrochemical Measurements

Figure 5.3 shows cyclic voltammetric i-E curves for a diamond electrode in
contact with a 100 ppb Au(lll) + 1 M HCI solution as a function of the As(lll)
solution concentration. Curves were recorded for standard additions of As(lll)
from 500 to 1000 ppb. The potential was scanned between -0.40 and 0.40 V at a

scan rate of 100 mV/s. Therefore, the cathodic potential limit was sufficient for

103

the co-deposition of both Au and As. The peak current at 0.16 V is due to the
oxidation of As to As(lll). The oxidation peak current increased linearly with the
As(lll) concentration in solution and linearly with the square root of the potential
sweep rate. The latter trend indicates that the oxidation reaction rate is limited by
semi-infinite linear diffusion of As(lll) away from the surface. The mass transfer
limitation was further confirmed by noting that the As stripping peak current
increased 4 to 5 times when convection was introduced by simple N2 gas
bubbling. The cathodic peak observed at about -0.38 V increased proportionally
with the As(lll) concentration and is attributed to the three-electron reduction of

As(lll) to As (i.e., formation of Au-As interrnetallic).18

 

 

 

Current mA)
_b c
01 O

I
—‘
O

 

 

 

-o.4 ‘ 9.2 ‘ 0:0 A o.L2 ‘ 0.4
Potential (V) vs. AglAgCl

Figure 5.3 Cyclic voltammetric i-E curves for a diamond electrode in contact
with a 100 ppb Au(||l) + 1 M HCI solution as a function of the As(lll) solution

concentration, which ranged from 500 to 1000 ppb. Scan rate = 100 mV/s.

104

A series of measurements was made to investigate the effect of the
deposition potential and time on the As stripping peak. Figure 5.4 shows how the
As stripping peak current density changed as a function of the deposition
potential at a Au-coated diamond. The current density was calculated by dividing

the peak current by the electrode’s geometric area (0.2 cmz).

 

 

 

 

A 5

N .

,5 3f //‘\

2 2 \

8 TV‘

8 0% \0—0
-0.6 -o.4 -o.2 o.o

Deposition Potential (V) vs. AgIAgCI

Figure 5.4 A comparison of the effect of the deposition potential on the As
stripping current density at a Au-coated diamond and Au foil electrode in a
solution of 200 ppb As(lll) and 50 ppb Au(lll) + 1 M HCI. DPASV parameters: tdep
= 110 3, pulse width = 50 ms, pulse amplitude = 50 mV, step height = 4 mV,
cycle period 200 ms and sampling time = 8.33 ms.

For the Au-coated diamond electrode, deposition of As commenced at -
0.2 V vs. Ag/AgCl. The stripping peak current increased as the deposition
potential was made more negative up to -0.5 V. For deposition at potentials

negative of this value, a decrease in the stripping peak current was observed.

105

This suggests that less elemental As was formed due to the competitive
generation of H2 or AsH3.2 This result is in agreement with the observation of
Feeney and Kounaves for a Au ultramicroelectrode array.2 Based on these

results, an optimum deposition potential of -0.45 V was selected for the Au-

coated diamond electrode.

 

0|

09

Current Density (pAIcmz)
N Is

 

 

1 - J A L a 4 - A . l .
0 50 100 150 200 250 300
Deposition Time (s)

 

Figure 5.5 Effect of deposition time on the As stripping peak current density at
a Au-coated diamond electrode in a solution of 200 ppb As(lll) and 50 ppb Au(|ll)
+ 1 M HCI. DPASV parameters: Edep = -0.45 V, pulse width = 50 ms, pulse
amplitude = 50 mV, step height = 4 mV, cycle period 200 ms and sampling time =
8.33 ms.

Figure 5.5 shows the relationship between the As stripping peak current
density and the deposition time at -0.45 V for a Au-coated diamond electrode.
The peak current increased proportionally with time between 35 and 160 3.

However, for deposition times greater than 160 s, the stripping peak current

106

leveled off. This is consistent with a saturation of the available Au surface sites
for As deposition. It is supposed that once As covers the entire Au particle
surface, no more As can be deposited.2 Based on these results, a deposition
time of 120 s was selected for use in the method; a balance between maximum

signal and excessive deposition time.

01

 

R2 = 0.999

.5

N
V I

Current Density (Wcmz)
d u

 

 

l A L A J

20 40 so A 80
Pulse Amplitude (mV)

0

 

D

Figure 5.6 Effect of the pulse amplitude on the As stripping peak current
density at a Au-coated diamond electrode in a solution of 200 ppb As(lll) and 50
ppb Au(lll) + 1 M HCI. DPASV parameters: Edep = -0.45 V, tdep = 120 5, pulse
width = 50 ms, step height = 4 mV, cycle period 200 ms and sampling time =

8.33 ms.

Figure 5.6 shows how the As stripping peak current density changed with
the pulse voltage amplitude at a Au-coated diamond electrode. According to the
equation derived by Lund and Onshus for the differential pulse voltammetric

response of a hanging mercury drop electrode (HMDE), the peak current should

107

scale proportionally with the pulse amplitude.57
Aip = knerEu1’2tdCb

In this equation, k is the electron transfer rate constant, n is the number of the
electrons transferred, r is the radius of the mercury drop, AE is the pulse
amplitude, u is the rotation speed of the stirrer, t, is the deposition period and 0.,
is the concentration of metal ion in the sample solution. The linearity between the
As stripping peak current and the pulse amplitude (10 to 75 mV) is shown in
Figure 5.6. Increasing the pulse amplitude caused an increase in the sensitivity
of the measurement but a decrease in the peak resolution.58 A pulse amplitude of

50 mV was selected for use in the method.

5.2.4 Au-Coated Diamond vs. Au Foil for As(lll) Detection

Figure 5.7A shows DPASV i-E stripping curves for As co-deposited with
Au from standard additions of As(lll) in 100 ppb Au(l|l) + 1 M HCI for Au-coated
diamond electrode. A response curve in the inset shows the proportionality
between the As stripping peak current and the As(lll) solution concentration
between 0.01 and 40 ppb. Figure 573 shows comparison curves for Au foil.
Each point is an average of 3 measurements with the bars representing the
standard deviation. The As stripping peak current increased at both electrodes
proportionally with the As(lll) solution concentration as both curves had linear

regression coefficients of > 0.990.

