Citation
The speciation of arsenic compounds by HPLC with electrochemical detection

Material Information

Title:
The speciation of arsenic compounds by HPLC with electrochemical detection
Creator:
Ramos, David Lee
Publication Date:
Language:
English
Physical Description:
xiii, 142 leaves : illustrations ; 29 cm

Subjects

Subjects / Keywords:
Arsenic ( lcsh )
Electrochemical analysis ( lcsh )
Arsenic ( fast )
Electrochemical analysis ( fast )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Bibliography:
Includes bibliographical references (leaves 139-142).
General Note:
Submitted in partial fulfillment of the requirements for the degree, Master of Science, Department of Chemistry
Statement of Responsibility:
by David Lee Ramos.

Record Information

Source Institution:
University of Colorado Denver
Holding Location:
Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
16855371 ( OCLC )
ocm16855371
Classification:
LD1190.L46 1985m .R35 ( lcc )

Full Text
THE SPECIATION OF ARSENIC COMPOUNDS BY HPLC
WITH ELECTROCHEMICAL DETECTION
Ly
David Lee Ramos
B.A.', University of Colorado, 1979
A thesis submitted to the
Faculty of the Graduate School of the
University of Colorado in partial fulfillment
of the requirements for the degree of
Master of Science
Department of Chemistry
1985


This thesis for the Master of Science Degree by
David Lee Ramos
has been approved for the
Department of
Chemistry
by
Robert Damrauer
Date ai ^ -feT


Ramos, David Lee (M.S., Chemistry)
The Speciation of Arsenic Compounds by HPLC
with Electrochemical Detection
Thesis directed by Associate Professor
John A. Lanning
The purpose of this work was to develop
an HPLC-EC method that would speciate arsenic com-
pounds. It is necessary that the HPLC separation
of the arsenic compounds and the electrochemical
(EC) detector used to detect the separated arsenic
compounds be compatible with one another. Prior
to the coupling of the HPLC system with the EC
detector, it was essential to determine if the
arsenic compounds could be separated and to char-
acterize the electrochemical behavior of As(III).
It was found that MMAA and PAA are separated quite
rapidly with fairly good resolution between the two
compounds. The separation of MMAA and PAA was
performed on a reversed-phase column using aceto-
nitrile as the mobile phase.
The data gathered from the reduction of
As(III) at the DME compared favorably with the
same data present in the literature. Reduction
data for the other arsenic compounds were also


iv
gathered using the DME. The reduction of As(III)
at the TMFE was nearly impossible to detect, dis-
missing any notion of its use in an EC detector.
Any attempt to reduce As(III) at the TMFE resulted
in the passivation of the Hg film surface. The
deterioration (or passivation) rate was measured
by performing successive DPP scans of the same
Cd(II) solution using a passivated Hg film. The
As(III) reduction, however, was detected at a single
Hg film on a gold-plated electrode. A flowstream
EC cell, employing an HMDE, was constructed in the
laboratory. This particular design, however,
failed to detect the reduction of Cd(II) and of
As(III) in a flowing stream. Nevertheless, the
development of an HPLC-EC method would be attractive
for the speciation of arsenic compounds.


V
ACKNOWLEDGEMENTS
I would like to thank Dr. John A. Lanning
for his assistance and guidance during ray course of
studies at the University of Colorado at Denver.
My thanks is also extended to Robert
McNelly of the Center for Environmental Studies,
University of Colorado at Denver for providing
access to the organoarsenic compounds used during
this project. I would like to also thank Dr.
Robert Meglen, Director of the Center for Environ-
mental Studies, and his staff for their courtesy
during this project.
I would like to thank Mr. James Crofter
of Electrical Engineering, University of Colorado
at Denver for-providing the brass and gold-plated
electrical connectors that were used as electrodes.
Dale Bethune of Metropolitan State College
also deserves thanks for the use of the four-chuck
lathe to help machine the flowstream EC cell.


CONTENTS
CHAPTER
1. INTRODUCTION............................. 1
1.1 Analytical Problem................. 1
1.2 Speciation Techniques.............. 2
1.2.1 HPLC-UV/VIS...................... 2
1.2.2 HPLC-atomic instruments.......... 4
1.2.3 HPLC-EC.......................... 9
1.2.4 Ion Chromatography............. 10
1.3 Research Problem..................JL1
1.4 Analytical Methods for Arsenic.....12
1.5 Arsenic Speciation Techniques......15
1.6 Research Proposal..................16
2. ELECTROCHEMISTRY.........................18
2.1 Basic Electrochemical Principles ... 18
2.2 Voltammetric Techniques............25
2.2.1 Conventions......................25
2.2.2 Instrumentation. .'............27
2.2.3 Mercury Electrodes...............28
2.2.4 Cyclic Voltammetry...............33
2.2.5 Differential Pulse Polarography..34
2.2.6 Amperometry......................38


Vll
CHAPTER
2.3 Electrochemical Detectors........f. 40
2.3.1 Detector Construction............. 40
2.3.2 Detector Operation................ 46
2.4 Electrochemical Behavior of Arsenic 5
3. EXPERIMENTAL............................. 54
3.1 Reagents............................ 54
3.2 Instrumentation..................... 55
3.3 TMFE Fabrication.................... 58
3.3.1 Wax Impregnation Technique........ 58
3.3.2 Mercury Film Techniques........... 60
3.4 Flowstream EC Cell.................. 62
4. RESULTS AND DISCUSSION................... 65
4.1 UV Spectra of Arsenic Compounds.... 65
4.2 Separation of MMAA and PAA by
HPLC-UV............................. 69
4.3 Differential Pulse Polarography
of As (III)........................ 79
4.3.1 As (III) Behavior at the DME...... 82
4.3.2 As(III) Behavior at the TMFE...... 94
4.3.3 As (III) Behavior at the HMDE...... 118
4.4 Flowstream EC Cell Design and
Performance.........................126
4.4.1 Flowstream EC Cell Design.........126
4.4.2 Flowstream EC Cell Performance...128


viii
CHAPTER
5. CONCLUSIONS.............................131
REFERENCES.....................................139


LIST OF TABLES
Table
1. Typical Detection Limits (in mg/L) for
AA, GFAA, ICP, and Hydride Systems........ 6.
2. Retention Times of MMAA and PAA in
Various Acetonitrile-Water Mobile
Phases.....................................7^
3. pK_ Values of Some Arsenic Compounds.......77
3.
The Reduction of As(III) in Various
Supporting Electrolyte Solutions...........86
5. Comparison of the Quality of Some
Hg-Film WIG Electrodes.....................97
6. Comparison of the Quality of Some
Hg-Film Brass Electrodes and Some
Hg-Film Gold-Plated Electrodes............10?


LIST OP FIGURES
Figure
1. Cross-sectional view of the all-Teflon
"well sampler" interface between the HPLC
and the GFAA............................. 8
2. A diagram illustrating the five arsenic
compounds to be used in this study....... 13
3. A schematic diagram of a three-electrode
cell...................................... 20
An illustration of the current vs
potential axes as set by convention. 26
5* An illustration of a hanging Hg drop
electrode.................'.............. 3
6. A diagram illustrating a typical cyclic
voltammogram.............................. 35
7. An illustration of the excitation
potential waveform for differential
pulse polarography........................ 37
8. An illustration of the output waveform
produced by differential pulse
polarography.............................. 37
9. A dc polarogram (i-E curve)............... 39
10. A diagram of the thin-layer cell design
used for amperometric EC detection.........^3
11. An illustration of a tubular-electrode
based amperometric EC detector............ ^5
12. Differential pulse polarogram of As(III)
in 1M HC1................................. 51
13. Oxygen Scrubber Unit used to remove
last traces of oxygen from nitrogen gas.. 57


xi
Figure
1^. Illustration of the flowstream EC cell
used for this study employing an HMDE
indicating electrode........................... 63
15. UV spectra of deionized water, arsenate,
and arsenite.............................. 67
16. UV spectra of PAA, MMAA, and DMAA......... 68
17. Individual chromatograms of MMAA and
PAA..................................... 71
18. Separation of Arsenic Cbmpounds........... 71
19. The separation of MMAA and PAA in
various acetonitrile-water mobile phases. 73
20. A graphical determination of the
detection limit for PAA as performed by
, HPLC-UV.................-................80
21. Chromatogram of 65 ppm of PAA using
HPLC-UV.................................. 81
22. The reduction of As(III)- in various
aqueous supporting electrolyte solutions. 8^
23. The reduction of As(III) in various
aqueous supporting electrolyte solutions. 85
2^. The reduction of As(V) and PAA in
1M HOAc.................................. 90
25. The reduction of MMAA and DMAA in
1M HOAc.........'........................ 9I
26. The reduction of PAA in various
aqueous supporting electrolyte solutions. 93
27. A graphical determination of the
detection limit for PAA as performed by
DPP at the DME............................ 95
28. The first appearance of the cathodic
background begins at a more negative
potential as the Hg plating time increase 98


Xll
Figure
29. The level of background current
generated at the TMFGE surface decreases
as the Hg plating time increases...........98
30. The rate of deterioration observed at
the TMFGE.................................101
31. The rate of deterioration resulting
from storage for a TMFGE..................101
32. The reduction peak for Cd(II) at the
TMFGE.....................................103
33. An attempted reduction of As(III) at
the TMFGE.................................103
3^. The rate of deterioration observed at
the TMFGE as followed by consecutive
DPP scans of Cd(II).......................105
35. The rate of deterioration for a Hg
film on a brass electrode and gold-
plated electrode...............................109
36. The rate of deterioration for a Hg
film on a brass electrode and gold-
plated electrode resulting from storage..IO9
37. The reduction peak for Cd(II) at the
Hg-film brass electrode...................112
38. An attempted reduction of As(III) at
the Hg-film brass electrode...............112
39* The rate of deterioration observed at
the Hg-film brass electrode as 'followed'
by consecutive DPP scans of Cd(II).........n3
^0. An attempted reduction of As(III) at
the first Hg-film gold-plated electrode..115
Al. The reduction peak for As(III) at the
second Hg-film gold-plated electrode.....11.3
k-Z. The rate of-deterioration of the As (III)
reduction peak at the second Hg-film
gold-plated electrode....................117


xiii
Figure
^3. The reduction peak for As(III) in
1M HC1 using the HMDE....................120
A graphical determination of the
detection limit for As(III) by DPP
using the HMDE...........................122
^5. The reduction of As(III) using non-
aqueous supporting electrolyte solutions.12^
k6. The reduction of As(III) using non-
aqueous supoorting electrolyte solutions.125
^7. A dc polarographic scan of Cd(II) in
a flowing stream....................-.....
130


