Title: Anodic stripping voltammetry at a glassy carbon electrode for the determination of platinum species derived from cis-diamminedichloroplatinum(II) /
CITATION PDF VIEWER THUMBNAILS PAGE IMAGE ZOOMABLE
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00098259/00001
 Material Information
Title: Anodic stripping voltammetry at a glassy carbon electrode for the determination of platinum species derived from cis-diamminedichloroplatinum(II) /
Physical Description: xii, 270 leaves : ill. ; 28 cm.
Language: English
Creator: Atherton, David Reed, 1952-
Publication Date: 1984
Copyright Date: 1984
 Subjects
Subject: Cisplatin   ( lcsh )
Voltammetry   ( lcsh )
Cancer -- Chemotherapy   ( lcsh )
Metals in the body   ( lcsh )
Chemistry thesis Ph. D
Dissertations, Academic -- Chemistry -- UF
Genre: bibliography   ( marcgt )
non-fiction   ( marcgt )
 Notes
Thesis: Thesis (Ph. D.)--University of Florida, 1984.
Bibliography: Bibliography: leaves 236-269.
General Note: Typescript.
General Note: Vita.
Statement of Responsibility: by David Reed Atherton.
 Record Information
Bibliographic ID: UF00098259
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: alephbibnum - 000500703
oclc - 12114993
notis - ACS0336

Downloads

This item has the following downloads:

PDF ( 10 MBs ) ( PDF )


Full Text














ANODIC STRIPPING VOLTAMMETRY AT A GLASSY CARBON ELECTRODE
FOR THE DETERMINATION OF PLATINUM SPECIES DERIVED
FROM CIS-DIAMMINEDICHLOROPLATINUM(II)









BY



DAVID REED ATHERTON


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN
PARTIAL FULFILLMENT OF THE REQUIREMENTS
FOR THE DEGREE OF DOCTOR OF PHILOSOPHY



UNIVERSITY OF FLORIDA


1984

























































Copyright 1984

by

David Reed Atherton



























This work is dedicated to the late Henry Ornstein,
who directed my natural curiosity into the field
of chemistry.


















ACKNOWLEDGEMENTS



I would like to thank all those whose efforts have made

my education possible and my stay enjoyable and enlight-

ening. Among those to be singled out are my parents for

their endless patience, understanding and enthusiasm as well

as support both moral and financial. Pam Vetro deserves a

special thanks for her encouragement and sacrifices toward

this goal. Gerhard Schmid has offered great assistance

throughout my stay. Susan Scherer did a fine job on the

artwork. Finally, I would like to thank Jeanne Karably for

the many hours of conversation that helped me keep my sanity

in Leigh Hall.



















TABLE OF CONTENTS


ACKNOWLEDGEMENTS . .


LIST OF TABLES . . . . . . .

LIST OF FIGURES . . . . . . .

ABSTRACT . . . . . . . . .

CHAPTER


. . . vii

. . . ix

. . . xi


I INTRODUCTION . . . . . . . .

II CIS-DIAMMINEDICHLOROPLATINUM(II):
ANTI-CANCER DRUG . . . . . . . .

III CIS-DIAMMINEDICHLOROPLATINUM(II):
CHEMICAL COMPOUND . . . . . . . .

IV .THE PLATINUM ELECTRODE IN ACIDIC SOLUTIONS. .

V ANODIC STRIPPING VOLTAMMETRY . . . . .

Brief Introduction and History . . . .
The Anodic Stripping Experiment . . . .
The Deposition Step . . . . . . .
The Stripping Step . . . . . . .

VI EXPERIMENTAL . . . . . . . . .

Material and Reagents . . . . . . .
Equipment and Instruments . . . . . .
Procedure . . . . . . . . . .

VII DATA, RESULTS AND DISCUSSION . . . . .

Preliminary Investigations . . . . .
Preliminary ASV Experiments . . . . .
Carbon Electrode Evaluation . . . . .
Stripping Solution Optimization . . . .
Conditioning of the Platinum Deposit . . .
Preliminary Deposition from Urine . . . .
Preliminary Chelation Experiments . . . .
Chloroplatinic Acid and its Dissolution . ..


Page

iv


32

47

63

63
67
72
83

129

129
130
133

143

143
153
158
161
175
182
188
189











Chapter Page

Chelation Procedure Optimization . . .. .206
Oxalic Acid as Stripping Solution Mcdifier. .. 216

VIII CONCLUSION . . . . . . . ... 233

REFERENCES . . . . . . . . . .. . 236

BIOGRAPHICAL SKETCH . . . . . . . ... 270


















LIST OF TABLES


Table Page

1. Chemical Speciation of Cisplatin in Blood
Plasma and Intracellular Environments ... .16

2. Equilibrium Constants and Forward Rate
Constants for Cisplatin and Related Species .36

3. Stability Constants of Some Pt2+ Complexes . 44

4. Standard Potentials for Some Platinum
Complexes . . . . . . . . ... 45

5. Status of the Platinum Electrode in Aqueous
Acid Solutions . . . . . . . .. 52

6. Correlation Between Current and Electrode
Rotation Speed . . . . . . . . 147

7. Cyclic Voltammetric Data for 10 ppm Pt at
Several Sweep Rates . . . . . . .. .149

8. Cyclic Voltammetric Peak Current Dependence
on Concentration of Cisplatin for Several
Scan Rates . . . . . . . . .. 151

9. Peak Currents Measured for Consecutive ASV
Experiments without Surface Renewal . . .. .154

10. Peak Currents Measured for Consecutive ASV
Experiments with Surface Renewal . . .. .157

11. Qualitative Evaluation of Stripping Media for
Platinum Redissolution . . . . . .. .159

12. Peak Potentials for Platinum Oxidation and
Reduction in H2SO4 Solutions of Various pH as
Determined by Cyclic Voltammetry. . . .. .172

13. Potential Program for Platinum Deposit Condi-
tioning in 0.05 M H2SO4 . . .. .... 177

14. Peak Heights and Peak Areas for a Series of
Identical Experiments . . . . . ... 178











LIST OF TABLES (continued)


Table Page

15. Deposition Currents and Relative Stripping
Peak Heights for a Series of Deposition
Potentials with Cisplatin as Analyte ... .180

16. Effect or Variation of Cisplatin Concen-
tration on Relative Stripping Peak Height . 183

17. Qualitative Cyclic Voltammetric Evaluation of
Electrolytes for Chloroplatinic Acid Disso-
lution and Subsequent Platinum Deposition . 195

18. Quantitative Evaluation of Electrolytes for
Chloroplatinic Acid Dissolution and Subsequent
Platinum Deposition . . . . . ... .197
O_
19. Evaluation of pH Dependence of the PtCl1
Reduction Process . . . ... . . .199

20. Aging Characteristics of Several Acidic
Buffers with 19.3 ppm Pt added . . ... .204

21. Aging Characteristics of Several Acidic
Buffers with 4.82 ppm Pt added . . ... .205
2-
22. Dependence of the PtC6- Deposition Current
on Electrode Rotation Speed . . . ... .207

23. Deposition Currents and Stripping Peak Heights
for a Series of Deposition Potentials with
PtCl1 as Analyte . . . . . . . 210

24. Results of Several Platinum Chelation
Procedures with Cisplatin in 0.15 M NaCl as
Evaluated by ASV . . . . . . ... .213

25. Results of Several Platinum Chelation
Procedures with Spiked Urine Samples as
Evaluated by ASV . . . . . . ... .215

26. Effects of Varying the Amount of Chelating
Agent on Platinum Recovery from Urine ... .217

27. Qualitative Results of Oxalic Acid Oxidation
for Quantitation of Platinum . . . ... .225



















LIST OF FIGURES


Figure Page

1. Structure of Cisplatin . . . . . .. .14

2. Deoxyribonucleosides . . . . . . 18

3. Hydrolysis of cislatin and reactions of the
hydrolysis products . . . . . . 35

4. Alignment of Pt(NH3)2Cl2 toward positively
charged electrode . . . . . . .. 42

5. Cyclic voltammogram for a typical platinum
electrode in pure aqueous 0.5 M H2SO4 . . 51

6. Differential pulse mode of the PAR Model 174 117

7. Cyclic voltammogram of 10 ppm Pt in
0.15 M NaCl . . . . . . . ... .145

8. jet reductive current in 10 ppm Pt as a
function of electrode rotation speed ... .148

9. Cyclic voltammetric peak current dependence
upon sweep rate for 10 ppm Pt in 0.15 M NaCl 150

10. Cyclic voltammetric peak current dependence
upon concentration of cisplatin for several
scan rates . . . . . . . ... 152

11. Consecutive cyclic voltammograms of 25 ppm Pt
solutions at a carbon paste electrode . . 156

12. Differential pulse stripping curves for an
untouched glassy carbon surface . . .. .163

13. Cyclic voltarmogram of platinum deposited on
glassy carbon . . . . . . ... 167

14. Cyclic and DP voltammograms of platinum
deposited from 10 ppm Pt for 3 different
times . . . . . . . . ... . 170











LIST OF FIGURES (continued)


Figure Page

15. DP voltammetric scans of platinum deposited
from a cisplatin solution after conditioning
CV scans . . . . . . . . ... 174

16. Effect of deposition potential on deposition
current and relative stripping peak height . 181

17. Cyclic voltammograms of H2PtCl6 in different
electrolyte solutions . . . . . . 193

18. pH dependence of stripping peak height,
pH 2 to 10 . . . . . . . ... 200

19. pH dependence of stripping peak height,
pH 1 to 5. . . . . . . . . ... 201

20. Net deposition current in 10 ppm Pt as a
function of electrode rotation speed . . 209

21. Effect of deposition potential on deposition
current and stripping peak height, PtCl as
analyte . . . . . . . . . 211

22. Sequential cyclic voltammograms in PtCl-
solution containing oxalic acid . . .. .218
2-
23. Sequential cyclic voltammograms in PtCl6
solution containing oxalic acid with
differing cathodic conditions . . ... .221

24. Cyclic voltammograms obtained in oxalic acid
solution with a glassy carbon electrode both
with and without a platinum deposit . .. .227

25. Anodic DP voltammograms obtained in oxalic
acid solution with a glassy carbon electrode
both with and without a platinum deposit . 230

26. Anodic DP voltammograms obtained in oxalic
acid solution with a glassy carbon electrode
both with and without a platinum deposit . 232


















Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

ANODIC STRIPPING VOLTAMMETRY AT A GLASSY CARBON ELECTRODE
FOR THE DETERMINATION OF PLATINUM SPECIES DERIVED
FROM CIS-DIAMMINEDICHLOROPLATINUM(II)

By

David Reed Atherton

August 1984

Chairman: Gerhard M. Schmid
Major Department: Chemistry

Following the administration of the anti-cancer drug

cis-diamminedichloroplatinum(II), platinum is found thrcugh-

out the body as a number of species. The lack of platinum

deposition from untreated urine or from urine denatured with

acetonitrile or mineral acids led to the use of chelation by

sodium diethyldithiocarbamate followed by solvent extraction

using chloroform to isolate the platinum from urine. Evap-

oration of the solvent and digestion of the residue in aqua

regia was necessary to obtain a readily deposited form of

platinum, chloroplatinic acid. Redissolution in a pH 1.7,

0.1 M bisulfate/sulfate solution resulted in both a strongly

acidic buffer needed to prevent hydrolysis and a non-

complexing medium with optimum deposition characteristics.

Deposition from the hexachloroplatinate(IV)/buffer

solution was performed at -0.2 V vs SCE, the most cathodic











potential possible without hydrogen evolution. A rotating

glassy carbon electrode was used because of the necessity

for an impermeable surface. Other forms of carbon were

evaluated and found unsuitable. After deposition, for times

of one to ten minutes, a transfer of the glassy carbon

electrode with deposited platinum to a pure aqueous 0.05 M

sulfuric acid solution prevented interference from compo-

nents of the deposition solution and provided an optimum

medium for subsequent steps.

An anodic treatment consisting of 2 s at 1.5 V with

electrode rotation desorbed and/or oxidized impurities in

the deposit, evolved oxygen, and initiated platinum oxide

formation. Continued conditioning at 1.25 V for 30 s with

rotation and for 90 s without rotation swept away oxygen and

completed the oxide formation resulting in maximum platinum

electrochemical activity. A cathodic treatment for 10 s at

0.1 V re-reduced the platinum. An anodic potential sweep

using differential pulse (DP) voltammetry re-oxidized the

deposit and allowed the quantitation of the deposited

platinum. The height of the DP peak, centered at +0.85 V,

was proportional to the original platinum concentration over

the range 0.1 to 50 ppm Pt.

The addition of 0.01 M oxalic acid to the stripping

solution extended the concentration range to as low as

10 ppb Pt. Oxalic acid was oxidized at platinum, but not at

glassy carbon or platinum oxide, resulting in a DP peak

centered at the same potential.


















CHAPTER I

INTRODUCTION



Cancer, the very word is capable of striking fear into

people everywhere. This disease, or more accurately group

of related diseases, and its causes and cures are common

news items. A 1975 National Institute of Health (NIH)

report stated that cancer is the number two cause of death

after heart disease (1). One in four Americans will even-

tually die from some type of cancer (1). The American

Cancer Society estimated that for 1982 there would be

835,000 newly diagnosed cancer cases and 430,000 deaths from

cancer (2). Estimates for 1983 postulate over sii hundred

thousand cases and over three hundred thousand deaths for

the eight most common cancers, including 117,000 deaths from

lung cancer, the most deadly form (3). A unified war on

cancer was begun in 1971 when the National Cancer Act became

law, resulting in over $6 billion being spent to date (4).

The fight has not been without its successes. During

the 1970s, five year cures were developed for nearly a dozen

forms of cancer that had no cure before (4). There are

three basic approaches to combating cancer once it has been

diagnosed. Radiotherapy, chemotherapy and surgery, either

alone or in combination, are used to help improve the










quality and increase the quantity of life of the cancer

patient.