108

 

P
on

 

 

 

 

50 ppb

A 0.4 A
3
ii 0.3
g .
_g 0-2 - 6 1'0 20 5070
3 As Concentration (ppb)
0.
2 0.1 -

0.0 i-

 

 

 

 

Potential (V) vs. AgIAgCI

 

 

 

 

 

 

 

 

20 “3)
A b 8'
g 15 3,,
5 ' 815‘
t 10 g ‘
8 ’2 ‘°‘
x ’ 8 5:
t! a.
8 5 - .. , a .
U) 0 200
< MW(wb)

o .
-o.4 -o.2 0.0 0.2 0.4

Potential (V) vs. AglAgCl

Figure 5.7 DPASV i-E stripping curves for As using standard additions of As(lll)

at (A) a Au-coated diamond electrode in a solution of 100 ppb Au(lll) + 1 M HCI

109

and (B) a Au foil electrode in a solution of 1 M HCI. The As(lll) concentration
ranged from 1 to 50 ppb. A response curve in the inset shows the proportionality
between the As stripping peak current and the As(lll) solution concentration.
DPASV parameters: Edep = -0.45 V. tdep = 120 s, pulse width = 50 ms, pulse
amplitude = 50 mV, step height = 4 mV, cycle period 200 ms and sampling time =

8.33 ms.

Table 5.2 Comparison of the ASV detection figures of merit for As at Au-

coated diamond and Au foil electrodes.

 

 

Au-Coated Diamond Au Foil
LOD (SIN = 3) 0.005 ppb 0.25 ppb
LOD (R2 = 0.99) 0.01 — 40 ppb 0.25 — 1200 ppb
Sensitivity 9.7 :l: 2.5 prm 48.2 :I: 10.2 prm
Response Precision (RSD) 4.4% 4.7%
FWHM at 50 ppb 38 mV 85 mV

 

Table 5.2 presents a summary of the As(lll) detection figures of merit for
the two electrodes. The LOD for diamond was estimated to be 0.005 ppb (SIN =
3) with a linear dynamic range (LDR) from 0.01 to 40 ppb and a slope or a
sensitivity of 9.7 i 2.5 prm. The lowest concentration actually measured was
0.01 ppb. The response reproducibility for both electrodes was excellent with
values less than 5% for 10 consecutive measurements. The As stripping peaks

for the Au-coated diamond were a factor of two or more narrow than the peaks

110

for the Au foil. This is attributed to As oxidizing from Au sites on diamond that are
more morphologically and electronically homogeneous than the sites on Au foil.
The larger sensitivity for the Au foil can be attributed to a higher Au surface area
available for As deposition.

The LOD and response sensitivity for the Au-coated diamond electrode
are as good or superior to other methods reported in the literature.13'"12°9259 A
summary of data reported for Au-coated carbon electrodes is presented in Table
5.3. The data from the present work are included. The Au-coated diamond
electrode has an LOD (0.005 ppb) lower than any of the other Au-coated carbon
electrodes (0.0096 ppb for Au-coated glassy carbon18 and 0.4 ppb for Au-coated
graphite”) with a smaller amount of Au (0.1 ppm) in solution and a similar
deposition time (120 s). The linear dynamic range from 0.01 to 40 ppb is wider
than other reports for both Au-coated glassy carbon (5 to 50 ppb)20 and Au-
coated graphite (0 to 20 ppb)”. The sensitivity for the Au-coated diamond
electrode (9.7 i 2.5 uA/ppm) is smaller than that reported for the Au-coated
glassy carbon (240 uA/ppm)18 and Au-coated graphite (22.5 :l: 0.9 x 106
pA/ppm)19, but this can be attributed to a reduced Au surface area available for
deposition. The precision and accuracy of Au-coated diamond electrode are
comparable to the other Au-coated carbon electrodes. In summary, the Au-
coated diamond electrode provides as good or superior detection figures of merit
for As(lll) as compared to other electrodes and can be used without any time
consuming pretreatment. There are additional advantages of this electrode that

are described in Chapter 6.

111

 

was. 3:395:85 ”asugfig ”DOA ”EEESaflagaujzuglosozuio
Egg; 5w; .Emgi ”mung-993338 Rte—559‘
gguaBQEEa—Saguegigefiggggfifimgta

an

 

I l t can: www.8— Nt our no. >m< :8 8. 28:85
#5.». .8. .2 x 8: to 8. n9 >m< :8 8... 28385
- .53 8.: a nun. ow - o .

.8: - I a: I . I 98.33
-..: 88 .8 8.8.: cu. m. 9 >82 2 size...
I .8. - .8 I 8: «no cm. 8.9 we 8: 8m 855.8
32.» - me. - 3w 88.6
. . - I 8: . . .888 8.88
.33: .8 e 8-... mm: 8 3. we do 8. 38:6
I I I c . 8:an
3m 88.: 8. 93. >9 9 m 5 3.8.9
H
.3. is: mu so 8-5.: 8:: ca. 9:. >w< do 3 .8388
€82... 3.5 83} E .58.
59
$8.82 . .85 3388 m9 8.. E 8. .9 a 8w 8.8: ...... .2 :0 083m

 

.gggggflabegboggn— iii-:0 Boom—03500 ado-nah

112

5.2.5 As(V) Detection

Due to kinetic limitations, As(V) is difficult to detect directly by ASV.
Therefore, its analysis is often performed by an initial chemical reduction to As(lll)
followed by ASV detection. Na2803 was used as the reducing agent because of
its high efficiency.” 14' 42 Several factors can affect the As(V) to As(lll)
conversion efficiency when using this reducing agent: heating temperature,
heating time, Na2803 concentration and solution pH because they determine the
amount of 8032' reactant available. 50 °C was used for heating because at this
temperature the 8032' production rate is high without great loss due to the
formation of 802. A 30 min heating period was found to provide a high
conversion efficiency (92.6 :t 10.9%, n=5). A N32803 concentration of 20 mM
was determined to be sufficient to reduce 50 ppb As(V) to As(lll) under such

heating conditions.

 

 

 

 

0.5
' R2=0.995
20.4-
i=1 .
20.3-
g .
50.2-
U
a ‘ ‘
00.1)-
o.
0.0
0 1o 20 30 4o 50

As (V) Concentration (ppb)

113

Figure 5.8 A response curve showing the proportionality between the As
stripping peak current and the As(V) solution concentration at a Au-coated
diamond electrode. The As(V) concentration ranged from 1 to 50 ppb. DPASV
parameters: Edep = -0.45 V, tdep = 120 s, pulse width = 50 ms, pulse amplitude =

50 mV, step height = 4 mV, cycle period 200 ms and sampling time = 8.33 ms.