CHAPTER 1
INTRODUCTION
1.1 Analytical Problem
The presence of heavy metals in the environ-
ment has been an issue of concern in recent years.
Though heavy metals are analyzed for on a routine
basis, a lack of comprehensive data on any given
metal exists. The absence of such data was
demonstrated by the numerous incidents with lead and
mercury, and has made the risk assessment process -
unreliable.
The various chemical forms a metal assume
(speciation) has become a subject of keen analytical
interest, particularly the mechanisms that lead to
speciation. Speciation may be based on the free
metal, various cationic and/or anionic oxidation
states, and/or organometals. Speciation studies,
however, have been slowed by difficulties with data
interpretation for complex samples, and by the lack


2
of analytical techniques. Nevertheless, the ability
to differentiate between metal species would provide
for a better understanding of metal behavior.
1.2 Speciation Techniques
Chromatography appears to be an excellent
analytical technique to perform metal speciation. It
provides for the separation of each species followed
by their detection. It is possible to perform
speciation studies by gas chromatography (GC), if
the sample is volatile. Since most metal species are
not volatile by nature, a derivatization reaction
must be performed to convert them into closely
related volatile compounds to be suitable for GC
analysis. Liquid chromatography, namely high-
pressure liquid chromatography (HPLC), appears to be
more suitable for speciation work. Most of the
existing methods for metal speciation and or-
ganometallic separations utilize HPLC.
1.2.1 HPLC-UV/VIS.
The use of UV-VIS spectrophotometric detec-
tors was quite common when HPLC instruments became
available on a commercial basis. The main disad-


3
vantage of UV-VIS detection suffers from is that a
chromophore must exist in the species of interest.
,In each case to be cited, the metal complexes that
were separated and detected met this requirement.
One class of metal complexes separated was
that of B-ketoamines (1-3). Gaetani, et al. (1)
showed that the Ni(II) and Cu(II) complexes could be
detected at the sub-nanogram (ng) level after
separation on a reversed-phase column. Uden, et al.
(2,3), working independently, used a silica column
to separate the Ni(II) and Cu(II) complexes. Another
class of metal complexes separated were that of
1,10-phenanthroline (4,5). The inert complexes of
Ni(II), Cu(II), and Ru(II) were detected after their
separation on a reversed-phase column, but the
labile Co(II), Zn(II), and Cd(II) complexes were not
detected. The detection of the Cu(I) complex was
found to be questionable due to resolution problems
(4). The work on the 1,10-phenanthroline complexes
was continued, but utilizing a strong cation ex-
change column to separate the inert and labile
complexes together (5).
Metal complexes of dithiocarbamates is
another class separated by HPLC and followed by UV-
VIS detection (6). Bond and Wallace (6) reported the


separation of the Ni(II) and Cu(II) complexes using
a reversed-phase column followed by UV-VIS detection
in conjunction with electrochemical (EC) detection.
Cassidy, et al. (7) used HPLC with UV-VIS to deter-
mine metal complexes of 4-(2-pyridylazo)resorcinol.
They reached detection limits on the parts-per-
billion (ppb) to parts-per-trillion (ppt) level.
1.2.2 HPLC-atomic instruments.
The use of atomic spectroscopy as a detector
for HPLC has several advantages over UV-VIS since it
is element specific. However, it has a formidable '
disadvantage its poor compatibility with HPLC. The
nebulizer flow rate of a burner is generally 2-6
mL/min; HPLC columns rarely run at this rate. To
overcome this interfacing problem, the liquid from
the column is collected in 100 jaL drops into a
Teflon cup attached to the nebulizer. Van Loon (8)
discusses, in detail, how HPLC systems are inter-
faced with atomic instrumentation.
The atomic instrumentation that have been
interfaced with HPLC are graphite furnace atomic
absorption (GFAA), hydride generation atomic absorp-
tion (HGAA), and inductively coupled plasma (ICP)
emission spectroscopy. GFAA utilizes a graphite


5
furnace in place of the burner assembly as found in
flame AA. HGAA is a technique that converts the
analyte into its hydride form prior to analysis by
AA. ICP is an atomic emission method utilizing an
argon plasma maintained by the interaction of a
radiofrequency field and an ionized argon gas. GFAA
and HGAA methods are usually more sensitive than ICP
as evidenced in Table 1 (9,10).
Krull, et al. (11) conducted a comprehensive
evaluation on metal speciation with the focus of
attention on the separation process. In Krull's
study, the separation was performed on a reversed-
phase column via ion-pairing mode. Detection was
done by ICP and refractive index.
Manahan, et al. (12) evaluated the use of
ICP as a detector for HPLC peaks containing specific
elements. They directly coupled- the ICP to the HPLC
column outlet by using a 3-cm length of narrow bore
Teflon flexible capillary tubing. The detection
limit obtained was on the ng level. Morita, et al.
(13) also performed a metal speciation study by
HPLC-ICP. This method also utilized a direct cou-
pling between the HPLC and the ICP via Teflon
tubing. The ICP detector was compared to the flame
AA and direct current plasma emission spectroscopy,


TABLE 1 TYPICAL DETECTION LIMITS (in mg/L) FOR HYDRIDE SYSTEMS AA, GFAA, ICP, AND
element AA GFAA a ICP Hydride System^
! aluminum 0.03 0.00001 0.02
antimony 0.03 0.00015 0.0001
arsenic 0.14 0.0002 0.05 0.00002
barium 0.008 0.00004 0.0005
bismuth 0.02 0.0001
boron 0.7 0.015 0.004
cadmium 0.0005 0.000003 0.004
chromium 0.002 0.00001 0.005
lead 0.01 0.00005
mercury 0.17 0.002 0.000001
a detection limits given are measured using 100 pL of solution.
detection limits given are measured using 5 pL of solution.
'Table 1 reproduced from reference 10. -


7
and they found that ICP was more sensitive than the
other two detectors. In both of these cases cited,
the nebulizer flow rate was adjusted to the HPLC
flow rate.
The use of HPLC-GFAA can be cited (14,15).
Brinckman, et al. (14) studied the speciation of
trace organometallic compounds by this process.
Figure 1 is an illustration of the HPLC-GFAA connec
tion. The detection of the compounds studied was on
the ng level. Vickery, et al.(15) used GFAA to
speciate organotin and organolead compounds. They
employed a postcolumn reaction to remove matrix
effects prior to detection. They also achieved
detection limits on the ng level.
Though not an example of HPLC-HGAA, the
speciation of Se compounds by column chromatography
HGAA demonstrates the excellent detection limits
provided by HGAA. Roden and Tallman (16) separated
the various Se compounds on XAD-8 resin. The detec-
tion limits they obtained for each Se compound was
on the sub-ppb level.


8
Figure 1. Gross-sectional view of the all-Teflon
"well sampler" interface between the
HPLC and the GFAA. The effluent from
the LC column continuously enters through
the bottom via the zero dead-volume port
fitting (A) into a 5 pL sample well, (B).
Excess liquid is drawn by suction through
outlet (C). From reference 14.


9
1.2.3 HPLC-EC.
In recent years, electrochemical (EC) detec-
tion for HPLC has been on the rise. As a speciation
technique, HPLC-EC appears to be a very attractive
and a very powerful method. Selectivity, sen-
sitivity, and economy are offered as advantages for
EC detection. EC detection, however, is restricted
to species (or their derivatives) that are
electroactive. Kissinger (17) outlined the basic
requirements for EC detection and for the design of
EC detectors. Ion-exchange and reversed-phase sup-
ports are preferred for EC detection because these
supports utilize polar solvents. Polar solvents have
the capacity to contain ions whereas nonpolar sol-
vents do not. Ionic strength, pH, and the
electrochemical behavior of the mobile phase are
important considerations as is the presence of the
electroactive species.
The basic requirements for an EC detector
are that it must be rugged in construction, provide
long term stability, and provide minimal dead
volume. Coulometric-based and amperometric-based EC
detectors are available for use. The amperometric EC
detectors are more sensitive, more efficient, and
less complex in design than their coulometric coun-


10
terparts. Because of these advantages, amperometric
EC detectors have found greater use (17). The
specifics and the principles employed.for EC detec-
tion will be discussed in the next chapter.
EC detectors have been used for metal
speciation and organometallic separations (6,18-21).
Bond and Wallace (6,18,19) have used HPLC-EC to
separate and detect metal-dithiocarbamate complexes.
The metals under study were Ni(II), Cr(III), Cr(VI),
and Cu(II). Detection limits on the ng to sub-ng
level were obtained. MacCrehan (20,21) has con-
centrated his speciation studies to various or-
ganotin and organomercury compounds. Detection
limits on the ng to sub-ng level were also obtained.
1.2.4 Ion Chromatography.
Ion Chromatography (IC) has become a very
prominent technique for the analysis of inorganic
ions. Naturally, IC has found its way into specia-
tion studies. The conductivity detector is commonly
used for IC, but UV-VIS, HGAA, and coulometry have
been used for speciation studies (22-26). Hansen, et
al. (22) determined metal species in environmental
samples using conductivity detection. Detection
limits on the |UM level were achieved. Sevenich and