Surgery usually meets with only partial success since

tumors often spread via lymph and blood circulation. Cancer

cells have lost shape specificity and thus lose adhesive

forces which normally bind cells. The result is that

cancerous cells break free and are carried from the tumor

site to lodge and multiply elsewhere. This process of

metastasis initiates new cancer growth and is responsible

for about 85% of cancer deaths (5).

Radiotherapy is the selective destruction of tumor

cells by either external or implanted radiation sources.

Since its inception in 1951 (6) cobalt-60 treatment of over

three million individuals has resulted in more than eleven

million patient-years of increased life expectancy (7).

Radiotherapy can cover a larger area of the body than

surgery but can not treat widespread metastases.

Chemotherapy for cancer began during World War II when

Dr. Charles B. Huggins showed that estrogen slowed the

growth of prostate cancer (8). Drugs can be distributed

throughout the body by using the cancer patient's circula-

tory system. In this manner the chemotherapeutic agent can

locate a single cell, since even cancer cells require a

blood supply. There are now about 40 government-approved

anti-cancer drugs in general useage. There is but one

member of this group that is based on a heavy metal (9).

This lone inorganic compound is the object of this study.











The compound cis-diamminedichloroplatinum(II), also

known as platinol and DDP but in the U. S. most commonly as

cisplatin, is a simple complex of divalent platinum. It can

be monitored by analytical techniques useful for platinum

itself as well as for the molecule. Atomic absorption

spectroscopy and high performance liquid chromatography are

two commonly used methods. This study was undertaken in

hopes of providing an alternative procedure.

The clinical monitoring of platinum in either blood

plasma or urine suggests an electrochemical technique

because of the high salt content of both solutions. The

electroanalytical methods' relative indifference to the

often confounding matrices of high salt content promises a

simpler procedure. Urine is an especially attractive

medium. It can be considered an almost protein-free ultra-

filtrate of the blood plasma (10). Thus urinary platinum

levels would be closely related to unbound, active drug

concentrations in the blood.

Electroanalytical chemistry consists of three primary

branches, potentiometry, coulometry and voltammetry (11).

Dating from the 1920s and Heyrovsky's invention of the

polarographic technique, voltammetry is conceptually quite

simple. As the name implies, the current-potential behavior

of an analyte is measured. In practice, the potential is

varied in a systematic manner so as to cause electroactive

species to be either reduced or oxidized. The resultant











current gives information concerning the amount of the

analyte present.

A wide variety of voltammetric experiments may be done,

depending on how the potential is varied and the current

measured. Sensitivity is limited to solutions of about 10-5

moles per liter (M) for the simplest analysis techniques
-7 -
(12, 13) or about 10-7 to 108 M with instrumentally more

complex procedures (12). Currents due to processes other

than those of the analyte(s) represent the lower bound to a

useful signal. The effect of most of this background noise

has been compensated for in the more instrumentally elab-

orate techniques, thus giving the lower detection limits.

The remaining method to increase the signal to noise-ratio,

thus improving the sensitivity or detection limit, is by

increasing the signal. Preconcentration of the analyte(s)

will increase the analytical signal (12, 14). In using

electrochemical methods to preconcentrate by factors of

about 100 to over 1000, stripping voltammetry has an extra

experimental step which distinguishes it from other electro-

analytical methods (15). One of the many available voltam-

metric techniques is subsequently used to electrochemically

strip the accumulated material from the electrode to com-

plete the analyses. Stripping voltammetry is divided into

anodic (ASV) and cathodic (CSV) variations depending on

whether the preconcentrated material is stripped by an

oxidative anodicc) or reductive cathodicc) path.











Accumulation by reduction and analysis by oxidation

will be employed in the quantitation of platinum. Anodic

stripping voltammetry is the technique of choice for several

reasons, including its sensitivity. Wang recently noted

. . there is no technique for trace metal analysis that

can compete with ASV on the basis of sensitivity per dollar

investment" (16, p. 104A). ASV is the most sensitive

electrochemical technique (17), determining concentrations

lower than any other electrochemical method (18). It was

recently stated by Bond that ". . there is no doubt that

this is the most sensitive polarographic [voltammetric]

method available . ." (19, p. 439). In addition to being
-10
capable of measuring concentrations as low as 101 M (12,
-11
20, 21) to 1011 M (15), ASV can offer, as well, a degree of

selectivity because of the preconcentration step. Instru-

mentation is also inexpensive to buy and operate, easy to

use and the results are precise, accurate and rapidly

obtained (22).


















CHAPTER II

CIS-DIAMMINEDICHLOROPLATINUM(II):
ANTI-CANCER DRUG



Cisplatin is now an established anti-cancer drug. The

work in the early and mid 1960s which led to this present

use is another example of serendipity in science and medi-

cine.

Barnett Rosenberg of Michigan State, at the time a

physicist, thought that the spindle fiber alignment during

mitotic cell division resembled lines of magnetic force like

those seen when iron fillings are placed around a magnet

(23). His initial study was on the effects of an external

electromagnetic field upon cell division. His fortuitous

use of platinum electrodes for their inertness and of a

1000 hz alternating current to prevent electrolysis or

electrode polarization led to interesting and unexpected

results (24). The Escherichia Coli bacteria used in prelim-

inary studies to check out the apparatus and culture media

did not divide normally, but continued to grow nonetheless.

Some eventually formed filaments up to 300 times normal

length (24). A long, careful series of control experiments

showed that the alternating current had not caused the

unusual growth but had caused the platinum electrodes to










dissolve to the extent of about 10 parts per million (ppm)

platinum (24). The ammonium chloride in the particular

growth medium had led to the formation of (NH4)2PtCl6 (24).

This compound was able to photochemically react to form any

of several compounds of the general formula
(2-n)-
[Pt(NH3)nCl_] (2n), n=l, 2 or 3 (25). Rosenberg and

coworkers showed that Pt(NH )2C14, formed after UV light

irradiation, was the active compound (25, 26). By testing

the synthetic isomers of Pt(NH3)2C14 and the related plat-

inum(II) salts Pt(NH3)2C12, it was established that only the

cis forms were active (25, 26).

It was reasoned that since cell division, but not

growth, was stopped perhaps these compounds would prove

effective against cancerous cells, the most rapidly dividing

cells in an organism. The first results were reported in

1969 by Rosenberg and associates using cisplatin (27). The

dosage level that killed half a sample (LD50) of Swiss white

mice was determined to be 14 mg kg- Levels below the LD50

were then tested for tumor growth inhibitory action against

transplanted Sarcoma 180 (one of about 100 standard test

tumors). Tumor growth was almost completely inhibited at

8 mg kg- which was less than LD10 (27). A further study

by Rosenberg and VanCamp showed even more encouraging

results (28). Even after the tumor had grown to a very

large size after eight days, a single intraperitoneal

injection of 8 mg kg-1 of cisplatin caused complete regres-

sion in nearly 100% of the test mice. These papers by











Rosenberg and coworkers trace the early development of this

now potent anti-cancer agent.

At this point in time, work with cisplatin had expanded

and other groups were involved in the late 1960s. Different

types of transplantable tumors (29, 30) as well as tumors

induced by carcinogens (31) or viruses (32) were studied and

the results reported. The outcomes of these investigations

were positive and spurred additional inquiries. Analogs of

cisplatin were soon considered (33-37). Scores of compounds

with either substitution of chloride by other halides,

nitrate, nitrite and other singly charged anions or substi-

tution of ammonia were investigated. Amines including C2-C4

n-alkyl, C3-C5 iso-alkyl, C2-C8 alicyclic and C3-C7 hetero-

cyclic amines and some diamino alkanes, aromatics and

cycloalkanes have all shown activity (9). These active and

many more inactive analogs have undergone scrutiny in the

search for second generation drugs based on platinum.

Generally the most satisfactory results were obtained with

the parent compound, which continued to be studied

primarily.

Summarizing nearly a decade of research on analogs of

his original compound, Rosenberg in 1979 gave the following

observations on platinum compound antineoplastic activity:

i. the complex should be electrically neutral,

ii. not all the ligands should leave or exchange,

iii. the 2 mono- or 1 bidentate leaving groups)

should be in the cis positions,











iv. the tightly bound ligand(s) should be a

substitutionally inert amine, and

v. the leaving ability of the labile ligand(s)

should be of intermediate liability (38).

Cleare has published a number of papers on the structure-

activity relationships of platinum compounds and summarizes

much of what has been done to this time in a recent review

(9). Seeking to find compounds with the same activity but

decreased toxicity, or with an increased spectrum of activ-

ity is the goal of studies of cisplatin analogs.

By the time the parent compound had been submitted to

the National Cancer Institute (NCI) for human clinical

trials, over 140,000 other compounds had been screened for

anti-cancer activity (39) of which only 20 were inorganic

(40). The Division of Cancer Treatment of the NCI approved

the start of human trials in 1971 (41). NCI drug research

is designed to run as a linear array (42). A detailed

explanation of the phases of human trials in drug research,

with emphasis on cisplatin, can be consulted (43). Follow-

ing the initial screening and preclinical investigations

with animals to determine formulations, toxicities and

pharmacology, Phase I studies begin.

Phase I studies are intended to learn how to use a

drug, find the maximum dose level tolerated at a given

schedule and evaluate any side effects (44, 45). Patients

in these trials have tumors which have shown resistance to

all other drugs and they are usually terminally ill. They











must have normal organ functions and have an estimated

survival of more than two months. Even so, promising

results are not expected due to the advanced stage of the

disease. Initial doses are developed from results of animal

tests. An example of a Phase I study with cisplatin used a

dose of one sixth the minimum lethal dose in dogs for the

45 human subjects (46, 47). The reports of this study show

that all the patients experienced nausea and vomiting, most

suffered kidney damage and nine of eleven with testicular

cancer had sone response to drug treatment. Kidney damage,

due to strong binding at the membrane of the proximal

tubules, slowed the early human research (48). Neverthe-

less, sufficient progress was made and significant responses

to the drug were seen so that Phase II trials were begun in

1975 (23).

Phase II trials are designed to screen for any clinical

activity when the drug is nearly optimally used (45). Five

slow growing and five fast growing types of cancer and about

15 to 30 patients with each type are evaluated for tumor

regression, cell cycle sensitivity and mechanism of the

drug's action (43). During these trials, careful controls

are made with regard to patient histories, treatments and

side effects encountered. Information obtained from Phase I

studies allows the response to the drug to be more reliably

monitored because adverse effects are generally understood

and held to a minimum.











Following, or often running simultaneously, are Phase

III studies. These trials are designed to compare the

responses to a new drug to those of established regimes

(45). The new agent is also evaluated in combination with

existing drugs. Combination therapy has the advantage of

several modes of action but generally decreased toxicities.

Upon successful completion of these trials, the drug is

recommended for general usage.

Success in the three phases of clinical evaluation led

to the approval of cisplatin as a human anticancer agent.

Government approval came in the United States by the Food

and Drug Administration (FDA) in December 1978 and in the

United Kingdom by the Department of Health and Social

Security in March 1979 (49). By this time the NCI had

screened 1055 different platinum complexes with about 18%

showing some activity (50). In the battle to control the

more than 100 different cancers over 300,000 compounds had

been screened by the time cisplatin was added to the approx-

imately 40 other drugs in general, approved usage (51).

In addition to the strictly clinical work which has

resulted in the present status of cisplatin, investigations

have also been carried out in related areas to broaden the

understanding of the drug. Studies designed to learn about

the mechanisms) of action, to understand the pharmaco-

kinetics of the drug distribution and to further the

knowledge of toxic side effects have all been undertaken in

conjunction with clinical work or as separate experiments.










In the course of these and future determinations concerning

cisplatin and second generation analogs, methods to analyze

for the amount of platinum present are needed.

It is important to establish a drug's mechanisms) of

action in order to obtain guidelines for optimal use of the

drug and to help in the design of more effective analogs.

The total action of the drug and the reaction of the body is

divided into three stages (52). First the drug must be

transported to the site of action. During this process, the

compound may undergo structural modification. Metabolic

processes or a change in the chemical environment can

contribute to these alterations. In addition the drug may

be reversibly or irreversibly bound, possibly rendering it

ineffective, or it may be excreted. The second stage is the

inhibition of biochemical pathways leading to the desired

and any side effects. The third stage of drug action is

recovery of the body from the drug's effects. This recovery

may be due to elimination of the drug, repair mechanisms

which excise the affected areas, or by-passing of the

altered pathway or drug action site. Drug actions differ

from the actions of pathogens primarily in that the net

result is beneficial to the organism acted upon by the

agent. Mechanisms may be quite similar.

The best mechanistic conclusions arise from a consider-

able body of information obtained from researchers in many

related fields. A thorough knowledge of chemical, physical

and physiological properties of the drug is required to










ascertain its mechanisms) of action. Physiological proper-

ties of cisplatin will be considered in this chapter with

details of the underlying chemical and physical properties

to follow in a subsequent chapter. As the relationships

between the physiological properties and the chemical and

physical characteristics are not completely separable, some

directly relevant characteristics of the compound will be

introduced here.

The structure of the cis-diamminedichloroplatinum(II)

molecule as determined by Truter and Milburn (53) may be

seen in Figure 1. The two ammonia ligands are tightly bound

while the chloride ligands are labile. In aqueous solution,

both chloride ions are slowly lost from the co-ordination

sphere of the central Pt2+ ion and are replaced by either

water or hydroxide. This equilibrium is effected by the

presence of chloride ion in the solution and the concentra-

tion of chloride ion determines the species distribution.

Due to the relatively high concentration of chloride ions in

the blood plasma (about 103 rai), the cis- platin molecule

retains its chloride ions and remains elec- trically neu-

tral. It has been shown that the principal form of free

platinum in the blood and urine is the unchanged drug (54).

Because of the concentration gradient and the neutral

molecule's lipid solubility, the molecule is able to diffuse

unaided by cellular mechanisms or energy through the cell

membrane into cells (55). This physical transport, without

cellular intervention, is called passive diffusion (56).
















2.75A 87 Pt 91.9 3.3A




H3N
So C -



Figure 1. Structure of Cisplatin (53).