Figure 5.8 shows that the stripping peak current for As (As -—+ As(lll)) increased
linearly with the As(V) solution concentration. This indicates the As(V) reduced to
As(lll) with high efficiency. The LOD was determined to be 0.08 ppb, the linear
dynamic range was from 0.08 to 50 ppb and the sensitivity was 9.1 prm. The
sensitivity is very close to that for As(lll) detection, as expected. A small error bar
was obtained for every concentration of As(V) tested (from 2 to 50 ppb), which

means the method is reproducible.

5.2.6 Real Sample Analysis

The Au-coated diamond electrode and the DPASV method were used to
determine the concentration of As(lll) in the local Red Cedar River water. Using
the standard addition method, the river water was found to contain 0.6 :l: 0.1 ppb
As(lll). This value was arrived at by making standard additions of As(lll) to the
river water and generating a response curve. Figure 5.9 shows DPASV stripping
voltammetric i-E curves and the corresponding response curve for standard
additions of As(lll) from 2.5 to 40 ppb to acidified river water containing 100 ppb

Au(lll). An electrode response sensitivity of 5 prm was obtained from the

114

 

curve. Higher baseline noise and lower sensitivity were observed for the

electrode in river water, as compared to the standard solutions.

 

 

 

 

     

 

 

0.20 - 45 PP5
A 0.20:
.- , A

E. 20.15
.. 0.15 g
C 3010
Q) ,0
t goes.
3 _ .
0010 0.00rr-.v.-.

0 10 20 30 40
g Arsenic Concentration (ppb)
0
o. 0.05
(D
<

 

 

      

0.00 -
-o.4 ‘ -o.2 0.0 0.2 A 0.4
Potential (V) vs. AglAgCl

 

Figure 5.9 DPASV i-E curves and the corresponding response curve for the
addition of 2.5 - 40 ppb As(lll) to Red Cedar River water acidified with 100 ppb
Au(lll) + 1 M HCl on a Au-coated diamond electrode. DPASV parameters: Edep =
-0.45 V, tdep = 120 8, pulse width = 50 ms, pulse amplitude = 50 mV, step height

= 4 mV, cycle period 200 ms and sampling time = 8.33 ms.

The As deposition charge was measured for both the standards and river
water samples containing 100 ppb As(lll) by holding the potential at -0.45 V for

120 s and integrating the current-time profile. It was found that the deposition

115

charge for the same concentration of A30") in the river water was about half of
that for the standard sample. This is the reason for the lower sensitivity and can
be attributed to the deleterious influence of other metal ion or organic matter
interferences present in the river water. For example, it has been reported that
Cu(ll), Hg(ll) and Pb(ll) can all cause a decrease in the As stripping peak
current.2 Pb(ll) poses a significant interference for As deposition, particularly
when its concentration is 5 times higher than that of arsenic. Hg(ll) also interferes
with As deposition when present at equal or greater concentrations. Cu forms a
complex with As in the solid phase, thereby reducing the As(lll) activity in
solution and decreasing the amount available for codeposition with Au.2
According to the lCP-MS analysis, the river water contained about 0.1 ppb Cu(ll)
and 5 ppb Pb(ll). Therefore, both contaminants are a cause for the lower
sensitivity. Organic matter in the water could also have been another reason
since these molecules can adsorb on the electrode, block reaction sites and
prevent As deposition. However, results described in Chapter 6 show that the
presence of up to 5 ppm of humic acid did not cause a significant decrease in the
Au-coated diamond electrode response magnitude and reproducibility. Therefore,
it is concluded that the main reason for the decreased electrode sensitivity for
inorganic arsenic in river water was the presence of interfering Cu(ll) and Pb(ll).
Results described in Chapter 6 show that the effect of metal ion interferences
can be mitigated by the use of solid phase extraction (cation exchange) in the
sample preparation.

The reproducibility and long-term response stability of the Au—coated

116

diamond was assessed using acidified Red Cedar River water (1 M HCI) spiked
with 20 ppb As(lll). Au-coated diamond and glassy carbon electrodes were
compared under the same conditions. A run-to-run reproducibility test was
performed over 10 consecutive measurements. Longer term stability studies
were carried out by leaving the electrode exposed continuously to the As(lll)
solution for 10 h and periodically measuring the As stripping peak current (each
hour). For the 10 consecutive measurements, the Au-coated diamond showed
RSD < 2% for both peak current and charge, while the Au-coated glassy carbon
yielded a 4.4% RSD in the peak current and a 7.5% RSD in the peak charge. For
the 10-h stability test, the Au-coated diamond showed superior stability with
response loss of only 9.1% in the peak current and 6.9% in the peak charge as
compared to the Au-coated glassy carbon, which showed a response loss of
>15% in both the peak current and charge. The reproducibility and stability

results are summarized in Table 5.4.

Table 5.4 Comparison of As Stripping Peak Reproducibility and Stability at

Au-Coated Diamond and Glassy Carbon Electrodes in River Water.

 

 

 

Peak Current Peak Charge
Au-coated Au-coated
RSD (n = 10) $332? Glassy 3:32;? Glassy
Carbon Carbon
Run-to-run test 1.5% 4.4% 1.9% 7.5%
10 h test 9.1% 15.5% 6.9% 15.7%

 

The accuracy of the method was assessed by comparing the

117

electrochemical results with those obtained by lCP-MS. A standard solution
containing 15 ppb As(lll) was analyzed using both methods. Response curves
were generated by measuring the signals of different concentrations of As(lll)
standard solutions. A value of 14.5 i 0.2 ppb was obtained by lCP-MS and a
value of 16.1 i 0.4 ppb was found electrochemically. This represents a difference

of 11%.

5.3 Conclusions

The Au-coated diamond electrode provided a sensitive, reproducible and
stable response for inorganic arsenic in both standard and real samples. The Au-
coated diamond electrode offers advantages over commonly used electrode
materials, such as Au and Au-coated carbon electrodes, in terms of the peak
shape and detection figures of merit. Au deposited on diamond from an Au(lll)
solution in the form of relatively small particles with a small variance (22 :l: 3 nm)
and high particle density (1010 cm'z). Sub-ppb LODs were obtained for As(lll) and
As(V) in both standard and local river water samples. The responses of Au-
coated diamond and a Au foil electrode for As(lll) detection were compared using
the same experimental conditions. Au-coated diamond was found to yield
narrower stripping peaks and a lower LOD than Au foil. This method was shown
to be practical for real sample analysis as 0.6 ppb of As(lll) was detected in local
river water with the electrode response variance less than 2% (RSD) for 10
consecutive measurements and a response stability over a 10-h period of better

than 94%.

118

5.4 References

(1)
(2)
(3)
(4)
(5)
(6)

(7)

(8)
(9)

(10)

(11)

(12)

(13)

(14)

(15)

(16)

Li, H.; Smart, R. B. Analytica Chimica Acta 1996, 325, 25-32.