11
Fritz (23) studied the effect of complexing agents
on the separation of polyvalent metal ions. They
found that the addition of a complexation agent may
improve selectivity and sensitivity.
Ricci, et al. (24) coupled HGAA with IC in a
metal speciation study. Detection limits of less
than 10 ppb were obtained in this study. Girard (25)
used coulometric detection after IC separations to
analyze for heavy metals, alkaline and rare earths,
and numerous inorganic anions. Williams (26) used
UV-VIS with IC to separate and detect various metal-
lic and nonmetallic anions.
1.3 Research Problem
As previously noted, a lack of comprehensive
data on any given metal exists. Arsenic and its
compounds fit into this category. Arsenic is a
semiconducting element that is not extremely abun-
dant. Arsenic occurs naturally as sulfide ores, and
has been used to harden lead alloys. It is widely
known that arsenic and its compounds are extremely
poisonous when consumed in high dosages (27).
Arsenic compounds are routinely associated
with oil shale and the resulting oil. The presence


12
of any arsenic compound in shale oil poisons the
catalysts used in the refining processes. However,
the effect from the release of arsenic compounds
into the environment is unknown, but it is con-
sidered to be a serious health and ecological
hazard. The various forms of arsenic found in the
environment are from their use as pesticides, espe-
cially the organoarsenic forms.
The arsenic forms that are the most
predominant in the environment's aqueous systems are
inorganic arsenite (As(III)) and arsenate (As(V)),
dimethylarsinic acid (DMAA), and monomethylarsonic
acid (MMAA). (Figure 2 is an illustration of the
structures of these compounds.) The distribution
among these various forms depend upon oxidation-
reduction reactions and biomethylation reactions.
The ability to speciate among these As forms would
yield a greater understanding of the toxicity,
carcinogenicity, mobility, and bioavailability of As
(13,28-31).
1.4 Analytical Methods for Arsenic
Arsenic has been determined on a routine
basis by GFAA, HGAA, and ICP (31-35). These methods


0 0 II U HoC-As-OH HoC-As-OH 3 | 3 1 OH CH3 monomethylarsonic acid dimethylarsinic acid MMAA DMAA 0 £> fs-0H OH phenylarsonic acid PAA
O^As-O" arsenite As(III) 0 n O-As-O i 0- arsenate As (V)
Figure 2. A diagram illustrating the five arsenic compounds to he
used in this study.


measure the total level of As present and not
specific As compounds present. Table 1 indicated
that As is best detected when converted into the
hydride form prior to AA analysis. These methods
obtained detection limits on the ppb level. Inor-
ganic arsenate may be determined by colorimetric
means (36). Inorganic arsenite is determined by
difference using this method provided the arsenite
form is oxidized to the arsenate form prior to
analysis. Colorimetry, however, tends to be rather
insensitive.
/
Arsenic has been determined by electrochemi-
cal means also. Only As(III) is electroactive, but
As(V) may be measured by difference if the As(V) is
reduced prior to analysis. Some electrochemical
methods employed to determine As(III) include dif-
ferential pulse polarography (DPP) (28, 36-40),
anodic stripping voltammetry (ASV), (41-43) ,
cathodic stripping voltammetry (CSV) (44,45), and
potentiometric stripping analysis (46). For all of
these methods, it has been found that As(III) is
best determined in highly acidic aqueous solutions.
The detection limits obtained by these various
methods range from the ppm to ppb level. Two
electroactive organoarsenic forms, namely DMAA and


15
MMAA, have also been determined in nonaqueous media
(CH-jOH and CH^CN) at the ppm level. The
electrochemistry of arsenic and the electrochemical
methods cited above will be discussed in the next
chapter.
1.5 Arsenic Speciation Techniques
Arsenic speciation has been performed
chromatographically with detection by various means.
These methods include GC (30,47,48), column
chromatography-GFAA (29), HPLC-ICP (12,13), IC (22),
IC-HGAA (24), IC-ICP (49), and column
chromatography-EC (38,39). These methods have detec-
tion limits ranging from the pg to sub-ng level. A
nonchromatographic method to speciate As forms with
a ppb level detection limit also exists (50). The
nonchromatographic method converts the As forms into
their respective arsines. The arsines are trapped in
a frozen U-tube which is eventually warmed. As each
arsine is volatilized, it is detected by dc arc
emission spectroscopy.
The column chromatography-EC methods cited
above are noncontinuous methods, i.e., the EC detec-
tor is not directly connected to the outlet of the


16
column. Rather, the fraction containing the As form
is collected then transferred to an electrochemical
cell for analysis by DPP. These methods clearly
demonstrate HPLC-EC may be a possible As speciation
technique.
1.6 Research Proposal
Presently, there are no arsenic speciation
techniques utilizing HPLC-EC. Speciation of arsenic
compounds by HPLC-EC could serve as an alternative
to existing techniques. As noted previously, the
column chromatography-EC methods used to speciate
arsenic compounds (38,39) demonstrate that a con-
tinuous HPLC-EC technique is a possible speciation
method for arsenic. HPLC-EC will be investigated as
a speciation technique for arsenic compounds.
An amperometric EC detector operating in the
differential pulse mode will be used in this study.
Differential pulse techniques have been used to
determine As(III). Detection limits on the ppb to
sub-ppb level have been obtained for these tech-
niques. The detection at this level is desirable
because arsenic levels in many environmental samples
are often at the ppb level (28,36-40). MacCrehan


17
(20,21) has used differential pulse techniques for
amperometric EC detectors with success.
In this project, an HPLC method will be
developed to separate the various arsenic compounds.
A review of the literature indicates there is a lack
of separation techniques utilizing HPLC for arsenic
compounds. The development of an HPLC method to
separate arsenic compounds will provide additional
chromatographic data about these compounds. It will
also provide another technique to separate arsenic
compounds.
For EC detection to be successful, addi-
tional information about the electrochemical be-
havior of the As(III)/As(0) couple must be gathered.
The behavior of this couple with respect to thin
mercury film electrodes (TMFE) is virtually unknown.
It must be determined if the As(III) reduction may
be detected by TMFE's. If TMFE1s are unable to
detect As(III) reduction, it must be determined what
type of electrode could be used to detect this
reduction.


CHAPTER 2
ELECTROCHEMISTRY
The success of this project depends .greatly
upon the ability to detect the various arsenic
compounds electrochemically. This chapter is
designed to better understand the various aspects of
electrochemistry as applied to this project. The
topics to be addressed in this chapter include basic
electrochemical principles, voltammetric techniques,
electrochemical (EC) detectors for HPLC, and arsenic
electrochemistry.
2.1 Basic Electrochemical Prinicples
The major concerns in electrochemistry are
the processes and the factors that influence the
transport of charge across interfaces between chemi-
cal phases. Typically, one phase of interest is the
electrolyte and the other is the electrode or
another solution. The electrolyte is a medium


19
through which charge transfer can take place by the
movement of ions, and the electrode is the interface
where charge transfer changes between the movement
of electrons and the movement of ions. To study
charge transfer, an electrochemical cell is used. An
electrochemical cell consists of at least two
electrodes, but three electrodes are commonly used
as shown in Figure 3. Electrochemical phenomena are
measured at a working electrode made usually of
inert material such as mercury, gold, or carbon. The
electrode potentials are measured against an un-
polarized reference electrode. The third electrode
is an auxiliary electrode that acts as a sink or
source for electrons so that current may pass
through the cell (51).
Twp types of processes may occur at
electrodes. One process is where charge is trans-
ferred across the electrode-solution interface
resulting in oxidation or reduction. Because these
reactions must follow Faraday's law (i.e., the
amount of chemical reaction caused by the flow of
current is proportional to the amount of electricity
passed), they are known as faradaic processes. If
charge transfer is not present, the application of a
potential to an electrode is able to attract ions of


20
Figure 3- A schematic diagram of a three-electrode
cell. WE is the working electrode, RE
is the reference electrode, and AE is
the auxiliary (or counter) electrode.


21
one charge sign. If such a migration occurs/ a
corresponding amount of electronic charge will move
in the electronic circuit (into or out of the
electrode) without a charge transfer across the
electrode-solution interface. The current flow
during the charging process is known as non-faradaic
or charging current. The effects of non-faradaic
processes must be accounted for when using
electrochemical data to evaluate the charge transfer
and associated reactions.
Kinetic factors greatly influence the
faradaic process. If the reaction 0 + ne-" <--> R
(where 0 is the oxidized species and R is the
reduced species) is controlled by the rates of
processes such as mass transfer, electron transfer
at the electrode surface, chemical reactions
(homogeneous or heterogeneous) before or after the
electron transfer, and other surface reactions
(adsorption, desorption, or electrodeposition). The
simplest reactions may involve mass transfer only
whereas the more complex reactions may involve
several steps until a steady-state current is
achieved. The magnitude of current obtained in the
latter case is limited by the rate-determining
step(s).