Within the cell, the molecule encounters a relatively

low chloride ion concentration of about 4mM. In this medium

the equilibrium is shifted and hydrolysis takes place (55,

57, 58). Following hydrolysis, deprotonation can take place

resulting in any of several species (38). Equilibrium

calculations considering the concentration of chloride ions,

a body temperature of 37C, and a pH of 7.4 show that 84% of

the drug remains intact in the plasma compared to only 9.23%

in cells (59). The total speciation can be seen in Table 1.

Once inside the cell and hydrolyzed, the compound is

capable of binding to the various molecules it encounters.

The DNA molecule, with its alternating deoxyribose sugars

and phosphates groups as the backbone and purine and pyri-

midine bases as linking groups, provides several possible

binding sites (60). Pascoe and Roberts showed that DNA is

the predominantly bound intracellular molecule when they

obtained the following binding data (61):

Macromolecule DNA mRNA rRNA tRNA proteins

Pt per molecule 22 0.125 0.033 0.00067 0.00067

No evidence of platinum binding to ribose rings at either

the ring oxygen or at the exocyclic hydroxyl groups has been

found in any investigation of DNA, RNA or their constituent

parts (59). There is also no evidence of platinum-phosphate

group interactions with natural DNA or RNA (59). The strong

nucleophilicity of the nitrogen atoms of the organic base

groups enables binding to selectively occur at these points

on the DNA strand (60, 62). Base site binding is not unique









Table 1. Chemical Speciation of Cisplatin in Blood Plasma and Intracellular Environments


cis[Pt(NH ),...
S..Cl12] ... (H20) (Cl) I . (OH) (Cl) ] ... (H20) 2]
Species


1 : 0.0178

(0.06%)





1 : 0.46

(4.6%)


. (H 0) (OH) I + ... (OH) ]


1.12

(3.9%)





2.90

(29.3%)


1.41

(4.9%)





3.63

(36.6%)


Note: Table from reference 29


Relative

Plasma

Abundance



Relative

Cellular

Abundance


24

(84%)





0.915

(9.2%)


1

(3.5%)





1

(10.1%)










to any nucleic acid site but occurs with all nucleic acids

with the exception of thymine (36). Formation constants for

the binding of hydrolyzed cisplatin to the remaining DNA

nucleic acids have been determined to be 1.06 x 10 for

guanosine, 6.6 x 104 for cytosine and 2.4 x 104 for adeno-

sine (62). The interaction is strongest and most rapid with

guanosine (G) (36, 58, 63) and takes place more slowly with

cytosine (C) and adenosine (A) (58, 64). Some evidence,

while confirming preferential guanosine binding, concludes

more binding at adenosine than cytosine (65, 66). The

preferential binding of the divalent platinum species to one

site of the guanosine molecule is the important biochemical

reaction. See Figure 2 for DNA component structures.

A potentiometric study using cisplatin and DNA demon-

strated that both chloride ions were released in the binding

process but only one from the inactive trans configuration

of diamminedichloroplatinum(II) (67). That the chloride

ions must be lost intracellularly was shown when little and

nonselective binding was seen when cis-Pt(NH3)+ was used

instead of cis-Pt(NH )2Cl2 (66). Many studies have been

conducted that conclude the N(7) site of guanosine to be the

platinum binding site. Spin-spin counting by means of
195Pt H NMR was used by Kong and Theophanides (68).

Using fast atom bombardment mass spectrometry, a reaction

product of [Pt(NH3)2(G)2]C12 was isolated with the binding

occurring at the N(7) sites (69). An X-ray structure was
2+
obtained for [Pt(en)(G)2]1 where en represents the











0
II
6 N
N 7\
2 4N
H2N N N
R



NH2
6 N



N N
R



NH2
-4

2 6
IN
R

0
HOCH2 "N


GUANOSINE










ADENOSINE










CYTIDINE






R=DEOXYRIBOSE
SUGAR
(N=NUCLEIC ACID)


Figure 2. Deoxyribonucleosides.










bidentate ethylenediammine molecule, which showed each

guanosine was bound at its N(7) site (70). Binding of

cisplatin to structurally similar adenosine has been shown

to be at the N(7) position and also at N(1) (71). The

cisplatin molecule is able to bind to two different sites as

a result of its two labile chloride ions. Completion of the

binding to the DNA molecule at a second site is the

determining factor in the anti-cancer activity.

It has been suggested that the 0(6) position in the

sane guanosine molecule could act as the second site for

platinum binding (72). This location is known to be

involved in the hydrogen bonding between base pairs of

associated single DNA strands (72). More evidence has been

gathered to suggest that the bidentate juncture is completed

when a second, adjacent guanosine is bound at its N(7) site

(73). The N(7) location is not involved in Watson-Crick

model base pairing (70) and so is more available than the

0(6) position.

Studies with polymers of G-C segments or long strings

of guanosine followed by strings of cytidine demonstrated

that more platinum incorporation occurred with sequential

guanosine molecules than with alternating guanosine mole-

cules (74). These buoyant density studies, when extended to

DNA, showed increased platinum binding with increased

guanosine content (74). NMR investigations showed adjacent

guanosines bound at their N(7) positions when cisplatin and

the CCGG unit were mixed (75). Enzymatic fractionation of











the DNA-cisplatin reaction product showed platinum bound to

adjacent guanosines (76, 77) and also to guanosines separ-

ated by an unbound nucleoside (77).

The binding of platinum to DNA results in several

molecular changes. The DNA double helix has been observed

to shorten, change its viscosity which implies a change in

shape, unwind and rewind, form microloops, change its

denaturation melting curves and increase in sensitivity to

single-strand specific enzymes (78). These results are

evidence of changes in the interstrand linkage (78).

Binding of the intracellular cisplatin species to the N(7)

position of a guanosine nucleoside results in the loss of

hydrogen bonding between the guanosine and its associated

cytidine on the opposite strand. From pyrolysis mass

spectrometry experiments, the G-C base pairing perturbation

is in evidence as the increase in magnitude of the guanosine

and cytidine peaks compared to the peaks due to adenosine

and its associated thymidine, which do not change (79). A

second, adjacent guanosine on the same strand has its N(7)
o
site approximately 3.4A from the first N(7) position (80).

As seen in Figure 1, the chloride-chloride distance and thus
0
the required binding site separation is 3.3A. This further

supports the supposition of bidentate binding of the cis-

platin species between N(7) atoms of adjacent guanosine

nucleosides on a single strand of DNA. Interstrand cross-

linking has been shown to occur in only one of 150 bound

platinum atoms in vivo (81). The intrastrand attachment of










the cisplatin species is able to prevent the crosslinking of

complementary strands necessary for the formation of the DNA

double helix and thus inhibit the synthesis of DNA (82-87)

Several reviews cover the subject of cisplatin binding to

nucleic acids and DNA in considerable detail (59, 87-89).

The first report of cisplatin inhibition of DNA syn-

thesis in vivo was in 1970 (90). Since that time, DNA

synthesis inhibition has become generally accepted as the

mode of antitumor activity. As Roberts said in a 1982

review on cisplatin, "Selective inhibition of DNA synthesis

appears to be the most likely biochemical lesion leading to

cell death" (91, p. 110). It has been reported that cis-

platin can kill cells in all stages of the cell cycle (92).
2+
The spontaneous dissociation of cis-Pt(NH )2 leaving DNA is

extremely slow at 37CC (93) and changing the surrounding

chloride ion concentration has no effect on the rate (94).

Although platinum is not lost from DNA in vitro (95), there

is evidence that platinum is removed from DNA in growing

cells following a single treatment with cisplatin (95, 96).

Nondividing cells can enzymatically excise platinum and

eliminate 50% of an initial dose in ,28 hours (95). Cell

recovery and survival depends on the ability to remove bound

platinum and thus repair DNA damage (91, 95). The exhibited

toxicity of the drug in cancer cells is manifested in the

differences between cancerous and normal cells. It is

widely held that the breakdown of processes to repair damage

to DNA, such as caused by cisplatin binding, is a major











factor in the transformation of normal cells to cancerous

cells. Cisplatin has been shown to be exceptionally toxic

to mutants with known DNA repair deficiencies (97-101).

This has been demonstrated using Chinese hamster ovary cells

in which 80% of normal cells compared to a small proportion

of mutant cells were able to fix DNA damage (101) and also

in human skin cells (99).

DNA replication is inhibited at therapeutic doses of

cisplatin while RNA and protein synthesis is affected only

at near lethal doses (102, 103). The 1970 paper by Harder

and Rosenberg is the first report of DNA synthesis inhibi-

tion in human cells in culture (102). One molecule of

cisplatin per 1000 DNA base sites is enough for biological

inactivation of the DNA molecule (104). Cytotoxicity of the

drug depends on the extent of reaction with DNA, the capa-

city of the particular cell to excise damage from the

template DNA and the ability of the cell to synthesize DNA

even with the damage. Any differences are reflected in the

sensitivity of various cell types to cisplatin. The drug

does not work equally well in all parts of the body (91).

The best results have been with testicular and ovarian

cancers. The initial approval in 1978 was for treatment of

these two forms of cancer (105). Very good results have

also been obtained in the treatment of bladder cancer (106)

and head and neck cancers (107). Durant described the

developments in chemotherapy for bladder and head and neck

cancers as a change from previously resistant to presently










responsive (108). The U.S. FDA approved the use of

cisplatin against advanced bladder cancer in early 1982

(105). Other significant outcomes with cisplatin treatment

have been with cervical and prostatic cancers, which have

progressed from unresponsive to resistant (108). Unfortun-

ately, cisplatin is the only highly active drug available

for the treatment of advanced cervical cancer (109). A

noncancer application was recently reported as 92.5% of mice

infected with a Trypanosoma parasite were freed of the

infestation (110).

Testicular cancers represent about 1% of all male

malignant tumors (111, 112), are the most common solid

tumors in 15 to 35 year olds (113) and the most deadly form

of cancer in 25 to 34 year olds (112). Before the availa-

bility of cisplatin, 90% of patients with advanced non-

seminomatous testicular cancer died (113). Cisplatin is

most commonly used in combination with the more established

drugs vinblastin and bleomycin against testicular cancers.

A group of 47 patients treated with this regime had 34 com-

plete and 13 partial remissions for a 100% response rate

(114). Results from a number of treatment centers in the

years 1975-1978 showed 63 complete and 53 partial remissions

for an 89% response rate in the 130 patients (115). The

same synergistic treatment produced 132 disease-free

patients out of 171 admitted with disseminated testicular

cancer in a study reported in 1981 (116). About 70% of all

testicular cancer patients and nearly 100% of those with











early or intermediate cancer can now be expected to live

(113).

Ovarian cancer is the most fatal gynecological malig-

nancy and estimates are that 1 in 100 American women will

die from it (23). Early treatments using cisplatin showed

significantly better results compared to previously used

drugs (117, 118). In the years 1975 to 1979, 235 women were

treated with cisplatin with 28% showing complete or partial

remission (119). With various combination chemotherapies,

all including cisplatin, 52% of 251 women responded to

treatment in the years 1977 to 1979 (119). In more recent

studies using cisplatin only for advanced ovarian cancer in

women with no prior treatment (a truer estimate of a drug's

effectiveness), 50% of the women responded for an average of

12.5 months (120).

As with virtually all chemotherapeutic agents used

against cancer, cisplatin has by virtue of its cytotoxic

nature undesirable side effects. Nausea and vomiting

accompany cisplatin use in most patients, commencing 1 to 2

hours after a dose (41, 83, 121) and lasting for 4 to 6

hours (121). This situation is often ameliorated by slow

infusion of the drug over a 24-hour period (122). These and

other adverse effects are dose dependent and often dose

limiting. In some situations the patient will refuse

further treatment because of the severity of nausea and

vomiting. In other cases, treatment must be discontinued

because of one of several cumulative side effects.











As was mentioned earlier, kidney damage was encountered

in preliminary testing on humans and slowed the initial

applications research. The damage to kidney function,

causing a decrease in the filtering capacity of the kidneys,

has now largely been controlled (41, 48, 83, 123, 124).

Hydration of the patient by infusion of normal saline for

several hours before drug administration followed by the use

of mannitol to force diuresis was discovered in 1977 to

markedly decrease kidney damage (123). Only 5 to 10% of all

patients experience nephrotoxicity under this treatment to

increase urine production (125). There are about 1 1/4 mil-

lion nephrons in each kidney (56). The proximal tubules are

the nephronic subunits responsible for the majority of the

reabsorption of water, electrolytes and nutrients from the

blood (56). Renal damage occurs in the area where the

platinum concentration is the highest, in the proximal

tubules (126). Although no change in kidney function or

urine composition is seen until after about 3 days (127),

light microscope evaluations of the damage in humans have

shown it to persist for at least 12 months following treat-

ment (128).

Hydration and diuresis before, during and after cis-

platin administration has reduced the incidence and severity

of nephrotoxicity by reducing the urinary platinum concen-

tration (124) and decreasing the renal residence time of the

drug (129). This mode of treatment does not affect the

total urinary excretion of platinum, the tissue distribution










of the drug or the rate of elimination (124). Because of

the life-threatening effects of compromised kidney function

such as excessive loss of magnesium and calcium (hypomag-

nesemia has been seen in 50% of patients in some cases

(130)) or inadequate elimination of urea (hyperuricemia),

other strategies have been advanced. Chelation therapy with

diethyldithiocarbamate administered shortly after the drug

appears to prevent binding of the platinum species in the

kidney (131-134). Sodium thiosulfate inactivates cisplatin

if mixed in solution but if injected one hour before the

drug it decreases renal damage (135). By using the circad-

ian rhythm of urine volume to naturally decrease platinum

concentration in the kidney, toxicity has been reduced (136,

137). A recent review covers the area of nephiotoxicity of

cisplatin (124).