Feeney, R.; Kounaves, S. P. Analytical Chemistry 2000, 72, 2222-2228.
Ferreira, M. A.; Barros, A. A. Analytica Chimica Acta 2002, 459, 151-159.
Sadana, R. S. Analytical Chemistry 1983, 55, 304-307.

Holak, W. Analytical Chemistry 1980, 52, 2189-21 92.

Barra, C. M.; Correia dos Santos, M. M. Electroanalysis 2001, 13, 1098-
1104.

Adeloju, S. 8.; Young, T. M.; Jagner, D.; Batley, G. E. Analytica Chimica
Acta 1999, 381 , 207-213.

Williams, D. G.; Johnson, D. C. Analytical Chemistry 1992, 64, 1785-1789.

Hignett, G.; Wadhawan, J. 0.; Lawrence, N. S.; Hung, D. Q.; Prado, C.;
Marken, F.; Compton, R. G. Electroanalysis 2004, 16, 897-903.

Wei, 2.; Somasundaran, P. Journal of Applied Electrochemistry 2004, 34,
241-244.

Simm, A. 0.; Banks, C. E.; Compton, R. G. Analytical Chemistry 2004, 76,
5051-5055.

Forsberg, G.; O'Laughlin, J. M.; Megargle, R. G. Analytical Chemistry
1975, 47, 1586-1592.

Davis, P. H.; Dulude, G. R.; Griffin, R. M.; Matson, W. R.; Zink, E. W.
Analytical Chemistry 1978, 50, 137-143.

Feeney, R.; Kounaves, S. P. Talanta 2002, 58, 23-31.

Prakash, R.; Srivastava, R. C.; Seth, P. K. Electroanalysis 2003, 15, 1410-
1414.

Kamenev, A. |.; Orlov, S. E.; Lyakhov, A. B. Jouranal of Applied
Electrochemistry 2001, 56, 850-854.

119

(17)

(18)

(19)

(20)

(21)

(22)

(23)

(24)

(25)

(26)

(27)

(28)

(29)

(30)

(31)

Simm, A. 0.; Banks, G. E.; Wilkins, S. J.; Karousos, N. G.; Davis, J.;
Compton, R. G. Analytical and Bioanalytical Chemistry 2005, 381 , 979-
985.

Dai, X.; Nekrassova, 0.; Hyde, M. E.; Compton, R. G. Analytical
Chemistry 2004, 76, 5924-5929.

Viltchinskaia, E. A.; Zeigman, L. L.; Garcia, D. M.; Santos, P. F.
Electroanalysis1997, 9, 633-640.

Munoz, E.; Palmero, S. Electroanalysis 2004, 16, 1528-1535.

Akhmedzhanova, G. M.; Tikhomirova, V. l. Experiment in Geosciences
1995, 4, 26-28.

Tikhomirova, V. l.; Akhmedzhanova, G. M.; Nekrasov, I. Y. Experiment in
Geosciences 1994, 3, 1-9.

Swain, G. M. Journal of The Electrochemical Society 1994, 141 , 3382-
3393.

Swain, G. M.; Ramesham, R. AnalyticalChemistry1993, 65, 345-351.

Muna, G. W.; Tasheva, N.; Swain, G. M. Environmental Science and
Technology 2004, 38, 3674-3682.

Haymond, S.; Babcock, G. T.; Swain, G. M. Electroanalysis 2003, 15, 249-
253.

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

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

Hupert, M.; Muck, A.; Wang, J.; Stotter, J.; Cvackova, Z.; Haymond, S.;
Show, Y.; Swain, G. M. Diamond and Related Materials 2003, 12, 1940-
1949.

Show, Y.; Matsukawa, T.; Ito, H.; Iwase, M.; lzumi, T. Diamond and
Related Materials 2000, 9, 337-340.

lniesta, J.; Michaud, P. A.; Panizza, M.; Cerisola, G.; Aldaz, A.;
Comninellis, C. Electrochimica Acta 2001, 46, 3573-3578.

120

(32)

(33)

(34)

(35)

(36)
(37)

(33)
(39)

(40)

(41)

(42)

(43)

(44)

(45)

(46)

(47)

(48)

Rodrigo, M. A.; Michaud, P. A.; Duo, l.; Panizza, M.; Cerisola, G.;
Comninellis, C. Journal of the Electrochemical Society 2001, 148, D60-
D64.

lniesta, J.; Exposito, E.; Gonzalez-Garcia, J.; Montiel, V.; Aldaz, A. Journal
of the Electrochemical Society 2002, 149, D57-D62.

Terashima, C.; Rao, T. N.; Sarada, B. V.; Tryk, D. A.; Fujishima, A.
Analytical Chemistry 2002, 74, 895-902.

Sonthalia, P.; McGaw, E.; Show, Y.; Swain, G. M. Analytica Chimica Acta
2004, 522, 35-44.

McGaw, E. A.; Swain, G. M. Analytica Chimica Acta 2006, 575, 180-189.

Cavicchioli, A.; La-Scalea, M. A.; Gutz, l. G. R. Electroanalysis 2004, 16,
697-711.

Arnold, J. P.; Johnson, R. M. Talanta 1969, 16, 1191-1207.

Kotoucek, M.; Vasicova, J.; Ruzicka, J. Mikrochimica Acta 1993, 111, 55-
62.

Sancho, D.; Vega, M.; Deban, L.; Pardo, R.; Gonzalez, G. Analyst 1998,
123, 743-747.

Lee, S. W.; Meranger, J. C. Analytical Chemistry 1981 , 53, 130-131.

Hamilton, T. W.; Ellis, J.; Florence, T. M. Analytica Chimica Acta 1980,
119, 225-233.

Henry, F. T.; Kirch, T. 0.; Thorpe, T. M. Analytical Chemistry 1979, 51,
215-217.

Rochette, E. A.; Bostick, B. C.; Li, G.; Fendorf, S. Environmental Science
and Technology 2000, 34, 4714-4720.

Greulach, U.; Henze, G. Analytica Chimica Acta 1995, 306, 217-223.

Kowalska, J.; Stryjewska, E.; Szymanski, P.; Golimowski, J.
Electroanalysis 1999, 11, 1301 -1 304.

Kowalska, J.; Golimowski, J. Electroanalysis 1998, 10, 857-859.

Kowalska, J.; Golimowski, J.; Kazimierska, E. Electroanalysis 2001, 13,
872-875.