22
The factors outlined above determines if the
reaction is electrochemically reversible or irre-
versible. If the rate of charge transfer is very
rapid in which the oxidized and reduced species are
at equilibrium and follows the Nernst equation, the
reaction is said to be reversible. An irreversible
electrochemical reaction occurs when there is a slow
charge transfer between the working electrode and
the reduction-oxidation species. The ability to
detect reversibility or irreversibility depends upon
the electrode reaction rates and the response of the
electrochemical instrumentation. For some systems,
the behavior may be reversible in one experiment and
irreversible in another, provided that the ex-
perimental conditions differ greatly.
Of the kinetic factors mentioned above, mass
transfer (or transport) is the most significant
because of its role in electrochemical dynamics.
Simply defined, mass transfer is the movement of
material from one location in solution to another
location. Mass transfer may be accomplished by
migration (i.e., a gradient of electrical
potential), diffusion (i.e., a gradient of chemical
potential), or by convection (i.e.,
hydrodynamic transport).
stirring or


23
Migration is an undesirable means of mass
transfer since it relies upon an electric field to
cause a movement of'ions. A number of factors, such
as charge, size, and the applied potential at the
working electrode, influence the rate of migration.
Another undesirable feature of migration is that the
analyte must also carry the largest fraction of
current as it undergoes simultaneous electrochemical
reaction (oxidation or reduction). To minimize
migratory mass transfer, a supporting electrolyte
solution is added. A supporting electrolyte solution
provides a large excess of inert particles of the
same charge and composition to reduce the electros-
tatic attraction of the analyte to the charged
electrode, and to reduce the fraction of current
carried by the analyte through solution to a negli-
gible value.
The selection of a supporting electrolyte
should be based on four main considerations. The
analyte should be electroactive in the supporting
electrolyte solution, and the current generated by
the supporting electrolyte solution should not
interfere substantially with the current produced by
the analyte undergoing electrochemical reaction. The
pH and the ionic strength of the supporting


electrolyte solution are the other main considera-
tions since they may affect the sensitivity of a
given electrochemical method, as will be shown for
arsenic.
Diffusion may be defined as the transport of
analyte from a region of high ionic activity to a
region of low ionic activity. As analyte is reacted
at the electrode surface, the concentration (or the
ionic activity) of analyte is reduced. In an effort
to maintain uniform ionic activity throughout the
solution, analyte from the region of high ionic
activity is transported to the region of low ionic
activity.
Convection is another means of mass transfer
which utilizes stirring or fluid flow to deliver
analyte to the electrode surface. Convection may be
accomplished by stirring the solution, by placing
the electrode in motion (i.e., rotating the
electrode in circular fashion), or by flowing the
solution past a stationary electrode (i.e., a
tubular, conical, or screen electrode). Methods that
utilize convective mass transfer of reactants or
products may be referred to as hydrodynamic methods.


25
2.2 Voltammetric Techniques
Voltammetric techniques are a group of
electrochemical methods based on the current to
potential behavior of the working electrode. In this
section, several voltammetric techniques will be
discussed. The conventions associated with voltam-
metry, and mercury electrodes will be discussed
also.
2.2.1 Conventions.
The conventions commonly employed in
electrochemistry are from the European system. The
main thrust of the European (or Stockholm) conven-
tions is to dictate sign, i.e., the more easily a
substance is reduced, the more positive it will be
in value (9). The convention that is the most impor-
tant to voltammetry is the display of current to
potential data (i-E curve). It is required that the
applied potential at the working electrode be
plotted from right to left and increases in a posi-
tive direction. A current which corresponds to a
reduction reaction is displayed upwards and is
positive by definition. Current below the zero
current line corresponds to oxidation reactions.


26
Figure k-. An illustration of the current vs
potential axes as set by convention.
i is cathodic current, i is anodic
c ' a
current. Cathodic current is positive
in value, and anodic current is negative
in value by definition.


27
Figure 4 is a diagram of the current versus poten-
tial axes.
2.2.2 Instrumentation.
An instrument which controls the voltage
across the working electrode and the counter (or
auxiliary electrode) is known as a potentiostat. The
potentiostat adjusts this voltage in order to main-
tain the potential difference between the working
and reference electrodes (which is sensed through a
high impedance feedback loop) with a program sup-
plied by a function generator. The potentiostat may
be seen alternatively as an active element whose
function is to send through the working electrode
whatever current is necessary to achieve the desired
potential. Since the current and the potential are
functionally related, that current is unique. The
potentiostat's response (current) is the measured
quantity (51).
The potentials applied at the working
electrode are often measured and quoted against a
reference electrode. The two main requirements for a
reference electrode are that it must remain un-
polarized (i.e., it does not participate in the flow
of electrons in the electrochemical cell), and that


28
its potential remains constant regardless of solu-
tion conditions or composition. The standard
hydrogen electrode is the internationally accepted
primary reference electrode, but it is not very
convenient experimentally since it utilizes hydrogen
gas. The saturated calomel electrode (SCE) is used
much more commonly as a reference electrode. This
type of electrode has a relatively high resistance
(2000-3000 ohms) and is limited in its current
carrying capacity before exhibiting severe polariza-
tion. The SCE is also limited by temperature since
it relies upon a solubility equilibrium. The SCE
becomes unstable at temperatures above 80C (9).
2.2.3 Mercury Electrodes.
One material commonly employed in voltam-
metry as the working electrode is mercury. The
mercury electrode may be classified into three
types: the dropping mercury electrode (DME), the
hanging (or static) mercury drop electrode (HMDE),
and the thin mercury film electrode (TMFE). The DME
and the HMDE both utilize liquid mercury whereas the
TMFE utilizes a thin mercury film deposited onto an
inert electrode substrate.


29
Of the three mercury electrode types, the
DME is the most commonly used in voltammetry. Liquid
mercury is sent through a capillary that has an
-3
internal diameter of approximately 5 X 10 cm from
a reservoir at a height of 20-100 cm. (The height of
the reservoir controls the rate of mercury flow.) A
nearly spherical drop is formed as the mercury flows
through the capillary. The mercury drop grows in
size (typical diameter of a mature drop is 0.1 cm)
until its weight can no longer be supported by the
surface tension between the mercury drop and the
solution. During the drop's growth, if electrolysis
occurs, the current produced will show a time de-
pendence that reflects the expansion of the mercury
drop and the depletion effects from the
electrolysis. When a drop falls, the solution is
stirred slightly and significantly reduces the
depletion effects to insure each drop is born into a
"fresh" solution. The advantages of the DME are that
the electrode surface is renewed with each drop, and
that each drop has a small but well-defined surface
area (51).
The HMDE is similar in most respects to the
DME, but with some notable differences. Figure 5 is
an illustration of the HMDE apparatus. Liquid mer-


30
Figure 5- An illustration of a hanging Hg drop
electrode utilizing a micrometer to
"dial" out a Hg drop. From reference
51.


31
cury is sent from a small "pool" located above the
capillary bore by a micrometer. As the micrometer is
turned the plunger exerts pressure on the mercury
"pool" until a spherical drop is formed hanging at
the end of the capillary. It is essential that the
mercury in the capillary bore is not segmented, and
that the hanging mercury drop remains connected with
the mercury in the capillary bore to maintain
electrical contact. The advantages of the HMDE are
the well-defined surface area, a new electrode
surface is generated with a turn of the micrometer,
and that the sensitivity of an electrochemical
method is increased. The disadvantage of the HMDE is
maintaining a hanging mercury drop; if it should
fall during the course of an experiment, the experi-
ment must be restarted from the beginning with a new
hanging mercury drop.
The TMFE utilizes a mercury film deposited
onto the surface of an inert electrode substrate.
There are two means in which the mercury film may be
deposited. The most common method involves the
immersion of the electrode substrate into a Hg(II)
solution. A potential is then applied at the
electrode substrate that is sufficient enough to
reduce the Hg(II) to Hg(0) and deposit the Hg(0)


32
onto the surface of the electrode substrate. The
thickness of the mercury film is dependent upon the
length of the electrolysis. The other method used
involves the immersion of the electrode substrate
into a pool of liquid mercury. This film deposition
technique requires that the liquid mercury adhere to
the surface of the electrode substrate. The thick-
ness of the film depends upon the length of time the
electrode substrate stays in contact with the liquid
mercury. The electrode substrates commonly used for
TMFE's include glassy carbon, wax-impregnated
graphite, platinum, gold, and silver electrodes. One
primary advantage of TMFE's is the increase in
sensitivity. The increase in sensitivity is due to
the large surface area associated with a TMFE. The
mercury film may be removed from the surface of the
electrode substrate by applying a potential which is
sufficiently positive enough to oxidize the mercury
film (51,52) .
The common advantage among these types of
mercury electrodes is that mercury is a very inert
substance. Mercury may be used from a potential of
+0.4V to approximately -2.8V versus SCE. At poten-
tials more positive than +0.4V, the Hg(0) oxidizes
to Hg(I). At potentials more negative than -1.2V,


33
the cathodic background begins to appear (from the
use of aqueous acidic supporting electrolyte
solutions), while at -2.0V supporting electrolyte
solutions made from alkali salts the cathodic back-
ground begins to appear. More negative potentials
may be reached by use of supporting electrolyte
solutions made from quaternary ammonium salts. For
example, the use of tetra-n-butyl-ammonium hydroxide
as a supporting electrolyte can provide negative
potentials up to -2.7V (9).
2.2.4 Cyclic Voltammetry.
Cyclic voltammetry (CV) is an electrochemi-
cal method which involves the generation of current-
potential (i-E) curves under diffusion-controlled
mass transfer conditions at the working electrode
utilizing a symmetrical triangular scan. The poten-
tial at the working electrode is scanned from an
initial value to a final switching value and back to
constitute a cycle. As the potential is scanned in
the positive direction, the working electrode be-
comes a stronger oxidant whereas if the potential is
scanned in the negative direction, the working
electrode becomes a stronger reductant. Stationary
electrodes or a hanging mercury drop may be used as


3^
the working electrode for CV. The triangular scan
rates may vary from a few millivolts per second to a
few hundred volts per second. By monitoring the
current at the working electrode, a cyclic voltam-
mogram is generated. Figure 6 is an illustration of
a cyclic voltammogram (9).
CV is a particularly useful method since it
may provide information about electrochemical sys-
tems. It may provide information about reaction
reversibilities. The investigation of stepwise
reactions and reactive intermediates can be made
possible by using CV. Electrochemical systems that
exhibit a wide range of rate constants may be
studied by varying the scan rate. Redox potentials
of metal ions, or of compounds, may be easily and
rapidly located with CV. Electrode product stability
may be studied by CV before using techniques such as
stripping analysis (anodic and cathodic), or before
using electrochemical detection for HPLC (9).
2.2.5 Differential Pulse Polarography.
Differential pulse polarography (DPP) is a
voltammetric technique in which a linearly increas-
ing dc ramp is applied to the electrochemical cell
and a fixed pulse height is superimposed on the


Figure 6. A diagram illustrating a typical cyclic
volt ammogram.