An often seen, non-life-threatening side effect is

hearing loss. Ototoxicity appears as hair cell loss in

rhesus monkeys (138) and hair cell loss with some scarring

in early as three days after dosage in guinea pigs (139). A

study in 1974 found hearing loss at 2 kHz in 4 patients, at

4kHz in 14 patients and at 8 kHz in 18 patients in a group

of 30 (140). Monitoring of the patient for any hearing loss

and discontinuation of cisplatin treatment if it is found,

is an effective deterrent. In about 30% of all patients

(41) hearing loss occurs in the range of 4 to 8 kHz, above

spoken language (41, 44, 83, 141). The loss is










irreversible, cumulative and can progress into the spoken

range of 1 to 4 kHz if the drug regime is not altered (44).

Physiological studies on a number of different types of

laboratory animals have shown no effect of cisplatin on

blood pressure or blood pressure regulation, heart rate,

respiration, EEG, neuromuscular transmission, muscle con-

traction, spinal reflex or body temperature regulation

(142). Toxicities which occasionally appear are related to

the drug's mode of action. The faster growing tissues in

the body are sometimes damaged by cisplatin. The intestinal

mucosa and villi are rapidly growing tissues; nausea and

vomiting are signs of gastrointestinal distress. The bone

marrow is another area of very active cell growth and anemia

can often result from cisplatin treatment (143). Other

areas, such as the lymphoid tissues, can be damaged as well.

Careful monitoring of body platinum levels is necessary to

best avoid these toxic side effects.

Cisplatin rapidly binds to blood proteins and is also

quickly eliminated, at least partially. The drug can enter

target cells and be an effective cytotoxin only if it

remains unbound. When a cisplatin/serum protein complex was

made and injected into rats, the tightly bound drug proved

to be an insignificant reservoir of the drug (144). The

concentration of free cisplatin is of primary importance.

Three hours after an IV bolus injection (410 minute infu-

sion) only about 10% of the drug remains unbound (145, 146).










Ultimately up to 97% of the serum concentration may be

protein bound platinum (84).

The total body load of platinum is reduced through a

two step or bi-phasic elimination. This can be expressed

mathematically as:

Rate of elimination = Ae- t + Be
0.693
where tl/2,a 0.
1/2,a
0.693
and t12/2 0
1/2,6

Reported values for the half-life of the initial, rapid

elimination of cisplatin from the blood plasma include about

30 minutes (146), 25 to 49 minutes (147), 23 minutes (148)

or 28 minutes (149). Other values reported for tl/2, fall

in the same range of one half to three quarters of an hour.

The second, much slower phase, follows the first and because

of much lower concentrations, data are imprecise. Values

of t /2, have been variously reported as 4 to 5 days (146),

58 to 73 hours (147), 67 hours (10) or 42.2 hours (148).

Traces of platinum have been found in the body up to

4 months after dosage (149). The first process is a combin-

ation of plasma protein binding, tissue absorption and

urinary excretion (of which little takes place initially).

The second process is primarily elimination from the body

through the urine, the other two courses being essentially

completed. Values measured for urinary excretion rates

include 17% in the first 4 hours (10) and 18 to 34% (147),

23% (10) and 20 to 30% (150) in the first day. Plasma










levels of platinum can range from a peak value of about

30 uM or 6 ppm (107) to zero. Blood plasma analysis can

lead to varying concentrations because of the rapid physical

and physiological processes.

There is not a selective uptake of cisplatin by any

particular type of tissue. Platinum has been found almost

everywhere in the body. A typical therapeutic dose of cis-

platin gives a concentration of platinum of about 1 ppm or

5 pM in tumor cells (151). Within normal body components,

the highest concentrations are in the liver (152-154),

kidney (152, 154) and intestines (152, 153). Very low

levels in the brain imply poor drug penetration of the

central nervous system (152, 154). Sensitive analytical

methods must be used to quantitate the small amounts of

platinum.

Several instrumental techniques have been used i'

cisplatin analysis by determining the concentration of

platinum. X-ray fluorescence, which gives total platinum

concentration, has been used and has a limit of detection

(LOD) of about 240 ppb (parts per billion) and a linear

dynamic range (LDR) of 570 to 5700 ppb (145). A molecular

fluorescence method to determine platinum bound to DNA uses

ethidium bromide, which also binds to DNA, as a fluorophore

(155). Because of the competition for binding sites, a

decrease in the fluorescence signal corresponds to an

increase in the amount of bound platinum. With a limit of

detection of about 1 platinum per 250 nucleic acids and a










linear relationship between fluorescence and bound platinum

up to 1 platinum per 5 nucleic acids, this is a very useful

technique for binding studies. High performance liquid

chromatography (HPLC) has been used in the determination of

several forms of platinum, using several detection methods.

Cisplatin and its metabolites were determined directly

from urine by complexation, extraction into chloroform,

injection onto the HPLC of the chloroform extract and UV

detection of the complex (156). Urine concentrations of 25

ppb were determined with 2.5% precision and 4% accuracy

(156). Platinum drug species have been determined by HPLC

with electrochemical detection (LCEC). Using reductive LCEC

for untreated urine samples at mercury electrodes (157) or

either reductive or oxidative LCEC for filtered plasma

samples at glassy carbon electrodes (158), platinum

concentrations as low as 100 ppb have been measured.

Another HPLC method used either on line UV detection or off

line atomic absorption (AA) for platinum quantitation (159).

AA was also used in conjunction with an ion exchange chroma-

tographic separation to obtain unbound drug concentrations

as low as 40 ppb, with an LDR up to 2 ppm (160).

Atomic absorption spectroscopy, in the flameless or

furnace mode, is probably the most widely used method for

platinum determination in cisplatin treatments. Depending

on sample preparation techniques, results can be total,

bound or filterable, or unbound platinum (161). Using

different degrees and methods of sample digestion, tissues






31



(162-164) and fluids (162, 165-167) have been successfully

analyzed with LOD's as low as 30 ppb (162, 163).

In addition to the specialized reviews of cisplatin

binding (59, 87-89) and nephrotoxicity (124), a number of

generalized reviews are available. A 1974 book (43) and a

more recent monograph (108) have appeared. An annual volume

on cancer chemotherapy has contained a section on cisplatin

each year since its inaugural issue in 1979 (91, 107, 119,

168). Reviews also periodically appear in journals (115,

121, 143, 151, 169).


















CHAPTER III

CIS-DIAMMINEDICHLOROPLATINUM(II):
CHEMICAL COMPOUND



In order to develop an analytical method for the

determination of platinum originating as the drug cis-

diamminedichloroplatinum (II), a survey of various proper-

ties of this compound was undertaken. Considering that

ultimately the drug and its metabolites were to be studied

in body fluids and that the analytical method was to be an

electrochemical technique, information deemed pertinent to

these areas was selectively obtained.

Dichlorodiammineplatinum(II) was first reported in 1845

by Peyrone (170). It was Werner in 1893 who first

distinguished two isomers by their distinct reaction pro-

ducts (171). He assigned each isomer a different name and

more importantly described the geometries we are familiar

with today. The cis form is the therapeutically active

compound and will be primarily considered in this discus-

sion.

Cis-diamminedichloroplatinum(II), or cisplatin, is

stable in powder form for over 2 years in a cool, dark place

and at least a year at room temperature (172). The solid is

only slightly soluble in water. Values reported in the










literature include 2.2 g L (35) and 0.2523 g per 100 g

water at 250C (173). In pure aqueous solution the compound

hydrolyzes, the labile chloride ions being replaced by water

molecules. A 500 ppm aqueous solution of platinum as

cisplatin will hydrolyze to the extent of 10% within 40 min-

utes; a 50 ppm solution takes 60 minutes (174). Isomeriza-

tion of cisplatin was measured at 200C in aqueous solution

and found to proceed by a first order process with a rate
-3 -1
constant o 1.06 0.08 x 103 min1 (175). This reaction

represents another mechanism for the loss of therapeutic

activity and suggests that aqueous solution of the drug must

be used within a short time period.

With the addition of chloride ions, hydrolysis slows

and solubility at first decreases. In a 0.006 M solution of

KC1, the cisplatin solubility has decreased by 15% compared

to pure water (176). The maximum decrease in solubility of

19% is seen at 0.015 and 0.030 M KC1, then the solubility

continually increases with increasing amount of chloride

(176). The solubility reaches that of pure water at 0.8 M

KC1 and increases to 0.385 g per 100 mL in 4 M KC1 solution

(176). The decrease in solubility was attributed to a shift

in the equilibrium of the hydrolysis reactions and the

increase to ion-dipole interactions, considering the cis-

platin molecule as a dipole (176). With 0.1% added chloride

a 500 ppm solution will take 6 hours instead of 40 minutes

to hydrolyze by 10% (174). In half normal saline (0.45%

chloride medium) less than a 5% loss will be noted for a










500 ppm solution (174). Using normal saline (0.15 M NaC1) a

1000 ppm solution will hydrolyze only about 3% at equilib-

rium, which is reached in less than an hour (177) and less

than 2% if protected from light (172). While the addition

of chloride protects the drug from hydrolysis, the chloride

slowly replaces one of the ammonia ligands resulting in an

increase in pH (176).

The hydrolysis of cisplatin proceeds in two steps,

consisting of the sequential replacement of the chloride

ions by water molecules. The cis configuration is retained

in the singly charged chloro-aquo compound as well as in the

doubly charged di-aquo compound. The water molecules bound

to the platinum can deprotonate resulting in three

additional species. A summary of the reactions of cisplatin

and accompanying thermodynamic data is shown in Figure 3 and

Table 2. Calculations based upon these equilibrium data

show that relatively little dissociation should take place

in the high chloride ion concentration and near neutral pH

of the blood plasma. Supporting data have been given in

Table 1. For preliminary studies, chloride ion concentra-

tions of 0.15 M, corresponding to normal saline, in neutral

solution should sufficiently mimic blood plasma.

During an anodic stripping voltammetry experiment, the

analyte is first preconcentrated by reduction onto or into

the working electrode. The standard potential demonstrates

the feasibility of the procedure under expected conditions,









cis Pt(NH3) 2C2


-Cl
+H20 (1)


+
+DNA -Ht^
products -- cis[Pt(NH3)2(H2O)Cl]+ cis Pt(NH3) (OH)Cl
(6) (3)



H-C2 (2)
+H2


+DNA -H+ -H+
products cis[Pt(Nfli)2(HB0)] 2 ii cis[Pt(NH3) 2(H2O)OH] cis Pt(NH3) (OH)
(7) (4) (5)







Figure 3. Hydrolysis of cisplatin and reactions of the hydrolysis products.







36



Table 2. Equilibrium Constants (K) and Forward Rate
Constants (k) for Cisplatin and Related Species as
seen in Figure 3.



Constant Temperature Reference

-5 -1
k1 2.5 x 10 s 250C 178
2.6 25 179
2.80 25 180
5.08 30 180
7.6 35 178
11 37 179
3.7 -- 177
-5 -1
k2 2.2 x 105 s1 25 93
3.3 25 178
4.0 25 180
10.4 35 178
K 3.63 x 10-3 mol L1 25 181
3.8 25 180
4.37 35 181
3.7 -- 177
K2 0.111 x 10-3 mol L1 25 181
0.188 35 181
K3 pK = 6.3 -- 182
K4 pK = 5.6 20 183
5.44 25 184
5.63 -- 180
K5 pK = 7.3 20 183
7.07 25 184
-1
k6 0.32 Ls mole ; t l2 = 6 hr 25 93
S143 Ls mole t = 0.8 min 25 93
k7 143 Ls mole ; tl/2 = 0.8 min 25 93










cis Pt(NH )2C12 + 2e Pt + 2NH3 + 2C1

E = +0.49V [1] (185).





A recent patent was issued for a process which utilizes

cisplatin in a NaCI solution to obtain electroplated plati-

num (186). Oxidation of the drug will also occur as seen

from the standard potential of a possible product,





cis Pt(NH )2C14 + 2e -cis Pt(NH3)2C12 + 2C1

E = +0.69V [2] (173,185).





This could interfere with subsequent anodic steps of the

experiment if the parent compound is present. In 1 M

chloride solution, the oxidation of cisplatin to cis

Pt(NH3)2C14 was observed to occur at +0.70 V and the reverse

reduction occurred at +0.62 V (187).

In early platinum electrodeposition studies, primarily

for the platinization of platinum electrodes, it was found

that small quantities of lead acetate improved the overall

characteristics of the deposit (188, 189). Addition of lead

acetate in the range of 10- to 10- M shifts the cathodic

current-voltage curve anodically by 100 to 200 mV (189),

increases the platinum deposition rate (189) and improves

the adherence and smoothness of the plate (188). One source










recommends a solution 3.5% (0.072 M) in platinum
(chlropatinc aid)-4
(chloroplatinic acid) and 0.005% (1.3 x 10-4 M) in lead

acetate (189). Perhaps the addition of similar amounts of

lead acetate will likewise enhance the present analysis.

A considerable number of electrochemical studies have

been done on complexes of platinum and in particular on

cis-diamminedichloroplatinum(II). The majority of the work

has been polarographic, but electrodes other than the

dropping mercury electrode (DME) have been used.

Chakravarty and Banerjea found that cisplatin in 0.1 M KC1

had a polarographic half-wave potential (E1/2) of

approximately -0.1 V vs SCE (saturated calomel electrode)

(190). They further noted that the reaction was

irreversible. Kivalo and Laitinen found an E of about

-0.2 to -0.25 V vs SCE in 1.0 M NaC1 (191). When they used

a rotating platinum electrode instead of a DME, the E/2 was

about -0.35 V. Sundholm, using stationary electrode polaro-

graphy or linear sweep voltammetry, determined that both

peak current and peak potential varied with scan rate (192).

He obtained the following data for a fresh 10"3 M cisplatin

solution in 1 M NaClO4,



scan rate i E
-p p
10 mVs 5.5 pA -0.15 V vs SCE

100 16.5 -0.25










This and additional data for different scan rates indicate

an irreversible electrode reaction.

These researchers all noticed a minimum in the reduc-

tion wave in the polarograms and voltammograms of cis-

diamminedichloroplatinum(II) but not for the trans form of

the complex (190, 191, 193). This observation led to

mechanistic studies and proposals by Sundholm. When gelatin

(a surface-active agent) was added in increasing amounts, a

continual decrease in the polarographic current before the

minimum was seen for the cis form (193). This inhibition of

the reduction process was also noticed more for the cis

complex when both cis and trans were studied in aged or

hydrolyzed solutions (193).