121

(49)

(50)

(51)

(52)

(53)

(54)

(55)

(55)

(57)
(53)

(59)

Holt, K. B.; Bard, A. J.; Show, Y.; Swain, G. M. Joumal of Physical
Chemistry B 2004, 108, 15117-15127.

Enea, 0.; Riedo, B.; Dietler, G. Nano Letters 2002, 2, 241-244.

Bennett, J. A.; Show, Y.; Wang, S.; Swain, G. M. Journal of The
Electrochemical Society 2005, 152, E1 84-192.

Honda, K.; Noda, T.; Yoshimura, M.; Nakagawa, K.; Fujishima, A. Journal
of Physical ChemistryB 2004, 108, 16117-16127.

Wilson, N. R.; Clewes, S. L.; Newton, M. E.; Unwin, P. R.; Macpherson, J.
V. Journal of Physical Chemistry B 2006, 110, 5639-5646.

Scharifker, B.; Hills, G. Electrochimica Acta 1983, 28, 879-889.

Vinokur, N.; Miller, B.; Avyigal, Y.; Kalish, R. Journal of The
Electrochemical Society 1999, 146, 125-130.

Dinan, T. E.; Jou, W.-F.; Cheh, H. Y. Journal of The Electrochemical
Society 1989, 136, 3284-3287.

Lund, W.; Onshus, D. Analytica Chimica Acta 1976, 86, 109-122.
Wang, J. Stripping Analysis; VCH publishers, Inc.: Deerfield Beach, 1985.

Jagner, D.; Josefson, M.; Westeriund, S. Analytical Chemistry 1981, 53,
2144-2146.

122

CHAPTER 6

TOTAL INORGANIC As DETECTION IN REAL WATER
SAMPLES USING DPASV AND A Au-COATED DIAMOND
THIN-FILM ELECTRODE

6.1 Introduction

In Chapter 5, a sensitive and reproducible method was developed for total
inorganic arsenic determination in real water samples using differential pulse
anodic stripping voltammetry (DPASV) and a Au-coated boron-doped diamond
thin-film electrode. In the analysis, the As(lll) present in the samples was first
detected by DPASV. The As(V) present was then reduced to As(lll) by reaction
with Na2303 and this was followed by a second detection of As(lll) by DPASV. In
general, Au-coated diamond showed lower LOD and higher stability compared
with Au and Au-coated carbon electrodes." 2 However, when applying Au
electrodes and ASV for real sample analysis, many researchers have reported
decreased sensitivity for As due to metal ion interferences commonly present in
ground water (e.g., Cu(ll), Pb(ll) and Hg(l|)).3‘12 For example, Feeney and
Kounaves reported that at a 1:1 concentration ratio Pb(ll) caused a 10%
decrease, Cu(ll) caused a 65% decrease and Hg(ll) caused a 30% decrease in
the As stripping current on a Au ultramicroelectrode array.5 Many other metal
ions, such as Fe(ll),9 Zn(|l),9 T|(|) and (lll)7'13 and Sn(ll) and (IV),7'13 can also
interfere with As detection and cause decreased sensitivity. There are two

mechanisms by which these metal ions interfere: (i) competition with As during

123

the deposition step or (ii) formation of an interrnetallic complex with As.3 In order
to accurately determine trace levels of inorganic arsenic in real water samples,
metal ion interferences must be removed.

Several approaches have been used to reduce metal ion interferences.
For example, Tanaka and Sato eliminated Fe(|ll) effects by reducing Fe(l|l) to
Fe(ll) with Kl.“ Hamilton separated Cu(ll) and Sb(|ll) from As(lll) and (V) by solid
phase extraction (SPE) using a Chelex 100 column.6 The Chelex 100 cation
exchange resin not only has high affinity for Cu(ll), but also for Fe(l|) and other
heavy metal ions, as well as good selectivity for divalent over monovalent ions
(approximately 5,000 to 1).15 Chelex 100 resin is a styrene divinylbenzene
copolymer containing paired iminodiacetate ions. The iminodiacetate ions are the
chelating groups binding polyvalent metal ions. At pH < 4, the resin acts as an
anion exchanger, while at pH > 4, the resin acts as a cation exchanger.15 In this
application, to separate inorganic arsenic from the interfering metal ions at pH =
4-7, use of the cation exchange resin is ideal. This is because inorganic arsenic
exists as a neutral or an anion in solution in this pH range. As(lll) exists as A5033”
(pH = 14), HAsOaz‘ (pH = 13), H2As03' (10 < pH < 12) and H3As03 (0 < pH < 9).
As(V) exists as [AsO4(H20)12]3' (pH > 13), [HAsO4(H20)5]2' (7 < pH < 11),
[H2AsO4(H20)2]’ (3.5 < pH < 6) and H3AsO4 (pH < 3.5).16 At pH 4 to 7, most of
the interfering metal ions will exist as cations and during passage through the
resin, they will be electrostatically attracted to the ion exchange sites while
arsenic will pass through unretained.

When using ASV for water analysis, another interference that can cause

124

decreased sensitivity is organic matter in the sample matrix. Humic material is a
form of organic matter often found in soil and ground water. Humic material is
produced by the decomposition or transformation of organic matter and consists
of free and bound phenolic OH groups, quinone functionalities, N and 0 bridge
units and acidic COOH functionalities located on aromatic rings.17 Humic
substances can be divided into three types: fulvic acid, humic acid and humin.
Each type is different in terms of color, molecular weight, functional group density
and extent of polymerization. From fulvic acid to humin, the molecular weight
increases from a few hundred to several hundred thousand, and the functional
group density and extent of polymerization also increase. Deprotonation of the
acid groups imparts negative charge to humic substances. Cationic metal ions
can bind at these negatively charged sites, thus reducing their activity in solution.
These high molecular weight substances have a tendency to adsorb on electrode
surfaces, driven by chemical and/or electrostatic forces. Adsorbed humic material
fouls the electrode and reduces the response stability and reproducibility for a
target analyte.18' ‘9

Boron-doped diamond possesses beneficial properties for water quality
monitoring. These properties distinguish it from the conventionally used metal
and sp2 carbon electrodes and include: (i) low and stable background current, (ii)
wide background window, (iii) an active response for many redox analytes
without extensive surface pretreatment, (iv) stable surface microstructure and

morphology in harsh electrochemical environment and (v) weak adsorption of

polar molecules due to the nonpolar, hydrogen surface termination. The last

125

property is particular important when applying the material for water quality
analysis.“"30

In this Chapter, we report on a sensitive, reproducible and accurate
method for total inorganic arsenic analysis in real samples by DPASV using a
Au-coated boron-doped diamond thin-film electrode. In the method, As(lll)
present in the sample was first determined by DPASV. The As(V) present was
then chemically reduced to As(lll) by Na2803 and this was followed by a second
As(III) determination. The difference between the two stripping currents is related
to the As(V) concentration in solution. Interfering metal ions (e.g., Cu(ll), Hg(ll)
and Pb(ll)) were removed by passing the water sample through a cartridge
containing Chelex 100 cation exchange resin (pH 4-7). The interfering metal ions
(e.g., Cu(ll)) were retained while the neutral arsenious acid (H3As03) and anionic
arsenate ([HzAsO4(H20)2]‘) passed through unretained. The effect of humic acid
was also studied by measuring the ASV response for As(lll) as a function of its

concentration from 1-5 ppm. Finally, the practical applicability of the method was

shown for the total inorganic arsenic in two real water samples.