36
ramp. The current flow is then sampled before the
application of the pulse and at the end of the
pulse. (Figure 7 is an illustration of the DPP
process.) The readout is then displayed as the
difference between the two currents. The DPP signal
produced closely resembles the derivative of a
polarographic wave, and yields a peak as shown in
Figure 8 (53).
The magnitude of the DPP signal strongly
depends upon the pulse amplitude applied and the
overall system gain. Typically, the pulse amplitude
applied should be less than the the overall width of
the rising portion of the step function. So when
large DPP signals are required, the pulse amplitude
on the order of the overall width of the wave should
be applied. A reduction in the pulse amplitude will
permit the resolution of the fine detail of the
wave, if present.
In DPP, the applied potential does not vary
slowly, but changes rather suddenly at the initia-
tion and at the end of the pulse. This sudden change
in the electrode environment at these points leads
to the establishment of new concentration gradients.
This results in an increase of sensitivity for this
method. Detection limits for DPP range from the ppm


37
Eapp.
INITIAL
POTENTIAL
Figure ?.
Figure 8.
An illustration of the excitation potential
waveform for differential pulse polarog-
raphy. From reference 53.
An illustration of the output waveform
produced by differential pulse polarography.
From reference 53.


38
to the ppb level. Another advantage DPP offers is
that more than one species may be determined simul-
taneously, provided the redox potentials for each
species do not overlap. As an added advantage of
DPP, a variety of working electrodes (from a DME to
a solid electrode) may be used.
2.2.6 Amperometry.
Amperometry is a voltammetric technique in
which the potential at the working electrode is
fixed with respect to the reference electrode. The
current passing through the cell is then measured.
Generally, the current measured is in the diffusion
current region of a current-potential curve. (Figure
9 is an illustration of a current-potential curve.)
In this region, the current is independent of the
applied potential at the working electrode due to
the presence of an extreme state of concentration
polarization. Since the concentration of the analyte
reacting at the surface of the working electrode is
sustained at a near-zero value, the current is
limited by the rate of diffusion. Thus, the rate of
diffusion (or current) is proportional to the con-
centration of analyte present in the bulk of the
solution (9).


39
limiting
current
plateau
E, V vs. SCE
Figure 9- A dc polarogram (i-E curve). Ideally,
for amperometry, the current is measured
in the limiting current plateau. This
particular dc pclarogram is of O.lmM
chromate in 0.1M NaOH. From reference 51.


40
2.3
ns
tio
pre
an
det
tioi
wil
sin
The
be
2.3
Electrochemical Detectors
Column selection and 'mobile phase considera
for HPLC-EC were discussed briefly in the
fious chapter as were the basic requirements for
lectrochemical (EC) detector. In this section,
iled discussion on the construction and opera-
of EC detectors will follow. The discussion
center itself on amperometric EC detectors
interface requirements for an EC detector will
iscussed also.
is
rec
mus
sta
two
us
whi
tha
broa
ed
1 Detector Construction.
The specific design of the EC detector cell
elatively unimportant, if it meets three basic
irements. These basic requirements are that it
be rugged in construction, have long term
ility, and have minimal dead volume. The first
requirements relate to the materials that are
for the working electrode in the EC detector
e the third requirement relates to a feature
all HPLC detectors must meet to prevent
dening of chromatographic peaks.


kz
use of such a bed of conducting particles are quite
unsatisfactory since these beds are suspectible to
flow disturbances (from the HPLC flow) and which may
change the response of the detector. This is the
main reason why coulometric EC detectors are
employed less often than their amperometric-based
counterparts. However, coulometric EC detectors have
been used for HPLC.
The amperometric-based EC detector is much
less complex in design than its coulometric counter-
part. An amperometric EC detector may employ either
a tubular electrode or a thin-layer cell. The thin-
layer cell is the more popular of the two am-
perometric EC detector configurations since low dead
volumes (less than 1 pL) are used and a greater
variety of materials may be used for the working
electrode. In the thin-layer cell configuration, a
very small volume of solution is "trapped" as a thin
layer (approximately 20-100 pm thick) against the
working electrode and the electrochemical measure-
ment made. Figure 10 is an illustration of an am-
perometric detector utilizing the thin-layer cell
configuration.
Amperometric EC detectors utilizing the
thin-layer cell design are constructed so that the


^3
Figure 10. A diagram of the thin-layer cell design
used for amperometric EC detection. W
is working electrode. From reference
17.


fluid flow is strictly parallel to the working
electrode imbedded in a rectangular channel, or the
fluid flow is directed to be perpendicular to the
surface of the working electrode followed by radial
dispersion. From the thin-layer cell design shown in
Figure 10, the channel wall forms part of the sur-
face of the working electrode. The wall is formed by
sandwiching a fluorocarbon gasket (50-125 pm thick)
between two machined blocks of Plexiglas, Kel-F, or
similar industrial plastic. A variety of materials
(such as platinum, gold, and glassy carbon) have
been used as the working electrode in the thin-layer
cell design.
Probably due to the great success of the
thin-layer cell configuration, tubular electrodes
have been used less often for amperometric EC detec-
tors. Figure 11 is an illustration of an am-
perometric detector utilizing a tubular electrode.
The fluid flow is directed through a bore that is
-drilled through the center of the working electrode
material. It is in the interior surface of the
tubular electrode where the electrochemical measure-
ment is made. Graphite and glassy carbon have been
commonly used as the material for the tubular
electrode. The main disadvantage of using tubular


^5
Figure 11.
An illustration of a tubular-electrode
based amperometric EC detector.
Courtesy of John A. Lanning.


electrodes for amperometric EC detectors is the
renewal of the interior surface is very difficult -
it must be assumed since there are no means for it
to be inspected thoroughly. The main advantage of
tubular electrodes is if the working electrode may
be easily replaced in the amperometric EC cell, thus
it becomes an attractive alternative to the thin-
layer cell configuration.
2.3.2 Detector Operation.
Regardless of the type of EC detector used,
both the amperometric and coulometric EC detectors
measure the current generated by the eluted
electroactive species undergoing reduction or oxida-
tion. Coulometric EC detectors are required to be
100% efficient (this is by definition) whereas the
amperometric EC detector at best can be 1-10% effi-
cient. However, amperometric EC detectors are more
sensitive than their coulometric counterparts. The
rhetorical question must be asked at this point -
how can the amperometric EC detector be more sensi-
tive when it converts (i.e., oxidize or reduce) less
material? It is the geometry of the EC detector cell
that leads to the increased sensitivity. The
geometry of the EC detector cell is required to


k?
achieve complete conversion in a flowing stream. The
efficiency of the amperometric EC detector is im-
proved by providing more surface area for the work-
ing electrode downstream; but with every new incre-
ment of surface area for the working electrode, a
lesser amount of material is converted. This leads
to detection at or near background current levels.
The amperometric EC detector may be operated
by using a two-electrode arrangement or by using a
three-electrode arrangement. In the two-electrode
arrangement, the reference electrode must also serve
as the auxiliary electrode. The disadvantages of
this type of arrangement are that an uncompensated
solution resistance is present and that the
reference electrode can be easily polarized.
However, the two-electrode arrangement is very
practical for trace analysis studies since very low
currents are generated. In the three-electrode
arrangement, as shown in Figure 10, the auxiliary
electrode is placed downstream. While the location
of the auxiliary electrode in this design is con-
venient, a nonlinear response is obtained with high
concentrations. The nonlinear response may be
remedied by using a mobile phase with a high ionic
strength, or by inserting a thicker gasket. Another


14-8
remedy is to relocate the auxiliary electrode to a
position opposite of the working electrode. Relocat-
ing the auxiliary electrode to this position reduces
the level of uncompensated solution resistance to a
negligible value. This then allows for the use of
low ionic strength mobile phases and provides for a
wider linear range (17).
An EC detector may be operated either in the
oxidation mode or in the reduction mode. The oxida-
tion mode is primarily used for species of interest
that are easily oxidized. At trace levels, however,
the oxidation mode is not particularly useful due to
the presence of significant background oxidation
current of many solvents. The reduction mode must
overcome some initial problems before it becomes
useful for EC detection. These problems are the
presence of dissolved O2, trace metal ions, and H+
ions, which all may be reduced very easily. These
problems may be easily reduced or eliminated by
taking the appropriate steps. To determine which
mode is to be used for the EC detector depends upon
the electrochemical data on the species of interest
gathered by CV.
The EC detector largely dictates what may be
done, instrumentally by the HPLC. Column selection is


k-9
solely influenced by the EC detector, which in turn
influences the composition of the mobile phase.
Nonpolar stationary supports (such as silica gel)
are not well suited to be combined with EC detectors
because the nonpolar mobile phases, which are com-
patible with these stationary supports, are not
compatible with many electrochemical reactions. The
very low dielectric constants associated with non-
polar solvents make it very difficult to conduct
current. Ion-exchange and reversed-phase stationary
supports are well suited to be combined with EC
detectors because they are compatible with polar
mobile phases. Polar mobile phases are able to
conduct current generated by the electrochemical
reactions due to their high dielectric constants.
The EC detector also influences the composi-
tion of the mobile phase. Since the EC detector is
essentially an electrochemical experiment, the
mobile phase also serves as the supporting
electrolyte solution. Ionic strength, pH, and the
electrochemical reactivity of the mobile phase are
very important factors since they may influence the
sensitivity of the EC detector. However, the mobile
phase that is used should not damage the stainless
steel and the Teflon parts of the HPLC.