Sundholm hypothesized that a direct interaction between

the unhydrolyzed cis compound and the electrode had taken

place (194). Inspection of the molecular structure, as seen

in Figure 1, leads to the conclusion that the cis configura-

tion is a dipole with the chloride side of the molecule the

negative end of the dipole. The trans molecule is not a

dipole. A study of the conductivity of cisplatin solutions

in NaCl showed a decrease for about one hour before stable

readings, 2 to 4% lower at 250C and 7 to 9% lower at 370C,

were achieved (195). No such change in conductivity was

observed for the trans molecule. The authors attributed the

decrease to ion-dipole association, with only the cis form

capable of it (195). Polarographically, at the potential

where reduction first begins to occur for cisplatin, the










mercury drop is positively charged since this potential is

positive of the zero point of charge (zpc) (196). The

negative end of a dipole, the chlorides of cisplatin, will

be oriented toward the positive electrode surface. Reduc-

tion of the central platinum ion could take place via

chloride ion bridges (197, 198).

By measuring the polarographic drop time as a function

of potential and plotting the resultant electrocapillary

curves for NaC104 solutions with and without added cis-

platin, Sundholm concluded that cisplatin is adsorbed (199).

The observed negative shift in the maximum of the curve

(electrocapillary maximum or ECM) with cisplatin present is

accounted for by adsorption of the molecule on the drop's

surface (199). Sundholm's studies with linear sweep vol-

tarmetry showed electrode coverages due to adsorption of

1 to 2% of a monolayer by trans Pt(NH3)2C12 (192) compared

to 60-70% for the cis isomer (194). As the potential scan

continues beyond the zpc and the electrode surface becomes

negatively charged, the cisplatin dipole reorients. Without

the chloride bridges to the electrode surface, reduction is

more difficult. The reduction rate and resultant current

decrease and a current minimum is observed due to the

hindered electron transfer to the platinum ion (194).

Most complexes of divalent platinum are square planar,
2
dsp2, with an unhybridized 6pz orbital perpendicular to the

plane of the ligands (200). This configuration holds for










Pt(NH3)2C12 and could provide an alternate means of platinum

reduction through the pz orbital as seen in Figure 4 (201).

This suggestion supports the observation that trans

Pt(NH3)2C12 does not experience the polarographic minimum

the cis compound does due to electrode surface charge

reversal. The lack of a preferential orientation toward the

electrode surface in the trans configuration explains why it

is not affected by surface charge or replacement of chloride

by water hydrolysiss). The reduction inhibition in

hydrolyzed cisplatin solutions observed by Sundholm (193)

could also be explained by lack of preferential orientation

of the pz orbital in a molecule without a dipole.

Investigations by Sundholm on the family of compounds
2-n
[Pt(NH 3)4-nCln]2, n = 0, 1, 2, 3, 4 showed that increasing

the number of chlorides present facilitated the reduction cf

the complex (194). Increased reduction currents for a given

concentration and shifts in E /2 s were the results (194).

Kinetics of the reduction reaction were ascertained by

Sundholm (193) using the method of Kuta and Smoler (202).

By analyzing the time dependence of the current on a single

drop of mercury from a DME, it was determined that the

reduction is kinetically controlled in the case of cispiatin

and diffusion controlled for the trans isomer (193).

Having obtained a reasonable understanding of the

reduction process of the drug (at least on mercury and

platinum electrodes) information pertinent to the stripping











\I

v'I


Cis


H3N ,CI





CNH3


trans





Figure 4. Alignment of Pt(NH3)2C12 toward positively
charged electrode, after Sundholm (194).


I










step can next be examined. In order to successfully com-

plete the ASV experiment, platinum must be oxidized in a

reproducible manner. Formation of a strong, stable complex

following oxidation is one convenient way to shift the

equilibrium between the reduced and oxidized forms of

platinum and to facilitate the stripping step. A variety of

completing agents were considered as to appropriateness for

this purpose.

Molecular orbital treatment of square planar Pt2+

complexes reveals a greater stability with ligands with a

r-orbital system (e.g. CN and SCN ) compared to those

ligands without (e.g. Cl Br H20, NH3) (203-205). This

stability was observed by Watt and Cunningham when they

reported no deposition from 0.01 M aqueous solutions of
2- 2-
Pt(CN) or Pt(SCN)4 before solvent reduction occurred
2-i
(206). The overall order of bond strengths of Pt+- ligand

bonds has been reported as CN > NO2 > NH3 1 OH > SCN >
2-
I > Br > C1 > SO0 > NO3 (173, 206, 207). Stability

constants have been reported for many complexes of Pt2+

Those of some of the stronger complexes are given in

Table 3. Some standard potentials of interest have been

reported and are listed in Table 4.

Previous work on the determination of platinum by

electrochemical methods has been scarce. Using a hanging

mercury drop electrode (HMDE), the decrease in the height of

the cathodic sulfide peak in CSV following analyte addition

was used to determine traces of Pt as well as Hg, Ag and Au















Table 3. Stability Constants of Some Pt2+ Complexes.


Ligand Complex log K Reference


2-
CN Pt(CN)4 41 208
4
2+
methylamine Pt(CH NH) 2+ 40.1 209
2+
ethylamine Pt(C2H5NH2)4 37.0 209
2+
ethylenediamine Pt(NHCCH2CH2NH42+ 36.5 209
2+
NH3 Pt(NH3)4 35.3 209
2-
OH Pt(OH)2 35 207

I PtI- 29.6 208
4
NH3, Cl cis Pt(NH )2C12 29.5 210

NH Cl trans Pt(HH3)2C12 28.4 210
2-
Br PtBr 20.5 208
2-
NO2 Pt2(NO2- 19.6 211
2Cl P(2-6
Cl- PtC1- 16.6 208
4


except ref. 207


Note: Values determined potentiometrically
and 211 which are of unknown methodology.









Table 4. Standard Potentials for Some Platinum Complexes.


Reaction Ligand

Cl Br SCN I- NH3 OH CN


PtL2-4x) + 2e--- Pt + 4LX- 0.75a,b 0.67b 0.40b,d 0.25a,e 0.15e 0.09a,e

0.72c 0.58d 0.34c 0.01c



PtL4-6x) + 4e--- Pt + 6Lx- 0.76ab 0.66b 0.40

0.33b


PtL(4-6x) + 2e---> PtL + 2LX- 0.77a 0.643f 0.468f 0.393f
6

0.758f 0.64a 0.39a,d

0.68e


aReference 185

eReference 215


bReference 212

Reference 173


Reference 213


dReference 214










(216). A polarographic method developed for ore analysis

makes use of catalytic hydrogen evolution as platinum

completed with ethylenediamine is reduced (217). The method

has been extended to the determination of platinum derived

from cisplatin chemotherapy (218, 219). In the determina-

tion of palladium by ASV in 1 M HC1, platinum interfered, as

its stripping peak at +0.7 V vs SCE was only 200 mV anodic

of the palladium peak (220). An ASV technique utilizing

electrochemical reduction of PtCl and PtCl in the
6 4
presence of lead acetate in an HC1 solution onto a graphite

electrode was developed for platinum analysis (221, 222).

The deposit was then electrochemically oxidized to provide

the quantification. Another ASV technique chemically
2- 2+
reduced PtC1 with electrogenerated Sn to form micro-

crystals of platinum on a carbon paste electrode (223). The

amount of deposit was determined from measurements of the

catalytic hydrogen evolution current. Similar work, from

the same group, codeposited platinum and tin to form a 1:1

compound which gave a stripping peak whose magnitude was

proportional to the platinum concentration (224). Another
2-
ASV method utilizing a graphite electrode determined PtC6-

with an excess of Hg2+ added (225). Platinum quantification

resulted from measurement of a peak at +0.2 V vs SCE that

was reportedly derived from mercury oxidation from a

mercury-platinum compound formed during the deposition at

-1.0 V vs SCE.


















CHAPTER IV

THE PLATINUM ELECTRODE IN ACIDIC SOLUTIONS



Obtaining a reproducible response from a limited amount

of platinum deposited on a graphite substrate is necessary

to successfully complete the electroanalytical method

developed in this study. The behavior of the platinum

deposit was expected to be similar in many respects to that

of a platinum metal electrode. The literature pertaining to

the platinum electrode is quite extensive, covering acidic,

basic and neutral solutions, both aqueous and nonaqueous

solvents and including cathodic and anodic conditions. Over

200 papers were consulted in order to better understand what

is still a topic of debate and often conjecture. The

following discussion focuses on the anodic activity of

platinum since anodic processes represent the final and

quantitative step in an ASV experiment. Acidic solutions

represent the majority of published work and are considered

primarily.

As early as 1924, experimenters were aware that the

response from a platinum electrode was dependent on its

prior history and treatment (226). Early studies on the

platinum electrode employed constant current techniques.

Cyclic voltammetry became an additional method of inquiry










starting with the work of Will and Knorr (227). The elec-

trochemical deposition and redissolution of hydrogen at a

platinum surface has become well delineated by this tech-

nique. The state of the platinum surface itself is often

characterized by investigating the behavior of hydrogen

because it acts in a thermodynamically reversible manner and

the equilibrium coverage of the platinum surface with

adsorbed hydrogen atoms is determined by the applied poten-

tial. The electrochemically accessible surface area and the

degree of cleanliness or freedom from impurities are easily

determined in a reproducible fashion by hydrogen atom

adsorption and desorption. The body of information

regarding hydrogen adsorption and desorption is thus useful

in qualifying and quantifying various aspects of the

platinum surface prior to studies under anodic conditions.

The reactions of the platinum surface at anodic potentials

are not nearly as well understood as those at cathodic

potentials. The formation and reduction of an oxide film

has several generally accepted features, but much remains in

debate regarding mechanisms and the exact status of the

platinum surface under differing anodic conditions.

Within the potential window of 0.0 to 1.229 V vs the

reversible hydrogen electrode (RHE), no continuous faradaic

oxidation or reduction occurs at a platinum electrode

surface in pure water at pH 0 because of the thermodynamic

stability of water in this region. Various electrolytes

that are also stable within this potential region are used










to make solutions which are sufficiently electrically

conductive for study. When either the potential is scanned

or a constant current is applied to a platinum electrode, a

variety of electrochemical processes occur as seen in

Figure 5. In a cyclic voltammetric experiment, hydrogen is

adsorbed in two peaks and desorbed in three. Oxygen is

deposited in three poorly defined peaks and a broad region

and is stripped in a single peak. These processes and a

number of other observations regarding the platinum surface

are summarized with references in Table 5.

The first indications of oxygen deposition on a plati-

num surface occur at % +0.8 V in pH = 0 solutions. This is

a generally accepted observation. This potential, where the

oxide film starts to form, shifts cathodically % 60 mV for

each unit increase in pH. The shift was noticed by Kolthoff

and Tanaka in 1954 (228). They also observed, in constant

current experiments, that the same results were obtained

with either a stationary or a rotating platinum electrode,

implying the process was not diffusion controlled. Bold and

Breiter observed the same pH dependence in the initial

oxidation of the surface as well as in oxygen evolution and

in the subsequent cathodic dissolution of the oxide film in

their early cyclic voltammetry experiments (229). They

suggested a two step process for the surface oxidation:

Pt + H20--->PtOH + H+ + e

and PtOH > PtO + H + e .






























Figure 5. Cyclic voltammogram for a typical platinum electrode in pure aqueous
0.5 M H 2SO4 at 25C; sweep rate is 50 mVs-. From Ref. 230.
0.5 M H2S04 at 25WC; sweep rate is 50 miVs .From Ret. 230.
















cc

0w
Oo

Oww
Z >
00
I u L


0


u --
<4U


tII


LI


o my
I- D
o _0


I







o
t L <











0 r IO ONV- I0O---



(EWujD VI) IN3In"D


z
0
I-

0


LI










Table 5. Status of the Platinum Electrode in Aqueous Acid
Solutions.


Potential or
Potential Range,
V vs RHE



<0.05 to 0.08

0.0 to 0.4

0.24




0.4 to 0.8


>0.8


0.85


<1.0


>1.07

1.10 to 1.60


1.25


<1.48 to 1.60


Observation (reference)


Hydrogen evolution (*)

Hydrogen adsorption and desorption (*)

Most cathodic potential for reproducible
reduction of oxide film without
hydrogen adsorption, chronopoten-
tiometry experiment (231)

Double layer region, no faradaric
reactions (*)

Oxygen adsorption, deposition, or film
formation (*)

Anodic limit of total reversibility of
oxide film formation (232)

Surface contaminated by adsorbed
impurities if present (233)

Platinum dissolves somewhat (234)

Indefinitely clean platinum surface,
only oxygen deposited (235)

Anodic limit for platinum on carbon
substrate without damage (236)

No oxygen evolution (*)


Note: Results in 1.0 or 0.5 M H2SO4 or 1.0 M HClO

*generally accepted, published my many authors.










Significant reversibility of these two processes would

explain the observed pH dependence of the anodic as well as

cathodic reactions. More recent researchers have disagreed

with this early interpretation. The mechanism and the

species involved have yet to be conclusively ascertained.

Reversibility of formation of at least a portion of the

oxide layer is indicated in Table 5. The 1973 paper by

Angerstein-Kozlowska, Conway and Sharp reported that total

reversibility of oxygen deposition up to 27% of a monolayer

was found (232). A monolayer in this sense is not neces-

sarily a one atom thick layer but an amount of deposited

atoms whose accumulated charge corresponds to a one atom

thick layer. Atoms may lie in clusters with bare metal in

other areas and the total charge could still define a

monolayer.

Rao, Damjanovic and Bockris had earlier used prere-

duced, oxide-free platinum electrodes in oxygen saturated

solutions to obtain equilibrium coverages at open circuit of

adsorbed oxygen corresponding to 135 pC cm-2 (237). Using a
-2
calculated monolayer coverage of about 500 pC cm-2, this

oxygen coverage also corresponds to 27% of a monolayer.