6.2 Results and Discussion

6.2.1 Effect of Metal Ions on As Detection

The effect of Cu(ll) on the As stripping peak current was studied using the

Au-coated diamond electrode and the results are presented in Figure 6.1A and B.

126

P
at

 

 

 
  
  
  
    
   

 

 

  

 

   

 

1:11am
,(A) ”my

0.4 . :' 1:2 M'WCUOI)
A . 1:3
é 0'3 ' 1:4
E l 1:5 As(lllyCual)
g 0.2 l
3
U

0.1 -

0.0 l l ...... Lu": ..... l

0.0 0.1 0.2 0.3 0.4 0.5
Potential (V) vs. AglAgCl

 

 

 

 

Normallzed Peak Current

11) 1:1 12 1:3 1:4 1:5
Ratio of As(llivCulii)

Figure 6.1 Effect of Cu(ll) on the As stripping peak current at a Au-coated
diamond electrode in a solution of 50 ppb As(lll) and 100 ppb Au(lll) + 1 M HCI.
(A) DPASV stripping voltammetric i-E curves. (B) Plot of the normalized As
stripping peak current as a function of the concentration ratio of As(lll) and Cu(ll)

in solution. DPASV parameters: Edep = 045 V, tdep = 120 3, pulse width = 50 ms,

127

pulse amplitude = 50 mV, step height = 4 mV, cycle period 200 ms and sampling

time at the end of the pulse = 8 ms.

Figure 6.1A shows the DPASV i-E stripping curves for As as a function of the
amount of Cu(ll) present in solution. The dotted line shows the i-E curve for the
original solution containing only 50 ppb of As(lll) and 100 ppb Au(lll) in 1 M HCI.
A peak for As oxidation to As(lll) is observed at 0.17 V. With the addition of Cu(ll)
into the solution at a As(ll|)/Cu(ll) concentration ratio of 1:1, the As stripping peak
current slightly increased and the peak potential shifted positive by 8 mV. The
increased peak current was caused by Cu(ll) as it has been reported that the
metal ion can facilitate As deposition on Hg and carbon electrodes by serving as
a deposition initiator via the formation of a Cu3As interrnetallic species.8 The Cu-
As interrnetallic species must be slightly more stable than the Au-As intermetallic
compound because a more positive potential is needed to oxidize the As. As the
concentration of Cu(ll) increased though, the As stripping peak current
decreased and the peak shifted slightly positive. A new peak also emerged at ca.
0.3 V that grew in magnitude with the Cu(ll) concentration. This new peak is
attributed to the oxidation of Cu metal (deposited on the Au particles and
diamond surface during the preconcentration step).31 The progressive
attenuation of the As stripping peak current with increasing Cu(ll) was caused by
the formation of a different stoichiometric Cu-As intennetallic compound. The
formation of this stable compound reduces the amount of As available for

stripping. The bar graph in Figure 618 shows how the As stripping peak current

128

was affected by the As(lll)/Cu(ll) solution concentration ratio. Except at a ratio of
1:1, the As peak current decreased with increasing As(lll)/Cu(ll) ratio. For
example, at As(lll)/Cu(ll) ratio of 1:3, the As peak current decreased by about
30%.

Without question, to accurately determine the inorganic arsenic
concentration in a water sample, Cu(ll) or other metal ion contaminants must be
removed. Chelex 100 is a cation exchange resin that is effective at removing
these interferences. Figure 6.2A shows a comparison of the DPASV i-E stripping
curves for As in a solution devoid of Cu(ll) and one that contained 1:3
As(lll)lCu(ll) concentration ratio but was passed through the cation exchange
resin prior to analysis. It is clear that solution passage through this resin
eliminates the interfering Cu(ll). The Cu(ll) ions are removed by electrostatic
binding with the negatively charged iminodiacetate sites on the resin. The
inorganic arsenic, which exists as neutral arsenious acid (H3A303) and anionic
arsenate ([H2AsO4(H20)2]') at pH 4-7, passed through the resin unretained. It can
be seen that there was no significant change in the stripping peak current and
peak potential with the Cu(ll) interference (compare with Figure 6.1). Recall that
a response attenuation of about 30% was seen when Cu(ll) was present at this
ratio. Figure 6.28 shows a comparison of DPASV i-E stripping curves for As in
the presence and absence of other interfering metal ions. Passing these
solutions through the cation exchange resin prior to analysis is effective at
removing these interferences when present originally at a 1:1:1z1

As(lll)/Cu(ll)/Hg(lI)/Pb(ll) concentration ratio. As was co-deposited with 100 ppb

129

of Au in all of these measurements. Important when using any solid phase for
sample preparation is the analyte recovery, which for As(lll) was determined to

be 94.8 s 2.4% (n = 5).

 

 

 

 

 

 

 

 

0.20
— 1 :3 AsICu .. (A)
I ....... A; onIy 3
0.15 -
i=1
E 0.10
2
8 0.05
0.00""; 4 - L + . - .
-0.1 0.0 0.1 0.2 0.3 0 4
Potential (V) vs. AgIAgCl
0'25 ------- 1:1 Aleu 3
l —1:1:1:1Aleu/Hgle ( )
0.20 f --- As only .A
a. , - ".
g..- 0.15 r
iii
: 0.10-
3
0 I
0.05 -
0.00» _ . . . z

 

 

-o.1 0:0 0:1 0:2 013 0.4
Potential (V) vs. AglAgCl

Figure 6.2 Comparison of the As stripping voltammetric i-E curves in solutions

with and without interfering metal ions after passage through the cation exchange

130

resin. (A) Curves for 25 ppb As(lll) + 75 ppb Cu(ll) and (B) Curves for 25 ppb
As(lll) + 25 ppb Cu(ll) and 25 ppb As(lll) + 25 ppb Cu(ll), Hg(|l), Pb(II) each.
DPASV parameters: Edep = -0.45 V, tdep = 120 s, pulse width = 50 ms, pulse
amplitude = 50 mV, step height = 4 mV, cycle period 200 ms and sampling time

at the end of the pulse = 8 ms.

6.2.2 Effect of Humic Substances on As Detection

The presence of organic matter can pose a serious threat to any
electroanalytical measurement due to molecular adsorption and electrode fouling.
A comparison of the effect of humic substance on the As stripping peak charge at
Au-coated diamond and Au foil electrodes is presented in Figure 6.3. A bar graph
of the normalized As stripping peak charge as a function of the humic acid

concentration from 0 to 5 ppm is presented.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

140.0%
§ 120.0% , IAu-Coated Diamond
a a, 100.0% , r U Planar AU
E g 30.0% - T
"E 5 60.0% - -
g 40.0% -
<23 20.0% -
0.0% A . A . . A . f _
0 1 2 3 4 5

Humic Acid Concentration (ppm)

Figure 6.3 Comparison of the effect of added humic substances (0 to 5 ppm)

on the As stripping peak charge in a solution of 50 ppb As(lll) and 100 ppb Au(III)

131

+ 1 M HCI at a Au-coated diamond and a Au foil electrode. DPASV parameters:
Edep = -0.45 V, tdep = 120 5, pulse width = 50 ms, pulse amplitude = 50 mV, step
height = 4 mV, cycle period 200 ms and sampling time at the end of the pulse = 8

ms.

Humic acid clearly had a greater fouling effect on Au foil than on Au-coated
diamond. At a concentration of 5 ppm, humic acid caused a 50% decrease in the
As stripping peak charge for Au foil but only about a 9% decrease for Au-coated
diamond. This observation indicates that the non-polar, hydrogen-terminated
diamond near the metal particles plays a role in terms of influencing molecular
adsorption on the metal by virtue of the hydrophobic chemical environment. As
discussed in the Introduction, the humic substances cause decreased electrode
sensitivity by binding with positively charged multivalent metal ions (the analytes)
and reducing their activity and/or by strong and irreversible adsorption on the
electrode surface (i.e., blocking effect). In these measurements, the solution
contained no added interfering metal ions so the reason for the response
attenuation is fouling by the humic acid. It is presumed that adsorbed humic acid
blocks the Au surface preventing the deposition of As. Based on these results,
the Au-coated diamond electrode provides a clear advantage compared with Au

foil in terms of fouling resistance and response reproducibility and stability.

6.2.3 Analysis of Inorganic As in Real Water Samples

Two contaminated water samples were analyzed for their total inorganic

132

arsenic content using the DPASV method and a Au-coated diamond electrode.
Four measurements were made with each sample using the sample preparation
and DPASV conditions described above. The results are summarized in Table
6.1 and compared with the concentrations determined by the US. Bureau of
Reclamation. For the UV plant influent water, 23.2 i 2.9 ppb of total inorganic
arsenic was found compared with the 24.2 ppb specified value; a difference of
4%. Well 119 water was found to contain 16.4 :I: 0.9 ppb of inorganic arsenic,
compared with the 16.8 ppb specified value; a difference of 2.4%. Clearly, for
both samples, the results provided by DPASV with the Au-coated diamond were

reproducible and accurate.

Table 6.1 Total Inorganic As Concentration in Real Water Samples.

 

 

UV Water (ppb) Tap 119 Water (ppb)

1 23.7 16.5

2 27.1 17.4

3 21.3 15.2

4 20.6 16.4

Average 23.2 i 2.9 16.4 i 0.9

Quoted Value 24.2 16.8
Difference% 4.0% 2.4%

 

6.3 Conclusions

The work reported on herein demonstrates that the Au-coated diamond

electrode, combined with DPASV, provides a sensitive, reproducible and

133

accurate method for determining total inorganic arsenic levels in real samples.
The properties of the diamond are key for obtaining high quality detection figures
of merit and impart a greater degree of response stability, as evidenced by the
humic acid studies. For example, a 5 ppm humic acid concentration caused a
50% decrease in the As stripping peak charge for Au foil and only a 9% decrease
for Au-coated diamond. Solid phase extraction (cation exchange) was employed
to remove interfering metal ions (e.g., Cu(ll)) prior to analysis. The As(lll)
recovery after passage through the resin was high at 94.8 1: 2.4%. For two real
water samples, the total inorganic arsenic concentration determined with the Au-
coated diamond differed from the specified value by 4%, or less. It is our
contention that this new electrode and method are practical for use in on-site
analysis of total inorganic arsenic, offering advantages over commonly used Au

and Au-coated glassy carbon electrodes.

134

6.4 References

(1)
(2)

(3)
(4)

(5)
(5)

(7)

(8)

(9)
(10)

(11)
(12)

(13)

(14)

(15)

(15)

Song, Y.; Swain, G. M. Submitted.

Hignett, G.; Wadhawan, J. D.; Lawrence, N. S.; Hung, D. Q.; Prado, C.;
Marken, F.; Compton, R. G. Electroanalysis 2004, 16, 897-903.

Dai, X.; Compton, R. G. Electroanalysis 2005, 17, 1325-1330.

Simm, A. 0.; Banks, C. E.; Compton, R. G. Analytical Chemistry 2004, 76,
5051-5055.

Feeney, R.; Kounaves, S. P. Analytical Chemistry 2000, 72, 2222-2228.

Hamilton, T. W.; Ellis, J.; Florence, T. M. Analytica Chimica Acta 1980,
119, 225-233.

Henry, F. T.; Kirch, T. 0.; Thorpe, T. M. Analytical Chemistry 1979, 51,
215-217.

Cavicchioli, A.; La-Scalea, M. A.; Gutz, I. G. R. Electroanalysis 2004, 16,
697-711.

Feeney, R.; Kounaves, S. P. Talanta 2002, 58, 23-31.

Simm, A. 0.; Banks, C. E.; Compton, ‘R. G. Electroanalysis 2005, 17,
1727-1733.

Sadana, R. S. Analytical Chemistry 1983, 55, 304-307.

Pretty, J. R.; Blubaugh, E. A.; Caruso, J. A. Analytical Chemistry 1993, 65,
3396-3403.

Jackson, K. W.; West, T. S.; Balchin, L. Analytical Chemistry 1973, 45,
267-271.

Tanaka, T.; Sato, T. Journal of Trace and Microprobe Techniques 2001,
19, 521-531.

Bio-Rad Laboratories, Chelex 100 and Chelex 20 Chelating Ion Exchange
Resin Instruction Manual, pp 1-24.

Tomilov, A. P.; Smetanin, A. V.; Chemykh, I. N.; Smimov, M. K. Russian
Journal of Electrochemistry 2001 , 37, 997-101 1 .

135

 

(17)

(18)

(19)

(20)

(21)

(22)

(23)

(24)

(25)

(25)

(27)

(28)

(29)

(30)

(31)

Stevenson, F. J. Humus Chemistry: Genesis, Composition, Reactions;
Wiley: New York, 1982.

Gattrel, M.; Kirk, D. W. Journal of The Electrochemical Society 1992, 139,
2736-2744.

Terashima, C.; Rao, T. N.; Sarada, B. V.; Tryk, D. A.; Fujishima, A.
Analytical Chemistry 2002, 74, 895-902.

Swain, G. M. Journal of The Electrochemical Society 1994, 141 , 3382-
3393.

Swain, G. M.; Ramesham, R. Analytical Chemistry 1993, 65, 345-351.

Muna, G. W.; Tasheva, N.; Swain, G. M. Environmental Science and
Technology 2004, 38, 3674-3682.

Haymond, S.; Babcock, G. T.; Swain, G. M. Electroanalysis 2003, 15, 249-
253.

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

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

Hupert, M.; Muck, A.; Wang, J.; Stotter, J.; Cvackova, 2.; Haymond, 8.;
Show, Y.; Swain, G. M. Diamond and Related Materials 2003, 12, 1940-
1949. ‘

Show, Y.; Matsukawa, T.; Ito, H.; Iwase, M.; lzumi, T. Diamond and
Related Materials 2000, 9, 337-340.

lniesta, J.; Michaud, P. A.; Panizza, M.; Cerisola, G.; Aldaz, A.;
Comninellis, C. Electrochimica Acta 2001, 46, 3573-3578.

Rodrigo, M. A.; Michaud, P. A.; Duo, l.; Panizza, M.; Cerisola, G.;
Comninellis, C. Journal of the Electrochemical Society 2001, 148, D60-
D64.

lniesta, J.; Exposito, E.; Gonzalez-Garcia, J.; Montiel, V.; Aldaz, A. Joumal
of the Electrochemical Society 2002, 149, D57-D62.

Bonfil, Y.; Brand, M.; Kirowa-Eisner, E. Analytica Chimica Acta 1999, 387,
85-95.

136

CHAPTER 7

KEY FINDINGS AND FUTURE WORK

7.1 Water Treatment

Boron-doped diamond thin-film coated on Ti (diamond/T i) was used for the

anodic degradation of atrazine. As stated, the electrode material has many

favorable attributes for this application, which include a large working potential

window, a chemically inert surface and good electrical conductivity. The key

findings from this dissertation on water treatment/decontamination are:

The diamond/Ti anodes are morphologically and microstructurally stable
under the reported polarization conditions as confirmed by SEM, Raman
spectroscopy and XRD. Residual stress, a common problem with diamond
film on Ti substrates, did not cause any pinholes or microcracks in the
diamond film under the deposition conditions used.

The diamond/T i electrode exhibited good, reproducible and stable
electrochemical activity for the electrochemical redox systems, Fe(CN)5&’4‘
and Ru(NH3)63*’2*.

The degradation of ATZ on diamond/T i was a first-order reaction as the
logarithm of the ATZ concentration changed linearly with the electrolysis
time. The reaction rate calculated from the slope of the curve was
proportional to the applied current.

The degradation of ATZ at diamond/T i anodes was reproducible from film-

to-film and stable long-term for a given electrode. From radio-labeling

137

studies, the major degradation product was found to be C02 at a high

yield of 81 .2% for a 120 min electrolysis period.

Compared with the other tests reported in the literature, the yield of 002
was substantially higher due to the dimensional stability of diamond and the large
overpotential for oxygen evolution. The results showed that diamond/T i anode is
effective for the oxidation of organic pollutants to C02.

lmportantly, though, was the fact that the calculated coulometric efficiency
(Qoxat’aZine/Qoxm‘a') was low. (0.1~0.2%) Since the conversion efficiency was
improved with added convection, we suppose that the efficiency could be further
improved by:

0 Conducting the electrolysis in a flow-through electrolysis cell so as to
improve the mass transport of ATZ to the electrode.

0 Making the electrode surface area larger (e.g., mesh electrode) so as to
increase the active surface sites and the electrode area available for

reaction.

7.2 Water Quality Monitoring

Au-coated boron-doped diamond thin-film electrode, combined with
differential pulse anodic stripping voltammetry (DPASV), was used for total
inorganic arsenic detection in water samples. The key findings in this dissertation
on water quality monitoring project are:

c The Au-coated diamond electrode provided a sensitive, reproducible and

stable response for inorganic arsenic in both standard and real samples.

138

As was codeposited (preconcentrated) with Au on the diamond surface.
When codeposited with As, Au was in the form of relatively small round
particles of small nominal size and variance (22 :l: 3 nm) and high particle
density (1010 cm'z).

Compared with commonly used electrode materials, such as Au and other
Au-coated carbon electrodes, the Au-coated diamond electrode offers
advantages in terms of the peak shape and detection figures of merit.
Sub-ppb LODs were obtained for As(lll) and As(V) using both standard
and local river water samples.

This method was shown to be practical for real sample analysis as 0.6 ppb
of As(lll) was detected in local river water with a response variance less
than 2% (RSD) for 10 consecutive measurements and a response stability
of better than 94% over a 10-h period.

The properties of the diamond are key for obtaining high quality detection
figures of merit. They render the electrode resistant to fouling as the
response attenuation caused by humic acid (up to 5 ppm) for Au-coated
diamond was smaller than that for a Au foil electrode. The method was
improved by employing solid phase extraction (SPE) for sample
preparation. Specifically, SPE (cation exchange) was used to remove
interfering metal ions (e.g., Cu(ll)). The As(lll) recovery after passage
through the resin was high at 94.8 1: 2.4%.

The method is accurate for inorganic arsenic detection in contaminated

water samples supplied by the US. Bureau of Reclamation. For two real

139

water samples, the total inorganic arsenic concentration determined with

the Au-coated diamond differed from the specified value by 4%, or less.

It is our contention that this new electrode and method are practical for
use in on-site analysis of total inorganic arsenic, offering advantages over
commonly used Au and Au-coated glassy carbon electrodes. Further research
could be conducted to design a field employable instrumental set-up for the on-
site As analysis. A portable potentiostat and sample preparation set-up would be
necessary. The Au-coated diamond deserves further study because it can have
many other applications (e.g., detection of peptide) due to the high sensitivity
provided by the nanometer size of Au particles and the high stability provided by

the inert diamond support.

140