50
The EC detector also dictates that the
elution of the species of interest by the HPLC be
done under isocratic conditions only (i.e., the
composition/ ionic strength, and the pH of the
mobile phase remains constant). If the elution of
the species of interest by the HPLC is done under
gradient conditions (i.e., a change in one of the
parameters specified above), considerable baseline
drift and noise will occur.
\
2.4 Electrochemical Behavior of Arsenic
As noted previously, As(III) shows
electroactive behavior. Inorganic As(III) is easily
reduced to As(0) at a potential of -0.43V vs SCE,
and the As(0) may be further reduced to AsH^ at a
potential of -0.84V vs SCE in a 1M HC1 solution. The
As(0) reduction step is difficult to detect due to
its close proximity to the cathodic background and
the volatility of AsH^ Figure 12 is an illustration
of the differential pulse polarogram of inorganic
As (III) showing its- reduction to As(0) and to AsH^.
The most information about the electrochemical
behavior of arsenic comes from the As(III)/As(0)
couple. The As(III) reduction to As(0) is an irre-
versible process based on the slow electron transfer


51
Figure 12. Differential pulse polarogram of As(III)
in 1M HC1. (a) is the reduction of
As (III) to As(0), (t>) is the polarographic
maximum, and (c) is the reduction of
As(0) to AsH^. From reference 5^.


52
for this reaction and the As(0) formed is adsorbed
onto the surface of the working electrode as a film
(36,54).
The pH of the supporting electrolyte solu-
tion greatly influences the behavior of the
As(III)/As(0) couple. Myers and Osteryoung (36)
conducted an extensive DPP survey on As(III) reduc-
tion behavior in various supporting electrolyte
solutions. It was found that highly acidic aqueous
supporting electrolyte solutions provide the best
results. A 1M HCl supporting electrolyte solution
yields the very best results in terms of providing
the best sensitivity. It was also found that as the
pH of the supporting electrolyte solution is in-
creased, the As(III) reduction wave is shifted
cathodically and may overlap with the cathodic
background. The pH also influences the degree of
irreversibility for the As(III)/As(0) couple. As the
pH is increased, the As(III) reduction reaction
becomes more irreversible.
The ionic strength of the supporting
electrolyte solution is another factor that may
influence the behavior of the As(III)/As(0) couple.
Megargle, et al. (43) reported in an ASV study of
the As(III)/As(0) couple that ionic strength was not


53
an important factor with respect to variations in
electrode response with pH.
As (V) has shown electroactive behavior, but
only in an ammoniacal 0.5 M pyrogallol solution.
Since As(V) is electroactive in one supporting
electrolyte solution, it is easier and more con-
venient to measure As(V) by difference if the As(V)
is reduced to As(III) prior to analysis. Reagents
commonly employed to reduce As(V) to As(III) are
SO2, acidic KI, and hydrazine salts (55).
The organoarsenic forms that are electroac-
tive exhibit electrochemical behavior which is very
similar to the electrochemical behavior of inorganic
As(III). More negative potentials (ranging from
-0.85V to -1.7V vs SCE), though, are required to
reduce the electroactive organoarsenic forms. The pH
of the supporting electrolyte solution also in-
fluences the degree of irreversibility and the
location of the reduction wave of a given electroac-
tive organoarsenic in the same manner as it affects
inorganic As(III). The reduced organoarsenic form is
also adsorbed onto the surface of the working
electrode as a film (40).


CHAPTER 3
EXPERIMENTAL
3.1 Reagents
Reagent-grade arsenic acid (H^AsO^), sodium
arsenite (NaAs02), dimethylarsinic acid (DMAA),
disodium monomethylarsonate hexahydrate- (MMAA) and
phenylarsonic acid (PAA) were employed in this
study. Aqueous supporting electrolyte solutions were
prepared from reagent-grade hydrochloric acid,
acetic acid, phosphoric acid, oxalic acid, tartaric
acid, potassium nitrate, and potassium chloride.
Nonaqueous supporting electrolyte solutions were
prepared also. They include concentrated acetic
acid-methanol (10:90), concentrated acetic acid-
acetonitrile (10:90), 0.1M tetra-ethyl-ammonium
bromide (TEAB) in acetonitrile, and 0.05M potassium
perchlorate in water-acetonitrile (10:90). The
concentrated acetic acid, TEAB, potassium
perchlorate, methanol, and acetonitrile used for the


55
nonaqueous supporting electrolyte solutions were
reagent-grade. To verify the working electrodes
utilized in this project performed properly, 10 ppm
solutions of Cd(II) and Pb(II) were prepared from
reagent-grade cadmium chloride and from reagent-
grade lead chloride. HPLC-grade acetonitrile (Bur-
dick & Jackson), with a maximum UV cutoff of 188 nm,
was used to make up various mobile phases. Redis-
tilled liquid mercury was used for the DME and HMDE
apparatus.
3.2 Instrumentation
The UV spectra of the arsenic compounds
listed above were taken with a Cary-14 UV-VIS
Spectrophotometer. HPLC separations were performed
by a Varian 5000 Liquid Chromatograph. The
chromatography was carried out on a Varian 30 cm X
4.6 mm (i.d.) C^g reversed-phase column and a Rainin
25 cm X 4.6 mm (i.d.) C18 reversed-phase column.
Samples were introduced by a Valeo injection valve
equipped with a 10 pL sample loop. UV-VIS detection
at 198 nm was performed by a Varian UV-50 variable
wavelength detector. This detector was linked to a
Houston Instruments OmniScribe strip-chart recorder.


56
All electrochemical experiments were per-
formed with a three-electrode arrangement consisting
of an SCE reference electrode, a platinum auxiliary
electrode, and a working electrode (the DME, HMDE,
and TMFE's were used). Home-made polarographic cells
were used for experiments employing the DME ap-
paratus while commercial polarographic cells from
Princeton Applied Research (PAR) were used for
experiments employing the HMDE and TMFE apparatus. A
PAR Polarographic Analyzer Model 174 Potentiostat
was used to generate current-potential (i-E) curves.
The i-E curves were recorded with a Houston Instru-
ments 2000 X-Y Recorder. Since the potentiostat was
operated in the reduction mode, it was necessary to
deaerate all working solutions with an inert gas. N2
was used to deaerate all working solutions. Because
N2 is extracted from air, an oxygen scrubber was
constructed. Figure 13 is an illustration of the
oxygen scrubber used. The oxygen scrubber used
consists of two gas towers one tower contains a
solution of vanadous chloride (which actually
removes the oxygen from the N2) and the other tower
contains the supporting electrolyte solution being
used in the polarographic cell (or deionized water
may be used if a number of supporting electrolyte
solutions are being used). The latter option was


57
Figure 13. Oxygen Scrubber Unit used to remove
the last traces of oxygen from nitrogen
gas. From reference56.


58
used for the second tower. The oxygen scrubber unit
was required to remove the last traces of oxygen in
the N2 since very low concentrations (50 ppm and
below) of analyte were being determined (56).
3.3 TMFE Fabrication
Several methods were evaluated to determine
which method yields the best thin mercury film on an
electrode substrate. A few electrode substrates were
also evaluated to determine which one is the most
suitable to accomodate a thin mercury film. In this
section, the process used to fabricate a wax-
impregnated graphite (WIG) electrode will be out-
lined as well as the methods used to deposit the
thin mercury film on the electrode substrate.
3.3.1 Wax Impregnation Technique.
A graphite rod, with a diameter of ap-
proximately 6 mm, was cut into several 7 mm sec-
tions. These graphite sections were then immersed in
molten parrafin wax. Impregnation using the parrafin
wax (Esso, Inc.) involves the melting of the wax in
a thick-walled vial suspended in a boiling water
bath. The molten parrafin wax, containing the


59
graphite sections, was placed under vacuum with a
water aspirator to outgas the graphite sections.
During the outgassing period* the vacuum was
released and reapplied several times to accelerate
the outgassing process. When the bubbling ceased
after 30-45 minutes, this indicated that the out-
gassing and impregnation process was completed.
Following the outgassing period, the impreg-
nated electrodes were allowed to solidfy on a
watchglass. A hole was then drilled into an end of
the electrode to accomodate a copper wire for
electrical contact. The copper wire was inserted
into the hole after the wire was coated with Duro
E-POX-E Quick. The sleeve for the impregnated
electrode was machined from a 15 mm diameter Kel-F
rod. The rods were cut into 12 mm lengths and a 6 mm
hole drilled through the rod to fit the electrode.
The electrode was cemented into the Kel-F sleeve
with Epoxy. After hardening, the electrodes were
machined to insure a smooth and uniform planar
surface.
After the electrode surfaces were machined,
they were rinsed with distilled, deionized water to
remove any loose material. Using a damp Kimwipe, the
electrode surface was polished by hand in a circular


60
motion until no darkening of the Kimwipe occurred.
3.3.2 Mercury Film Techniques.
Four mercury film techniques were evaluated
to determine which technique produces the best
mercury film. Three of the four mercury film tech-
niques utilize plating solutions prepared from
reagent-grade mercuric nitrate monohydrate while the
fourth technique utilizes liquid mercury. Several
gold-plated electrical contacts and several brass
electrical contacts were evaluated as TMFE electrode
substrates as were WIG electrodes.
Two mercury film techniques for WIG
electrodes were evaluated. The first technique
employed a 5 ppm Hg(II) solution prepared from
HgCNO-j^* This solution was made slightly acidic
(approximately pH 5) with reagent-grade concentrated
nitric acid. A potential of -0.4V vs SCE was then
applied for 30 minutes at the WIG electrode to
deposit the mercury film. The mercury film was
removed by applying a potential of +0.6V vs SCE at
the WIG electrode. The second technique used a
plating solution of 0.006M HgCNO^^ in 0 5M KNO^. A
potential of -1.2V vs SCE was applied at the WIG
electrode for approximately 4 minutes to deposit the


6l
mercury film. The mercury film was removed with a
damp Kimwipe (57,58).
The two remaining mercury film techniques
were applied to the gold-plated and brass electrical
contacts serving as electrode substrates. One method
utilized a plating solution of 0.25M Hg (NO^) 2 an<^
0.5M KCl in 0.1M HC1. A potential of -0.7V vs SCE
was applied at the electrode substrate for 2 minutes
to deposit the mercury film. The electrode with the
mercury film was then rinsed with deionized water. A
solution of 0.5M KCl in O.lM HC1 was used to condi-
tion the electrode at a potential of -1.2V vs SCE
until the dull gray film becomes shiny. The other
method involves the immersion of the electrode
substrate into a pool of liquid mercury. The
electrode substrate is left in the mercury pool
until it adheres to the electrode substrate (ap-
proximately 5-10 minutes). The removal of the mer-
cury film from the electrode substrate for both
techniques was accomplished by applying a potential
of +0.4V vs SCE for 30-45 minutes (52,59).