This coverage was attributed to one oxygen atom sharing two

electrons with four platinum atoms each sharing on the

average all its 0.55 unpaired d electrons (237). The one

oxygen per four platinum atoms was also the formulated

reversible step in oxide film formation proposed by

Angerstein-Kozlowska, Conway and Sharp (232). Similar










results were obtained by Appleby who considered the elec-

trode surface as a close packed plane of platinum atoms

(238). This arrangement can only accommodate a coverage of

30% before nearest neighbor sites must be used for addi-

tional oxygen atoms (238). Oxygen atoms on the platinum

surface beyond 6 = 0.3 must then bind by a more energetic

and less reversible process (238).

At more anodic potentials, the surface film continues

to grow. Reversibility of the anodic film is decreased.

This is generally attributed to a change in the mechanism of

the continuing growth. At coverages of less than a mono-

layer of oxygen, changes in the mechanism of the surface

oxidation process are indicated by several results. The

cyclic voltammograms of Angerstein-Kozlowska, Conway and

Sharp show three anodic current peaks (239). They are due

to the aforementioned process at +0.89 V vs NHE and to two

previously unmentioned processes occurring at +0.94 to

0.95 V and at +1.04 to 1.05 V. These result in a combined

coverage of 78% of a monolayer of oxygen (232). Biegler had

earlier reported two peaks at +0.95 V and +1.07 V (235).

One general interpretation is that of different lattice

sites with associated differences in energy for the binding

processes. A stepwise formation postulated to fit the

overall mechanism is:










Pt4 + H20 -Pt40H + H + e 1 < .12 to .16,

Pt40H + H20,-2Pt2OH + H' + e 81 < 02 < .37 to .40,

PtO2H + H201 2PtOH + H+ + e- 82 < 83 < .78 (232).



This proposed scheme of lattice occupation is not supposed

to represent true species, merely stoichiometric ratios.

However the actual film formation occurs, it is generally

agreed that a monolayer is not reached until the potential

has increased to 1.1 V vs NHE, in the so-called broad

region of oxidation.

In the approximate potential range of +1.1 V to

+1.5 V the anodic current remains fairly constant and film

growth continues. Unlike the submonolayer oxidation, film

formation will continue at these potentials even when the

potential is held constant (239). Several spectroscopic

techniques have supplied corroborating evidence. An ESCA

study by Kim, Winograd and Davis furnished information about

the status of oxygen on a platinum surface held at various

potentials in 1 M HC104 (240). After 3 min at +0.94 V vs

NHE the surface corresponded to 56% Pt, 39% PtOads and
ads
< 5% PtO (240). This agrees well with a second lattice

site occupation of 82 = .37 to .40 with a peak potential

of +0.94 V already mentioned. Similar treatment, but at

+ 1.44 V, resulted in a surface composition of 39% Pt,

39% PtOads and 25% PtO (240). The species PtO2 was not

detected until o +1.8 V (240). Using Auger electron spec-

troscopy Johnson and Heldt discovered that between +0.8 V










and +1.0 V oxygen was weakly adsorbed, but at more anodic

potentials oxygen seemed to be absorbed into the surface

(241). They found about 45 atomic percent oxygen under the

more anodic conditions, but could not distinguish oxidation

states of platinum.

A third technique, ellipsometry, has been used to

measure the film thickness on platinum. It was determined

that the film remained constant up to 1.12 V vs NHE at u 1 A

(242). At 1.22 V, the film was 2.5 A thick and slowly

thickened to 4.7 A at 1.62 V. This concept of oxygen

absorbing into the outer layers of platinum resulting in the

"place exchange" of platinum and oxygen atoms or ions has

been accepted by many. While it is not known whether oxygen

is absorbed, adsorbed, or if a true oxide is formed, the

explanations fit much of what is known.

Considerable evidence has been accumulated to support

the contention that upon subsequent reduction of the anodic

film, the resultant platinum is more active than before. A

review by James covers several proposed explanations (243).

One of the two most tenable theories purports that during

anodic treatment at sufficiently positive potentials,

impurities on the platinum surface desorb and/or oxidize.

This then allows the oxide film to form on a clean surface.

Reduction of the oxide film results in a clean platinum

surface free of impurities with an increased activity due to

an increased number of reaction sites. The other defensible

theory contends that upon reduction, a form of platinized










platinum results from either the rearrangement of platinum

atoms after the oxide film is decomposed or the redeposition

of platinum dissolved in the anodic treatment. The paper by

Rand and Woods contains a number of references to both

schools of thought (234). Numerous other theories exist.

The relative importance of the two most accepted theories is

debated, but both have considerable support and are

important to this study because maximum activity and repro-

ducibility are necessary.

In partial confirmation of the latter idea, Biegler

found platinum in solution after repeated oxidation and

reduction (235). Separate groups calculated platinum

dissolution rates of 5 ug cm-2 cycle-I when cyclic voltammo-
-2
grams extended to +1.4 V vs NHE (234) and 4.5 ug cm

cycle-1 when carbon supported platinum was cycled to +1.25 V

(236). Untereker and Bruckenstein measured platinum dis-

solving in cycles between -0.25 V and +1.55 V by means of a

rotating ring disk electrode (RRDE) (244). They made an

excellent case for the notion that platinum actually dis-

solved and redeposited rather than merely rearranged on the

surface during the oxide formation-reduction sequence. The

finding that roughening of the surface decreased with

increasing rotation rates supports the redeposition hypothe-

sis (244). Two other RRDE studies found that all anodic

current up to +1.6 V was due to disk surface reactions since

no detectable current was obtained at the ring until oxygen

evolved at +1.7 V at the disk (245, 246). This result











contradicts the platinum dissolution hypothesis. The

suggestion that platinum atoms rearrange to accommodate

subsurface oxide films of greater than a monolayer has other

proponents (232, 247-249). In any case, an increase in

activity resulting from an increase in the number of plati-

num atoms accessible to the solution is obtained and should

increase any signal generated during the course of this

work. The activation of a platinum deposit by oxide film

formation and reduction should play an important role in

procedural optimization.

Adsorption of organic compounds and inorganic ions at

the platinum surface is another mechanism of decreasing the

electrochemical activity. Breiter studied halide adsorption

in 1 M HC104 at platinum electrodes and also made reference

to a considerable number of earlier papers on the subject

(250). He found that at potentials anodic of the zero point

of charge (zpc), 0.2 V vs NHE in 1 1! acid, a monolayer of

chloride adsorbed when the solution concentration was
-2
102 M. Chloride noticeably altered the cyclic voltammo-

grams obtained by Breiter at solution concentrations as low
-6
as 106 M (250). Chloride displaced the oxide film forma-
-4
tion reaction anodically at 104 M and decreased the total

amount of oxide formed and subsequently reduced (250).

A paper examining the adsorption of Cs and SO- by

radiochemical methods found identical amounts of both ions

adsorbed at +0.17 V in 0.1 M H2SO4 (251). Cesium ion

adsorption was minimal anodic of this potential, the zpc,










and sulfate ion adsorption increased until oxide film

formation commenced (251). The potentials of maximum

sulfate ion adsorption were +0.75, +0.68 and +0.60 V in 0.1,

0.01 and 0.001 M H2SO4 respectively (251). This illustrates

the weak adsorption of sulfate ion since the oxide film

formation was not hindered. It also demonstrates the pH

dependence of the initiation of oxide film formation. In a

paper by Conway, Angerstein-Kozlowska, Sharp and Criddie

detailing a method for preparing high purity water, platinum

behavior in sulfuric acid solutions made with this water was

used as a check on solution purity (230). Organic adsorp-

tion was considered the primary problem encountered. One of

their criteria for solution purity was no change in the

hydrogen adsorption and desorption peaks after cycling to

only +0.75 V, thus not allowing oxygen to deposit and clean

the surface. Phosphate, sulfate and perchlorate ions are

only slightly adsorbed on platinum, explaining why the

cyclic voltamniograms obtained in acid solutions of these

anions are so similar (252). The proper electrolyte and

sufficient solution purification play a significant role in

the collection of optimal data using a platinum electrode.

A clean and active platinum surface is necessary for

optimum performance in an electrochemical system. It is

recognized that several anodic-cathcdic treatments will

generally result in a clean surface with increased activity.

Maximum activation of a fresh platinum electrode may take a

considerable number of such cycles. Various authors using











cyclic voltammetry have suggested a myriad of possibilities.

Inclusion or avoidance of hydrogen adsorption or evolution

and oxygen evolution, faster cathodic than anodic scans, and

varying scanning speeds, times and number of scans have all

been suggested. The common element is the deposition and

removal of oxygen. Approximate calculations by Gilman point

out the necessity of maintaining clean conditions. Con-

sidering diffusion controlled adsorption and average values
-6
for adsorbable impurities of about 10- M total concentra-

tion from reagents and solvent, a platinum surface can

become covered to the extent of 1% in about 50 s if the

solution is not stirred (252). Only 2 s are needed if the

solution is stirred (252). He suggests that either the

electrolyte solution be shown to be sufficiently clean or

the platinum electrode should be cleaned before each use.

Biegler reproducibly activated a platinum electrode

by repetitive cyclic voltammetric scans in 1 M H2SO4 and

prepared tor each new experiment by cleaning the surface by

pulsing the potential from cathodic to anodic potentials

(253). Gilman has published a number of papers on pulsing

as a means of attaining reproducible surface states. His

first two papers on the multipulse potentiodynamic (MPP)

method detailed the preparation of platinum surfaces for CO

adsorption studies (254, 255). He found he could attain

reproducible active areas and maintain the cleanliness for

> 100 s (253). He also introduced the method for the study

of chloride and phosphate adsorption and desorption (256),










hydrogen adsorption (257) and platinum surface oxidation

(258). In each case a series of steps ensured the

reproducible surface conditions necessary for comparative

studies.

The method of Gilman used an initial applied potential

of +1.8 V vs NHE with stirring of the solution to desorb

and/or oxidize adsorbed impurities and evolve oxygen gas to

help sweep away any products. The oxide film also began to

form in this 2 s step. At the less anodic potentials of

+1.2 V or +1.5 V, the oxide layer was established and

maintained and molecular oxygen no longer evolved. During

30 s of stirring, products were removed and during 90 s

without stirring, solution quiescence and double layer

equilibrium conditions were reestablished. No current was

observed to flow after the first few seconds. A final

potential step to +0.4 V for 10 s reduced the oxide layer

(mostly within the first 0.1 s) and produced a clean plati-

num surface. A subsequent step or steps were sometimes used

to study various species and procedures. Initiating an

anodic scan to study the formation of the oxide film is an

example of one of Gilman's followup procedures (258).

This discussion has highlighted some of the work done

with the platinum electrode in acidic solutions under anodic

conditions. The results which seemed relevant to the

present investigation have been emphasized. A great deal of

material on mechanisms has been omitted as not directly

applicable. Additional information can be obtained from







62



several reviews (239, 252, 259), a trilogy by Conway and

colleagues (232, 249, 260) and a recent dissertation which

studied single crystal planes of platinum (261).


















CHAPTER V

ANODIC STRIPPING VOLTANMETRY



Brief Introduction and History

Anodic stripping voltammetry (ASV) is the most commonly

used of a family of electrcanalytical techniques collec-

tively known as stripping analysis (SA). The features

common to all members are the preconcentration step and the

stripping step. Preconcentration is a common analytical

procedure. It is useful in increasing the response due to

hde analyte while ideally not affecting any response from

non-analytes. In other words, the goal is to increase the

signal but not the noise. Risks of sample loss and impurity

introduction, amongst others, accompany conventional sample

concentration methods. In stripping analysis, the sample

preconcentration is accomplished by electrochemical means at

a suitable electrode within the sample itself. The strip-

ping step or analytical measurement is usually carried out

within the same solution, minimizing the aforementioned

risks.

Cathodic stripping voltammetry (CSV) is comprised of an

oxidative preconcentration and a reductive cathodicc)

stripping. This technique has been applied primarily to

inorganic anions which form sparingly soluble salts or











complexes with mercury ions. The mercury electrode used is

oxidized in the preliminary step to provide needed mercury

ions. Examples of analytes include Cl-, Br I S2-
2- 2- 2- 2-
CrO4 WO MoO and VO (262) as well as oxalate, suc-

cinate and dithiozonate (263). Organic molecules and ions,

in drug analysis, have been determined in the same manner

(264).

Anodic stripping analysis comprises the other branch of

stripping analysis. In this method, a reductive preconcen-

tration step and an oxidative anodicc) stripping step are

employed. The most generally used form of anodic stripping

analysis determines metal ions by reducing them at a mercury

electrode to form an amalgam and subsequently stripping the

accumulated amalgam to complete the experiment (17). Solid

electrodes such as gold, platinum and several forms of

carbon or graphite are also occasionally used (18). Analy-

sis of molecules is possible by adsorpticn of the molecule

or an electrochemical reaction product of the molecule onto

an electrode surface followed by anodic stripping of the

concentrated material (264, 265). For further information

on stripping analysis, the reader is referred to several

introductory articles, review articles and books (11, 12,

14, 16, 17, 19, 21, 262, 263, 266, 267).

Stripping analysis originated with G.G. Grower in 1917

(268). He determined the thickness of a tin coating on

copper wire by electrochemically stripping it off. The

coated wire was the anode and a platinum cathode completed











the cell. A gas coulometer and a voltage source were in

series with the cell, while tin was being removed from

copper wire, the potential of the wire compared to a pure

tin reference wire was monitored. The first change in

monitored potential was interpreted as the completion of the

removal of tin. The second change was equated with the last

of tin-copper alloy being stripped. The coulometer reading

was related to the thickness of the tin coating. Similar

film or coating thickness experiments were reported by Evans

and Bannister in 1929, who determined AgI films on silver

(269). Campbell and Thomas in 1939 reduced oxide films on

copper to discern coating thicknesses from a few monolayers

to about 1000 A (270). Ogarev in 1946 (271) and Francis in

1948 (272) analyzed coatings on a variety of metals.

The first reported use of a deposition step prior to

stripping is credited to Zbinden in 1931 (273). Attempting

to determine copper electrogravimetrically, he found the

weight of the deposit too small to be accurately measured.