62
3.4 Flowstream EC Cell
A Lucite rod, with a diameter of 33 mm, was
cut into a 45 mm length to fashion the flowstream EC
cell. A port of approximately 6 mm in diameter and
10 mm deep was drilled at one end of the Lucite rod
and was threaded to accomodated a male connector
made from Kel-F. The male connector was ap-
proximately 6 mm in diameter and 25 mm in length. At
the other end of the Lucite rod, a bore of ap-
proximately 6 mm in diameter and approximately 24 mm
deep was drilled. Approximately 10 mm of this bore
was threaded to accomodate a male connector of the
same dimensions mentioned above. Into opposite sides
of the Lucite rod, ports of approximately 6 mm in
diameter and 10 mm deep were drilled and threaded to
accept male connectors of 6 mm in diameter. These
ports were used to accomodate the reference
electrode and the working electrode. A hole of
approximately 6/32 of an inch was drilled into the
side of the Lucite rod to accomodate a brass screw
that would serve as an electrical contact with
tubular graphite electrode. The tubular graphite
electrode, with a 5.5 mm diameter and 6 mm in
length, served as the counter electrode in this
design. The tubular electrode had a bore size of


63
1 in
V
Kel-F male
connector
graphite counter
electrode
cur
Teflon^
spacers
Figure 1^.
Illustration of the flowstream EC cell
used for this study which employs an
HMDE indicating electrode.


6k
approximately 1.5 mm in diameter. A Teflon spacer of
approximately 5.5 mm in diameter and 8 mm in length
was inserted insure water tightness between the
graphite electrode and a male connector. Teflon
washers were used at the other port connections to
insure water tightness. The finished flowstream EC
cell is shown in Figure 14.


CHAPTER 4
RESULTS AND DISCUSSION
4.1 UV Spectra of Arsenic Compounds
It is a commonly held assumption that inor-
3-
ganic anions, such as AsO^ and ASO2 lack
suitable chromophores to be detected by UV spectros
copy. This assumption is incorrect since many inor-
3- 3-
ganic anions, including AsO^ and AsO^ have
absorption bands in the 190-220 nm region (26). In
order to investigate if NaAsC^ and H^AsO^, as well
as MMAA, DMAA, and PAA, indeed have absorption band
in the 190-220 nm region, a UV study was conducted.
Aqueous solutions of each arsenic compound with
' _4
concentrations of 1.0 X 10 M were prepared.
Deionized water was selected as the solvent of
choice since it readily dissolves the five arsenic
compounds and has a wider spectral transparency
range.


66
Figures 15 and 16 show the spectra of the
five arsenic compounds and of deionized water. The
common feature in the spectra of the five arsenic
compounds is the presence of absorption bands in the
185-230 nm region. The spectrum of H^AsO^ (in Figure
15) shows two absorption bands in the 185-190 nm
region whereas the literature indicate the two
absorption bands lie in the 195-200 nm region. The
difference in the location of the absorption bands
could be due to differences in the experimental
conditions (such as instrumentation, sample prepara-
tion, and solvent selection). In Figure 15, the
spectrum of NaAsC^ shows an absorption band at 189
nm and another at 228 nm. No comparison spectrum
could be found in the literature for NaAsC^.
The location of an absorption band at 198 nm
for MMAA and PAA (in Figure 16) corresponds closely
3_
to the absorption band located at 200 nm for AsO^
The location of this absorption band may be due to
the structural similarities between MMAA, PAA, and
AsO^-. The additional absorption bands found for
PAA at 254, 261, and 267 nm are due to the phenyl
ring. The spectrum of DMAA (in Figure 16) closely
resembles the spectrum of NaAsC>2 due the presence of
the absorption band at 189 nm. The location of this


67
Abs
Figure 15.
UV spectra of deionized water, arsenate,
and arsenite (top to bottom).


68
Figure'16. UY spectra of PAA, MMAA, and DMAA (top
to bottom).


69
particular absorption band may also be due to the
structural similarities between DMAA and AsC^-*
4.2 Separation of MMAA and PAA by HPLC-UV
With the UV spectral information, it is
possible to develop a speciation technique for
arsenic compounds using HPLC-UV. The use of HPLC-UV
as a speciation technique is limited since only MMAA
and PAA both absorb UV radiation at a common
wavelength (198 nm). The detection of these two
species at 198 nm limits the choice of solvents for
the mobile phase to acetonitrile and water. Despite
these limitations, the separation of MMAA and PAA by
HPLC-UV was performed to obtain data on their reten-
tion characteristics, and possibly obtain data on
the retention characteristics of the other arsenic
compounds.
The separation of MMAA and PAA was performed
on a C-^g reversed-phase column. The selection of a
reversed-phase column over an ion-exchange column
was based on fact that an ion-exchange column needs
to go through the same preparation and regeneration
process as used for ion-exchange column chromatog-
raphy. Additionally, the reversed-phase column was


70
selected because arsenic compounds have been pre-
viously separated using this type of column (14).
Before the separation of MMAA and PAA was
performed, MMAA and PAA both were individually
eluted to determine their respective retention
times. The mobile phase, at a flow rate of 1.0
mL/min, used for these elutions was 100%
acetonitrile. Figure 17 shows the individual
chromatograms for MMAA and PAA. MMAA and PAA both
yielded very sharp peaks with retention times of 3.6
minutes and 4.7 minutes, respectively. The separa-
tion of MMAA and PAA was then performed using the
same conditions as that used for the individual
elutions. Figure 18 is the chromatogram showing the
separation of MMAA and PAA. The retention times from
the separation for MMAA and PAA, which are 3.4
minutes and 4.4 minutes, respectively, compare very
favorably with the retention times observed for the
individual elutions. Though MMAA and PAA were
rapidly separated, both components were fairly well
resolved from one another. To determine if the
resolution between MMAA and PAA could be improved,
the flow rate was reduced from 1.0 mL/min to 0.7
mL/min at 0.1 mL/min intervals. Reducing the flow
rate did not improve the resolution between MMAA and


71
Figure 17. Individual chromatograms of (a) MMAA
and (b) PAA. Mobile phase- acetonitrile,
flow rate 1.0 mL/min. Column- C-^g
reversed-phase.
11
XO.OI AU
6
1
min
(a)
t
Jo:005 au
Figure 18. Separation of arsenic compounds -
(a) the separation of MMAA and PAA
(b) the separation of arsenate,
arsenite, MMAA, and PAA. Mobile
phase- acetonitrile, flow rate 1.0
mL/min. Column- C^g reversed phase.


72
PAA, but rather broadened both peaks. The flow rate
at 1.0 mL/min provided the best resolution between
MMAA and PAA.
The mobile phase composition was varied to
determine if it affected the retention of both MMAA
and PAA, as well as the resolution between the two
arsenic compounds. The mobile phase composition was
varied from 100% acetonitrile to 80% acetonitrile/
20% water while maintaining the flow rate at 1.0
mL/min. The retention times of MMAA and PAA, and the
resolution between the two arsenic compounds, in
various mobile phases are compared in Table 2. The
separation of MMAA and PAA in three different mobile
phases is shown in Figure 19. Table 2 and.Figure 19
both show that as the amount of water is increased,
the retention times of MMAA and PAA are increased
significantly. Apparently, the increase in retention
observed for MMAA and PAA may be attributed to their
greater affinity for the stationary phase than for
the mobile phase. Changing the mobile phase composi-
tion, however, did not improve the resolution be-
tween MMAA and PAA. The chromatograms in Figure 19
show reduced peak heights for both MMAA and PAA as
well as broadened peaks. The flow rate was adjusted
to improve peak shape, but no noticeable improvement


73
JO.01 AU
J0.01 AU
J_____I_____i_____L
10 20
min
Cb)
(c)
Figure 19. The separation of MMAA and PAA in.
(a) 9076 acetonitrile, 10$ water,
(b) 85$ acetonitrile, 15$ water, and
(c) 80$ acetonitrile, 20$ water at a
flow rate of 1.0 mL/min using a C-,0
' reversed-phase column. 10


TABLE 2 RETENTION TIMES OF ACET ONITRILE-W ATER MMAA AND PAA IN MOBILE PHASES VARIOUS
jMobile Phase T (in min) T^ (in min) Ra
Composition MMAA r PAA s
100% ch3cn 3.6 4.4 1.25
90fo CH^CN 14.2 15.8 0.64
10$ water
85$ CH^CN
J 15% water 18.7 19.5 0.68
80$ CH3CN 24.0 2^.9 0.51
20$ water


75
was observed.
Reversed-phase columns may separate com-
pounds either by their relative polarities (i.e.,
the less polar a compound is, greater is its
retention), or by their relative hydrophobicities
(i.e., the more hydrophobic a compound is, greater
is its retention). The class of compounds being
separated dictates whether the retention mechanism
is based on polarity or hydrophobicity, or by the
mobile phase used. The retention behavior exhibited
by MMAA and PAA appears to be more consistent with
the retention mechanism based on polarity. The
greater retention exhibited by PAA is somewhat
expected since the phenyl group appears to be more
effective in diffusing the polarity of compound than
the methyl group in MMAA. This implies that the
retention of the organoarsenic forms would be dic-
tated by the type, of organic group and by the number
of organic groups attached to the central arsenic
atom. This would also imply that the inorganic
3-
arsenxc forms (AsO^ and AsC^ ) would be the least
retained since they are the most polar. Based upon
these implications, an elution order for the five
arsenic compounds may be proposed. Inorganic ar-
3_
senate (AsO^ ), the most polar of the five arsenic


76
species, would elute first. Inorganic arsenite
(AsC>2_) would then rapidly follow. The next species
to elute would be MMAA, then followed by PAA and
DMAA, respectively. The proposed elution order
coincides with an increase in pK values for the
a
five arsenic compounds, except for As02. This would
imply that as the pK value increases, the retention
cl
of the arsenic compounds will become greater.
Moreover, it indicates that the pKa values of the
arsenic compounds (as shown in Table 3) may play an
important role in dictating the elution order. Using
the pK values as the primary consideration to
cl
establish an elution order, it would remain the same
except that As02_ would elute after DMAA and not
3_
after AsO^ as initially suggested. It would appear
that methods such as HPLC-GFAA or HPLC-EC may be
more appropriate to determine the elution order of
the five arsenic compounds than by HPLC-UV.
Though HPLC-UV may not be the appropriate
method to determine the elution order for the five
arsenic compounds, nevertheless, it was used to
separate the five arsenic compounds. The UV detector
was adjusted to 193 nm to utilize the end absorp-
tions of the five arsenic compounds. It should be
noted that the elution data gathered from this


77
TABLE 3 pKa VALUES OF SOME ARSENIC COMPOUNDS
arsenic compound PKa aQ at pH 1 a at pH ^ o c
As(III)a 9.23 1.00 1.00
As(V)b 2.25 0.9^8 0.01?
MMAA 2.60 0.975 0.038
DMAA 6.19 1.00 1.00
PAA 3.59 1.00 0.280
a source of As (in; - HAs02
^source of As(V) - - H^AsO^. pK^ value cited.
'a is the fraction of the acid in the un-
dissociated form, 1.00 is equal to no dissoci-
ation.
Table from reference 50.


78
separation may be unreliable since the UV detector
was operated very near its physical limitations
(i.e., the lowest wavelength the UV detector may be
operated at is 190 nm). The separation of the five
arsenic compounds, as shown in Figure 18, also
utilized acetonitrile as the mobile phase. The
3_
AsO^ eluted at approximately 2.0 minutes with
AsC^- eluting as a shoulder at 2.2 minutes. The MMAA
and PAA eluted at 3.6 minutes and 4.4 minutes,
respectively. No chromatographic peak could be found
for DMAA; this could be due to a strong interaction
between the DMAA and the stationary phase or that it
could not be detected by the UV detector at this
wavelength. The separation of the four arsenic
compounds were quite rapid and that three of the
four arsenic compounds are fairly well resolved from
one another. The AsC^- may be resolved from the
3_
AsO^ by adjusting chromatographic conditions.
To determine the detection limit for this
HPLC-UV method, a calibration curve was constructed.
The calibration curve was constructed only for PAA
since only a rough estimate was needed, and to take
advantage of its absorption maximum at 261 nm. The
UV detector was then reset to 261 nm. Individual
elutions of PAA, at concentrations ranging from 65


79
ppm to 1000 ppm, were then performed. From the
calibration curve, as seen in Figure 20, a detection
limit of approximately 20 ppm may be graphically
determined. However, the chromatogram of the 65 ppm
of PAA (see Figure 21) suggests that the detection
limit may lie in the 55-65 ppm range since a very
small peak was observed. Compared with the detection
limits for arsenic in Table 1, the detection limit
obtained by this method is several orders of mag-
nitude higher due to the poorer sensitivity of UV
detection as opposed to atomic methods.
4.3 Differential Pulse Polarography of As(III)
Differential pulse polarography (DPP) was
employed to study the electrochemical behavior of
As(III) at the DME, HMDE, and at the TMFE. Though
the electrochemical behavior of As(III) has been
cited in the literature, this study was conducted to
supplement the data presently in the literature. The
electrochemical data gathered from this study, along
with the data from the literature, would then be
useful in the construction and operation of the EC
detector.


peak
height (in mm)
Figure 20. A graphical determination of the
detection limit for PAA, as per-
formed by HPLC-UV. A detection
limit of 20 ppm was obtained
' from 'this graph for PAA.


0
2
min
Figure 21. Chromatogram of 65 ppm of PAA using
HPLC-UV. The very small peak found .
for PAA at this concentration indicated
that the detection limit is in this
c o nc e ntrat ion range.


82
4.3.1 As(III) Behavior at the DME.
As earlier noted, Myers and Osteryoung (36)
performed an extensive DPP survey on As(III) reduc-
tion behavior in various supporting electrolyte
solutions at the DME. It was the objective of this
part of the DPP study to reproduce, in part, the
Myers-Osteryoung results, and to determine the
reduction potentials of the other arsenic compounds
in the supporting electrolyte solutions used for
As (III) (arsenite).
Five supporting electrolyte solutions from
the Myers-Osteryoung work were selected for use in
this study. The selection of the supporting
electrolyte solutions (i.e., 1M HC1, 1M HOAc, 1M
H3PO4, lM tartaric acid, and 0.5M oxalic acid) was
to characterize the the reduction behavior of
As(III) in various supporting electrolyte solutions.
The supporting electrolyte solutions prepared from
the weak acids are of special interest since they do
not attack stainless steel (HPLC system) as does
HC1. DPP scans of each supporting electrolyte solu-
tion were taken to determine if potential inter-
ferents (Pb(II) and T1(I) are the most common) that
would mask the As(III) reduction peak were present.
These scans showed that no potential interferent was


83
present that would mask the As(III) reduction peak.
The DPP scan of the 1M H-^PO^ supporting electrolyte
solution, however, revealed a small reduction peak
at -0.98V vs SCE. This small peak may be attributed
to the reduction of Zn(II) which occurs in this
general area. The probable source of the Zn(II) is
from the concentrated H^PO^ used to prepare this
particular supporting electrolyte solution.
DPP scans of 10 ppm of As(III) in each
supporting electrolyte solution (as seen in Figures
22 and 23) were then taken. The potential at which
As(III) reduces in each supporting electrolyte
solution compares very favorably with the values
cited in the literature (see Table 4), except for 10
ppm of As (III) in 0.5M oxalic acid. No reduction
potential value could be determined for As(III) in
oxalic acid since it appeared that the As(III)
reduction peak was shifted into the cathodic back-
ground (see Figure 23). To determine if the As(III)
reduction peak was shifted into the cathodic back-
ground, a DPP scan of 100 ppm of As(III) in 0.5M
oxalic acid was taken. The DPP scan of this solution
resembled the DPP scan of the 10 ppm of As(III) in
0.5M oxalic acid. The shift of the As(III) reduction
peak into the cathodic background is probably due to


8k
The reduction of (a) As(III) in 1MHC1,
(b) As(III) in 1M HOAc, and (c) As(III)
in 1M H-POh. The reductions were per-
formed ^at^the DME.
Figure 22.


85
E (in mV)
E (in mV) E (in mV)
Figure 23. The reduction of (a) 10 ppm As(III) in
1M tartaric acid, (b) 10 ppm As(III)
in 0.5M oxalic acid, and (c) 100 ppm
As(III) in 0.5M oxalic acid. All
reductions were performed at the DME.


86
TABLE k THE REDUCTION OF As(III) IN VARIOUS SUPPORTING ELECTROLYTE SOLUTIONS 1
supporting P a a b b electrolyte § ^b -§- 11 b
1M HC1 -k60 -850 -430 -900
1M HOAc -820 -1250 -800 -1100
1M H3P0^ -660 -1100 -630 -850
1M tartaric acid -690 -1150 -660 -850
3.5M oxalic n/ac -750 -580 -600
acid
E_i is reduction potential for As (III), E^
2 is first appearance of the cathodic
background, potential values obtained
experimentally and measured against
SCE. potentials reported in mV.
is reduction potential for As(III), E,
is first appearance of the cathodic
background, potential values from
reference 36 and measured against SCE;
potentials reported in mV.
c As(III) reduction peak could not be found
with 0.5M oxalic acid as supporting
electrolyte solution.


8?
the strong complexation of As(III) by the oxalic
acid. It is known that a more negative potential
(i.e., greater energy) is required to reduce a metal
complex than to reduce its corresponding simple
(uncomplexed) metal ion. The extent of the cathodic
shift observed for a metal complex as opposed to its
corresponding simple metal ion depends upon the
stability of the metal complex, i.e., the more
stable a given metal complex is, the greater will be
the cathodic shift (9). No formation constant could
be found in the literature that would indicate the
stability of a complex between As(III) and oxalic
acid.
The DPP scan of As(III) in 1M HC1 obtained
in this study was compared with that in the litera-
ture (Figure 12). The differences between the two
scans are the presence of a polarographic maximum
and of the reduction of As(0) to AsH^ in Figure 12,
and are probably due to differences in experimental
conditions. The width of the As(III) reduction peak
obtained in this study (Figure 22) compares very
favorably with that in Figure 12. No comparison DPP
scans of As(III) in the other supporting electrolyte
solutions could be found in the literature. The DPP
scans of As(III) in each supporting electrolyte


88
solution, however, were compared with one another.
The reduction peak for As(III) in 1M HC1 was the
narrowest while the As(III) reduction peak in the
other supporting electrolyte solutions were broader.
The broadening of the As(III) reduction peak prob-
ably indicates that the reduction of As(III) to
As(0) is becoming more irreversible. Broad reduction
peaks are indicative of electrochemically irrevers-
ible systems such as the As(III)/As(0) couple (51).
The reduction peak for As(III) in 1M HC1 was
slightly taller than the reduction peak for As(III)
in the other supporting electrolyte solutions. This
indicates that 1M HC1, as a supporting electrolyte
solution, does offer the better sensitivity than the
other supporting electrolyte solutions to determine
As(III) electrochemically.
DPP scans were then taken of the other
arsenic forms (i.e., As(V) (arsenate), MMAA, DMAA,
and PAA) to determine their respective reduction
potentials. The DPP scans of these arsenic forms
were taken using 1M HOAc as the supporting
electrolyte solution. The selection of 1M HOAc as
the supporting electrolyte solution was based on
that the cathodic background begins at a more nega-
tive potential than the other supporting electrolyte