He reversed the current flow and stripped the deposited

copper from the platinum substrate. By manually adjusting

the potential to obtain an approximately constant current,

the stripping continued until the current suddenly dropped.

Measuring the elapsed time of current flow enabled the

amount of copper deposited to be calculated. Zbinden

measured deposits of 6.36 pg of copper with errors as low as

3.3% (273). In 1938 Zakhar'evskii performed similar experi-

ments to determine lead and copper (274). By following the











current with an ammeter, he was able to measure as little as

10 pg of lead and 20 pg of copper. Elema in 1947

followed Zbinden's method in an attempt to routinely analyze

for copper (275). He found he could determine 10-100 ug of

copper by a technique that even poorly trained personnel

could follow.

It wasn't until the 1950s that ASV was studied

seriously by several groups of researchers. Rogers and

co-workers investigated silver on platinum electrodes in

1952 (276) and later used a mercury film on platinum to

study silver (277) and lead (278). Hickling and Maxwell in

1955 examined mercury pool electrodes and the dissolution of

amalgams formed by constant potential stripping (279). With

Shennan, they studied Cu, Pb, Tl, Cd, Zn and Mn by nearly

exhaustive electrolysis into the mercury pool (280). A

small mercury pool contained in a well at the end of a

"U-shaped" glass tube was used by Kalvoda in oscillopolaro-

graphic stripping (281). Mercury drop electrodes were

employed by Barker in 1952 (282), by deMars and Shain (283)

and Nikelly and Cooke (284) in 1957 and by Kemula and Kublik

in 1958 (285). The latter authors with Rakowska, studied

impurities in uranium salts and were able to measure

5 x 1010 M cadmium (286). A review in 1961 covers these

and other early anodic stripping experiments and also

discusses several types of mercury electrodes (287).

The large mercury pool electrode used by Cooke had a

slow rate of electro-oxidation (288). Nikelly and Cooke










noted that platinum, gold, silver amalgam and mercury on

platinum electrodes required lengthy conditioning to give

reliable results (284). It was the success with mercury

electrodes that permitted anodic stripping to become an

established analytical technique (11). Theoretical treat-

ments of the hanging mercury drop electrode (HMDE) and the

thin film mercury electrode (TFME) were derived in 1960

(289-291). Earlier theoretical attempts were of limited

utility, but laid the groundwork (292). As these two terms

of mercury electrodes gained acceptance and improvements in

electronics and instrumentation came about, a more wide-

spread use of the technique arose in the early 1970s (11).



The Anodic Stripping Experiment

A typical anodic stripping experiment is comprised of a

series of separate steps in a fixed sequence. Care must be

taken in performing each stage because of the low concentra-

tions encountered. These individual tasks must be reproduc-

ibly done for optimum results to be attained. Unintentional

variations in any procedure could result in loss of accuracy

and/or precision.

The first step is sample preparation. This may be done

by any conventional method, depending on the sample, bearing

in mind the final form will be a solution with several

necessary properties. The analyte should generally not be

more concentrated than 1 ppm (17). A variety of analytical











techniques, including polarography, are adequate for solu-

tions greater than this concentration. As in other electro-

chemical techniques, a supporting electrolyte is added for

several reasons. The movement or mass transfer of molecules

and ions in a solution may occur by several means. Diffu-

sion, mass transfer resulting from a concentration differ-

ence between two regions, and convection, resulting from

mechanical stirring or density differences, affect both

molecules and ions. Migration of charged particles due to

an applied electric field is a third method of mass trans-

fer. An indifferent or supporting electrolyte is added in

large excess to carry the majority of any current. Analytes

thus move by convection (if any) and diffusion only. By

decreasing the contribution of migration to total mass

transfer of the analytes, mathematical treatments are

simplified, and reproducibility is improved. Supporting

electrolytes also ensure a low solution resistance. The

supporting electrolyte may originate in the sample or sample

preparation procedure or may be added prior to analysis.

Purity of any added reagent is important, as the analytical

usefulness of the overall method is often limited by impuri-

ties in the supporting electrolyte (20).

The suitably prepared sample is added to the electro-

chemical cell. Typical cell volumes range from 100-200 mL

to as low as 5 or 6 uL in some special cases (293-295). The

necessary electrodes are introduced, if not already incor-

porated into the cell. Electroactive oxygen is removed










prior to analysis, commonly by bubbling purified nitrogen or

helium through the solution. Oxygen removal is necessitated

by the fact that it can be electrochemically reduced in two

steps which will alter the pH of the solution and generate

new chemical agents (296). The reductions shown below

produce a current flow over most of the negative potential

range (297).



In acid solution: 02 + 2H' + 2e--->H202 [3]



H 02 + 2H+ + 2e--->2H20 [4]



In neutral or

base solution: 02 + 2H20 + 2e--> H202 + 20H [5]



1202 + 2e 20H [6]



Hydrogen peroxide (H202) is a strong oxidizing agent and

could react with any analytes present. Hydroxide ion (OH

when formed may precipitate many metals present. The

removal of H+ or generation of OH will increase the pH

unless the solution is adequately buffered. Oxygen may

oxidize deposited metals and thus decrease the analytical

signal.

After ten to fifteen minutes of inert gas flow, oxygen

is removed and the preconcentration or deposition is per-

formed. Deposition is nearly always accomplished by










using a constant potential. This method enables a degree of

selectivity in the overall analysis since more noble metals

may be reduced in the presence of more difficultly reduced

metals. To increase the availability of analyte at the

working electrode surface, to increase the deposition rate,

convection is employed. This is achieved by either rotating

the working electrode, rotating the cell itself or using a

magnetic stirring bar and motor. A more fully detailed

discussion of the deposition step will be given later.

When the deposition step has been completed, a rest

period follows. A potential is usually still applied (12,

21, 263), although it may be at a more anodic value than

during deposition, in the case of ASV (263). The stirring

is stopped to allow the solution to become quiescent (12,

15, 21). If deposition continues, the reduction current

rapidly decays to the diffusion limited value (12, 21). At

least thirty seconds is chosen (14, 298) to ensure that

convection has ceased (298). In cases where mercury

electrodes are used, this rest period allows the

concentration distribution in the amalgam to become

sufficiently uniform (12, 15, 18, 19, 21). It has been

shown that two seconds is enough time for the thin film

mercury electrode (TFME) (299) and thirty seconds is

sufficient for the hanging mercury drop electrode (HMDE)

(291, 298). This rest time is especially important if

theoretical studies are undertaken, as stripping theories

require a uniform amalgam concentration distribution.










Although any deposition that may occur during this time

contributes only a small amount of material to the total

deposit (19), the rest or quiescent period should be repro-

duced accurately.

In comparisons of available stripping techniques in

1957, Nikelly and Cooke (284) and deMars and Shain (283)

concluded that the best results were obtained when stripping

was done by linear potential sweep chronoamperometry, also

called linear sweep voltammetry (LSV). This technique

became the procedure of choice for many researchers because

of its instrumental and methodological simplicity (12). The

name anodic stripping voltammetry was applied to the overall

technique. Since LSV is merely polarography at other than a

DME and simple polarographic instrumentation is widely

available, this remains a common means of completing the

anodic stripping experiment.

With LSV and related methods, a current peak should

occur in the current-voltage curve for each substance

deposited and subsequently stripped. The potential at which

this peak is centered generally allows the identity of the

analyte to be determined. The height of the peak above the

baseline (current) and the area under the peak (charge) are

related to the original solution concentration of the

analyte. A more detailed discussion of various stripping

procedures and data handling methods will be presented

later.










The Deposition Step

The deposition or preconcentration step begins the

anodic stripping experiment. Sample preparation will have

been done according to the demands of the sample and experi-

ment. A supporting electrolyte will usually be present

along with the sample. The specifics of the deposition

procedure are dependent on the analyte, which also influ-

ences the choice of the working electrode. Prior knowledge

of the nature of the analyte(s) is required in order to

effect a useful deposition.

The working electrode most commonly used and most

practical for routine determinations is the mercury elec-

trode (17). This may take the form of a mercury drop as in

the HMDE or in the sessile MDE which sits upon a support.

Mercury may be used as a thin film supported on a substrate

in the TFME. The HMDE was reportedly used in

electrochemistry as far back as 1939 by Antweiler (300). A

review discusses many applications of the HMDE in

electroanalytical procedures, including stripping analysis

(20).

The TFME was introduced to stripping analysis by

Gardiner and Rogers in 1953 (277) and Marple and Rogers in

1954 (278). The reasoning leading to this particular

geometry was the desire to decrease the volume of the

working electrode without substantially decreasing the

surface area. This would lead to a greater preconcentration

under identical conditions. The TFME has been produced










using many metals such as Au, Pt, Ag and Ni as substrates.

Several forms of carbon have also been used since this

substrate was introduced in 1965 (301). The thin mercury

film can either be formed separately before the analysis is

begun or as part of the deposition step as shown by Florence

in 1970 (302). With mercury film thicknesses between 1 and

100 um when preformed (12, 18) and less than 10 nm when

formed in situ (15), the high surface area to volume ratio

desired and improved preconcentration have been obtained.

Problems can occur with mercury electrodes because of

interactions with the supporting substrate (15). The

substrate can dissolve in the mercury or form intermetallic

compounds with mercury or with the analytes, leading to

additional stripping peaks. Intermetallic compounds can

also form between metals concentrated at or within the

mercury (15, 17, 20, 21). Intermetallic compounds present

more of a problem with liquid mercury electrodes than with

the solid electrodes (18). Because of the greater concen-

trations in the TFME, intermetallic compounds are a more

severe problem with this type of mercury electrode. Com-

pounds of Au-Mn and Au-Sn (21), Ag-Cd, Ag-Zn, Au-Cd, Au-Zn,

Cu-Cd, Cu-Mn, Cu-Ni, Cu-Zn, Fe-Mn, Mn-Ni, Ni-Sb, Ni-Sn and

Ni-Zn (20,21) and Pt-Sb, Pt-Sn and Pt-Zn (20, 21, 285) have

been documented. The reproducible and easy to make HMDE and

the sensitive TFME are nevertheless popular in stripping

analyses (19).










However, the use of mercury demands not only that the

analyte be capable of forming an amalgam with or a film or

mercury, but it must also be stripped before the electrode

material itself oxidizes. This limits the mercury elec-

trodes to studies of the less noble metals and precludes

their use in this investigation. Besides the generally used

mercury electrodes, Au, Ag and Pt among other metals have

served as solid electrodes. Various forms of graphite or

carbon have seen application as both supports for the TFME

and as electrodes directly. The use of carbon as a working

electrode has extended the range of anodic potentials

accessible and has enabled the more noble metals to be

determined as a result. Determinations of Ag, As, Au, Ba,

Bi, Cd, Cu, Ga, Ge, Hg, In, K, Mn, Ni, Pb, Pt, Rh, Sb, Ti

ana Zn (18) as well as Co, Fe and Sn (303), Ir (304), Pa

(305, 306) Ta (306) and Te (307, 308) have been reported

using carbon electrodes. Several sources provide informa-

tion for specific analyses (12, 21, 262, 263, 309, 310).

The ideal working electrode has a reproducible surface

composition, surface area and/or volume, and a low residual

or background current (17). Mercury has been used exten-

sively in electroanalysis because it has these properties.

Solid electrodes, including carbon, can compete with mercury

in many ways and offer advantages in several areas. The

first suggestion for carbon as an electrode material in

voltammetric work by Rogers and Lord in 1952 (311) led to

the experiments of Gaylor, Elving and Conrad with carbon










rods brushed with an insulating layer of melted wax before

use (312). A fresh surface (akin to a new mercury drop trom

a DME) could be obtained by breaking off the end of the rod.

A high residual current was present and the investigators

decided the carbon needed to be totally impregnated with

wax. The filling of all pores eliminated oxygen interfer-

ence and memory effects. The first wa:: impregnated graphite

(WIG) electrode was reported in 1954 by Lord and Rogers

(313). The WIG electrode was successfully used by Gaylor,

Conrad and Landeri in 1957 (314). A voltage range of -1.3 V

to +1.3 V vs SCE was reported in 1959 by Horris and Schempf

(315).

During this period, Adams was attempting to develop a

non-metallic analog to the DME (316). He was experimenting

with suspensions of carbon particles in fluids that were

sufficiently immiscible in water to prevent the resultant

mixture from dissolving in the analytical solution. The

results were far from satisfactory and he concluded it would

be better to pack the "carbon paste" into a holder and use

it as a stationary or a rotating electrode. His initial

report on the carbon paste electrode was in 1958 (317).

Olson and Adams used carbon paste electrodes in ASV in 1963

(303) as did Jacobs (318).

These early investigations and successes with carbon as

a working electrode spurred an interest in carbon and

additional formulations were proposed and tested. Carbon in

packed beds, in silicone rubber, mixed with polypropylene










and polyethylene, in Teflon, in Kel-F polymer wax and in

various pure states such as low temperature isotropic,

pyrolytic, glassy, reticulated vitreous, and cloth have all

been investigated for general and specific applications

(319). Epoxy bonded graphite has been used as well (320).

Many anodic stripping investigations using carbon

electrodes have involved the analysis of mercury. Useful

stripping experiments were performed as early as 1964 (321)

with following works (322, 323) leading to results that

showed a thiocyanate medium best for mercury analysis (324).

Nickel also presented special problems because it could not

be determined with mercury as the amalgam stripping was

obscured by mercury oxidation (325) and it partially com-

bined with platinum (326). Graphite proved a suitable

electrode material (303, 327). Gold (318, 328) and silver

(318) were successfully analyzed by ASV using carbon elec-

trodes. In the mid 1970s, a series of papers described the

use of co-deposited mercury in the stripping analysis of

several noble metals using graphite electrodes (225, 304,

306-308, 329-331). An interesting aspect of these studies

was that the mercury stripped off prior to the more noble

analytes and often the amount of mercury stripped was used

to quantitate the noble metal.

In 1962, the first reported use of pyrolytic carbon as

an electrode material by Laitinen and Rhodes (332) and the

initial manufacturing of gas impermeable glassy carbon

occurred (333). Miller and Zittel used pyrolytic carbon for










aqueous voltammetry in 1963 (334) and reported on the

general usage of glassy carbon in 1965 (335). Yoshimori and

collaborators determined gold by anodic stripping coulometry

using glassy carbon (336). Florence chose glassy carbon as

the supporting material for his in situ formed TFME in 1970

because of its hardness, good electrical conductivity and

chemical inertness (302). A recent review of glassy carbon

as an electrode material stated that glassy carbon, along

with pyrolytic carbon, possesses the largest useful poten-

tial range of any presently used form of carbon (337). The

kinetics of a number of thoroughly studied redox systems

were investigated at glassy carbon (GC), platinum (Pt),

carbon paste (CP) and wax inpregnated graphite (WIG) by

Taylor and Humffray (338). The rate constants they obtained

were in the order kpt > kGC > kWIG > kCp. A study by Panzer

and Elving reported on the general characteristics of glassy

and pyrolytic carbon in aqueous and nonaqueous solutions

(339). Glassy carbon appears superior to pyrolytic carbon

and is an excellent choice for electroanalytical electrodes.

No other material can compete in terms of hardness, chemical

inertness, low porosity, electrical conductivity, low

residual current and high overpotential for oxygen and

hydrogen evolution (337, 340, 341).

Problems do occur with glassy carbon, some common to

solid electrodes in general and some unique to carbon. As

is often observed with solid electrodes, it is difficult to

obtain good, homogeneous, analyte deposits (17). Several










monolayers may form if sufficient material is deposited and

this can lead to multiple peaks for a single analyte (337).

The extra peak(s) have been attributed to the stripping of

the individual monolayers, the first one deposited being

bound more strongly to the electrode surface than subsequent

layers are bound to it and to each other (342, 343). This

problem can limit an anodic stripping analysis to only one

analyte in some cases (17, 21). A problem unique to carbon

surfaces is that of the formation by either chemical or

electrochemical reactions of functional groups which contain

oxygen. These groups can be repeatedly oxidized and reduced

and can often only be removed physically (337). However,

abrading a hard surface or wiping a softer one is often

enough to remove the unwanted effects.

In the anodic stripping experiment, preconcentration is

the result of electrochemical reduction of metallic ions at

the electrode surface. The resultant metal forms a film on

the surface of solid electrodes. Ions in the bulk of the

solution replace the ions at the electrode surface that have

been reduced and the process continues. The ion transport

process may be either diffusion, in unstirred solutions, or

convection, as is commonly done. Constant potential elec-

trolysis is the predominant means of obtaining the required

deposition. Constant current electrolysis can be used

instead, but without potential control it lacks the selec-

tivity often necessary. The reproducibility also suffers











when constant current electrolysis is used for deposition

and for these reasons, it is seldom employed.

Knowledge of the reduction behavior of the analyte(s)

at the particular working electrode used must be obtained

prior to the ASV experiment. The deposition potential,

Edep, is chosen on the basis of polarographic E1/2 values or

from peak values from cyclic voltammetry. Tabulated results

of previous investigations can also prove beneficial. In

any case, Edep must not be less than 120/n mV (n = number of

electrons involved in the reduction) negative of the poten-

tial indicated by other techniques (14). Deposition poten-

tial values recommended are usually 300 to 400 mV negative

of the E1/2, Ep or reduction potential determined for the

analyte (21). This allows for an adequate safety margin to

ensure satisfactory deposition. Care must be taken not to

exceed the decomposition or reduction potential of any

interfering substances. Generally it is not possible to use

a deposition potential more negative than -1.5 V vs. SCE in

neutral solution or about -1.2 to -1.0 V in acidic solution

because of hydrogen evolution (21).

By proper choice of the deposition potential, the

reduction may be limited by mass transport processes and not

by the charge transfer reaction rates. If dep is not
dep
sufficiently cathodic, the kinetics of the charge transfer

reaction can limit the deposition rate. If electron transfer

is slower than mass transport, the process is said to be

irreversible under the given conditions. For the case where











electron transfer is fast enough to maintain a reaction rate

limited only by the mass transport process, the reaction is

said to be reversible. Reversibility or irreversibility and

reaction kinetics are often studied on their own merits.

More detailed discussions and treatments of the fundamentals

of electrode reactions and kinetics may be found elsewhere

(12, 15, 18, 19, 210, 344-349).

In a system controlled by mass transport processes an

increase in the reaction rate may be achieved by increasing

the mass transport, e.g. by stirring the solution or

rotating the electrode. An increase in the stirring or

rotation rate will give an increased flux of material to the

electrode surface and will yield an increased deposition

rate. This is one way to decrease deposition times and

increase the extent of preconcentration within given physi-

cal and kinetic limitations. The Levich equation for

rotating electrodes is useful to show how improvements in

the deposition may be accomplished (19, 266, 350).



i(t) = 0.62 n FAD2/3 1/2 v- /6 C(t) [7]



where i(t) is the deposition current as a function of time,

F is the Faraday constant, A is the electrode area, D is the

diffusion coefficient of the analyte, w is the angular

velocity, v is the kinematic viscosity of the solution and

C(t) is the solution concentration of the analyte. More

complete discussions of rotating electrodes and











electrochemistry in moving solutions can be found in

other sources (350-354).

In order to increase the reduction current, several

variables can be altered. As mentioned, increasing the

rotational speed of the electrode will have the desired

effect. The viscosity of the solution is difficult to

change so as to significantly alter the reduction current

because of the one sixth power dependence as well as the

narrow ranges of values accessible in a given analysis (19).

It can also be inferred that larger electrode surfaces will

increase the amount of the deposit. This modification is

limited practically because of the need for small surface

areas for the low concentrations encountered. A small

electrode is needed to minimize the residual current which

is proportional to the electrode area and to ensure coverage

of the surface by the analyte to avoid overvoltage and

undervoltage effects (14).

In the Levich equation the reduction current and the

analyte concentration are both given as functions of time.

Because as the deposition proceeds the bulk analyte concen-

tration decreases, the reduction current will necessarily

decrease as well. This process will follow Lingane's

equations (355),



i(t) = i0e [8]



where i is the current at the start of the deposition and
0










AD
k VS 9]


where V is the solution volume, S is the diffusion layer

thickness and A and D are as before (353). Although these

equations pertain strictly to diffusion limited experiments

only, the same general conclusions can be drawn for stirred

solutions. To increase the deposition current at any given

time, increasing the electrode area, decreasing the solution

volume or decreasing the diffusion layer thickness will have

the desired effect. The results are similar to those of

Levich if one realizes that stirring the solution results in

a decrease of the diffusion layer thickness.

In ASV experiments, complete deposition is not neces-

sary (12, 18, 277, 278, 288). Commonly only 2-3% of the

total amount available will be reduced during the deposition

step (12). As a result, the deposition current remains

approximately constant and the amount deposited can be

calculated from (15)



i t
number moles = dep dep [10]
nF [0]


This result shows that the stripping process will be

enhanced by longer deposition times. Long times are not

always advantageous as eventually the concentration will

begin to decrease as given by Lingane. The primary disad-

vantage of long deposition times is long total analysis

times and therefore decreased experimental throughput.











Because the deposition is not exhaustive yet the

technique was developed to be quantitative, reproducibility

in this procedure is critical. Every variable that can be

controlled, must be. The stirring rate or rotation speed,

the geometric placement of the stirring bar or rotating

electrode and other electrodes and the solution volume must

all be carefully reproduced (14). The deposition time must

be accurately duplicated from experiment to experiment.

Times of the order of 2 to 3 minutes up to 45 minutes or an

hour are used depending on analyte concentration, cell

geometry, etc. Generally speaking, every facet of the

deposition should be regulated carefully to attain maximum

precision.



The Stripping Step

The actual electroanalysis occurs during the stripping

phase of the experiment. Historically, the early stripping

analyses used the simplest available stripping method, i.e.,

controlled current. Either a carefully controlled constant

current with the stripping time as an indication of the

amount of analyte, or any applied current in conjunction

with a coulometer was used in these experiments (268-275).

As more sophisticated methods were developed for polaro-

graphy and electrochemical instrumentation became more

widely available, an increased variety of stripping methods

also appeared. These include the classic electrochemical










procedures as well as variations of the polarographic

technique.

Methods that have found application include those with

controlled current, constant potential, linearly changing

potential as well as alternating current and pulse methods.

Among the electrochemical techniques discussed in the

general literature of stripping analysis are chronoampero-

metry, chronopotentionetry and sine wave voltarmetry (12,

20, 21), pulse, coulostatic and oscillographic stripping

(12, 21), coulometry with both constant potential aLr

controlled current (12, 20), differential pulse voltammetry

(11, 34i) and square wave voltammetry (11, 12, 21). Chemi-

cal stripping (21) phase sensitive alternating current

detection (11) and fundamental as well as second harmonic

alternating current methods (341) are used occasionally.

This listing is by no means complete, but illustrates the

diversity present.

The classic techniques of electrochemistry were the

first to be applied to stripping analysis because the

theoretical aspects were well founded and suitable instru-

mentation was readily available. As mentioned, coulometry

was first used, essentially without instruments. Batteries,

gas coulometers and ammeters were all that was available and

all that was necessary. With the advent of reliable voltage

and current sources, oscilloscopes and recorders, the field

of electrochemistry improve.










Chronoamperometry is one of the earliest electrc-

analytical techniques developed. In this method, the

potential of a working electrode in an unstirred solution is

changed from a value where the reaction of interest does not

occur to a value where it occurs at the limiting rate.

Following this potential step, a large current begins to

flow, decaying with time as the amount of available electro-

active material at the electrode surface decreases.

Cottrell, in 1902, derived the equation for this process

(356),



nFAD /2 C
i(t) = *"_-"- 1111
i(t) 1/2 1/2 11]
1 t



(all symbols have been previously defined).

This equation was experimentally verified forty years

later by Laitinen and Kolthoff (357) and by Laitinen (358).

The use of chronoamperometry in stripping analysis was

reported in 1957 (359). Because of the preconcentration

step, the Cottrell equation was not used directly to quan-

titate the analyte, although the experimental current did

follow the expected decay. A stripping theory for the

chronoamperometric method was developed in the same paper.

Shain improved the method by analyzing with both a long and

short deposition time to compensate for the charging current

present in both experiments (360). He determined iodine to
4 x -8 M with a silver electrode in a CSV experiment
4 x 10 M with a silver electrode in a CSV experiment










(360). He also derived an equation for use with a HMDE

(361). Chronoamperometry has found few applications in

stripping analysis because of the superiority of other

methods. A recent paper by Brainina and Vdovina contains an

equation for the stripping of metal films from solid elec-

trodes by chronoamperometry (362).

Chronopctentiometry is similar to chronoamperometry in

that a diffusion limited process in an unstirred solution is

used. However the potential step is replaced by a current

step trom zero applied current to some fixed, finite value

in the milliamp range. The potential of the working elec-

trode versus a suitable reference electrode is monitored as

a function of time. As the current is applied, the poten-

tial rapidly assumes a value indicative of the most easily

reduced (or oxidized) species. The potential slowly changes

as the electroactive species reacts at a constant rate ano

the concentration ratio of the product to reactant changes.

When the analyte has completely reacted at the electrode

surface and the amount of reactant replenished by diffusion

from the bulk solution is insufficient to maintain the

applied current, the potential again rapidly changes. This

time a value associated with the next most easily reduced

(or oxidized) species is attained.

The time required for the electrolysis reaction, the

time needed for the concentration of the particular reactant

at the electrode surface to drop to zero, is called the

transition time. The term was introduced, although not in










the presently accepted sense, by Butler and Armstrong in

1933 (363, 364). An equation which relates this transition

time to the applied current and the concentration of the

species undergoing the electrode reaction was developed by

Sand in 1901 and bears his name (365, 366),



1/2 = nFAD1/2 1/2 C [12]
i 2


T = transition time

i = applied current.



Sand was studying diffusion coefficients, continuing work by

Weber who had been using constant current electrolysis to

study the diffusion process (367). Karaoglanoff developed

an equation which defined the potential vu time curve in

1906 (368). Rosebrugh and Miller in 1910 reported on a

number of applied current functions and derived equations to

describe the results of these processes (369).

The analytical value of chronopotentiometry was little

appreciated until instrumentation improved and Gierst and

Juliard demonstrated its utility, reviving the technique in

the early 1950s (370, 371). Berzins and Delahay in 1953

derived an equation for the chronopotentiometric stripping

of a deposit accumulated by constant current electrolysis

(372). Delahay and coworkers reported an early use of

anodic stripping chronopoteiitiometry and derived appropriate

theoretical expressions in 1957 (359). Chronopotentiometry










has found many applications in stripping analysis, consider-

ably more than chronoamperometry. It is limited to solu-

tions in the millimolar range and is used primarily to study

reaction kinetics. Excellent reviews covering chronopoten-

tiometry (373-375) and a discussion of anodic stripping

chronopotentiometry (12, p. 115) are available. A variation

of anodic stripping chronopotentiometry wherein a chemical

oxidant replaces the constant current source has been

increasingly used of late and will be discussed later.

Linear sweep voltammetry is the simplest of the more

commonly used stripping methods. It originated with the

development of polarography, credited to Heyrovsky who in

1922 reported the first use of it (376). An instrument

capable of linearly scanning the potential was built and

used in 1925 (377). The first general publication on

polarography to appear outside Heyrovsky's native Czecho-

slovakia was in 1936 (378).

Using the DME and collecting current-voltage data on a

point by point basis with a manual polarograph as well as

with the early recording polarographs was tedious and time

consuming. Matheson and Nichols in 1938 applied a continu-

ously varying potential to a rapidly dropping DME and used

an oscilloscope to monitor the current-voltage behavior

(379). Laitinen and Kolthoff, because they used a platinum

electrode rather than a DME, coined the term "voltammetry"

in 1940 for the general technique of current-voltage curve

acquisition and reserved "polarography" to describe




University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs