PB82-257817
Speciation of Arsenic Compounds in Water Supplies
Kurt J. Irgolic
Texas A&M University
College Station, Texas
June 1982
U.S. DEPARTMENT OF COMMERCE
National Technical Information Service
-------
PBS2-257817
EPA 600/1-82-010
June 1982
SPECIATION OF ARSENIC COMPOUNDS
IN WATER SUPPLIES
by
Kurt J. Irgolic
Department of Chemistry, Texas ASM University
College Station, Texas 77843
R 8047 740 10
Project Officer
Frederick C. Kopfler
Toxicology and Microbiology Division
Health Effects Research Laboratory
Cincinnati, Ohio 45268
HEALTH EFFECTS RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, NORTH CAROLINA 27711
. • •
NATIONAL TECHNICAL
INFORMATION SERVICE
IS. OEPAimiEfir OF COMMERCE
SFtMGFIUO. VA. 22161
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/1-82-010
ORD Report
3. RECIPIENT'S ACCESSION NO.
PBB7 ' 9 g -7 a 1 7
4. TITLE AND SUBTITLE
Speciation of.Arsenic Compounds in Water Supplies
5. REPORT DATE '
June 1QS?
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Kurt J. Irgolic
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Texas A&M University
Department of Chemistry
College Station, Texas 77843
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
R-804774010
12. SPONSORING AGENCY NAME AND ADDRESS
USEPA
Toxicology & Microbiology Division
Health Effects Research Laboratory
Research Triangle Park, North Carolina 27711
13. TYPE OF REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE
13. SUPPLEMENTARY NOTES .
\
16. ABSTRACT .''•'.-••'
The objectives of this project were to develop and test analytical, methods that;would
allow the qhemical form (i.e. valence state or compound) of arsenic in drinking waters to
be determined, and to use the methods to analyze samples of drinking water from.sources
where.adverse health effects in consumers had been .attributed to arsenic. Analytical
techniques were developed for the determination of arsenate: (differential pulse
polarography), for inorganic and organic arsenic compounds (high pressure 'liquid
chromatography with graphite furnace atomic absorption spectrometry as element-specific
detector) and for the detection of arsenocholine, arsenobetaine, and iodoarsines (mass
spectrometry). These techniques, -inductively coupled argon plasms emission spectro-
metry, and hydride generation/DC-helium ar« emission were used for the characterization
of water samples from Utah, Alaska, Antofagasta, Taiwan and Nova Scotia. The. total
arsenic concentrations were in the range 18 ppb to 8 ppra with arsenic/arseriate ratios
between 0.007 and 3.4. No organic arsenic compounds were detected in any of the water
samples. The1 trace elements A1;,B, Ba, Ca, Cu, Fe, Li., Mg, Mn, Na,. P,. S, Si and Sr were
present in most of the water samples.. The results show that, the various physiological
effects, observed in populations exposed to the arsenic-containing water supplies could
not be caused by arsenic compounds other than arsenite or arsenate. . Other, trace elements
acting in concert with arsenite and/or arsenate might,produce these symptoms. However,
sufficient data are not yet available to evaluate th4sg hypothesis. ..
17.
. KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
c.. COSATI Field/Group
18. DISTRIBUTION STATEMENT
Release to Public.
19. SECURITY CLASS /Tliis Report!
Unclassified
21. NO. OF PAGES
124
20. SECURITY CLASS (This page)
Unclassified
22. PRICE
EPA Form 2220-1 (Rev. 4-77) PREVIOUS EDITION is OBSOLETE
-------
NOTICE
THIS. DOCUMENT" HAS BEEN REPRODUCED
FROM, THE; BEST COPY FURNISHED us.BY
THE SPONSORING AGENCY,,/A.LTH,OUGH: IT
IS RECOGNIZED' THAT. CE.RTAIN PORTIONS
ARE: ILLEGIBLE;,, ix is; BEING RELEASED
IN THE. IN.TE.REST: OF..MAKING; AVAILABLE
AS MUCH INFORMATION AS POSSIBLE,
-------
NOTICE
Mention of trade names or commercial products does not
constitute endorsement or recommendation for use.
-------
FOREWORD
The many benefits of our modern, developing, industrial society are
accompanied by certain hazards. Careful assessment of the relative risk of
existing and new man-made environmental hazards is necessary for the
establishment of sound regulatory policy. These regulations serve to
enhance the quality of our environment in order to promote the public health
and welfare and-the productive capacity of our Nation's population.
The-complexities of the environmental problems originate in the deep
interdependent relationships between the various physical arid biological
segments of man's natural and social world. Solutions to these en-
vironmental problems require an integrated program of research and deve
lopment using input from a number of disciplines. The Health Effects
Research Laboratory conducts a coordinated environmental health research
program in inhalation toxicology, genetic toxicology, neurotoxicology,
developmental and experimental biology and clinical studies using human
volunteer subjects. These studies address problems in air pollution, water
pollution, non-ionizing radiation,'environmental- carcinogenesis and the
toxicology of pesticide-s and other chemical pollutants. The Laboratory
participates in and provides data for the development and revision of
criteria documents on pollutants for which national ambient air quality and
water quality standards exist or are proposed, provides the data for
registration of new pesticides or proposed suspension of those already in
use, conducts research oh hazardous and toxic materials, and is primarily
responsible for providing the health basis for non-ionizing radiation
standards. Direct support to the regulatory function of the Agency is
provided in the form of expert testimony and preparation of affidavits as
well as expert advice to the : Administrator to- assure the adequacy of
.environmental regulatory decisions involving the protection of the health
and welfare of all U.;;S.: inhabitants. ;.,,..;:'.,,•/ '••"• : .-•.•--. .•-, ..••.".
This report contains a description of methods that can be used to
determine the specific arsenic ions and compounds that may be present in
drinking waters and have heretofore been reported as total arsenic. It also
contains the results of the application of these methods to samples of
several drinking waters known to contain arsenic, which in earlier studies
has been implicated as the cause of disease .in:consumers of these^waters.
F. G. Hueter, Ph.D. •"".'.•'•"
''..,. . , ' ' : Director ••'.'' .':• .• " - ". •••'"•••••'.•-
Health Effects. Research Laboratory ;
ill
-------
ABSTRACT
The objective of this project was to determine qualitatively and
quantitatively the arsenic compounds present in drinking water supplies which
are known to have caused adverse health .effects..
Water samples from Hinckley, Utah; Delta, Utah; the Barefoot Site,
Alaska; the Mauer Site, Alaska; Antofagasta, .Chile; Yenshei, Taiwan; and two
sites in Nova Scotia were; analyzed for total arsenic-, arsenite, arsenate,
arsenic compounds reducible to methylarsines and other organic arsenic
compounds. Graphite furnace atomic absorption spectrometry, differential
pulse polarography, hydride generation/DC-helium arc emissionj high pressure
liquid chromatpgraphy with a. graphite furnace atomic absorption spectrometer
as an element-specific detector (HPLC-GFAA) and simultaneous inductively
coupled argon plasma emission spectroscopy (ICP) were employed as analytical
techniques for the determination of. total arsenic and arsenic compounds. ICP
was also used to determine other trace elements in the water samples. The
total arsenic concentrations ranged from 18 ppb to 8 ppm with arsenite/
arsenate ratios between 0.007 and 3.4. No organic arsenic compounds reduci-
ble, to methylarsine or dimethylarsine were, detected by the hydride generation
technique, which has a detection limit of 1 ppb for these compounds. The
HPLC-GFAA system gave no indication of the presence of methylated or other
organic compounds. , The following trace elements were found in almost all
water samples: Al, B, Ba, Ca, Cu, Fe, K, Li, Mg, Mn, Na, P, S (probably
present as sulfate), Si and Sr.
Various physiological effec'ts (melanosis, keratosis, skin cancer,
vascular ailments,, pneumonia) have been attributed to arsenic. These effects
cannot be caused by arsenic compounds other than arsenite and/or arsenate.
The presence of arsenite and arsenate in different ratios in the water
samples or one or more of the other trace elements acting in concert with
arsenite and/or arsenate might produce the various symptoms. Sufficient
data are not yet available to evaluate these hypotheses.
The water samples from Yenshei, Taiwan, were investigated to determine
the nature of-the dissolved fluorescent compounds. High pressure liquid
chromatography employing UV- and fluorescence-detectors indicated that
ID-lysergic acid and/or ergometrine, ergotamine and perhaps calciferol might
be present in these water samples. Several unknown, UV-detectable and
fluorescent compounds were also found in these samples. These results call
for a reevaluation of the long-held belief that arsenic alone caused the
adverse health effects in the population in Taiwan exposed to these waters.
As part of this project the following techniques were developed for the
iv
-------
determination of arsenic compounds: differential pulse polarography for
arsenate; a high pressure liquid chromatography system with a Hitachi Zeeman
graphite furnace atomic absorption spectrometer as an element-specific
detector for arsenate, arsenite, methylarsonic acid, dimethylarsinic acid,
arsenocholine and arsenobetaine; mass spectrometry for organylarsonic acids,
diorganylarsinic acids, organyldiiodoarsines, diorganyliodoarsines,
arsenocholine and arsenobetaine.
X
This report was submitted in fulfillment of Grant No. R 8047 740 10 by
Texas A&M University under the sponsorship of the U.S.. Environmental
Protection Agency. This report covers the period'November 1, 1976 to
February 28, 1981, and work, was completed as of March 31, 1981.
v
-------
CONTENTS
Foreword . . iii
Abstract . iv
Figures vii
Tables x
Abbreviations and Symbols xiii
Acknowledgments ........... xv
1. Introduction .»• . . . . . . . . . . .-•'•.• . ... . . . ., .... 1
2. Conclusions 2
3. Recommendations .;.... 3
4. Arsenic Compounds in Natural Water Samples ... . . . • . • • 4
5. Preservation of Arsenic Compounds in Aqueous Solutions .... 7
6. Development of Analytical Techniques for the Determination
of Arsenic Compounds 12
Hitachi Zeeman Graphite Furnace Atomic
Absorption Spectrometer as an Element-
Specific Detector for High Pressure
Liquid Chromatography . ... . 14
The Speciation of Arsenite, Arsenate,
Methylarsonic Acid, Dimethylarsinic
Acid,. Arsenocholine and Arsenobetaine
. Using the HPLC-Hitachi Zeeman GFAA
System ................ 24
Differential Pulse Polarograhic / . .
Determination of Arsenate and
Arsenate . . ... • 32
Mass Spectrometry of Organylarsonic
Acids, Diorganylarsinic Acids,
Organyliodoarsines, Arsenocholine
and Arsenobetaine 53
7. Analysis .of Water Samples . 65
Water Samples from Hinckley and
Delta, Utah 68
Water Samples from Antofagasta, Chile 74
Water Samples from Alaska 80
Water Samples from Taiwan . 84
Water Samples from Nova Scotia 98
Discussion 100
References 103
vi
-------
FIGURES
Number Page
1 Transformation of arsenic compounds in the
environment . 5
2 Block diagram for the HPLC-GFAA system 15
3. Schematics for the HPLC-GFAA interface 17
4 The sampling sub-system of the HPLC-GFAA
interface . . . ....... . . . . . . . . . ... . 19
5 The injection system for the HPLC-GFAA
; .interface ••.,,. . . . . . . . . ..... . . .. . ... . . . ... 21
6 Separation of inorganic arsenic (arsenite r
and arsenate), arsenocholine and :
arsenobetaine by HPLC-GFAA.on.a C-18 .
reversed-phase column . . . . . . . . . . . ... . . 25
7 Separation of arsenite and arsenate by .
HPLC-GFAA on a C-18 reversed-phase .
column in the absence and presence of
tetrabutylammonium phosphate . . . . .... . ..,..-. . . . ... 27
8 The influence of phosphate on the GFAA
arsenic signals ...... 28
9 The separation of arsenite, arsenate, ........ , ,
dimethylarsinic acid and. methylarsonic
acid by HPLC-GFAA on a C-18 reversed-:phased
column in the presence of tetraheptylammonium
nitrate . . . . . . . . . . . . . . . . ... 29
10 HPLC-GFAA calibration curve for sodium arsenite . . . . . . . / . 30
11 HPLC-GFAA calibration curves for sodium arsenite,
dimethylarsinic acid, methylarsonic acid and
disodium hydrogen arsenate . . . . . :.: . . .... . . . . . . 31
12 Differential pulse polarograins of arsenate
solutions two molar with respect to perchloric
acid in the presence of ^D-mannitol ... '. ••;...... ........ 34
vii
-------
13 The dependence of the height of the differential
pulse polarographic curve of arsenate at
-0.55 V on the concentration of I>-mannitol 35
14 Differential pulse polarographic curves for
10 ppm As (arsenate)in solutions 0;5 M with
respect to &-mannitol at various perchloric
acid concentrations . . . . 37
15 Calibration curves for the determination of
arsenate by differential pulse polarography
in aqueous 20 M perchloric acid in the
presence of IKmannitol 39
16 Differential pulse polarograins for arsenite
in 2.0 M aqueous perchloric acid in the
presence and absence of D-mannitol 40
17 Calibration curve for the determination of
arsenite by differential pulse polarography
in aqueous 2.0 M perchloric acid in the
presence of J>-mannitol . . 41
18 Differential pulse polarograms of mixtures
of arsenite and arsenate in 2.0 M
perchloric acid in the presence of D-mannitol 42.
19 Differential pulse polarograms of solutions
containing 20 ppb As(arsenite) and 40 to
100 ppb As(arsenate) in 2.0 M perchloric
acid in the presence of D-mannitol 43
20 Differential pulse polarograms of arsenite,
and of arsenate obtained by Ce(IV)
oxidation of arsenite 45
21 The dependence of the peak height of the
differential pulse polarographic maximum
at -0.34 V (arsenite reduction) as a
function of the amount of cerium(IV)
ammonium nitrate added ........ 46
22 Differential pulse polarograms of 500 ppb
As(arsenate) obtained by cerium(IV)
oxidation of arsenite in 2.0 M perchloric
acid, followed by addition of hydroxylamine
hydrochloride, addition of J>mannitol and
standard addition of arsenate 47
viii
-------
23 Calibration curves for the determination of
arsenate obtained by oxidation of arsenite
with cerium(IV) 48
24 Flow chart for the determination of arsenite
and arsenate by differential pulse polarography 51
25 The mass spectrum of methylarsonic acid . . 54
26 The mass spectrum of dimethylarsinic acid 55
27 The mass spectrum of butyldiiodoarsine 61
28 The mass spectrum of dioctyliodoarsine ............. 62
29 The mass spectrum of arsenocholine chloride . .63
30 The mass spectrum of arsenobetaine chloride ..... . . ....... . . . 64
31 Cation exchange liquid chromatogram of the ...,,.,.. .
concentrated Yenshei'water' sampled . . .'f; . . ... ..... 94
32 Ion-pair reversed phase chromatogram of the
concentrated Yenshei water sample and
alkaloid standards employing an UV-detector .... 95
33 Ion-pair reversed phase chromatogram of the
concentrated Yenshei water sample and
alkaloid standards employing a fluorescence
detector . . . . . . . 96
34 Ion-pair reversed phase chromatograms of the
aged, concentrated Yenshei water sample :
employing a UV-detector at 254 nm and a
fluorescence detector -'., 97
35 Rapid scanning fluorescence analysis of
unconcentrated Yenshei well water ............... 98
";- ix
-------
TABLES
Number Pa8e
1 Specifications and Sources for Electronic
Components for the HPLC-GFAA Interface 18
2 Peak Heights of the Differential Pulse
Polarographic Curves at -0.55 V Obtained
with 2.0 M Aqueous Perchloric Acid
Solutions of 100 ppm As(Arsenate) in the
Presence of Various Polyhydroxy Compounds 33
3 Heights of the First Differential Pulse
Polarographic Maximum in the Reduction of
9.5 ppm As(Arsenate) in Aqueous Acidic Media
in the Presence of D-Mannitol at 0.5 Molar
Concentration 36
4 Relative Abundances of Ions in the Electron
Impact Mass Spectra of Organylarsonic Acids ..... 57
5 Relative Abundances of Ions in the Electron
Impact Mass Spectra of Diorganylarsinic Acids . , 58
6 Relative Abundances of Ions in the Electron
Impact Mass Spectra of Organyldiiodoarsines -... ......... 59
7 Relative Abundances of Ions in the Electron
Impact Mass Spectra of Diorganyliodoarsines 60
8 Total Arsenic Concentrations in ppb in the
Hlnckley and Delta Water Samples ..... 69
9 Concentrations of Arsenite and Arsenate in
ppb in the Hinckley and Delta Water Samples 71
10 Trace Element Concentrations in the Hinckley
Water Samples Determined by .ICP 72
11 Trace Element Concentrations in the Delta Water
Samples Determined by ICP 73
-------
12 Total Arsenic Concentrations in ppm in the
Antofagasta Water Samples 75
13 Concentrations of Arsenite and Arsenate in ppm
in the Antofagasta Water Samples 76
14 Trace Element Concentrations in the Untreated
Antofagasta Water Samples Determined by ICP 77
15 Trace Element Concentrations in the Treated
Antofagasta Water Samples Determined by ICP 78
16 • Results of Total Element Analyses for Water
Samples from Antofagasta, Chile, Collected
and Analyzed in 1977 79
17 Results of Total. Arsenic Analyses and Arsenite
and Arsenate Determinations in Alaska Water
Samples . . ........... 81
18 Trace Element Concentrations Determined by ICP
in the Barefoot Water Samples from Alaska .
(Collected in 1979). . . . . . .... ...... . . . ,82,
19 Trace Element Concentrations Determined by ICP
in the Barefoot Water Samples from Alaska.
(Collected in 1980) . . . ... . ,.' 83
20 Trace Element Concentrations Determined by ICP
in the Mauer Water Samples from Alaska
(Collected in 1979) . . ". . . . .v:. ..".-. . . .". . . ';' . . . 85
21 Trace Element Concentrations Determined by ICP
in the Mauer Water Samples from Alaska
' (Collected in 1980) . ....... ..'.............. .",...86
22 Analysis Results for the Pei-Men and Pu-Tai
Water Samples 87
23 Results of Total Arsenic Analyses and Arsenite
and Arsenate Determinations in the Yenshei
Water Samples 89
24 Trace Element Concentrations in the Yenshei <
Water Samples, Well. I, Determined by ICP ............ .90
25 Trace Element Concentrations in the Yenshei
Water Samples, Well II, Determined by ICP ........... 91
xi
-------
26 Analysis Results for the Nova Scotia Water Samples 99
27 Summary of Total Arsenic, Arsenite, Arsenate and
Trace Element Concentrations in Drinking Water
Samples 101
28 Arsenic-Containing Water Supplies and Their
Physiological Manifestations in Man 102
xii
-------
LIST OF ABBREVIATIONS AND SYMBOLS
ABBREVIATIONS
A — ampere
AA — atomic absorption
AAS — atomic, absorption spectrometry
AC — alternating current
Ag-DDC — silver diethyldithiocarbamate
cone.. — concentrated
DC — direct current
DMAA — dimethylarsiriic acid
DPP — differential pulse polarography (polarographic)
g • — gram.' ' • ' .-• • • •• • '" . "' •.. ••' -••••:• ' ' - •-.-•• • •• •-•-•••'
GFAA — graphite furnace-atomic absorption spectrometry •
(spectrometer, spectrometric)
HG — hydride generation ; •
HPLC-GFAA — high pressure:liquid chromatography-graphite furnace atomic
absorption spectrometer system
ID • — inner diameter ..'."• ".-•' • '\-'''•'•.'. :'':.;;. •' • 7 •-•••.•
ICP —inductively coupled argon plasma emission spectrometry
(spectrometer, spectrometric), •. v-• . .
L • . .. . ,'T-..liter •',' ._ •-; ' -• •• .-.•••; -; '••-... .. .•'.-/:.'•." -';-;••-'•".'
M. , —molar ' . • . . ' ' '
mA — milliampere T
'MAA —methylarsonic acid.,,,. ';;.)::-•,:•,, : :• "" ." ;
MeOH —methanol
mg — milligram . •.. -^ , ,
min — minute(s)
mL —milli-liter ..•••••''"• ;
nm ~ millimeter (lxlO~3 meter) , .. ....
y —micron (IxKT*5' meter), micro (xlO~6)
yA — microampere
yC — microcurie
yF —microfarad
yg — microgram
yL — microliter .-...." • .,.,
nm — nanometer.(lxlO~9 meter)
OD — outer diameter '
ppb — part per billion; the concentrations of arsenic compounds
are expressed in .terms of arsenic, e.g., 1 ppb As(arsenate)
means that the solution contains 1 ppb of arsenic in form of
arsenate _
pH —negative logarithm of hydrogen ion concentration
xiii
-------
pKa — negative logarithm of the acid dissociation constant
ppm — part per million (see ppb for note)
psi — pounds per square inch
RI — refractive index (detector)
TBAP — tetrabutylammonium phosphate
THAN — tetraheptylammonium nitrate
UV — ultraviolet
V ' —.volt
W -- watt
xiv
-------
ACKNOWLEDGMENTS
The assistance of the following' individuals in collecting water samples
is gratefully acknowledged: Mr. E. Western, Utah State Department of Health,
Division of Environmental Health (samples- from Hinckley and Delta, Utah); Dr.
J. M. Borgono, Servicio Nacional de Salud, Ministerio.de Salud Publica,
Santiago, Chile; Mrs. Alicia Araya Fuentes, Acting Director, Laboratorio
Bromatologico,. Antofagasta, Chile; Mrs. Stefania Razmilio Valdes, Antofagasta,
Chile (samples from the Antofagasta drinking water supply); Ms. Jane L.
Valentine, Associate Professor of Public Health, University of California,
Los Angeles (Antofagasta water samples); Mr. Pete McGee, Alaska Department
of Environmental Conservation, Fairbanks (Alaska water samples); Dr. W.
Tseng, Department of Medicine, National Taiwan University, Taipei (Pu-Tai
and Pei-Men samples); Dr. Fung-Jou Lu, Associate Professor of Biochemistry,
College of Medicine, Taipei (Yenshei samples); and Mr. C. E. Tupper,
Administrator, Environmental. Health,. Department of Public Health, Nova Scotia
(Nova Scotia, samples).
The following postdoctoral associates and graduate students participated
.in the .analytical work: Dr. R.. Bess,. Mr. •• R. Nichols (differential pulse
polarography) ; .Dr. M.. West (graphite furnace AA) ; Mr.. D. Aylmer (adsorption
of arsenic compounds by .container walls);; Mr. R. Stockton (HPLC^GFAA system,
ICP, arsenocholine/arsenobetaine separation) ;. Dr.. P. Clark and Ms^__P. Micks
(hydride generation); Mr. K.. Ehrhardt (mass spectrometry, arsehite/arsenate/
methylarsonic acid/dimethylarsinic acid separation) and Dr. D. Chakraborti
(graphite furnace AA) . ••.....'
Special, thanks: are owed to Mr. D. Heinrich, Radian Corporation, Austin,
and Ms..Denise Painchaud, Applied Research Laboratories, Sunland, California
for performing ICP analyses on the- water samples. •
Dr. I. Warner, Assistant Professor, Department of Chemistry, Texas A&M
University and Mr. Dennis Shelly carried out the work on the fluorescent
compounds in the Taiwan well" water.
Elsevier Scientific Publishing Company, and Gordon and Breach Science
Publishers granted permission to use figures from references 57 and 58 and
quote passages pertinent to these figures.
xv
-------
SECTION 1
INTRODUCTION.
Arsenic is an element which possesses a rich chemistry. Inorganic
and organic arsenic compounds may contain trivalent or pentavalent arsenic..
The trivalent arsenic compounds are generally more toxic than the pentavalent
derivatives (1). Many inorganic and organic arsenic compounds are linked in a
cycle with chemical and biologically mediated-reactions changing the compounds
into each'other. The input, of arsenic into this-cycle:is suppliedby
weathering of arseiiic-containing rocks and1human.use and disposal of .various
.arsenic compounds. ... -.,' . . ..,,-... ... .
Since arsenic is ubiquitous, man consumes small,amounts1 of arsenic
compounds with the food he eats and the water he drinks. Life- developed in
the presence, of arsenic. Organisms are, therefore, expected to tolerate a .....
certain, not yet clearly defined dosage of arsenic. Certain geographically
limited groups of people have taken into their-systems arsenic compounds
present in their drinking water supplies over extended periods'of time. The
most famous- localities where, arsenic-containing water" has been consumed' are
certain regions in. Taiwan (71^78) and the city, of Antofagasta in Chile (69).
Hyperpigmentation, skin cancer, vascular problems and other ailments have
been attributed to the arsenic present in the drinking water. Other groups,
for which the people of Fallen, Nevada, may serve as examples, have been
exposed to similar arsenic levels in their drinking water without any ill
effects (82). This project was undertaken in'order to determine the arsenic
compounds and other trace elements present in arsenic-containing water-
supplies and to check whether these drinking water supplies contain the same
or different arsenic -compounds.
-------
SECTION 2
CONCLUSIONS .
As the result of this project, together with previously developed
analytical techniques, reliable methods are now available to determine
arsenite, arsenate, methylarsonic acid, dimethylarsinic acid, trimethylarsine
oxide, methylarsines, arsenocholine and arsenobetaine at the low ppb level.
The high pressure liquid chromatography system employing a Hitachi Zeeman
graphite furnace atomic absorption spectrometer (HPLC-GFAA) as a sensitive,
element-specific detector is specially useful for the detection of yet
unidentified arsenic compounds present in water samples. 'In order to increase
the confidence in and the reliability of the analytical data, total arsenic
concentrations and the concentrations of individual arsenic compounds must be
determined by at least two independent methods.
The analyses of water samples from Taiwan, Antofagasta, Alaska, Utah
and Nova Scotia produced total arsenic concentrations in the range 18 ppb
to 8 ppm. The arsenite/arsenate ratios were in the range 0.007 to 3.4. No
indications of the presence of methylated arsenic compounds, reducible to
methylarsine or dimethylarsine, at levels above 1 ppb have been found.
Experiments with the HPLC-GFAA system showed only the presence of arsenite
and arsenate in the water samples. Most of the water samples contained
Al, B, Ba, Ca, Cu, Fe, K, Li, Mg, Mn, Na, P, S (probably as sulfate), Si and
Sr (determined by ICP).
Various physiological effects observed in populations exposed to arsenic-
containing drinking water supplies could have been caused by the presence of
varying amounts of arsenite and arsenate. It is also conceivable that one
or more of the trace elements present in the water supplies acted in concert
with arsenic to cause the observed effects. Many more samples need to be
analyzed and the results of these analyses correlated, with epidemiological
studies before ^a definite statement can be made about the interactions''of
trace elements with arsenite.or arsenate.
The chromatographic work on the fluorescent compounds in the Taiwan
well waters strongly suggests the presence of alkaloids such as ^-lysergic
acid, ergometrine and calciferol. These findings call for a reexamination
of the presumed cause-effect relationship between arsenic and vascular
diseases in Taiwan.
-------
SECTION 3
RECOMMENDATIONS
The availability'of the hydride generation technique,, differential pulse
polarography, high pressure liquid chromatography coupled to sensitive,
element-specific detectors and colorimetric methods for the determination of
total arsenic concentrations and concentrations of arsenic compounds, and the
availability of simultaneous, inductively coupled argon plasma emission .:••'••
spectrometers for trace element determinations and of ion chromatography for
anion analyses;make the thorough characterization of water samples a
relatively easy and not too-time-consuming task. Many more arsenic-
containing water samples must be.analyzed in support of or in preparation for
epidemiological studies. . ,/ -
Arsenic has been declared a priority pollutant. Many environmental
samples will have to be analyzed for arsenic compounds before final ,
conclusions can be drawn about the health effects of arsenic compounds.
Experience has shown.that .analysis by one method is not sufficient-to produce
reliable results. At least two independent methods should;be used for' the : ?:
determination of; arsenic-'-.compounds. . : - - ,
-------
SECTION 4
ARSENIC COMPOUNDS IN NATURAL WATER SAMPLES
Arsenic is associated with igneous and sedimentary rocks in form of
inorganic compounds. The average concentration of arsenic in these rocks
varies between 0.1 ppm and several hundred ppm (1). Arsenic is,found in
high concentrations in sulfidic ores.. Special conditions created by recent
or geologically ancient volcanic, or hydrothermal activities have also caused
high concentrations of arsenic in localized areas. Weathering of arsenic-
containing rocks, produces arsenic trioxide, arsenites and arsenates which
might be retained in soils or dissolved in water and transported to the ocean.
In the absence of living matter water most likely will contain only arsenite
and/or arsenate. •
Many reactions are known.which produce organic arsenic compounds from
inorganic arsenic compounds (2).. Such reactions generally do not occur in
aqueous systems unless organisms capable of transforming arsenic compounds
are present. The conversion.of inorganic to methylated arsenic compounds
mediated by various organisms is well established (3,4,5).- Methylarsine,
dimethylarsine and trimethylarsine may thus be formed from arsenite or
arsenate. The formation of methylarsonic and diinethylarsinic acid was also
observed (6). Microorganisms can also demethylate methylarsenlc derivatives
(7). These methyl-transfer reactions, the reduction of arsenate, the oxida-
tion of arsenite,. the'-reduction of methylarsenic acids to arsines and the ox-
idation of these arsines to the arsenic acids set up an arsenic cycle (Fig.
1). The arsenic cycle in nature ,is: much more complex than the one shown in
Mg. 1. Arsenic compounds are distributed among the various compartments of
the environment including air,.water, soil, sediments, rocks and plant and
animals (7). Living matter seems to be a rich source of many organic arsenic
compounds. Arsenic has been shown to be incorporated into both marine and
freshwater organisms in the form of water and lipid soluble organic arsenic
compounds (8). The suggestion was made (8) that arsenic is incorporated into
lipids by replacing nitrogen in lecithins. A phospholipid containing 0.5
percent of arsenic (9), which might have formula^ (R = H), was isolated from
the alga Tetvaselmis chui. Hydrolysis of the arsenic-containing lipid JL^would
yield water-soluble arsenocholine 2^ (R = H). Arsenobetaine ^3, an oxidation
product of arsenocholine 2.(R = H), was isolated from the Australian rock
lobster (10). Chaetpceros aoneauicoimis grown in arsenate-containing, axenic
media is reported (11) to have synthesized 0-phosphatidyltrimethylarsoniolac-
tic acid 2 (R = COOH). The arsenic-containing sugar ^ was isolated from
Eaklonia radiata (12). Most of these investigations employed marine
organisms. It is very likely, that freshwater organisms are capable of
synthesizing the same or similar arsenic compounds. The isolation and
-------
CH3AsH2 .^p—±: CH3As03H2
(CH3) 2AsH
) 2As02H
methylation
demethylation
H3AS04:
I
As-containing sugars;
other, yet unidentified
arsenic compounds;
As-containing lipids
I
(CH3)2As02H.
*''*'
[(CH3)3AsCH2CH2OH]n
) 3AsCH2COO"-*
Figure 1. The transformation of arsenic compounds in.the environment.
CH—COR*
H,. COOH
!3As-CH2-CH-OH
- II, COOH
.'- 2
(cH3)3AsCH2COO-
.3- ' '
CH3
O-AS-CH
CH,
CH2CHOHCH2R
R = OH, S02H
-------
and identification of organic arsenic, compounds has just begun. Many more,
yet unknown arsenic compounds will be discovered and their properties deter-
mined as analytical techniques are improved for the speciation of such
derivatives.
In. natural fresh water not polluted with arsenic compounds through
human activities, arsenite and/or arsenate are expected to be found. Based
on the current knowledge of the arsenic cycle in nature methylarsenic
compounds, arsenobetaine and arsenocholine might also be present.
Man-made arsenic compounds are used as pesticides, herbicides, desic-
cants,. wood preservatives, feed additives and drugs (13). Most of these
compounds are either inorganic or belong to the class of methylated arsenic
derivatives and will be detected and quantitatively determined upon routine
analysis. Pollution by the. aromatic arsenic compounds employed as feed
additivies and wood preservatives is unlikely unless there is a manufacturing
or processing plant for such compounds near the origin of water supplies.
-------
SECTION 5
PRESERVATION OF ARSENIC COMPOUNDS .IN AQUEOUS SOLUTIONS
In contrast to most metal ions, which do not change valence easily and
are not transformed into organic compounds in aqueous solution, arsenic
possesses a complicated chemistry characterized by equilibria, steady state
conditions or situations approaching such conditions between trivalent and
pentavalent inorganic arsenic compounds and various derivatives containing
arsenic-carbon bonds. A brief raview of the chemistry of arsenic is avail-
able in the literature (14). The analysis for arsenic compounds in water
samples, for which the term speciation shall be used, will correctly give the
composition of the sample at the time of collection only if no arsenic com- '
pounds were lost, interconversions of arsenic compounds present in the sam-,
pies did not occur and no new arsenic compounds were formed during the time
between collection and analysis.' The determination of. total arsenic:concen-
trations in samples requires only that no arsenic shall be lost and that the
analytical techniques to be employed shall be applicable to all the arsenic-
compounds present in the sample.
Water samples can generally not be analyzed -for trace elements immedi-
ately after collection.. Several hours or more often several days elapse
between collection and analysis. During this time the chemical nature of a
particular trace element can change, trace elements can be lost by volatili-
zation and/or can be absorbed on cqntainer walls. The adsorption of many
metal ions (15, 16) and of phosphate ions (17) were studied but arsenic was
rarely included. ' • ., • ' :; '." ""."'. .... '", "
Disagreements exist in the literature as to the extent of loss of ar-
senic from solutions stored in various containers. These disagreements are
probably caused by differences in composition of the samples and sample con-
tainers, the species of arsenic present and the sensitivity of analytical
methods.
Acidification of samples is usually considered necessary when trace ele-
ments are to be determined (15, 16). Several acids (nitric, perchloric,
hydrochloric and acetic acid) were suggested as perservatives. Guimont and
co-workers (18) mention the use of acetic acid as a preserving agent for
arsenic solutions but give no details. Ray and Johnson (19) report losses of
arsenic up to 70 percent within one week from seawater acidified with 9 raL
concentrated hydrochloric acid per liter of sample. Talmi and Bostick (20)
observed no losses of arsenate from distilled-water solutions containing
15 percent nitric acid or 5 percent perchloric acid upon storage for three
weeks in polyethylene or soft glass containers. Whitnack and Brophy (21)
-------
reported no loss of arsenite from samples of well water kept for one week
in polystyrene vials with snap caps of polyethylene. Al-Sibaai and Fogg
(22) found that the total arsenic concentrations in dilute standard solutions
containing only arsenite or arsenate (20 yg. mL"1) remained unchanged for at
least 50 days when stored in borosilicate glass, soda glass or polyethylene
in light or in the dark. Experiments with seawater containing 2 yg L of
arsenic gave different results (23). When such filtered seawater samples
equilibrated with 5 yC of arsenic-74 were placed 'in high density polyethylene,
soda glass or Pyrex bottles and stored in the dark over a period of five
weeks, losses of arsenic occurred with all types of bottles (23). With soda
glass bottles sixteen percent had been adsorbed by the 15th. day when equili-
brium was attained.. The uptake by polyethylene and Pyrex bottles was com-
plete within ten days and amounted in both cases to approximately six percent
(23).
Storage of water: samples in the frozen state has been recommended (16,
23, 24, 25). Harrison and co-workers (26) found that frozen solutions (1 ppm
As) did not lose arsenic upon freeze-drying. The average retention was
96.3±1:8 percent- Natural water samples stored below -15° or under dry ice
did not lose arsenite (27). Seawater samples kept in a refrigerator at 4°
showed no measurable change in the arsenite/ar senate ratio within about ten
days (28). For longer storage the samples had to be frozen quickly with
crushed dry ice within 15 minutes. They could then be stored frozen. In
samples frozen in a freezer,, spurious losses of arsenite were observed (29) .
Sndreae ~(27y~s~tated~tfiat~ cono^/'di^ and" tfimethylarsihe~ were" slfable~in .
aqueous' solutions for a f ew~days~whenT"storie3~in aTF-tightPcontainers . Me thy 1
arsonic acid and dimethylarsinic acid, were lost measurably after a period of
approximately three days from untreated water samples. .Acidification with
hydrochloric acid to make the sample 0.05 M prevented these losses.
It is not possible to recommend at this time a method for the preser-
vation of water samples applicable to all the various arsenic compounds which
may occur in water. Not enough is ;known, for instance, about the adsorption
behavior of inorganic and organic arsenic derivatives in the presence of
other compounds generally found in water. Even less . knowledge exists about
the conditions under which arsenic compounds are converted to other species.
Arsenite and arsenate are rather easily changed into each other. Ar-
senite in natural water samples is slowly oxidized to arsenate. Acidifica-
tion of the sample increases the oxidation rate (27). Fifty percent of the
arsenite in an aqueous standard solution (20 yg mL ) was oxidized to arse-
nate after storage in a glass container for 33 days (22) . Arsines were
oxidized by traces of air (27) . An arsenic cycle operates in nature (7)
connecting inorganic with organic arsenic compounds. The chemical nature of
many of the organic arsenic compounds participating in this cycle are still
unknown .
EXPERIMENTAL
Conditions which will prevent adsorption are likely to favor transfor-
8
-------
mation of arsenic compounds and vice versa. A universal preservation agent
which would prevent adsorption and transformation is not' known and: will
probably not be found. It will be necessary, to develop preservation techni-
ques for each sample dependent, upon the type of arsenic compounds present.
Investigations were carried out to find conditions under which arsenic'
compounds would not be adsorbed on container walls and arsenite would not be
converted to arsenate.. .
Most of"the water samples, which are the subject of these speciation
studies,.contained arsenic at concentrations of approximately 1 ppm. To ob-
tain data on the amount of arsenic lost to container walls, the adsorption
of arsenate, arsenite and dimethylarsinic acid from aqueous 1 ppm As solu-
tions on polyethylene and Pyrex was studied.
New polyethylene Cubitainers (soft polyethylene manufactured by Kimberly)
and Pyrex glass containers..were filled with 7.5 M nitric acid and kept for .
24 hours. They were then emptied and rinsed thoroughly with deionized water.
The arsenite and arsenate solutions were prepared using either distilled,
deionized water of pH 6.5 or tap water of pH 8 with a sodium concentration
of 211 ppm. Neither water contained arsenic.
Weighed quantities (required for the preparation of stock solutions) of
Na2HAsO^-7H20, NaAsO^ or &S2°5 were irradiated with neutrons at the Texas
A&M University Nuclear Science Center to produce 76As, a gamma-emitter with.
a half life of. 24.6 hours:. For- the experiments with Pyrex flasks the stock
solution was prepared by dissolving irradiated As2<35 in distilled water.
The stock solutions for the polyethylene experiments were made; from Na2HAsO^-
7H20 or NaAsC>2. Working solutions of 1 ppm arsenic (As) were prepared by
appropriate dilution. . .
Acid-cleaned 25 mL volumetric .flasks made from Pyrex were filled to the
mark with the 1 ppm As solution (prepared from As2^5)• One flask each was
emptied after one, two, four,, five, six and ten days :had elapsed. The flasks
were rinsed, then filled, to the mark with distilled water and placed into the
gamma-ray counter well equipped with a Li--drifted germanium detector. The
entire gamma spectrum was recorded.. The areas under the.arsenic peaks were
computed and corrected for the decay of the 76As-isotope. A calibration
curve was obtained by counting 25 mL-volumetric flasks containing 0.5, 1.0,
1.5 or 2.0 mL of the 1 ppm As solution in 25 mL.
For the adsorption studies in polyethylene bottles, 1 mL of a solution
containing 100 ppm As of either arsenate or arsenite was pipetted into
100 mL of distilled water or tap water stored in the plastic bottles. These
bottles were treated and counted as described above for the Pyrex studies.
Samples of 1 ppm As solutions containing dimethylarsinic acid in poly-
ethylene bottles were prepared similarly. The arsenic concentrations in
solution as a function of time were determined by flameless atomic absorp-
tion spectrometry after a ten-fold dilution of an appropriate aliquot.
For the polarographic determination of arsenite in solution, 45 mL were
'"..'••• 9 - • ' '
-------
withdrawn from the plastic bottles and transferred into the polarography
cell. After addition of 5 mL of cone. HC1 nitrogen was bubbled through the
solution for five minutes before performing the analysis. To determine
arsenate, 30 mL of the solution, 10 mL cone. HCIO^ and 3 g pyrogallol were
placed into a 50 mL volumetric flask. After filling to the mark with dis-
tilled water, the cell was charged with this solution. Calibration curves
were obtained by standard addition.
The following systems were investigated:
Arsenate in distilled water - Pyrex
in distilled water - Polyethylene
in tap water - Polyethylene
Arsenite in distilled water - Polyethylene
in tap water — Polyethylene
Dimethylarsinic Acid in distilled water - Polyethylene
in tap water - Polyethylene
RESULTS
When 1 ppm As (arsenate) solutions were kept in 25 mL acid-cleaned Pyrex
volumetric flasks activity corresponding to 0.22 ug of arsenic was found
on the walls after one day. The amount of arsenic adsorbed did not change
during the ten-day experiment. The samples could not be counted after more
than ten days had elapsed because of decay of the '^As-isotope.
In all the experiments in polyethylene bottles no adsorption of arse-
nate, arsenite or dimethylarsinic acid was detected. The results for dime-
thylarsinic acid are tentative because the results of atomic absorption
spectrometry are accurate only within ±5 percent (30). ^As-Dimethylarsinic
acid was not available for these experiments. After two days all the arse-
nite in the NaAs02 solutions had been converted to arsenate (checked by
polarography). These studies indicate that polyethylene bottles are the con-
tainers of choice to store aqueous solutions of arsenite and arsenate.
Among the chemical changes to which arsenic compounds may be subjected
in aqueous solutions, the oxidation of arsenite to arsenate has' been re-
ported repeatedly in the literature (7, 22, 27, 31). Preliminary analyses of
water samples to be speciated for arsenic compounds indicated the presence
of arsenite and arsenate. To keep these two ions from converting into each
other a search for an appropriate preservation technique was made.
The most convenient and experimentally tested procedure for the pre-
servation of arsenite was reported by Feldman (31). Ascorbic acid at a con-
concentration of 1 mg mL prevented the oxidation of arsenite to arsenate in
aqueous standard solutions at room temperature. Arsenate was not reduced
to arsenite under these conditions (31). However, special care must be
taken, not to expose arsenite solutions preserved by ascorbic acid to
10
-------
elevated temperatures. Ascorbic acid has been used to reduce arsenate to
arsenite in boiling seawater (23). Whether ascorbic acid scavenges oxygen
from the solution, combines with arsenite producing a species less reactive
toward oxidation than arsenite or acts both ways in preventing oxidation of
arsenite, is not known. Natural water samples contain many other metal
ions. Comprehensive experiments to elucidate the influence of such metal
ions on the ability of ascorbic acid to preserve arsenite have not yet been
carried out. Metal ions do catalyze the oxidation of ascorbic acid fay
oxygen (32).
The water samples which were analyzed as part of this project contained
arsenite and arsenate (see Section 7). Additional water samples were analyzed
during the past six months. On the basis of the experience gained with these
samples the following procedure is recommended for the analysis of arsenite
and arsenate in water samples:
• Collect three samples for each water supply to be analyzed in
one-quart polyethylene Cubitainers.
• Preserve-the first sample by addition of 0.1 percent by weight
of ascorbic acid.
• Preserve the second sample by addition of 7 mL ultrapure,
concentrated nitric acid to. one quart of sample.
••Leave the third sample unpreserved.
•'Fill all,the containers to the brim;to avoid air space's. /
• Analyze the samples as soon as possible after collection.
• The samples should be kept at or below room temperature during
shipment and storage. •
If the concentrations of arsenate and arsenite in the three samples deter-
mined by at least, two independent methods agree, arsenite and arsenate were
not chemically changed during the period between collection and analysis.
Among the fifteen different water samples analyzed by this procedure thus far,
the arsenite and arsenate values obtained from a,set of preserved and un-
preserved samples did not agree in only one :case.
Since methylated arsenic compound's were not.present in any of these
samples, no statement can be'made about the effect of this procedure on .
organic arsenic derivatives.
11
-------
, SECTION 6
X
DEVELOPMENT OF ANALYTICAL TECHNIQUES'FOR THE
DETERMINATION OF ARSENIC COMPOUNDS
Several methods are available to determine total arsenic concentrations:
colorimetry (silver diethyldithiocarbamate - AsH, complex; arsenomolybdic
acid); neutron activation analysis employing the gamma-active isotope As;
polarography of arsenite; coulometry; X-ray fluorescence spectroscopy;
atomic emission spectrometry; plasma emission spectrometry; and gas chrom-
atography of volatile arsenic compounds. Recent reviews of these techniques
are available (33, 34). The detection limits may be as low as picograms of
arsenic per milliliter.
Whereas many methods exist for the determination of total arsenic, the
choice of techniques for the determination of individual arsenic compounds
is limited. The following methods were available for the determination of
arsenic compounds when this project was initiated. The arsenic compounds,
determined by these methods, are listed as part of the heading. It is
likely that other arsenic compounds not yet investigated may also be
determinable by these methods.
Silver diethyldithiocarbamate (Ag-DDCl. / colorimetry for inorganic
arsenic and methylarsonic acid: Arsine generated by reduction of arsenite
and/or arsenate and methylarsine obtained by reduction of methylarsonic
acid, reacts with Ag-DDC in pyridine to produce colored complexes. Zinc/
hydrochloric, acid serves as,the reducing agent. The absorbances of the
complexes are measured at 540 and 440 nm. Detection limits as low as 20
ppb are achieved (35).
*
Ammonium molybdate for arsenite and arsenate: Arsenate (but not
arsenite) reacts with ammonium molybdate to form blue arsenomolydate,
the absorbance of which is measured at 865 nm. Arsenite can be oxidized
to arsenate for the determination of total inorganic arsenic. Reduction
of arsenate to arsenite with sodium disulfite and sodium thiosulfate in
aqueous sulfuric acid allows the determination of phosphate (36).
Polarography for arsenite; Arsenite can be determined polarographically
in acidic aqueous media by single-sweep polarography (21) (sulfuric acid,
detection limit 5 ppb) or by differential pulse polarography (37) (1 M
hydrochloric acid, detection limit 0.3 ppb). Arsenate is polarographically
inactive under these conditions. Dimethylarsinic acid, methylarsonic acid
(38), and other aliphatic (39) and aromatic (40) arsinic and arsonic acids
are reducible on the dropping mercury electrode. Whether inorganic arsenic
12
-------
compounds and organylarsenic acids can be determined simultaneously by
polarography must yet be decided by experiment.
/
Hydride generation for arsenite,?;. arsenate; methyl-, ethyl-, propyl,
butyl-, phenylarsonic acid; dimethylarsinic acid; trimethylarsine oxide:
Arsenite, arsenate, organylarsonic acids, diorganylarsinic acids and tri-
methylarsine oxide are reduced in aqueous medium by sodium borohydride to
arsines (27, 31, 41, 42, 43). Speciation of these compounds is accomplished
by their pH selective reduction to arsines, transferring the arsines into
an organic solvent (43), or adsorbing the arsines on Chromosorb 101 (43),
selective volatilization of arsines from a cold trap (27, 41, 42, 43) or
gas chromatographic separation of the arsines (27, 43) and detection of the
separated arsines by helium-DC discharge spectrometry (31, 41, 42), argon
microwave emission spectrometry (43), atomic absorption spectrometry using a
hydrogen/air flame (27, 44), electron capture or flame ionization detectors
(27). Detection limits in the part per trillion range have been reported
(27). Great care must be exercised when arsenic compounds are reduced with
sodium borohydride. Some arsenate may be reduced to arsine under conditions
.under which only arsenite should react, and cleavage of methyl groups from
methylarsenic compounds has been observed during their, reaction with sodium
borohydride (31). Improvements••concerning the reduction,conditions (31) and
modifications of the analytical system (31, 45) made the hydride generation
technique a more reliable method for the determination of arsenic: compounds.
The-disadvantages.of the hydride generation method include the1 require-
ment that the arsenic compounds must be reducible,.to arsines, and the
necessity that the .arsines,. possess, sufficient volatility to allow their trans-
fer to the detectors. Many arsenic compounds exist which do not fulfill
these requirements. : ; . ,-.;
Derivatization/gas chromatography. for-methyrarsonic acid and dimethy1-
arsinic acid: Methods which digest samples to convert arsenic compounds to
inorganic, derivatives (46) and subsequently transform these into organic
derivatives (20, 47) suitable for gas chromatography cannot be employed
for speciations. All information about the chemical nature of the original
compounds is lost during digestion. Volatile arsenic compounds, such as
methylarsines, can be determined without prior chemical modification by gas
chromatography (27, 48)..
Chemical modification has been successfully employed to convert methyl-
arsonic and dimethylarsinic acid to volatile derivatives. Methylarsonic
acid combines with ethylene glycol (49), dimethylarsinic acid reacts with
hydriodic acid (50), and both acids interact with allylthiourea (51) to
form arsenic derivatives which can be determined by chromatography. The
reactions.of arsenic compounds with trimethylsilyl chloride produce volatile
arsenic compounds. Unfortunately,.the method is non-selective. Several
arsenic compounds gave the same product (52).
Gas chromatbgraphy/atomic absorption spectrometry for arsenite, arsenate,
methylarsonic acid, dimethylarsinic acid, trimethylarsine oxide and
13
-------
methylarsines: Arsenic compounds, which possess adequate volatility or can
be converted to volatile compounds, can be separated by gas chromatography
and quantitated employing graphite furnace (53) or flame (27) atomic absorp-
tion spectrometers as element-specific detectors. This method has not been
used extensively; . '
Gas chromatography/mass spectrometry for methylarsines and ethyl, propyl-
and butylarsine: After separation of volatile arsenic compounds by gas
chromatography a high-resolution mass spectrometer may serve as an element-
specific and compound-specific detector which can also provide the elemental
compositions for the compounds (43).
All the methods available for the speciation of arsenic compounds at
the time this project was initiated had severe limitations. The methods
were applicable only to the determination of a few arsenic compounds or
required their conversion to volatile derivatives. Practically, the quanti-
tative determination of arsenic compounds was thus limited to arsenite,
arsenate, methylarsonic acid, dimethylarsinic acid, trimethylarsine oxide,
ethyl-, propyl- and butylarsonic acid, and the arsines obtainable from these
compounds. Non-volatile arsenic compounds and arsenic compounds not reducible
to volatile arsines could not be determined with the existing methods.
Therefore, an analytical system wi.th an element-specific detector had to be
developed capable of separating volatile and non-volatile arsenic compounds
from complex matrices.
The development efforts produced a high pressure liquid chromatography -
Hitachi Zeeman graphite furnace atomic absorption system, a differential
pulse polarographic method for the determination of arsenite and arsenate,
and the elucidation of the mass spectral behavior of organylarsenic
acids, organyliodoarsines,' arsenocholine and arsenobetaine.
HITACHI ZEEMAN GRAPHITE FURNACE ATOMIC ABSORPTION SPECTROMETER AS AN ELEMENT-
SPECIFIC DETECTOR FOR HIGH PRESSURE-LIQUID CHROMATOGRAPHY
Liquid chromatography and specially high pressure liquid chromatography
(HPLC) with the great resolving power of its micro-particulate.column
materials are potentially the best techniques for the separation and
simultaneous determination of arsenate, arsenite and organic arsenic com-
pounds. A water sample may contain many substances in addition to arsenic
compounds, which can be separated by HPLC techniques. The common detectors
(refractive index, UV-visible spectrometer) will not respond specifically
to arsenic compounds. The identification of arsenic-containing fractions is,
therefore, difficult if not impossible unless element-specific detectors
with high sensitivity are available.
Flame atomic absorption spectrometers have been interfaced with
chromatographic columns for the element-specific detection of several metal
ions (54). The detection limits were relatively poor primarily because of
peak spreading in the chromatographic column. Cantillo and Segar described
a system in which the effluent flow is stopped at fixed time intervals
14
-------
for introduction of a sample into the graphite furnace (55). This system
gave better sensitivity but the total analysis-time was very long. A
graphite furnace atomic absorption spectrometer (GFAA) combines the advantage
of element-specificity with high sensitivity for many elements. Two automated
systems combining HPLC with GFAA as an element-specific detector were
simultaneously and independently developed. The first system uses a Perkin-
Elmer GFAA with a specially adapted autosampler (56). The second system,
developed for this project, employs a Hitachi-Zeeman GFAA with a sample
valve, an injector and associated electronics to control the analysis sequence
(57). .
Description of the HPLC-Hitachi-Zeeman GFAA System
In order to determine arsenic quantitatively in the HPLC column
effluents an automated HPLC-GFAA interface had to be developed capable of
controlling the following functions sequentially and reliably.
• Lowering .the*injection tube into the cool graphites cup.
• Turning the-sample valve into the inject position.
• •Actuating the pump to inject nickel'solution into the graphite
cup (if desired) and controlling time of its operation to
deliver a predetermined volume of this solution. :
• Turning on the Start/Stop switch to. initiate the AA analysis,
sequence-. • '•••• •' ' '''- • .' . ' :-''• '•'.""; >•"' ''.'"'•• •
• Raising: the injectipn tube out of,.the graphite cup and the -,,:, ...
. furnace. ....... .. '.'..' '
• Returning the sample valve to the-'load: position in preparation -•.••••••-
for the next injection.. ' V
•Allowing the GFAA-to perform the analysis and the graphite
, furnace to cool.sufficiently to permit injection of. the next . '
sample without splattering. . . . ;x • ' '••••••
• Actuating the Start/Stop switch to terminate the graphite
furnace cooling cycle. . :
• Returning to the first function to repeat.the sequence of events.
Although there are several possibilities to accomplish .the required
sequencing and timing of events, the design chosen divides the interface
into the sampling sub-system, the injection sub-system and the control and
timing circuit. A block diagram for the HPLC-GFAA system is given in Fig. 2
and schematics for the interface are presented in Fig. 3. The specifications
and sources for the electronic components are summarized in Table 1.
The Sampling Sub-System: The sampling sub-system consists of an Altex, eight
port, Teflon Slider Injection Valve (Altex catalog #201-14) with two attached
pneumatic actuators and connecting tubings as shown in Fig. 4. One of the
eight ports (Fig. 4) is closed in one of the two possible slider positions.
To this port the nitrogen line is connected. When the nitrogen port is in -.
the open position.it will connect to the. "bottom sample loop port" (BSLP).
The BSLP is connected by a 0.6 mm ID Teflon tubing to the port directly
above it (top sample loop port, TSLP). The length of the Teflon tubing is
adjusted to deliver the volume (40 uL in the case; of the HPLC-GFAA system
15
-------
SOLVENT
RESERVOIR
1
HPLC
PUMP
I
SAMPLE
INJECTION
PORT
—
Flow
Elec
RECORDER
UV :
DETECTOR
.;"
HPLC COLUMN '
of Liquid
trie Connections
RECORDER
DIFFE
REE
INDEX
' 1
SLID
INJE
VALV
;.
RENTIAL
RAQTIVE
DETECT,
;
ER ••
CTION
E
.
.
Y
X
T—
•
^
NITROGEN
RESERVOIR
Mi SOLUTION
RESERVOIR
AND PUMP
TIMING
AND
CONTROL
CIRCUIT
1 '/
SAMPLE
INJECTION
SYSTEM;
LINEAR
ACTUATOR
FRACTION
COLLECTOR
.u— Jj
HITACHI
ZEEMAN
GEAA
RECORDER
Figure 2. Block diagram for the HPLC-GFAA system.
(From reference 57; redrawn with permission of Gordon and Breach Science Publishers)
-------
i* -E
Spll
^ to I .-
Boara 1
.(•c'T
j_
T
0
S
CD
in ,
m
•
OJ
o
a
ftnqrfl
X
T
CM
U
-2.
1
7
M
U
1
O
1
in
u
a •
-a. "i
0
a.
Trioaer/Carner Culoll ConlroB
LOAD. SAMPLE LOOP
1. DETECTORS '•'. .; '.t
NITROGEN
FRAC. COLLEC TOR
4. A.A. ' v ,'• ';
SAMPLE LOOP " 4
6 BVPASS .LOOP i
1 P: Match points
i • B: Interconnecting cables between boards and other components
Figure 3. Schematics for the HPLC-GFAA interface.
-------
TABLE 1. SPECIFICATIONS AND SOURCES FOR ELECTRONIC COMPONENTS FOR THE
HPLC-GFAA INTERFACE
COMPONENTS DESIGNATION*'
SPECIFICATIONS AND SOURCES
C 1,2
C 3,7,9,22,36
C 4,5,15,29,43-49
C 6
C 9,10,11,18,19,25,26
32,33,39,40
C 12,17,20,24,27,31,34,
38,41
C 13,14,16,21,23,28,30,
35,37,42
D 1-26
F 1
1C 1
1C 2
1C 3-7
R 1,4,5,8,9,12,13,16,
17,20
R 2
R 3,6,7,10,11,14,15
18,19
Relay 1-6,9,10
Relay 7
Relay 8
S 1,4,5
S 2,3
SO 1-4
SD 5
T 1
VR 1-4
dpdt
100 vf, 50 V electrolytic capacitors
0.1 uF, 500 V ceramic disc capacitors
0.01 uF, 1000 V ceramic disc capacitors
10 yF, 25 V tantalum electrolytic
capacitors
0.01 yF, 50 V ceramic disc capacitors
100 uF, .50 V electrolytic capacitors
0.001 uF, 50 V ceramic disc capacitors
600 V PRV, 1.0 A silicon diodes,
Sylvania model ECG116
1.5 A quick blow fuse
30 A, 70 PRV epoxy bridge rectifleer,
type VK 148 distributed by Allied
Electronics
+12 V- voltage regulator, type LM 340-
12
LM 556 CN dual monostable multi-
vibrators (timers)
10 kn, 1/4 W resistors
5 Mfl, 1/4 W variable resistors
2 MSJ,. 1/4 W variable resistors
12 V-DC, 60 mA SPOT relays; Calectro
part #974
SPDT mercury-wetted reed relay, series
W133MPC, Allied Electronics
5 V-DC, 2.3 mA SPDT relay type R10S-
E1-Y1J, Allied Electronics
Single pole, single throw switch
Single pole, single throw switch
110 V-AC, 9W solenoid valves, NC model
#61P18C3, Valcor Engineering Copr.,
Kenilworth, N.J.
110 V-AC, 7 W'three-way solenoid valve
model 0V55KJ8NGV, Peter Paul
Electronics, New Britain, Conn.
35 V, 1.5 A Centertap transformer
150 V RMS metal oxide varistors, model
#750, Workman Electronics,
Sarasota, Florida
Double pole double throw switch
*Coraponent designation as used in Figure 3.
18
-------
sample loop
TSU>A
to fraction
collector
VWVWWWVWVW
. pressure relief loop
Figure A. The sampling sub-system of the HPLC-GFAA interface.
to GFAA
from HPLC column
-------
described here) required by. the graphite furnace spectrometer. When the
nitrogen port is open, the TSLP is connected to .the port leading to the
injection device. The bypass loop uses the .two ports opposite the BSLP and
the TSLP. The volume of the bypass loop should be made as .small as
conveniently possible. To the port opposite the nitrogen port the tubing
coming from the HPLC column is connected. The port above the nitrogen port
and opposite the ports leading to .the injection device is the outlet port
for the HPLC eluent going to a fraction collector or to waste. Several feet
of a 0.3 mm ID Teflon tubing (pressure relief loop, Fig. 4) connect the
tubing coming from the column with the line leading .to the injector.. This
restrictor tubing is required for relief of any pressure built up during the
time the valve switches between load and injection postions. Pressure
build-up may damage the windows in the refractive index detector.
During the load cycle the nitrogen port, the bypass loop and the
injection port are closed. The effluent passes from the inlet line through
the sample loop to the fraction collector. The valve is switched from
loading to injection by two pneumatic actuators, which .are controlled by the
pressure control system shown in Fig. 3. ^Nitrogen at 60 psi is introduced
via the inlet to selenoids SD 2 and SD 3, which determine whether nitrogen
is routed to outlet 1 or outlet 2. Outlets 1 and 2 are connected to the
actuators on the slider injection valve. Outlets 3 and 4 are exhaust ports
controlled by selenoids SD 1 and SD 4, respectively. Selenoids SD 1 and SD
3 operate simultaneously as do selenoids SD 2 and SD 4 releasing pressure
from one actuator and applying it to the other thereby changing the slider
injection valve from the loading to the inject position. During the
injection cycle the column effluent travels from the inlet line through the
bypass loop to the fraction collector. The sample contained in the sample
loop is pushed .out from the loop into the injection line by nitrogen pressure.
The nitrogen at approximately: 5 psi comes from a nitrogen tank. To deliver
nickel solution to the graphite cup a peristaltic pump with a capacity of 200
pL per minute is activated for the required amount of time to push the
appropriate volume of nickel solution (10 yL in the.case of the HPLC-GFAA
system described here) into the nitrogen line just before the nitrogen
port. Nitrogen pressure then transports the nickel solution through the
_sampleL loop and theinjection line into_the ..graphite, .cup_. The entire cycle
begins \ag-ain after the slider injection, valve .is__switched._b.ack_.to__the_load
position. .
The Injection Sub-System: The injection system consists of an American
Design Components 2-% inch stroke 24 V-DC linear actuator, a filter
capacitor (0.2 uF, 200 V), injection tubing and mounting brackets (Fig. 5).
The positioning bracket is slotted for exact positioning of the injection
probe on the mounting bracket. The delivery tube is a section of 1/16
inch OD 316 Stainless Steel tubing inside a section of 1/8 inch OD 316
Stainless Steel tubing (guide tube). The back support plate has a slot into
which the injection bar extends holding it straight and providing stops at
the top and the bottom for the actuator stroke. The mounting bracket has a
viewing port through which the injector can be exactly positioned. The
injection sub-system must be very carefully positioned and adjusted. The
samples must be deposited into the center of the cup without splattering
and without touching the sides of the graphite cup. The delivery tube must
20
-------
electrical
connections
filter-capacitor
delivery tube-
positioning bracket...
guide tube-
to
sampling
valve.
o
u
tj
OJ
0)
B
injection bar'
1/8" pipe to
1/8" tube
fitting
mounting bracket
•walls of GFAA furnace and. magnet cavity-
Figure 5. The injection sub-system of the HPLC-GFAA interface.
21
-------
just touch the deposited sample but not the sides of the cup.
The Timing and Control Circuits: The timing and control circuits are divi-
ded into three printed circuit boards (Fig. 3). Board 1, the power supply
board, consists of the. transformer T 1, which converts 110 V-AC to 17.5
V-AC, and>the rectifier IC'l, which changes 17.5 V-AC to 24.7 V pulsating
DC. Ripple is removed by.capacitor C 1.
Board 2. contains additional filtering and despiking capactitors C 2-5
and relays 1-6. The despiking capacitors remove line transients, which
might damage the voltage regulator 1C 2 on board 3. The SPDT relays 1-6
control the GFAA, the injector and the actuators on the slider injection
valve.
Relay 1 is closed by a signal from: 1C, 3B set at 1 second. Closing of
relay 1 actuates relay 7 which is connected in parallel with the start/stop
switch on the AA and thus terminates the cooling cycle, the length of which
is controlled by 1C 3A.
Relay 2 activated by timer--IC 4A switches-on the motor of the linear
actuator lowering the delivery tube into the graphite,cup. 1C 4A is set to
allow the injection bar to reach the bottom stop in the back support plate.
Relay 3 controlled by timer 1C 4B activates solenoids SD 1 and SD 3
which cause the slider injection valve to change from the load to the inject
position. At the same time the nickel pump is activated and kept on by 1C
4B long enough for 10 yL nickel solution to flow into the injection system.
Relay 4 controlled by timer 1C 5B switches on the motor of the linear
actuator to raise the delivery tube out of the graphite cup.. 1C 5B is
set to allow the injection bar to just reach,the top stop in the back support
plate.
Relay 6 controlled by timer 1C 6A set at one second activates solenoids
SD 2 and SD 4 which return the slider injection valve to the load position.
The silicon diodes D 21 - D 26 and the capacitors C 41 - C 49 minimize line
transients propagating into the timing circuit.
Board 3 contains the timing circuit for the HPLC-GFM interface
consisting of five cascaded LM 556 CN dual, monostable multivibrators (Fig.
3, Table 1) and one 12 V-DC LM 340T 12 voltage regulator, 1C 2. 1C 2
converts the 24.7 V-DC coming from Board 1 to a regulated 12 V-DC providing
power for the timing circuit and the relays 1-6. Each one of the dual
multivibrator packages (1C iA and 1C iB; i=l-7) controls two events. The
timing in each multivibrator is controlled by a capacitor and a variable
resistor. This arrangement makes it possible to time each multivibrator
independently of the other nine multivibrators. The first multivibrator 1C
3A (Fig. 3) is triggered by relay 9, which is activated through relay 8 by
the timer (1C'3.4-9). in the GFAA determining the length of the atomization
cycle (Trigger/Carrier Cutoff Controls, Fig. 3). 1C 3A remains on until
the graphite cup has cooled. The output pin of each multivibrator (with
the expection of the last multivibrator 1C 7B) is connected to the input pin
22
-------
of the following multivibrator through a 0.001 uF capacitor. When a multi-
vibrator has timed out, the output pin returns to ground triggering the next
multivibrator. This arrangement assures that the events occur in the intend-
ed sequence, because a timer cannot start until the preceding one has timed
out. The pin configurations and other pertinent specifications of the multi-
vibrators are described in the manfacturer's literature (59). ••
The timers 1C 6B, 7A and 7B are not utilized presently. They provide,
however, the capability of dual injections of sample into the graphite cup
to increase the sample size to be analyzed. The second injection will come
after the drying cycle for the first injection has been completed.
The diodes D 1 - D 10 sharpen the triggering pulses to the respective
timers 1C 3 - 1C 7. The bypass .capacitors C 10,11,18,19,25,26,32,33,39 and
40 reduce the susceptibility of the timers 1C 3 - 7 to noise., The
despiking capacitors C 8,15,22,29,36,43 remove line transients.
The switch S 2 interrupts the timing sequence and resets'all timers.
The switch S 3 allows the manual starting of the timing sequence. ,,...
The Trigger/Carrier Cutoff Controls of the interface are mounted inside
of the power supply of the GFAA- Relay 8is connected to 1C 34-9" in the. ' ;
GFAA power unit- When 1C 34-9 initiated, the atomization .cycle relay 8 is.
activated, which in turn activates relays 9 and 10. Relay 9 triggers timer
1C 3A. Relay 10 turns on solenoid,SD 5 which terminates the carrier gas flow
through the graphite furnace during the atomization cycle. Capacitor C 43
and varistor VR 4 control voltage transients.: " " - :
Relay 7 is mounted outside the GFAA power unit together with the switch.
dpdt (double pole-double throw). This switch isolates the GFAA from the
interface allowing normal operation of the spectrometer.
The HPLC-GFAA system has functioned almost flawlessly during the past
three years.. The major problems encountered.initially were;caused,by line
spikes propagated through the building wires' to the electronics of the
_interface, where tphey triggered^relays ^.^
events. The capacitors installed as shown in the. schematics (Fig. 3) .elim-
inated this interference.
The time interval between two analyses can be made as short as 30
seconds. In most cases analyses carried out in one minute intervals pro-
duce satisfactory chromatograms. The1 firing of the graphite furnace Pvpry
_?^e_^aHsld_lhA..PA^^^^ ._
fore, a refrigerated glycol/water mixture;was circulated through the magnet's
cooling system replacing tap water used under normal operating conditions.
Because the glycol/water mixture was pumped at.much lower pressure than the
tap water, the water pressure switch in the GFAA had to .be deactivated. The
instrument is still protected by an overheat shut down switch. The time
between analyses can be further shortened by using cooled gases to bring
the graphite cup to room temperature after completion of the atomization
cycle and by shortening the drying cycle.
. •• 23 "••'•'' ' •'•'• '"•
-------
The Hitachi-Zeeman GFAA Model 170-70 has a detection limit for arsenic
of 10 picograms. This sensitivity is, of course, retained in the HPLC-GFAA
system for each separate analysis. Upon migration through the chrbmatographie
column the arsenic compounds are separated and spread out into bands. Ali-
quots of 40 pL are withdrawn from the effluent. The 40 \iL aliquots taken
from the center of a band, where the arsenic concentration is highest, must
each contain at least 10 picograms of arsenic. The detection limit of the
HPLC-GFAA system is therefore strongly dependent on the degree of band
spreading. Under optimized conditions ten nanograms of an arsenic compound
placed on the column should produce discernible GFAA signals. The sensitivity
of the HPLC-GFAA system can be increased by optimization of the chromato-
graphic parameters, by using multiple injection - drying cycles before
atomization or by pre-concentration of the arsenic compound on the top of
the chromatographic column.
THE SPECIATION OF ARSENITE, ARSENATE, METHYLARSONIC ACID; DIMETHYLARSINIC
ACID, ARSENOCHOLINE AND ARSENOBETAINE USING" THE. HPLC-HITACHI ZEEMAN GFAA
SYSTEM .
Arsenite, arsenate,, methylarsonic acid, dimethylarsinic acid, arseno—
choline and arsenobetaine are arsenic compounds which might be present in
natural water-not polluted by human activities. Methods were developed for
the separation and determination of these compounds by HPLC employing a
Hitachi Zeeman graphite furhance atomic absorption spectrometer as an
arsenic-specific detector. The HPLC-GFAA system will not only indicate the
presence or absence of the arsenic compounds for which the separation methods
were designed but will also show the presence of other arsenic compounds.
The possibility that an unknown.arsenic compound has the same retention
time (volume) as one of the arsenic compounds named above does exist.
However, chromatograms using several mobile phases and mass balances should
reveal, the presence of additional compounds.
Separation of inorganic arsenic, arsenocholine and arsenobetaine:
Arsenocholine, arsenobetaine arid inorganic arsenic (arsenite and
arsenate) were separated on a micro-particulate (10 y) C-18 reverse phase
column with the sodium salts of heptanesulforiic acid or dodecylbenzenesul-
fonic acid as counter ions for the arsonium salts (57 i 58).
Arsenocholine, arsenobetaine and inorganic arsenic required mixtures of
acetonitrile, water and counter ions as mobile phases, which are made 1.0 M
with respect to acetic acid to suppress the dissociation of the carboxylic
acid group in arsenobetaine and thus achieve greater retention on the
reverse phase column. By varying the ratio acetonitrile/water in the eluent
the retention of arsenocholine can be changed from total retention to co-
elution with arsenobetaine (Fig. 6). Arsenobetaine cannot be totally
retained with heptansulfonic acid as the counter ion. However, total
retention on the top of the column was achieved with dodecylbenzenesulfonate.
This procedure is useful for the pre-concentration of arsenobetaine from
very dilute solutions. Arsenite and arsenate were not retained at all with
either of the sulfonates on the C-18 reversed phase column. These inorganic
24
-------
mobile phase:
0.005 M
water 20 mL
acetonitrile 30 mL
acetic acid 6 mL
Na dodecylbenzenesulfonate
C-18 reversed-phase column
arsenite
arsenate
1 Ug
injection
/
.... 1
arsenobetaine
1 ug
1 - •' '
i i i i 1 i i i i 1 i i
1
flow rate: 0.5 mL/min
arsenocholine
1 Ug
> : * '
mobile phase
changed to
. 100% acetonitrile .
i 1 I 1 i i 1 t i i I
o :
"inutes
15
2O
25
30
' I ' 1 1 1 I I 1 j 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
"-^injection
mobile phase: water 95 mL
• -.,
?i.....J!\,
I:
arsenite ji
arsenate j
1 Ug j
i
'*
acetonitrile 5 mL
^
|
. • • acetic. acid 6 mL.
0.005 M heptanesulf onic acid
• '.i:.-.:'. flow rate: 0.5 mL/min
C-18 reversed-phase column .
If
I
arsenobetaine
1 ug
"
arsenocholine
1 ug
1
Figure 6. Separation of inorganic arsenic (arsenite and arsenate), arseno-
choline and arsenobetaine by HPLC-GFAA on a C-18 reversed-phase
column.
(From reference 57; redrawn with permission of Gordon and Breach Science
Publishers)
25
-------
arsenic compounds eluted with the solvent front. Peaks caused by arsenic-
free impurities in. the solvent or the samples were recorded by the refractive
index (RI) and UV detectors. These detectors were not sensitive enough to
respond to the arsenic compounds. However, the GFAA detector has the required
sensitivity to show the presence of the arsenic compounds (Fig. 6). No
peaks other than those indicating the presence of arsenic compounds will
appear in the GFAA chromatograms facilitating the quantitation and identifi-
cation of arsenic compounds even in complex matrices which would produce a
large number of RI and UV signals.
The separation and determination of arsenite, arsenate, methylarsonic acid and
dimethylarsinic acid: A C-18 reversed-phase column can be used to separate
ionic species provided the ions are paired with a suitable lipophilic counter
ion to produce a rather hydrophobic ion pair. Tetrabutylammonium phosphate
(TBAP) at 0.005 M concentration in the appropriate mobile phase has been
recommended as the reagent for such paired-ion chromatography. Arsenite and
arsenate migrated almost with the solvent front through a C-18 reversed-
phase column with water-methanol (95:5 v/v) at pH 7.3 as the eluent in the
absence of TBAP but were completely separated in the presence of TBAP
(0.005 M) with retention times of 19 min and 11 min for arsenate and
axsenitE^ respectively (Fig. 7). When a mixture of arsenate, arsenite,
dimethylarsinic acid (DMAA) and methylarsonic acid (MAA) was chromatographed
with 0.005 M TBAP in pure water as eluent, which should increase the
retention time of lipophilic ion pairs, the peaks caused by arsenite and
DMAA overlapped. The MAA signals were located under these conditions between
those belonging to arsenite and arsenate. TBAP does not cause sufficient
retention and separation, even with water as the mobile phase and has the
additional disadvantage, of containing phosphate. Phosphoric acid used to
adjust the pH of the mobile phase and ammonium phosphates added as a source
of counter ions depressed the arsenic signals (Fig. 8) Such interferences
have been mentioned in the literature (60). Although phosphates have been
recommended for use in HPLC because of their compatibility with the
stainless steel columns, nitrates or perhaps organic anions should be used
for work with arsenic compounds.
A much better separation was achieved with tetraheptylammonium nitrate
(THAN). Utilizing water-methanol (75:25 v/v) of pH 7.6 saturated with THAN
as the mobile phase all four arsenic compounds were satisfactorily separated
(Fig. 9). Arsenite eluted almost with the solvent front, followed by DMAA,
MAA and arsenate. The retention times of these compounds increase with
increasing pKa values of the arsenic acids. Arsenous acid (pKa = 10) remains
undissociated under these conditions, does not pair with the tetraheptyl-
ammonium cation, retains its polar character and is, therefore, not adsorbed
onto the organic, reverse-phase column material. The weak acid DMAA (pKa6.3)
owes its retention time of 10 min to the two methyl groups rather than to its
presence in solution as the tetraheptylammonium dimethylarsinate ion pair.
The moderately acidic MAA (pKa3.6) is very likely combined with the ammonium
ion and adsorbed by the column material as the tetraheptylammonium methyl-
arsonate resulting in a retention time of 23 min. Arsenic acid with a pK
value of 2.2, the most acidic compound among the four arsenic derivatives
investigated, is strongly retained on the column head probably as the ion
26
-------
RI
mobile phase: water 95 mL \
methanol 5 mL ';
flow rate: 0.5 mL/mln \
C-18 reversed—phase column \
injection
arsenite 5' ug
arsenate 4.5 ug
without
tetrabutylamraonium phosphate
minutes
12
16
2O
inj ec tion
mobile phase: water 95 mL
methanol 5 mL
flow rate: 0.5 mL/min
C-18 reversed-phase column.
arsenite
5.0 ug
with .'•'•'•
•0.005 M tetrabutylammonium
phosphate
arsenate
4-5 pg:'-
Figure 7. Separation of arsenite and arsenate by HPLC-GFAA on a C-18 reversed-
phase column in the absence and presence of tetrabutylammonium
phosphate.
(From reference 58; used with permission of Elsevier Scientific Publishing
Company) .
27
-------
pH 7.3
mobile phase: water 95 mL
methanol 5 mL
flow rate: 0.5 mL/min
C-18 reversed-phase column
0.005 M tetrabutylammonium phosphate
pH 5.0
/I
1 \
pH adj.usted with phosphoric
acid
arsenate 5 ug
1
pH 4.0
h
1 1 (
19
21
16 18
retention time in minutes
16
18
Figure 8. The influence of phosphate on the GFAA arsenic signals.
pair (R.N)'t' HAsO,. The mobile phase had to be changed to pure methanol to
elute the arsenate.
The GFAA signals produced by the arsenic compounds eluting from the
columns can be quantitated either by calculating the area under the curves
obtained by connecting the maxima of the GFAA signals or by summing the
heights of the signals corresponding to a particular compound. A comparison
of the two methods to convert the GFAA signal clusters corresponding to a
particular arsenic compound to numerical values to be plotted against the
amounts of arsenic, proved that the summation of peak heights produces the
same results as the determination of the areas under the signal clusters
(58). Examples of calibration curves are given in Figs. 10 and 11.
The sensitivity and general response of the GFAA instruments is rather
strongly dependent on the characteristics of the graphite cups and other
parts of the furnace system. The cups, for instance, tend to deteriorate
with use and reduce the sensitivity. A new cup does not necessarily produce
28
-------
4JD
t
1 3JB
2 2 . ' ' ' •: '
.y
1.,,
' , ' ' \ ..,
20
CH
3/
•
,""""
03H2
Na2HA&04
-
• —
. —
1
1 , 1
30 40
RETENTION TIME (MINUTES)
Figure 9. The separation of-arsenite, arsenate, dimethylarsinic acid and methylarsonic acid by
HPLC-GFAA on a reversed-phase column in the presence of tetraheptylammonium nitrate.
(From reference 58; used with permission of Elsevier Scientific Publishing Company)
-------
120
UJ
O
to 6tQ
4.0
200
• (SO®
ng ARSENIC -4
Figure 10. HPLC-GFAA calibration curve for sodium arsenite.
(From reference 58; used with permission of Eisevier Scientific Publishing Company)
-------
12.O
I
8.0
£ §
^ I—I
to
tn
4.0
0.0
NaAsOz x x>—-'
.^•-"
[CHj)2AsQzH
I
100
200 300
^ng ARSENIC— j
400
500
Figure 11. HPLC-GFAA calibration curves for sodium arsenlte, dimethylarsinic acid, methylarsonic
acid and disodium hydrogen arsenate.
(From reference 58; used with permission of Elsevier Scientfic .Publishing Company)
-------
the same response as one subjected to a number of analysis cycles. This
behavior necessitates the frequent use of standards. The Hitachi Zeeman
GFAA with a sensitivity of 10 picograms of arsenic serving as a detector
for the effluent from a C-18 reverse phase column is capable of detecting
several nanograms of arsenic under ideal conditions (Fig. 10) and approxi-
mately 10 nanograms of sodium arsenite in the presence of DMAA. The
determinations are accurate with ±5-10 percent and are intrinsically
background-corrected. The causes of the' changes in the slopes of the calibra-
tion curves are not clear yet and need to be .investigated further.
Various arsenic compounds produce different GFAA responses for the same
amount of arsenic (58^. As much as a two-fold difference in intrinsic GFAA
sensitivity was: ojjseryedjwith: DMAA and MAA as well as between their sodium
salts. The less volatile species give a more intense signal than the more
volatile compounds. Onrthe basis of these results quantitative information
for unknown arsenic compounds obtained'with the'HPLC-GFAA system must be
regarded to be qualitative, until the compounds are identified and calibra-
tion curves constructed with pure samples.
The HPLC-GFAA system has been shown to be applicable to the qualitative
and quantitative, determination of arsenite, arsenate,- methylarsonic acid and
dimethylarsinic acid in the" ppb range with detection limits approaching 5 ppb
under favorable conditions. The GFAA detector responds only to a specific
element and is better suited for trace element compound determination than
UV detectors. Development of column materials will lead to better separa-
tions. Recently the use of a proprietary special anion-exchange column for
the separation of arsenic compounds (61) and an automated system for-the
analysis of arsenite, arsenate,. methylarsonic acid, 4-aminophenylarsonic
acid and dimethylarsinic acid 'based;on -the separation with a Dionex-anion
exchange column (62) was reported. , . . . :
DIFFERENTIAL PULSE POLARQGRAPHIC DETERMINATION OF ARSENATE AND ARSENITE
Arsenite is reducible at the dropping mercury electrode. A very de-
tailed review of the literature about the polarographic behavior of arsenite
appeared in 1969 (63). In hydrochloric, perchloric, sulfuric or acetic acid
media arsenite is reduced polarographically to elemental arsenic and then to
arsine. The best supporting electrolyte among those thus far investigated
is hydrochloric acid. In 1 M hydrochloric acid the detection limit is 0.3
ppb with a linear response up to 60 ppm (37).
Arsenate is polarographically inactive under the conditions used for
the polarographic determination of arsenite. However, arsenate produced
polarographic waves in '11.5 M hydrochloric acid (64) and in 2 M perchloric
acid in the presence of pyrogallol (65). Arsenate gave three polarographic
waves in pyrogallol, the heights of which increased linearly with concentra-
tion. Solutions with concentrations lower than 37 ppm arsenic were not in-
vestigated (65). Pyrogallol is difficult to handle, is easily oxidized by
air and may pose health hazards. The use of a 11.5 M hydrochloric acid
medium for routine analysis is untenable. Several other polyhydroxy
32
-------
compounds were investigated as reagents to make arsenate polarographically
active and develop a method to determine arsenate and arsenite simultaneously
at ppb levels.
Differential Pulse Polarographic(DDP) Determination of Arsenate in the
Absence of Arsenite
Pyrocatechol, pyrogallol, 1,2-ethanediol, 1,2-propanediol, 3-chloro-
1,2-propanediol, 1,2-butanediol, 2,3-butanediol, D-ribose, sucrose, fructose
and D-mannitol were investigated.. All. of these polyhydroxy compounds made
arsenate polarographically active.. The peak heights of the DPP curves for
the reduction of 10 ppm As(arsenate) in solution 2.0 Q with respect to
perchloric acid and 0.5 >1 with respect to the polyhydroxy compound are
listed in Table 2. An example of the DDP curve produced by arsenate in the
presence of mannitol is given in Fig. 12. Under these conditions D-mannitol
TABLE 2. PEAK HEIGHTS OF THE DIFFERENTIAL PULSE POLAROGRAPHIC CURVES
AT -0.55 V OBTAINED WITH 2.0 M AQUEOUS PERCHLORIC ACID ,
SOLUTIONS OF 100 ppm As(ARSENATE) IN THE PRESENCE OF .
VARIOUS POLYHYDROXY COMPOUNDS AT 0.5 MCONENCTRATION
Polyhydroxy Compound Peak: Height at -0.55 V in mm
3-Chloro-l,2-propanediol 45.7
Pyrocatechol 68.6
Pyrogallol . " 71.1
1,2-Ethanediol 93.9
1,2-Propanediol : ."'"" , 132.1
1,2-Butanediol . 142.2
D-Ribose : • : 170.2
Sucrose 172.7
Fructose 182.8
2,3-Butanediol 185.4
D-Mannitol 203.2
produced the largest signal.
The heights of the reduction waves are influenced by the concentration
•of mannitol. As shown in Fig. 13 these heights become independent of the
mannitol concentration after three grams of mannitol had been added to 50 mL
of the sample. To make certain that, an excess of mannitol is always present
,••..•"•'•'•'•'•• 33 •••'"••''"''• ,. • ' . •
-------
AE
current: 0.1 mA
scan rate: 2 mV/sec
drop time: 2 sec
ref. electrode: satd. KC1
mod. amplitude: 100 mV
initial potential: 0.0 V
scan direction: negativ
supporting electrolyte: 2 M
mannitol: 0.5 M
1. 2 ppm As(arsenate) '
2. 4 ppm As(arsenate)
3. 6 ppm As(arsenate)
4. 8 ppm As(arsenate)
5. 10 p'pm As(arsenat'e)
I
I
I-
-QA Potential, V -Q.6
-O.8
I
I
I
Ai
current: 0.01 mA
other parameters same as
given above
500 ppb As(arsenate)
with mannitol ..••''
..X
500 ppb As(arsenate)
without mannitol
Figure 12. Differential pulse polarograms of arsenate solutions two molar
with respect to perchloric acid in the presence of D_-mannitol.
-------
2O
0 1§
10
o
current: 0.01 mA
scan rate: 2 mV/sec
drop time: 2 sec
ref.. electrode: satd. KC1
mod., amplitude: 100 tnV
initial potential: 0.0 V
scan direction: negative
supporting electrolyte:
2. M HC104
20 ppm-AsCarsenate)- —
1
o
2 3 4
Grams of D-Mannitol; Added to 50 mL
Figure 13.
The dependence of the height of the differential pulse
polarographic curve of arsenate at -0.55 V on the concen-
tration of D-mannitol.
in the solutions to be polarographed and thus assuring the independence of
the heights from the mannitol concentration, all further experiments were
carried out with solutions prepared by addition of 4.5 g of mannitol to 50 mL
of the acidified sample. These solutions are approximately 0.5 M with
respect to mannitol.
Sulfuric, hydrochloric, acetic, phosphoric and perchloric acid were
investigated as supporting electrolytes in the DPP determination of arsenate
in solutions containing mannitol at 0.5 M concentration. The solutions were
all 2.0 M with respect to the acid to be investigated. The results of these
35
-------
experiments are summarized in Table 3. Perchloric acid produced the largest
signal among the acids tested and was used in all further experiments.
TABLE 3. HEIGHTS OF THE FIRST DIFFERENTIAL PULSE POLAROGRAPHIC
MAXIMUM IN THE REDUCTION OF 9.5 ppm As(ARSENATE) IN AQUEOUS
ACIDIC MEDIA IN THE PRESENCE OF D_-MANNITOL AT 0.5 MOLAR
CONCENTRATION _^
Acid Peak Height in mm Potential of Maximum
Sulfuric
Hydrochloric
Acetic
Phosphoric
Perchloric
7.6
15.2
17.8
20.3
22.8
-0.45 V
-0.55 V
-0.90 V
-0.45 V
-0.44 V
*A11 solutions were 2.0 M with respect to the acid.
The reduction waves and the current maximum shifted anodically and the
heights.of the shoulders on the anodic and cathodic side of the current
maximum increased with increasing perchloric acid concentration (Fig. 14)..
For example, an increase of the perchloric acid: concentration from 0.94 M
to 5.8 M shifted the first reduction wave from -0.57 V to ^-0.5CTV lind ~""~
.doubled the peak height. . Because of the difficulty of adjusTing"water
samples to a perchloric acid concentration of five molar, two molar
perchloric acid was chosen as a compromise providing satisfactory sensitivity
'and relative ease of sample preparation.
The DPP curve of arsenate in an aqueous perchloric acid/mannitol
solution is characterized by maxima at -0.55 V and -0.75 V (Fig. 12). Above
an As(arsenate) concentration of 500 ppb a current maximum appears at
-0.59 V which increases in intensity with increasing concentration. The
peak at -0.55 V merges at As(arsenate) concentrations of approximately 5 ppm
into the current maximum and becomes a shoulder. Another rather low
intensity peak between -0.7 V and -0.8 V might be obscured at low arsenate
concentrations by the solvent breakdown and at high arsenate concentrations
by the current maximum. Since arsenite is reduced in 2.0 M perchloric acid
in the presence of mannitol at -0.34 V, it is likely that the first arsenate
reduction wave at -0.55 V is caused by the reduction As(V) -* As(0) followed
by the conversion of As(0) to AsH3 at approximately -0.7 V. Experiments
which would verify these postulated steps have not been carried out yet.
The differential pulse polarographic maximum or shoulder at -0.55 V
generated by arsenate in 2.0 M perchloric acid in the presence of mannitol
is suitable for the quantitative determination of arsenate. The current
maximum could perhaps be used for the determination of arsenate at
36
-------
Ai
AE
molar concentration of HC10.
5.8 2.8 . 2,0 0.94
current: 0.1 mA
scan rate: 2 mV/sec
drop time: 2 sec
ref. electrode: satd. KC1
mod. amplitude: 100 mV
initial potential: 0.0 V .
scan direction: negative /
A 1. A /
- y >. ;v \/
10 ppm As(arsenate)
0.5 M mannitol
-Q2
-0.4
-Q6
Potentially
-08
Figure 14.. Differential pulse polarographic curves for 10^ppm As(arsenate) in solutions
0.5 M^with respect to D-mannitol at various perchloric acid concentrations.
-------
concentrations above 1 ppm As(arsenate).
Calibration curves for the determination of arsenate in the absence of
arsenite were obtained by addition of 20 yL aliquots of an arsenate solution
of appropriate concentration to 50 mL of sample solution in the polarographic
cell. After each addition the differential pulse polarogram was recorded.
An example of a calibration curve using the -0.55 V maximum (shoulder), is
given in Fig. 15. By proper adjustment of the current, linear calibration
curves were obtained for As(arsenate) concentrations ranging from 20 ppb to
900 ppb at 2 yA, 1 ppm to 18 ppm at 50 yA and 20 ppm to 160 ppm at 0.5 mA.
Differential Pulse Polarographic Determination of Arsenite in the Absence of
Arsenate
Arsenite does not require mannitol for its reduction at the dropping
mercury electrode. In 2.0 molar perchloric acid solution arsenite is
reduced to As(0) at -0.425 V. As(0) is converted to arsine at approximately
-0.75 V. A current maximum appears at an As(arsenite) concentration of
approximately 300 ppb. The addition of mannitol to 2.0 M perchloric acid
solutions of arsenite shifts the As (III) ->• As(0) reduction anodically to
-0.34 V, but does not affect the position of the current maximum and the
wave at -0.75 V (Fig. 16). The heights of all the maxima are increased.
The location of the maxima, is dependent on the perchloric acid concentration.
The As (III) -»• As(0) peak moves from -0.49 V at 0.7 M to -0.31 V at 4.7 M
perchloric acid concentration.
An example of a calibration curve for the determination of arsenite in
the range 0.2 to 2.4 ppm As(arsenite) in the presence of mannitol is given
in Fig. 17.
Differential Pulse Polarographic Determination of Arsenate and Arsenite in
Solutions Containing Both of these Arsenic Compounds
When arsenate and arsenite are present in the same solution, differen-
tial pulse polarography employing 2.0 M aqueous HClOi^ as supporting
electrolyte, which is 0.5 M with respect to D-mannitol, will produce
polarograms with maxima at -0.34 V [As(III) -»• As(0)],. -0.55 V (arsenate
reduction), -0.59 V (current maximum) and -0.75 V [As(0) ->• As (-III)]
(Fig. 18). The arsenite reduction peak at -0.34 V [As(III) •*• As(0)] is well
separated from the other maxima and can be used to quantitatively determine
arsenite using the standard addition technique. The maximum at -0.55 V
caused by the reduction of arsenate can be used for the determination of
arsenate with some confidence only when the concentration of As(arsenite)
is between 100 ppb and approximately 500 ppb, the current is not lower than
2 yA and the arsenate concentration is equal to or greater than the arsenite
concentration. At As(arsenite) and As(arsenate) concentrations lower than
100 ppb (Fig 19) the polarograms must be obtained with a current of 0.5 yA.
Although the current maximum at -0.59 V does not appear at these low
concentrations, the arsenate reduction peaks occur as shoulders on the anodic
side of the steep solvent breakdown signals. The shoulders can be used to
ascertain the presence of arsenate and to semi-quantitatively determine
arsenate. At an As(arsenite) concentration of 100 ppb and As(arsenate)
38
-------
5O ppm As(arsenate) •JQQ
150
-ill
15
:*>
MS
T*
01
i
r
current: 0.5 mA
scan,race: 2 mV/sec
drop time:.2 sec
ref. electrode: satd. KC1
mod. amplitude: 100 mV
initial potential: 0.0 V
scan direction: negativ
supporting electrolyte: 2.0 M HC10
0.5 M mannitol
current: 0.05 mA
other parameters
as given above
I I I I
I I I I
I I
5 ppm As (arsenate) 1Q
15
15
10
current: 0.002 mA
other parameters same
. as given above
500
ppb As(arsenate)
1000
Figure 15. Calibration curves for the determination of arsenate by
differential pulse polarography in aqueous 20 M perchloric
acid in the presence of jD-mannitol.
39
-------
Aj
AE
current: 0.01 mA
scan rate: 2 mV/sec
drop time: 2 sec
ref. electrode: satd. KC1
mod. amplitude:; 100 mV
initial potential: 0.0 V
scan direction: negativ
supporting electrolyte:
2.0 M HC10,
__ 20 ppm As(arsenite)
-O.2
-0.4
Potential, V
-O.6
-O.8
Figure 16. Differential pulse polarograms for arsenite in 2.0 M aqueous perchloric acid in the
presence and absence of I)-mannitol.
-------
25
2O
e
o
•>
sr
oc
•H-
0)
1:
15
10
i i i i
i i i i i i i i
i.i i
current: 0.01 mA .
scan rate:. 2 mV/sec
drop time: 2 sec
ref.. electrode: satd. KC1
mod. amplitude: 100 mV
scan direction: negative
supporting electrolyte:. 2.0 M'HCIO,
0.5 M mannitol
I I I I I I I I I I I I I I
I I I
0.5
2O
2.5
10 15
.ppm As(arsenite)
Figure 17. Calibration.curve for the-determination of.' arsenite by differential
pulse polarography in aqueous .2.0 MHCi04 in the presence of D-manhitol.
41
-------
Ai
AE
.0.5 ppm As(arsenate)
1.5 ppm As(arsenite)
with mannitol
•v
\ i
\ \.
current: 0.01 mA
scan rate: 2 mV/sec
drop time: 2 sec
ref. electrode: satd. KC1
mod. amplitude: 100 mV
initial potential: 0.0 V
scan direction: negative
supporting electrolyte: 2.0
0.5 M mannitol
~ / "V.'
HCIOV
/ /
/
— /
11
II
; I
0.5 ppm As(arsenate)
1.0 ppm As(arsenite)
with mannitol
0.5 ppm
As(arsenate) with mannitol
0.5 ppm As(arsenate)
-0.2
Figure 18.
-0.4
-0.8
-0.6
Potential, V
Differential pulse polarograms of mixtures of arsenite and arsenate in 2.0 M
perchloric acid in the presence of ID-mannitol.
-------
M
AE
current: 0.5'yA
scan rate:' 2.mV/sec
drop time: 2 sec
ref.. electrode: satd. KC1
mod., amplitude: 100 mV
initial potential: 0.0 V
scan direction:, negative
supporting-electrolyte:.
.2.0 M HC10A
0.5 M mannitol:
77
1
-i
' /
I
'
.•'• i
-------
concentrations in the range 200 to 300 ppb the polarogram can be obtained
at 2 yA. Under these conditions the solvent breakdown curve is shifted
cathodically, a current maximum is not present, and arsenate, therefore,
may be determined by measuring the peak height of the -0.55 V maxima.
At As(arsenite) concentrations of 500 ppb with a current setting of
10 yA the solvent breakdown curve rises from the baseline beyond -0.8 V and,
therefore, does not interfere with the -0.55 V maxima. However, at
As(arsenite) concentrations of 500 ppb the current maximum at -0.59 V
appears. The arsenate reduction peaks at -0.55 V are not resolved from the
current maxima and appear as shoulders which make the quantitative determi-
nation of arsenate difficult.
At a concentration higher than approximately 0.5 ppm As(arsenate plus
arsenite) the current maximum at -0.59 V, which is produced by arsenate as
well as arsenite,. could perhaps be used to determine total inorganic arsenic.
Additional experiments are needed to clarify this issue.
Oxidation of Arsenite Followed by Polarography of Arsenate
The difficulties encountered in the direct-determination of arsenate in
the presence of arsenite can be overcome by quantitatively oxidizing
arsenite to arsenate and determining the total inorganic arsenic as arsenate
by differential pulse polarography with I)-mannitol as reagent. As reported
in the literature, the oxidation of arsenite to'arsenate was achieved by
addition of an approximately 200-fold molar excess of cerium(IV) ammonium
nitrate in two equal proportions (66). An example of a.polarogram of
arsenate obtained by cerium(IV) oxidation of arsenite is presented in Fig. 20.
The arsenite. polarographic wave (curve 1) disappears completely upon
oxidation of the arsenite'. The arsenate-containing solution does not
produce a polarogram in the absence pf mannitol (curve 2). .After addition of
marinitol the arsenate reduction peak at -0.55 V can be used to determine the
concentration of arsenate (curve 3).
The amount of cerium(IV) required for complete oxidation of 2xlO~ moles
of arsenite was ascertained by successive addition of ; 40 uL.ofxa. 0.58 M
cerium(IV) ammonium nitrate solution to the solution in the polarographic
cell. After each addition the polarogram was recorded and the peak height
at -0.34 V measured. A plot of the peak height versus the amount of
cerium(IV) added (Fig. 21) shows that at least a 30-fold molar excess is
required to completely oxidize arsenite to arsenate; To ascertain complete
oxidation of arsenite, a 200-fold excess of cerium(IV) was employed in all
experiments as recommended in the literature (66).
The excess cerium(IV) must be reduced to cerium(III) to avoid inter-
ferences in the polarographic determination of arsenate. It has been
reported in the literature that the mercury on the bottom of the cell causes
this reduction (66). However, erratic polarograms were obtained several
times, which were caused by cerium(IV) still present in the solution.
-------
Ai
AE
current: Oi05 mA
scan rate: 2 mV/sec
drop time: 2 sec •
ref. electrode: satd. KC1
mod. amplitude: 100 mV
initial potential: 0.0V
scan direction: negative •
supporting electrolyte: 2.0 M HC10,
5 ppm 4™
As(arsenite) /
X
jl '\
J. . l\
/! \\
/i \
'
5 ppm As(arsenite)
oxidized by Ce(IV)
plus mannitol /
''•••.. /
/
/
.
5 ppm As(arsenite)
bxidlzed by Ce(IV)
i
-0.2
-0.4
, V
-06.
-as
Figure 20. Differential pulse polarograms of arsenite^ and of arsenate obtained by
Ce(IV) oxidation of arsenlte.
-------
current: 0.01 mA
scan rate: 2 mV/sec
drop time: 2 sec
ref. electrode: aatd. KC1
mod. amplitude: 100 mV
20 ppm As(arsenite), 50 mL
1.33 x 10 moles
supporting'electrolyte:
2.0 M HC104
0.5 M mannitol
1.9 2.8
Moles of Cerium(IV) x
Figure 21. The dependence of the peak height of the
differential pulse polarographic maximum
at -0.34 V (arsenite reduction) as a
function of the amount of cerium(IV)
ammonium nitrate added.
The addition of approximately 0.35 g of hydroxylamine hydrochloride
lO"3 moles) to a solution containing excess cerium(IV) assures that all
the cerium is reduced to the trivalent state. Excess hydroxylamine hydro-
chloride does not interfere with the polarographic reduction of arsenate
(Fig. 22).
This procedure allows the determination of arsenite in the form of
arsenate in the absence of arsenate and the determination of total inorganic
arsenic (the sum of the arsenite and arsenate concentrations).
Fig. 23 shows calibration curves for the determination of arsenate in
the ranges 20 to 160 ppb and 500 to 1500 ppb As(arsenate). These curves
were obtained by oxidizing 20 ppb As(arsenite) or 500 ppb As(arsenite) by
cerium(IV), reducing excess cerium(IV) by hydroxylamine hydrochloride and
recording the polarograms after addition of D-mannitol. Aliquots of
46
-------
Aj
AE
.1. 0.5 ppm As(arsenite) + Ce(IV) +
+ NH2OH-HC1
2. 0.5 ppm As(arsenite) + Ce(IV) +
+ NH2OH-HC1 + mannitol
/8\
\
standard addition pf arsenate
to solution 2. !
3.
4.
5.
6.
7.
1 ppm As (arsenate)
2 ppm As(arsenate)
3 ppm As (arsenate)
4 ppm As (arsenate)
current: 0.01 mA
scan rate: 2 mV/sec
drop time: 2 sec
ref. electrpde: satd. KC1
mod. amplitude: 100 mV
initial potential: 0.0 V
scan direction: negative
0.5 ppm As(arsen.ate)
_ 8. 1.0 ppm As (arsenate)
supporting electrolyte:
2.0 M HC10,
-0.2
-0.4
-O.6
Potential, V
-O.8
Figure 22. Differential pulse polarograms of 500 ppb As(arsenate) obtained by cerium(IV) oxidation
of arsenite in 20 M perchloric acid, followed by addition of hydroxylamine hydrochloride,
addition of D-mannitol and standard addition of arsenate.
-------
00
5O
ppm As(Arsenate)
1OO
ISO
B
o
o
•s
•H
01
X
IT)
?
00
•H ,
0)
21
1.5
Figure 23. Calibration curves for the determination of arsenate obtained by oxidation of arsenite
with cerium(IV).
-------
appropriate standard arsenate solutions were then pipetted into the polaro-
graphic cell and the solutions polarographed after each addition. The
detection limits for arsenic under these experimental conditions were
calculated to be 6 ppb at the 95 percent confidence level and 10.5 ppb at
the 99 percent confidence level.
To check, the reproducibility of the method,, a 200 ppb As(arsenate)
solution was polarographed seven times with 2.0. M: HClOi^ as supporting
electrolyte in the. presence of mannitol. The peak heights were measured in
inches. The mean of the peak heights was 0.543 inches with a. standard
deviation of 0.012 and a variance of 0.0013:.
Based on the experimental results described in the previous sections,
the following procedure is recommended for the quantitative determination
of arsenate and arsenite in water samples by differential pulse polarography.
1.. Pipette 200 mL, of the water sample into a 250 mL volumetric flask.
2. Add 42.5 mL of 70 percent perchloric acid and fill to the mark
with distilled water. ; .,vv: • -;;!. ;.,.;.:. ,;.-•
3. Pipette 50 mL of this solution into the polarographic.. cell and
deaerate for 15 minutes.
4. Determine the arsenite concentration by standard addition
employing the maximum at -0.425 V. -
5. To another 50 mL aliquot of the .sample add 4.5 g D_-manriitol,.
record: the polafogram1.and check for an arsenate reduction
peak-at -0.55 V..,;..;..:...•.„',:'. .- :; •'•'•;•;• ''-....-• ''.''.'•' ";.V--'..;..;. .,;... .
6. Determine the arsenate concentration by standard addition
employing the maximum at -0.55 V if the. arsenite and. arsenate
concentrations permit it.
7. Oxidize arsenite to^arsenate with a .200-fold molar excess of
cerium(IV) ammonium nitrate by addition of the appropriate
number, of microliters of a 0.58 M cerium(IV) solution.
8. Reduce the excess cerium(IV) by addition of 0.35 g of
hydroxylamine hydrochloride.
9. Record the polarogram and determine arsenate employing the
standard addition technique. Measure the peak height at
-0.55 V.
10. Subtract the concentration of As(arsenite) obtained in Step 4
from the As(arsenate) concentration determined in Step 9 to
find the( As(arsenate) concentration originally present .in
the water sample.
49
-------
A flow sheet for the recommended analysis procedures is given in
Fig. 24. If desired, arsenite.can also be determined in another aliquot
after addition of D_-mannitol using the peak at -0.34 V.
Lacking a natural water sample containing both arsenite and arsenate,
Texas A&M University tap water was spiked with 100 ppb arsenite and 300 ppb
arsenate solutions. In 2.0 M perchloric acid 103 ppb As(arsenite) were
found in the absence and in the presence of 4.5 g D-mannitol. The
As(arsenate) concentration was determined as 300 ppb. After oxidation of
50 mL of the spiked tap water sample with cerium(IV) and addition of
hydroxylamine hydrochloride 392 ppb total arsenic were detected as arsenate.
Experimental .
Reagents
The sources of the reagent grade chemicals used in these investigations
are listed below.
1,2-Butanediol: Aldrich Chemical Company
2,3-Butanediol:. Aldrich Chemical Company
3-Chloro-l,2-propanediol: Aldrich Chemical Company
1,2-Propanediol: Aldrich Chemical Company
Vanadium(III) Chloride:. Alfa Division Ventron Corporation
Acetic Acid:.Allied Chemical
Hydroxylamine Hydrochloride: American Drug and Chemical Company
Fructose: Eastman Organic Chemicals
Ceric Ammonium Nitrate: Fisher Scientific Company
Ethylene Glycolr Fisher Scientific Company
Hydrochloric Acid: Fisher Scientific Company
Nitric Acid:. Fisher Scientific Company . ,
Phosphoric Acid: Fisher Scientific Company
Pyrogallol: Fisher Scientific Company
Disodium Hydrogen Arsenate Heptahydrate:. Fisher Scientific Company
Sodium Arsenite (NaAs02): Fisher Scientific Company
Sucrose: Fisher Scientific Company
Sulfuric Acid: Fisher Scientific Company
Pyrocatechol: J. T. Baker Chemical Company
Zinc: Mallinckrodt, Incorporated .
D-Ribose: Pfanstiehl Laboratories, Incorporated
Mercury: The Instrument Grade, triple-distilled mercury was obtained
from D. F. Goldsmith Chemical and Metal Corporation.
Instrumentation
All polarograms were recorded employing a PAR (Princeton Applied
Research, Incorporated) Model 174A Polarographic Analyzer, with a PAR Model
172A Drop Timer in the differential pulse mode on a Houston Instrument 1000
x-y Recorder.
50
-------
{Sample Containing Arsenite and/or Arsenate |
Determination of Arsenite
Simultaneous Determination
of Arsenite and Arsenate
Determination of Total
Arsenic as Arsenate
Acidify Sample to 2.0 M HC10. Acidify Sample to 2.0 M HC10, Acidify .Sample to 2.0 M HC10,
t ••'•''•' i ••"•' — «t —
.*' 1 I
Add D-Mannitol Add 200-Fold Molar Excess of
Ce(IV), then NfH-HCl
.
DPP of Arsenite
Peak Height at -0.43 V
Followed by Mannitol
DPP of Arsenite and Arsenate* DPP of Total Arsenic as
Peak Height at -0.34 V [As(III)] Arsenate; Peak Height at -0.55 V
Peak height at -0.55 V [As(V)]
* '.'•••'
Determination of arsenate in the presence of airsenite possible only in a restricted range of arsenite
and arsenate concentrations and arsenite/arsenate ratios.
Figure 24. Flow chart for the determination of arsenite and arsenate by differential pulse polarography.
-------
The polarographic cell had a side arm and Luggin probe for the
reference electrode. The cell was equipped with a medium porosity gas
dispersion tube, a platinum wire counter electrode, and the dropping mercury
electrode. A Coleman Model. 3-711 Asbestos Wick Saturated Calomel Electrode
was used in the side arm as the reference electrode. The calomel electrode
must frequently be soaked in distilled water to avoid plugging of the
electrode tip by a potassium perchlorate precipitate.
Degassing of Solutions
Nitrogen was passed through two scrubbing towers each containing an
aqueous solution of vanadium(III) chloride over amalgamated zinc. Before
entering the cell the nitrogen was passed through a presaturation tower
containing the supporting electrolyte. The cell contents were degassed for
15 minutes by passing the scrubbed nitrogen through the solution. During
the polarographic determination nitrogen was passed over the surface of the
solution. .
General Procedure for Sample Preparation
The stock solutions of arsenite and arseriate were prepared fresh for
each analysis using NaAsC>2 and Na2HAsO^'7H;>0, respectively. The concentra-
tions of these stock solutions were chosen in such a manner that the
addition of 20 microliters to the contents of the polarographic cell would
produce the desired concentration. For example, 20 microliters of a 2500 ppm
As(arsenite) solution pipetted into, a 50 mL solution in the polarographic
cell produced a 1 ppm As(arsenite) solution. The pipette used was a fixed.
volume Eppendorf micropipette. •
Standard Addition
The method of Standard. Additions was used in the analyses for arsenite
and arsenate. The additions were made by pipetting microliter aliquots of
the arsenite and arsenate solution into the side arm of the polarographic
cell. All of the solution in.the side arm was then forced into the main
chamber of the cell by applying air pressure by means of a pipette bulb.
Preparation of Natural Water Samples for Analyses
A 2QO-mL aliquot of the natural water sample was pipetted into a
150-mL_volumetric~ flask". Then~42T5 mlTof"~70% :HCTOiT~wefe" addedT Distilled","
deionized water was used tofill~the"~flisk"t6~'the mark."'"'
Supporting Electrolytes
The following acids were used as supporting electrolytes: sulfuric,
hydrochloric, acetic, phosphoric, and perchloric. The concentration of the
supporting electrolytes was made 2.0 M with the single exception of
perchloric which was prepared in the range from 0.7 M to 7 tl. The supporting
electrolytes were prepared by adding the appropriate amount of acid to
deionized distilled water. For example, 8.5 mL of 70% (12 M) perchloric
acid were pipetted into a 50 mL volumetric flask containing distilled,
52
-------
deionized water, the contents mixed, and the flask filled to the mark with
distilled, deionized water.
Polyhydroxy Compounds
The following polyhydroxy compounds were used as reagents: 3-chloro-
1,2-propanediol, pyrocatechol, pyrogallol, 1,2-ethanediol, 1,2-propanediol,
1,2-butanediol, D_-ribose,, sucrose, fructose, 2,3-butanediol, and D_-mannitol.
The concentrations of the polyhydroxy compounds were 0.5 M_. They were
weighed only with one-tenth gram accuracy. For example, 4.5 g of D_-mannitol
were added to 50 mL of solution in the polarographic cell. The mixture was
stirred until all the mannitol had dissolved. The resulting solution was
approximately 0.5 M with respect to mannitol.
MASS SPECTROMETRY OF ORGANYLARSONIC ACIDS, DIORGANYLARSINIC ACIDS,
ORGANYLIODOARSINES, ARSENOCHOLINE AND ARSENOBETAINE
Mass spectrometry has the advantage of high sensitivity, of requiring
only small samples and, of providing both elemental composition and
structural information. High resolution mass spectrometry will also
distinguish between substances containing only lighter elements and those
having one or more heavy elements in their-molecules. ; '.'.",':.
Arsinic and. arsonic acids could perhaps be determined by mass spectral
analysis in.the residues obtained by evaporation of water samples. The mass
spectral behavior of'.five organylarsonic'acids and fourteen diorganylarsinic
acids was studied using a.CEC 21-110B mass spectrometer.
At probe temperatures between 110° and 250° required to. obtain
satisfactory mass spectra,., the arsinic and arsonic acids form anhydrides and
decompose. The products of these thermal .reactions are then ionized and
fragmented yielding complicated mass spectra with many peaks at m/e values
higher than those expected for the molecular ions of the arsinic and
arsonic acids (Figs. 25,. 26), . .
There are several reports in the literature concerning the dehydration
of arsonic acids. Differential thermal and vacuum thermogravimetric
experiments indicated that arsonic acids,;RAsO^H^, are first converted to
diarsonic acids, H02(R)As-0-As(R)02H, and then via polyarsonic acids to the
polymeric arsonic acid anhydridesj (RAs02)n> which decompose on further
heating to arsenic trioxide (67). Arsinic acids, R2AsOOH, can only form
the anhydrides, [R2As(0)]20.
The mass spectra of the arsonic acids do not contain molecular ions
for the acids, the diacids or the anhydrides. The highest mass peak in
the spectrum of methylarsonic acid corresponds to Cl^As^Og. In the spectra
of most of the acids As^Og is the highest mass peak. Characteristic
fragments are formed by loss of arsenic and oxygen from As^Og and by loss
of methyl, oxygen and hydrogen from CHsAsOsH. In all the spectra peaks of
considerable intensity are observed for Asn (n = 1-4). The arsenic is
. 53 - '
-------
TOO
S
-S
i
so
s
8
0
• .
—
—
=C
™~ ' o
\
\
- j
1
Cj
_f I
eo \
tT ^ .
i fi ^io !:
i I*!* \
S j \
l!
1
111
1
1
° CHiAsO,H, %
x1" ^>j ^ ----- PEAK HSIGHT jc 0 1 • " **
" ' § § o" • «AK HtlGHT»O.OS —
<
i
J|
1 III i
i'g" =~ J, 0~
. • cT* 5
i t
I
a1
1
s
f
tl Uu
C -C ' ' a"
If. |
iii ,
1 . , ... i . i . i . 1 .
3 5O 10O 15O 20O 25O
c?
' . 9
j * cS" ^,
. . | 1 . - C 3*
0^
-------
1UU
01
c
to
. -o
c
3
J3
«T
50
- OJ
•r-
«O
i
°c
- (CH3)2AsOOH
_ — — — DFAIf MFIfXMT v (11
~~ R^CH3
CO
3:
o
— - . in
3:
CM
CJ
— •
r«*
or
co w»
0 <
1
I
I i
,1 , ill 1.
) 50
J i
o
I/I
-------
probably formed by decomposition of arsenic-containing materials on the hot
filaments in the ion source. The most important ions in the mass spectra
of arsonic acids are summarized in Table 4.
Diorganylarsinic acids- appear to produce molecular ions (Fig. 26,
Table 5) . The signals for these ions decrease with increasing molecular
mass of the arsinic acid. The ions with m/e values equal to or lower than
those of the molecular ions and the ions containing arsenic and oxygen only
are listed in Table 5. Molecular ions formed from arsinic acid anhydrides
have not been observed in any of the mass spectra. However, intense peaks
corresponding to R3AS203 are present in the spectra. The signals with m/e
values higher than those of the molecular ions of R,AsOOH are generated by
fragments arising through loss of alkyl groups, oxygen, hydroxyl groups and
hydrogen from R3As203,
Mass spectrometry is of little value for the identification of arsonic
acids, but can be used to establish the presence of diorganylarsinic acids
in the residues obtained from water samples. Exact mass measurements by
high resolution mass spectrometry might be necessary to distinguish
arsenic-containing from' arsenic-free fragments.
Organyliodoarsines, RjjAsIj^ (n = 1,2) are much more volatile than
arsinic or arsonic acids, are not capable of forming anhydrides and are
easily prepared by treating arsinic and arsonic. acids with hydriodic
acid (68). These reactions should be- useful for the conversion of organyl-
. arsenic, acids, in environmental samples to Organyliodoarsines, which can then
be determined by mass spectrometry. Arsenite and. arsenate are both
transformed to arsenic triiodide under these conditions.
Purified samples of Organyliodoarsines were employed to obtain their
mass spectra at probe and source temperatures given in Tables 6 and 7: All
the iodoarsines gave intense molecular ion peaks . Fragmentation proceeds
by loss of alkyl groups, iodine and hydrogen abstraction.. Peaks of lower
intensity corresponding to AsI3, R2 As I. and Asn (n = 1-4) were observed in
many of the spectra. Thermally induced disproportionation of RasI into
R2AsI and Aslj is a probable mechanism for the formation of these compounds.
Interesting series of fragments represented by the formulas CnH,n+1AsI,
CnH2a+2As anc* cnH2nAs are Part °^ tlle mass spectra of Organyliodoarsines.
It appears that methylene groups are lost generating signals fourteen mass
units apart from n = 0 to n equal to the number of carbon atoms in the
Organyliodoarsines. Additional experiments are in progress to verify this
unusual cleavage of methylene. As examples for the mass spectra of Organyl-
iodoarsines the spectra of butyldiiodoarsine and dioctyliodoarsines are
presented in Fig. 27 and Fig. 28, respectively. Organyliodoarsines are
well suited for the mass spectrometric identification of organic arsenic
compounds which can be converted to their iodo derivatives .
Arsenocholine chloride [(CH3) 3AsCH2CH20H]+Cl~, (Fig. 29) and
arsenobetaine chloride, [(CH3)3AsCH2COOH]+Cl, (Fig. 30) produce rich mass
spectra, which, however, do not contain molecular ion peaks. The highest
mass peaks correspond to (CHa^AsCI^CI^O and (CH3)2AsCH2COOH, which are
56 .
-------
ABLE 4. RELATIVE ABUNDANCE OF IONS IN THE.ELECTRON IMPACT MASS SPECTRA OF
ORGANYLARSONIC ACIDS
Ion
R
RAs
RAs02
RAs02H
RAs03
RAs03H
RAs03H2
RAs°3H3
R2As
TJ A oC\
*\fj£\9\s - ,
R2As02
R2As02H
13 AsO "H
As
AsO
As62" ''•••'• :
As03
AS202
As2°3
'^394
As406
Probe Temp PC
Source Temp °C
CH,
0.00
3.60
1.15
100.00
24.04
0.96
0.00
8.30
1.63
7.20
.53.90
24.0.0
4.00
4.40
75.00
5&. 70
—
4.70
0.96
40.38
75.96
200
210
R in RAsO
11.32
0.40
3.40
100.00
0.00
1-21
0.16
8.29
. 0.00
0.58
18.29 .
7.63
3.95
1.84
53.95
0.53
0.89
3.80
0.76
19.00
21.60
165
235
3H
100.00
0.57
0.64
29.90
0.50
0.64
0.00
2.36
0.00 ,.-
0.00
5.36
1.57
-1.64
2.86
54.29
3.29
4.79
3.43
0.57
9.29
8. 57
180
210
C9H19
0.38
7.30
1.03
0.23
0.00
o.oo
0.00
0.00
2.40
0.34
0.00
0.00
0.00
0.34 ~"
0.91
0.11
1.14
0.11.
5.14
0.80
2.40
210
230
C6H5- '
31.43
21.43
0.00
0.00
0.00
o.oo,
0.00
0.00
19.29
, 3.29
,'11.43
3.00
1.71
4.71
80.71
0.00
0.43
9.29. *
0.00
62.86
92.86
255
210
57
-------
TABLE 5. RELATIVE ABUNDANCES OF IONS IN THE ELECTRON IMPACT MASS SPECTRA OF DIORGANYLARSINIC ACIDS
01
00
Ion
R
RAs
RAsO
RAs02
RAs02H
RAs03H
RAs03H3
R2As
R2AsO
R2As02
R2As02H
R2As02H2
As
AsO
As02
As03
As202
As 2 03
As30^
As4°6
Probe Temp °C
Source Temp °C
™3
5.56
7.22
4.56
8,33
6.67
0.56
0.77
40.56
51.11
0.67
15.56
8.89
2.33
36.67
26.67
6.67
4.56
3.33
0.00
0.00
184
210
C2H5
o.oo
0.96
0.93
0.00
0.37
0.00
0.00
1.00
2.11
1.15
7.41
0.44
0.33
10.37
0.11
0.00
20.37
0.33
2.22
1.89
110
200
S*7
3.48
1.96
0.43
0.30
0.30
0.07
0.20
0.50
0.39
1.17
0.20
5.87
0.02
11.02
1.24
0.04
2.61
0.28
0.91
1.30
120
200
R in R_
C4H9
5.00
3.11
2.30
0.22
0.32
0.00
11.89
0.22
1.62
0.00
0.11
0.49
0.03
2.30
1.11
0.00
2.30
0.24
1.62
0.92
130
220
AsOOH
C5H11
4.00
5.00
3.17
0.67
0.50
0.00
0.00
0.33
1.98
0.00
0.00
0.00
0.25
3.00
1.42
0.00
2.08
0.25
1.42
10.00
165
210
'"tfu
100.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
9.77
4.65
4.65
23.26
1.16
1.63
1.86
0.00
1.16
0.00
0.93
0.70
150
210
C8H17
1.19
17.84
4.83
12.27
4.46
0.52
1.86
1.64
1.78
0.74
6.15
4.01
0.37
5.58
3.94
0.52
0.89
0.97
2.68
8.18
202
210
C6H5
18.80
25.56
2.33
0.33
0.00
0.00
0.00
46.94
23.89
14.44
3.33
5.28
1.28
10.56
0.00
0.00
3.89
0.00
1.83
6.33
220
220
C6WH2
100.00
24.27
0.78
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
•;"••' o.oo
0.58
9.90
0.78
0.00
3.50
0.00
12.62
18.45
„ 220
/** 220
-------
TABLE 6. RELATIVE ABUNDANCES OF IONS IN THE ELECTRON IMPACT MASS SPECTRA OF ORGANYLDIIODOARSINES
Ion
R2AsI
RAsI2
RAsI
AsI2
Asl
R2As
RAs
AsI3
I
As
R
I2
Probe Temp QC
Source Temp °C
CH3
0.19
100.00
51.58
11.84
1.42
0.08
10.53
0.10
13.95
3.42
0.47
2.92
40
220
c2
0
100
48
24
31
0
3
d
16
7
9
1
V
.15
.00
.75
.88
.86
.06
.25
.31
.86
.81
•69 i
.31
45
190
C3H?
0,14
100.00
23.88
32.24
41.63
0.05
2.45
0.86
16.73
4.29
61.22
1.69
30
190
R
0
57
13
19
20
- -
4
0
9
3
100
0
in RAsI2
.24
.58
.33
.39
.61
r— '_ .
.24
.18
.09
.21
.00 ^
.72
40
210
0.30
65.57
21.52
24,78
20.87
0.04
3.48
0.83
7.83
2. 83
63.04
0.69
35
210
C7H15
21.05
7;02
7.02
6.14
7.50
2.10
37.74
4.34
0.87
1.66
0.46
60
190
C14H29
3.23
35.48
100.00
85.48
30.65
0.09
6.77
48.39
5.00
14.35
0.32
85
250
C6»5
0.03
31.03
100.00
2.53
2.64
0.14
30.34
2.76
4.60
0.92
0.60
1.84
50
200
Cx-Hg-CHnCHn
0.26
2.30
57.45
2.77
3.19
0.18
31.06
0.91
5.74
0.98
100.00
1.96
55
: 20°
-------
TABLE 7. RELATIVE ABUNDANCES OF IONS IN THE ELECTRON IMPACT MASS SPECTRA.OF
DIORGANYLIODOARSINES
Ion
R^I
RAsI2
RAsI
AsI2
Asl
R2As
RAs
As
R
I
*2
CH3
100.00
0.41
63.84
0.32
16.07
1.09
61.61
8.04
3.57
15.18
0.15
C-C6H11
25.24
0.10
12.38
0.14
1.31.
1.21
3.33
0.21
100.00
2.00
0.29
R in R2AsI
C8H17 C10H22
35.96
1.05
3.93
2.13
7.30
3.47
100.00
1.01
4.83
5.39
0.20
44.96
3.88
8.37
4.42
6.36
5.81
100.00
0.62
4.19
4.57
0.23
C6H5CH2CH2
6.82
1.36
0.07
1.23
97.73
97.73
0.93
100.00
2.05
0.07
formed by thermal cleavage of HC1 and CHsCl from the arsonium salts.
Fragmentation proceeds by loss of methyl groups, cleavage of hydroxyl groups
and loss of ethylene.. Fragments containing arsenic-oxygen or arsenic-
chlorine bonds were either formed by thermal decomposition of the arsonium
salts or through intramolecular rearrangements. In spite of the absence of
molecular ions mass spectrometry can provide an indication of the presence
of arsenocholine or arsenobetaine in a sample.
60
-------
1UU
O)
u
03
c
a
•f
so
01
.^
fO
O)
a:
co-
r* <
2: z
m
•
in
CVJ
o
j
J>
a*
^ •
QK
1 A 1
bAsh
PEAK
HEIGHT x 0.1
•
z
(fl
eC
CT»-
U
f
"^n
Kj <^
Z
ro
o
•
I
1 ,I.I,L
ll
K.J..-I-..I ....1.
100
i
«
i 1 i
"LO "%x "^n
c*j « «•
00 0
fl
c
15O 2OO
• .,-•• ... .... -, m/e ,. .., .....'
100
so
<—• VI (/I l/l
J2 < < . <
00 O
11.1
<: vi
3?!
O
i
i
J....1.-.L..I....I I....I....I. I I L..I....I....I....I.... l....i....l....i....l.-.i....l-.i..
20O
25O
30O
m/e
35O
4OO
Figure 27. The mass spectrum of butyldiiodoarsine.
61
-------
PEAK HE KMT ix 0.1
150
200 m/e 2SO
300
350
400
Figure 28. The mass spectrum of dioctyliodoarsine. The peaks marked by * and ** belong to the
series CnH2n+1AsI and cnH2n+2As, respectively.
-------
1OO
OJ
0
CO
TJ
-Q
SO
OJ
>
((CH3)3AsCH2CH2OH)c
" CO
DC
,_ —
>M_
^"T
^
•HW
"•^T"
_
.11
h
JL
-r-
O
cv
3C
(_J
*•
3
™
H
• c_
' 3
c_
CJ
1
1
J....I....1 I....
31
i
•
|o
i
i ,
i i
'*.
'
5
NJ
5
CM
3
.
U L
50
'f—.
o
CO
3C
O
5 ' PEAK HEBGHTx 0.1
o .......
CO CO :—
. =t =t O
: CM co to
to
^^
CO
"'.'.- DC
O
\
10
•a
";
...1
...J....I....I....
;
\
1'
.
.1
i
i
U
M
n
ii
^•^ ^C "3T T"
CO CO CO CM CD
ac ac — » oc • ac CM —
CJ CJ CO O O '3C
• / 3= CM CM 0
U 3Z 3: CM
«— • • • O O 3C
1
1
1 i 1 i
CO CM O
i - .- • ~f— •
VI ^ " ^t—
O
\
_,
..-— ^
3C
(
,t- nc A i/ LJCI/^
PEAK HEJu
HTxO.05
HTxttl
01
^C
C^
nz
t_5
»~~
CJ
Nl
O
^
1 \
o
— o
|
L
,...,...,1 .,.,( i....
j
1
i
I
CO
oc
CO
•*
i
i • '
• •
: I 1 •
Li) ilJiLjiliL
li
CO
CO
*~^
CO CO r-
< Z C
CM O •— C
-— ~ <_> <
CO : CO
n: i «a: ^~
CJ • CO
^__^«
JihllllilLJ
...i ...i, .,i,,,i...,i....l....i....l.........i....:....l.......,,
or n
o c.
; •
/ ;
;
g
g
I
Jl
....I....I...J,...
j
0
C
CM
^
CO
5
_^
_
3C
O
O
C_)
CM
t_>
CO
=t
CO
:n
O
,
l_5 _
"-•-. co
CO
CO
a:
o
T*
0
o
t_3 — —
CM
CJ
CO __ _
CM
CO
^C <~w«
o
.
• • — ,
/ -'••''
•
, 1
IidiQtQiln
O
5O
1OO
200
Figure 30. The mass spectrum of arsenobetaine chloride.
-------
SECTION 7
ANALYSIS OF WATER SAMPLES
Samples of arsenic-containing drinking water supplies selected.by the
project officer were collected and shipped to College Station as quickly as
possible. Total, arsenic concentrations and the concentrations of arsenite
and arsenate' were, determined by several methods. Each water- sample was
checked for the presence- of. methylated arsenic compounds and. other organic
arsenic derivatives. Water, samples from Hinckley, Utah; Delta, Utah.;
Barefoot Site, Alaska;. Mauer Site, Alaska; Antofagasta, Chile; Yenshei,
Taiwan; Hartlin Site,. Nova. Scotia; and Sullivan Site, Nova Scotia were
investigated.
The following methods were: employed for the determination of total
arsenic and arsenic compounds.
® Graphite furnace atomic absorption spectrometry: A
Hitachi Zeeman graphite furnace atomic absorption
spectrometer with 'graphite cups was employed for total •
arsenic determinations. . Forty microliters were injected
in most' cases.. - Since GFAA is subject' to many inter-
ferences, arsenic was extracted from water samples.
The extract was then injected into-the graphite cup.
A suitable aliquot of the water sample is made
0.1 M with respect to hydrochloric acid. Arsenate is
reduced to arsenite by addition of 5 mL of a solution
containing NaHSOs, NaaSgOs and HC1. The solution is
extracted with 5 mL of 0.01 M (CifH90)2PS2H in hexane..
The extraction is repeated four times with 2 mL each
of the reagent solution. The combined organic layer
is washed first with 4.5 M HC1 and then with water.
The organic layer is extracted twice with 2 mL of
bromine water. The aqueous phases are transferred
to'a.volumetric.flask, adjusted to 0.1" M acid with
nitric acid, brought to 200 ppm with respect to
nickel by addition of nickel nitrate and treated with
50 vL of 30% hydrogen peroxide. After filling to the
mark with distilled water, 20 yL are injected into the
graphite cup for analysis. The standards and blanks
are prepared in the same manner.
65
-------
© Differential pulse polarography: Arsenite, arsenate and
total arsenic were determined by DDP as described in
Section 6.
High pressure liquid chromatography/GFAA as element-
® specific detector: Concentrated water samples were
analyzed by the HPLC-GFAA method to check for the
presence of organic arsenic compounds as described in
Section 6.
@ Hydride generation technique: Arsenite, arsenate and
methylated arsenic compounds are reduced by sodium
borohydride at pH 1.. At pH 6.5 only arsenite is reduced.
The reductions are carried out in a chamber capable of
holding 10 mL of sample. After the addition of the
appropriate buffer'solution a measured volume of sodium
borohydride solution is pumped into the chamber. A
stream of helium transports the arsines through a U-trap
cooled in an isopropanol/dry ice bath to remove water.
The arsines are collected in a U-trap cooled by liquid
nitrogen. One leg of this trap is packed with silanized
glass beads. After the arsines had been transferred to
the.U-trap, they are rapidly volatilized by electrically
heating the trap. A stream of helium transports the
arsines through a short tube packed with sodium hydroxide
pellets to remove carbon dioxide into the helium discharge
created, by passing direct current between two helium-arc
welding electrodes. The radiation from the discharge
passes through the quartz window of.the emission chamber
and is focused on the. entrance slit of the monochromator
set at the 228.8 nm arsenic emission line. A gas
chromatograph can be attached to the exit of the arsine
trap to improve separation of AsH3 from methylarsines.
The hydride generation technique has detection limits
of 0.2 ppb for arsine, 0.3 ppb for methylarsine and
1 ppb for dimethylarsine.
® Inductively coupled argon plasma emission spectrometry:
ICP is the most convenient method to determine the trace
element concentrations including arsenic concentrations in
water samples. An ICP spectrometer became available at
Texas A&M University during the last months of the
project. The recently collected samples were analyzed
for 48 elements by the simultaneous ICP Model 34000
manufactured by ARL. Prior to this time ICP analyses
were carried out with ARL instruments at Radian
Corporation in Austin, Texas or at the ARL plant in
Sunland, California.
As an illustrative example for the sampling procedures the sampling
instructions for the Taiwan well waters are cited. Sampling instructions
66
-------
were pasted to each bottle. The correct amounts of preservations in vials
were also attached to the bottles (with exceptions of nitric acid).
Rinsing of all bottles before filling:
1. Fill bottle half full with well water, close bottle with
cap and shake thoroughly. Empty bottle.
2. Repeat procedure of item 1 three times.
3.. Fill the bottle as described below.
4. Tightly close the cap, wipe it dry and tape securely with
the tape provided.
Glass container for unpreserved sample (Well I):
Fill the rinsed bottle with water to the brim, close the. cap tightly and
secure by taping the cap with the tape provided. Provide the information
requested on the label.
Glass container for sample preserved with HgCl^ (Well I):
Fill the rinsed bottle with water almost to the brim. Add ,all of the
mercuric chloride in the vial to the sample. Fill to the brim, close the
cap tightly and secure by taping the cap. Shake the closed bottle to
dissolve the mercury salt.. Provide the information requested on the label.
Plastic containers for unpreserved sample (2 containers for Well I;
2 containers for Well II): .
Fill the rinsed bottles with water to the brim. Make sure no air space
remains in the bottle. Close the cap tightly and tape. Provide the
information requested on the label.
Plastic containers for sample preserved with 0.1 weight % ascorbic acid
(2 containers for Well I; 2 containers for Well II):
Fill the rinsed bottles with water almost to the brim. Add all the
ascorbic acid from the vial to the sample. Fill to the brim, close the
cap tightly and secure by-taping. Turn the bottles several times to aid
the dissolution of ascorbic acid. Provide the information requested on the
label.
Plastic containers 'for samples preserved with nitric acid
(2 containers for Well I; 2 containers for Well II):
Fill the container almost to the brim. Add 7 mL of ultrapure nitric acid.
Fill to the brim, close the cap and secure by taping. Shake the bottle
several times to make the sample homogeneous. Provide the information
requested on the label.
67
-------
Nitric Acid Control Solution;
Measure 7.0 mL concentrated nitric acid from the nitric acid bottle used
for the preservation of the water sample into a clean 1.0 L volumetric flask.
Fill to the mark with distilled water and shake to homogenize the solution.
Fill the two 125 mL-plastic bottles with this solution, tighten the cap
securely and tape the cap.
X
The results of the analyses of the water samples are presented in the
following sections..
The samples were shipped in sturdy cardboard or wooden boxes. None of
the samples were shipped frozen.
Water samples from Delta and Hinckley, Utah were collected by
Mr. E. Western, Utah State Department of Health, Division of Environmental
Health on July 16, 1979. Pyrex glass bottles and Cubitainers were employed
as containers. Some of the samples were preserved by addition of 0.1 weight
percent of ascorbic acid. The samples were shipped to College Station as
air freight and received July 17, 1979. The analyses were performed during
the period July 17 to 25, 1979.
The following analytical methods were employed: graphite furnace
atomic absorption spectrometry (GFAA) using a Hitachi-Zeeman GFAA Model
170-70; differential pulse polarography (Princeton Applied Research Model
174A); hydride generation with atomic emission spectrometric. detector
(Braman method); inductively coupled argon plasma emission spectrometry
(ARL 34,000). ~~
Hinckley Water Samples
The results of the total arsenic analyses of the Hinckley water samples
are listed in Table 8. The average arsenic concentration in the Hinckley
water samples obtained from 36 determinations (twelve each by graphite
furnace atomic absorption spectrometry, differential pulse polarography and
hydride generation) is 185±8 ppb.
The analyses of the water samples by ICP were carried out several
weeks after the collection, because the instrument had not been available
earlier. The samples stored in plastic containers gave ICP arsenic
concentrations of 170 ppb (without ascorbic acid) and 160 ppb (with ascorbic
acid), which are in fair agreement with the average value from the other
methods. The ICP values in glass containers (140 ppb and 120 ppb) were
considerably lower than the values obtained from samples stored in plastic
containers. Glass containers in contrast to polyethylene containers are
known to adsorb arsenic on prolonged storage (see section 5). Samples
stored in glass containers are, therefore, expected to have less arsenic in
solution than fresh samples or samples stored in plastic Cubitainers.
68
-------
TABLE 8. TOTAL ARSENIC CONCENTRATIONS IN ppb IN THE HINCKLEY AND DELTA WATER SAMPLES
)H H
y <0 (U
•H O Ci C!
' •? "1 'd * "3
M 4J n) X id
O TJ 0) 4-1
-------
The determinations of arsenic compounds were performed employing the
hydride generation (HG) technique and differential pulse polarogrphy (DPP).
The results are summarized in Table 9. Only arsenite and arsenate were
found. The average concentration of arsenate was 172±7 ppb (by HG) and
187±8 ppb (DPP). The ars'eni.te concentrations averaged 11 ppb. The arsenite
concentrations (by HG) in the samples preserved with ascorbic acid were
approximately four times higher than in the unpreserved samples. Ascorbic
acid might have reduced some arsenate (see section 5) at these low arsenic
concentrations. The sum of the concentrations of arsenite and arsenate agree
well with the values of the total arsenic concentrations which were deter-
mined separately.
Organic arsenic compounds, if present at all, cannot occur at concen-
trations larger than a few part per billion. The search for methylated
arsenic compounds by the hydride generation method, which has detection
limits for- such compounds of approximately 1 ppb did not produce any
indication of their presence.
The results of the ICP analyses of the Hinckley water samples are
summarized in Table 10. With the exception of the concentrations of As, B,
Ba and Si the listed values are averages obtained from the concentrations of
the following four samples: Pyrex container with 0.1% ascorbic acid;
Pyrex- container without ascorbic acid; plastic Cubita'iner with 0.1% ascorbic
acid; plastic Cubitainer without ascorbic acid. The values for As, B, Ba
and Si are averages of the samples stored in Cubitainers. The elements
present in concentrations larger than the detection limits (Table 10) are
As, B, Ba, Ca, Cu, Mg,. Na, Si and Sr. These average concentrations have
average deviations of ±5 units in the last digit of the listed numbers.
The boron, barium and silicon concentrations in the samples stored in
Pyrex containers were 2.25 ppm (B), 0.13 ppm (Ba) and 15.5 ppm (Si),
significantly higher than in the samples stored in Cubitainers. According
to information supplied by the Wheaton Glass Co., the manufacturer of
the Pyrex containers, a- soda ash-barium-borosilicate melt was; .used in the
manufacture of the containers. The sodium concentration averaging 233±3 ppm
was the same in all four samples.
No significant changes, in arsenite, arsenate and total arsenic
concentrations were observed during the one week period, during which the
initial analyses were performed.
Delta Water Samples
The results for the Delta water samples are summarized in Table 8
(total arsenic concentrations), Table 9 (arsenite and arsenate concentra-
tions) and Table 11 (trace elements by ICP). The total arsenic concentra-
tion is approximately 21 ppb. Arsenite and arsenate concentrations are
comparable in magnitude and close to the detection limits of the methods
employed. According to ICP results (Table 11) Ba, B, Ca, Cu, Fe, Mg, Na,
Pb, Si and Sr are present in the water samples. The Ba and B concentrations
in the samples stored in glass containers are again higher than those of
samples kept in plastic containers. The Delta water samples are harder
70
-------
!TABLE 9. CONCENTRATIONS OF ARSENITE AND ARSENATE IN ppb IN THE HINCKLEY AND DELTA WATER SAMPLES
U M '
0 a) Q)
! -H o d rt . • ,
•° 'H -rl °d TT/mat
, • l-i 4-1 ctt • X ni . : HG**
O TJ • COW CU-W
o -H •• a a ^ o ;
, Location
-------
TABLE 10. TRACE ELEMENT CONCENTRATIONS IN THE HINCKLEY WATER SAMPLES DETERMINED BY ICPf
H
•°ff
5.5
Na
233
42
K
Rb
Cs
Fr
0.46
Be
34
Mg
1.83
45
Ca
3.24
0.44
Sr
0.10
0.58
Ba
).026
Ra
Concentration of the
Element in ppra Found
in the Sample*
*The absence of a yyyy
concentration is at 01
8
Sc
2
Y
La
Ac
16
Ti
4
Zr
Hf
8
V
Nb
Ta
8
Cr
8
Mo
18
w
E
•yyyy
JJCLCI.I-J.IJII IJ-LUU.U in yy
figure indicates that the
r below the detection limit.
Mn
Tc
Re
8
Fe
Ru
Os
6
Co
Rh
Ir
10
Ni
Pd
25
Pt
4
Cu
0.09
2
Ag
48
Au
10
Zii
6
Cd
32
Hg
u
9
B
1.2
50
A!
Ga
55
In
91
Tl
C
16
Si
13.4
Ge
120
Sn
84
Pb
•H
150
P
57
As
0.165
32
Sb
49
Bi
O
S
84
Se
103
Te
Po
H
F
a
Br
1
At
He
Ne
Ar
Kr
Xe
Rn
Ce
Pr
Nd
Pm
Sm
Eu
Gd
Tb
Er
Th
Pa
64
U
Np
Pu
Am
Cm
Bk
Cf
Es
Fm
Md
No
Lr
''"The average deviation of the listed concentrations is ±5 units of the last digit.
-------
T^.BLE 11. TRACE ELEMENT CONCENTRATIONS IN THE DELTA WATER SAMPLES DETERMINED BY ICP1
H
?ff.
5.5
Na
65
42
K
Rb
Cs
Fr
?'
0.46
Be
34
Mg
6.63
45
Ca
1.5.4
0.44
Sr
0.52
0.58
Ba
0.045
Ra
Concentration of the
Element in ppm Found
in the Sample*
*The absence of a yyyy
concentration is at o
8
Sc
2
Y
La
Ac
16
Tf
4
Zr
Hf
8
V
Nb
Ta
8
Cr
8
Mo
18
w
E
•yyyy
figure indicates that the
r below the detection limit.
Mn
TC
Re
8
Fe
0.007
Ru
Os
6
Co
Rh
Ir
10
Ni
Pd
25
P*
4
Cu
0.26
2
Ag
48
Au
10
Zn
6
Cd
32
Hg
9
B
0.01
50
AS
Ga
55
in
91
Tl
C
16
^
Ge
120
Sn
84
Pb
1.17
N
150
P
57
As
32
Sb
49
Bi
0
S
84
Se
103
Te
Po
H
F
Ci
Br
1
At
He
Ne
Ar
Kr
Xe
Rn
Ce
Pr
Nd
Pm
Sm
Eu
Gd
Tb
Dy
Ho
Er
Tm
Yb
Lu
Th
Pa
64
u
Np
Pu
Am
Cm
Bk
Cf
Es
Fm
Md
No
Lr
average deviation or the listed concentrations is ±5 units of the last digit.
-------
(15 ppm Ca, 6.6 ppm Mg) than the Hinckley samples. The sodium concentra-
tion is 65 ppm. The relatively high Pb (1.3 ppm) and Sr (0.5 ppm)
concentrations are noteworthy.
Organic arsenic compounds were not detected in the Delta water samples.
WATER SAMPLES FROM ANTOFAGASTA, CHILE
Water samples from the drinking water supply for the city of
Antofagasta, Chile were collected by Dr. K. J. Irgolic at the water treat-
ment plant on October 21, 1980. Three samples were taken from the untreated
water just before it entered: the treatment plant. The faucet attached to
the large delivery pipe was opened and water was allowed to run for five
minutes. Then the one-quart plastic Cubitainers were rinsed with the water,
filled to the brim and closed tightly. No air was left in the container.
One of these samples was preserved with 0.1% by weight of ascorbic acid.
Two samples were similarly taken from the pipe leaving the water treatment
plant. All samples were collected in plastic Cubitainers. The analyses
were performed during the period October 25 to November 10, 1980.
The following.analytical methods were employed: graphite furnace
atomic absorption spectrometry (GFAA), differential pulse polarography;
hydride generation and inductively coupled argon plasma emission spectron-
etry.
The results of the total arsenic analyses on the five water samples
from Antofagasta are summarized in Table 12. The average arsenic concen-
tration (average deviation) obtained from all values for the untreated
samples listed in Table 12 is 0.910.1 ppm. This, value is close to the
concentration of 0.8 ppm reported earlier for this water (69). The results
for the untreated samples with and without ascorbic acid as preservative are
in good agreement when the values obtained with a particular method or
instrument are considered. The ICP analyses were carried out on three
different instruments and with three different standards. The treated
water samples gave arsenic concentrations in the range of 0.25 ppm to
0.51 ppm (Table 12). After passing through the alum treatment for arsenic
removal the treated water has generally an arsenic concentration of
approximately 80 ppb (70). The high arsenic values in the treated water
samples collected at Antofagasta on October 21, 1980 could have been caused
"by a malfunction of the water treatment plant. The removal of arsenic by
alum coagulation is highly pH dependent. The percent arsenic removed from
arsenite-containing, chlorinated well water and from arsenate-containing
tap water decreased from 90 percent at pH 7 to approximately 20 percent at
pH 8.5 (70a). The treated water sample could have had a pH above 7. The
pH values of the treated and untreated samples after nine months of storage
in Cubitainers were 7.8 and 8.2, respectively. Another cause for the high
arsenic values of the treated water samples could have been a local region
of trapped arsenic-rich water or soluble deposit in the vicinity of the
faucet from which the samples were taken.
74
-------
TABLE 12. TOTAL ARSENIC CONCENTRATIONS IN ppm IN THE ANTOFAGASTA WATER SAMPLES
Sample
Untreated I
Untreated II
Untreated
Treated I
Treated II
Ascorbic Acid HG*t GFAA*t DPP*t ICP*t
No 0.71±0.02 0.75+0.02 0.78*0.03 0.90, 0.78, 1.06
No .'••!- 0.73+0.02 0,78*0.03 0.91, 0.81, 1.02
i
Yes 0.72±0.06 0.70±0:.02 - 0.91, 0.82, 1.05
No O.;43±0,02 0.40+0.01 0.41±0.01 0.51, 0.42, 0.33
No - 0.36+0.01 ' - 0.32, 0.33, 0.25
HG: Hydride generation.
DPP: Differential pulse polarography.
ICP: Inductively coupled argon plasma emission spectrometry.
GFAA: Graphite furnace atomic absorption spectrometry.
Average + average deviation of at least three determinations for each sample.
-------
The arsenic compounds present in the Antofagasta water samples were
determined by the hydride generation technique and by differential pulse
polarography. Most of the arsenic is in the form of arsenate (Table 13).
The arsenite concentrations in the preserved and unpreserved, untreated
water samples are at most 16 ppb. The treated.water sample contained 3 ppb
arsenite. No-methylated or other arsenic compounds were detected.
TABLE 13. CONCENTRATIONS OF ARSENITE AND ARSENATE IN ppm IN. THE ANTOFAGASTA
WATER' SAMPLES .
Sample Ascorbic Acid Arsenate** Arsenite**
Untreated
Untreated
Treated
No 0.70±0.02*
0.78±0.03t
Yes 0.7010.02*
0.78±0.03f
No 0.43±0.02*
0.41+0. Or1"
0.016±0.001*
0.016±0.001*
0.003+0.001*
* Hydride generation method 1" Differential pulse polarography
**Average + average deviation from at least three determinations given.
Trace elements in the Antofagasta water samples were determined by
inductively coupled argon emission spectrometry employing three different
instruments. Among the 41 elements on the analytical arrays of these
instruments Li, Na, K, Mg, Ca, Sr, Ba, Fe, B, Al, Si, P, and As were present
in the untreated (Table 14) as well as the treated (Table 15) water samples
above the detection limits, which are also listed in these Tables. The
values obtained for Mn and Cu are very close to the detection limits.
Selenium was not found in the samples (detection limit 84 ppb). The
beryllium content (7.4 ppm) of the untreated samples is noteworthy.
Beryllium is a proven carcinogen (70b, c) , and beryllium sulfate has been
shown to cause enhancement of SA7 transformation of fetal cells of the
Syrian hamster (70d). Additional analyses for beryllium in the untreated
water samples should be carried out. The treated water samples did not
contain beryllium.
The trace element concentrations obtained for the untreated, not
preserved, samples did not differ from those found in the untreated sample
preserved with 0.1% by weight of ascorbic acid with the exception of the
values for Fe and Mn. Therefore the nine values available for each of the
other elements were averaged. These average values with their average
deviations are listed in Table 14. The Fe and Mn concentrations in the
76
-------
TABLE 14. TRACE ELEMENT CONCENTRATIONS IN THE UNTREATED ANTOFAGASTA WATER SAMPLES DETERMINED BY ICP"1"
H
°tf
.62/2
5.5
Na
102/2
42
K
L3.3/7
Rb
Cs
Fr
Be
7.4/2
34
Ma
7.4/2
45
Ca
20JD/2
0.44
Sr
.28/2
0.58
Ba
008/1
Ra
Concentration of the-—
Element in ppm Found
in the Sample* :
*The absence of a yyyy
concentration is at 61
8
Sc
2
Y
La
Ac
16
Ti
4
Zr
Hf
8
V
Nb
Ta
8
Cr
8
Mo
18
w
E
•yyyy
figure indicates that the
r below the detection limit.
Mn
.002
Tc
Fie
8
;f&
Ru
Os
6
Co
Rh
Ir
10
Ni
Pd
25
Pt
5M
2
Ag
48
Au
10
Zn
6
Cd
Hg
9
B
2.8/2
50
.0873
Ga
55
In
91
T!
C
16_
3.8T/6,
Ge
120
Sn
84
Pb
.N
150
P-
.'34/3
57
As
.91/8
32
Sb
49
Bi
O
S
84
Se
103
Te
Po
H
. F-
Cl'
Bf
I
At
He
Ne
• Ar
Kr
Xe
Rn
'- ' ' ' • ' • "
Ce
Pr
Nd
Pm
Sm
Eu
Gd
Tb
Dy
Ho
Er
Tm
Yb
Lu
Th
Pa
64
u
Np
Pu
Am
Cm
Bk
Cf
Es
Fm
Md
No
Lr
'''Average concentrations/average deviations are given in the form 20.0/2 to be read as 20.010.2.
-------
TABLE 15. TRACE ELEMENT CONCENTRATIONS IN THE TREATED ANTOFAGASTA WATER SAMPLES DETERMINED BY ICP
H
OA3
• Li
.64/1
5.5
Ha
103/1
42
K
112/8
Rb
Cs
Fr
0.46
Be
34
Mg
7.572
45
Ca
20.3/;
0.44
Sr
.29/2
0.58
Ba
008/1
Ra
Concentration of the ••••
Element in ppm Found
in the Sample*
*The absence of a yyyy
concentration is at 01
8
Sc
2
Y
La
Ac
16
Ti
4
Zr
HI
8
V
Nb
Ta
8
Cr
8
Mo
18
w
xxxx-
E
•yyyy
Detection Limit in pp
figure indicates that the
r below the detection limit.
Mn
.006
Tc
Re
8
Fe
.3/1
Ru
0s
6
Co
Rh
8r
10
Ni
Pd
25
Pt
4
Cu
.007
2
Ag
48
Au
10
Zn
6
Cd
32
Hg
b
9
B
2.8/1
50
A\*
Ga
55
In
91
TI
C
16
Si
363/9
Ge
120
Sn
84
Pb
N
150
P
.26/5
57
As
.36/7
32
Sh
o
49
BJ
O
s
84
S@
103
Te
Po
H
F
Cl
Br
,|.=
At
He
HQ
Ar
Kr
Xe
Ro
Eu
Gd
64
t
LIT
Average concentrations/average deviations are given in the form 20.0/2 to be read 20.0+0.2.
-------
untreated samples with ascorbic acid, were five- and twenty-times higher,
respectively, than the values in the untreated samples without ascorbic acid.
It could be that Fe and Mn were introduced by the ascorbic acid.
Another significant' difference between the trace element, concentrations.
in the untreated and treated water samples was noted for aluminum and
arsenic. The untreated water had 0.08 ppm Al and 0.91 ppm As, whereas the
treated water gave 3.1 ppm Al and 0.36 ppm As. The; water treatment process
uses potassium aluminum sulfate to produce aluminum hydroxide. The higher
aluminum concentration in the treated water is, therefore, not unexpected.
Another set of samples from Antofagasta had been received by mail from
the EPA Laboratories in Cincinnati during October 1977. The six samples
were provided by Dr. Jane L". .Valentine,;Assistant Professor of;Public -
Health, University of California:, Los Angeles. The samples had been
collected sometime before August 30, 1977. The exact collection date was
not stated in the transmittal letter. The samples were received in
thin-walled polyethylene containers (Cubitainers) which were closed by
polyethylene caps with paper liners. All samples, were colorless and
contained little if any suspended matter. All samples were filtered through
0.45 um Millipore filters. The filtrates were stored in 250 mL acid-cleaned
polypropylene bottles in a refrigerator after acidification with the acids
listed in Table 16. The analyses of the samples began on.October 14, 1977.
The results of total element analysis for As, Na, Fe, Mn, Zn and Cu by
.atomic absorption spectrometry are summarized in Table 16.
TABLE 16. RESULTS OF TOTAL ELEMENT ANALYSIS FOR WATER SAMPLES FROM
ANTOFAGASTA,. CHILE COLLECTED AND ANALYZED IN 1977
Sample
Untreated
»
Treated
2 mL Cone.
Acid Added
H2S04
HN03
none
HC104
HN03
H2S04
As*
ppm
0.52
0.89
0.91
0.91
0.088
0.081
Na**
ppm
54
106
106
106
107
107
Fe**
ppm
0.18
0.18
0.15
0.55
t
t
Mn**
ppm
J-
I
t
t
t
t
t
Zn**
ppm
0.042
0.10
0.047
0.058
0.10
0.10
Cu**
ppm
t
t
t
t
t
t
* Graphite furnace atomic absorption spectrometry.
** Flame atomic absorption spectrometry.
t Less than the detection limit (Fe 0.075 ppm, Mn 0.075 ppm, Cu 0.10 ppm).
79
-------
The total arsenic concentrations in the untreated Antofagasta water
samples ranged from 0.5 to 0.9 ppm as determined by flameless atomic
absorption spectrometry, in good agreement with the values found in the
samples collected in 1980. The two samples which had passed through the
water treatment plant had an arsenic concentration of 0.08 ppm in agreement
with the values obtained in Antofagasta (70).
Arsenite and-arsenate were determined only in the sample-which had not
been acidified. Only arsenite (0.12. ppm) and arsenate (0.80 ppm) were
detected~~with no evidence for organic arsenic compounds. These arsenate and
arsenite determinations were carried out by differential pulse polarography
employing pyrogallol as the reagent to make arsenate electroactive. These
values are considered less reliable than the ones obtained with the hydride
generation technique (see Table 13).
WATER SAMPLES FROM ALASKA
Water samples from Alaska with arsenic concentrations in the ppm range
were collected and shipped to College Station by air freight on
October 17, 1979 and July 29, 1980 by the Alaska Department of Environmental
Conservation in Fairbanks, Alaska. The samples came from the Barefoot and
Mauer sites. The samples were collected in Pyrex. contaJLners_and plastic
~Cu6ftainers. Ascorb~ic "acid or "nitric aci"d7wer'e~"ad'de
-------
TABLE 17. RESULTS OF TOTAL ARSENIC ANALYSES AND ARSENITE AND ARSENATE DETERMINATIONS IN ALASKA
WATER SAMPLES ** \
oo
•o
3 , . « .
•^j ' ^5
o .y • • H
.^ . »-*
.•' .'J3w.il
^ i M 4J Tj S -
Sample to to 8 *s
fO.fi O H i-H
O A4
-------
TABLE 18. TRACE ELEMENT CONCENTRATIONS DETERMINED BY ICP IN THE BAREFOOT WATER SAMPLES FROM Af
(COLLECTED IN 1979)
H
°ff
5.5
Ma
47/2
42
K
Rb
Cs
Fi-
le6
34
Ma
L13A
45
Ca
291/8
0.44
Sr
.47/3
0.58
Ba
.23/1
Ra
Concentration of the — •
Element in ppm Found
in the Sample*
*The absence of a yyyy
concentration is at 01
8
Sc
2
Y
La
Ac
16
Ts
flOS/2
4
Zr
HI
8
¥
.15/1
Nb
Ta
8
Cr
.033/y
0
mo
18
w-
E;
•yyyy
figure indicates that the
r below the detection limit.
Wn
.44/4
Tc
Re
8
Fe
53/f
Ru
Os
6
Co
Rh
IT
10
Ni
Pd
25
Pt
4
Cu
.010? f
2
Ag
48
Au
10
fa
6
Cd
.010/f
32
Hg
9
B
50
Al
<0.06
Ga
55
In
91
• Tl
C
16
Si
11/1
Ge
120
Sn
84
Pb
N
150
P
57
2^3
32
Sb
49
Bi
O
S
84
Se
.103
Te
Po
H
F
CI
Br
1
At
He
Ne
Ar
Kr
Xe
Rn
oo
N>
Ey
Gd
64
^Average concentrations/average deviations are given in the form 20.0/2 to be read as 20.010.2.
-------
TABLE 19.
TRACE ELEMENT CONCENTRATIONS DETERMINED BY ICP IN THE BAREFOOT WATER SAMPLES FROM ALASKA1
(COLLECTED IN 1980)
H
°tf
5.5
Na
37/1
42
12/1
Rb
Cs
Fr
0.46
Be
34
Mq
52/1
45
180/9
0.44
Sr
ND
0.58
Ba
.17/2
Ra
' , • '- • E •
Concentration of the-— -yyyy
Element in ppm Found .
iii the Sample* :
*The absence of a yyyy figure indicates that the
concentration is at or below the detection limit.
8
Sc
2
Y
La
Ac
16
Ti
4
Zr
Hf
8
V
Nb
Ta
8
Cr
8 '
Mo
18
w
Mn
.40/1
Tc
Re
SR
Ru
Os
6
Co
Rh
Ir
10
Ni
Pd
25
Pt
4
Cu
2
Ag
48
Au
10
Zn
6
Cd
32
Hg
9
B
50
Ai
.50/3
Ga
55
In
91
Tl
C
16
Si
ND
Ge
120
Sn
84
Pb
N
150
P
.27/9
57
As
3.4/1
32
Sb
49
Bi
O
s
28/1
84
Se
103
Te
Po
H
F
Cl
Br
1
At
He
Ne
Ar
Kr
Xe
Rn
00
OJ :
Ce
Pr
Nd
Pm
Sm
Eu
Gd
Tb
Dy
Ho
Er
Tm
Yb
Lu
Tn
Pa
64
u
Np
Pu
Am
Cm
Bk
Cf
Es
Fm
Md
No
Lr
^Average concentrations/average deviations are given in the form 20.0/2 to be read 20.0±0.2.
ND = Not determinedi •
-------
Zn, Cd, Al, Si, P and As. Comparison of Table 18 (results for samples
collected in 197-9) with Table 19 (results for samples collected in 1980)
shows considerable differences between the concentrations of Na, Mg, Ca, Fe
and Al in the two samples. Since some concentrations in the 1979 samples
are higher and others lower than in the 1980 samples a systematic instrument-
related error is unlikely. Errors in preparing standards should not be so
frequent. The possibility that the composition of the water supply changed
must be considered and should be checked on additional samples.
Three of the Barefoot samples collected in 1980 (one preserved with
ascorbic acid, two preserved by acidification to 0.1 M HNC^) were analyzed
again by ICP eight/months after collection. The analyses of samples
preserved with nitric acid agreed fairly well with the results obtained
shortly after the samples had been collected. The sample preserved with
ascorbic acid gave considerable lower values'for As, Ca, P and Fe.
The Mauer samples contain 5.1±0.7 ppm of arsenic when all the values in
Table 17 are averaged. Arsenite is the predominant form. The arsenite
concentration of one of these samples collected in a Cubitainer and
preserved by ascorbic acid (bottom line in Table 17) found by the hydride
generation technique was much lower than for the other samples. Repetition
of. these analyses produced the same value.
Trace element analyses by ICP (Tables,20 and 21) showed that Na, K, Mg,
Ca, Sr, Ba, Ti, V, Cr, Mn,, Fe, Zn, Al, Si and As were present. Differences
similar to the ones pointed out for the Barefoot samples were noted between
the trace element concentrations of the Mauer samples collected and analyzed
one year apart. Analyses of the eight months old nitric acid-preserved
samples gave results in general agreement with those reported in Table 21.
The arsenic concentration in the ascorbic acid-preserved samples was 0.7 ppm,
one-fourth of the value found in the samples acidified with nitric acid.
Samples collected in 1979, which had been preserved with ascorbic acid
and had not been opened since the collection, had formed a black precipitate
after two to three weeks. Similar samples in containers, which had
frequently been opened for removal of aliquots for analyses, did not form
this black precipitate. More than 90 percent of the arsenic originally in
solution was present in the black precipitate. It is possible that these
samples contained microorganisms which grew in the ascorbic acid medium
reducing sulfur compounds to sulfide. Sulfide.then would react with iron to
produce black iron sulfides. Arsenic sulfides would coprecipitate with the
iron compound. Approximately 20 ppm sulfur were found in these water
samples by an ICP analysis.
WATER SAMPLES FROM TAIWAN
Two sets of water samples from Taiwan were analyzed. One set
collected by Dr. Tseng, Department of Medicine, National Taiwan University,
Taipei, arrived in College Station toward the end of May 1977. These
samples came from wells at Pu-Tai and Pei-Men. The second set was collected
84
-------
TABLE 20. TRACE ELEMENT CONCENTRATIONS DETERMINED BY ICP IN THE MAUER WATER SAMPLES FROM ALASKA
(COLLECTED IN 1979)
H
•°ff
5.5
Na
50/2
42
K
118/1!
Rb
Cs
FT
0.46
Be
34
Mg
13979
45
Ca
309/
0.44
Sr
.60/3
0.58
Ba
.35/2
Ra
Concentration of the
Element in ppm Found
in the Sample*
*The absence of a yyyy
concentration is at 61
8
Sc
5
2
Y
La
Ac
16
Ti
009/2
4
Zr
Hf
8
V
.17/2
Nb
Ta
8
Cr
.03
8
Mo
18 '
w
xxxx-
E
yyyy
Detection Limit in pp
figure indicates that the
r below the detection limit.
Mn
.76/1C
Tc
Re
8
Fe
.63/3
Ru
Os
6 -
Co
Rh
Ir
10
Ni
Pd
25 •
PI
4
Cu
2
Ag
48;
Ail
10
Zrs
.3/1
6
Cd
32
Hg
b
9
B
50
.#
Ga
55
In
91
Tl
C
16.
10376
Ge
120
Sn
84
Pb
N
150
P
57
As
4.5/1
32
Sb
49
Bi
O
S
84
Se
103
Te
Po
H
F
CI
Br
'J."
At
He
Ne
Ar
Kr
Xe
Rn
00
Ol
Ce
Pr
Nd
Pm
Sm
Eu
Gd
Tb
Dy
Ho
Er
Tm
Yb
Lu
Th
Pa
'64
u
Np
Pu
Am
Cm
Bk
Cf
Es
Fm
Md
No
Lr
Average concentrations/average deviations are given in the form 20.0/2 to be read 20.0+0.2.
-------
TABLE 21. TRACE ELEMENT CONCENTRATIONS DETERMINED BY ICP IN THE MAUER WATER SAMPLES FROM ALASKA1"
(COLLECTED IN 1980)
H
°ff
.007/1
5.5
Ma
35/1
42
K
Rb
Cs
Fr
Go
34
MQ
ssri.
45
Ca
200/9
0.44
Sr
0.58
Ba
29/1
Ra
Concentration of the -—
Element In ppm Found
in the Sample*
*The absence of a yyyy
concentration is at 01
8
Sc
2
Y
La
Ac
16
Ti
4
Zr
Hf
8
V
Nb
Ta
8
Cr
8
Mo
18
w
XXXX--
E
•yyyy
Detection Limit in pp
figure indicates that the
r below the detection limit.
Mn
.56
Tc
Re
8
Fe
25/1
Ru
Os
6
Co
Rh
Ir
10
Ni
Pel
25
PI
4
Cu
2
Ag
48
Au
10
Zn
6
Cd
32
Hg
b
9
B
50
Al
.54/1
Ga
55
In
91
Tl
C
16
Si
Ge
120
Sn
84
Pb
N
150
P
57
As
6.0/1
32
Sb
49
Bi
O
17.5/3
84
Be
103
Te
Po
H
F
C3
Br
I
At
He
Ne
Ar
Kr
Xe
Rn
oo
cr>
Gd
64
u
Cm
id
'Average concentrations/average deviations are given in the form 20.0/2 to be read 20.0±0.2.
-------
by Dr. Fung-Jou Lu, Associate Professor of Biochemistry, College of
Medicine, Taipei in the Yenshei-Provlnce from two wells and shipped by air
mail to College Station on July 9, 1980.
Pei-Men and Pu-Tai Samples .'.'•:
The samples from Pei-Men and Pu-Tai were received toward the end of
May 1977 in Chlorox containers. Later, a second set of Taiwan samples was
received in one gallon Cubitainers with the date of collection unknown. The
samples had a yellowish hue with some flocculated matter present. Analysis
of the Taiwan waters was initiated on June 24, 1977. All analyses were
performed on the original, unfiltered, unpreserved water as received except
that in some cases dilution with water from a sub-boiling still, or with
0.1 £ Ultrex HNC>3 was necessary to bring the analyte- concentration into the
linear region of the analytical working curve. Data relating to the
speciation of arsenic, total arsenic, sodium, copper, manganese, zinc, and
iron concentrations are given in Table 22.; Tseng £t_ al. (71) reported that
the total arsenic concentrations for drinking water .taken from artesian
wells in 37 villages located in the southwest ••-'.of. Taiwan fell within the range
0.017 to 1.097 ppm. The total arsenic concentrations,, determined by
flameless atomic.absorption, spectrometry and-, neutron activatipn analysis for
the Pei-Men and Pu-Tai samples, fell within this range. Only arsenite
TABLE 22. ANALYSIS RESULTS FOR THE PEI-MEN AND PU-TAI WATER SAMPLES
Element Pei-Men Pu-Tai
Arsenite, ppm As (Hydride technique) 0.05 0.09
Arsenate, ppm As (Hydride technique) 0.52 0.63
Arsenite and Arsenate . . 0.57 0.72
total As, ppm (GFAA) 0.72 0.76
total As, ppm (Neutron activation) 0.76 ...,.
Sodium, ppm (Flame AAS) 282 223
Copper, ppm (Flame AAS)
Manganese, ppm (Flame AAS)
Zinc, ppm (Flame AAS)
Iron, ppm (Flame AAS)
(^0.07 ppm) and arsenate (^0.57 ppm) could be detected. Since some of the
well waters are known to be anaerobic, one can expect substantial chemical
changes to have occurred in the samples since, the time of collection. The
sum of the arsenite and arsenate concentrations is somewhat less than the
total arsenic level found by flameless AAS. The sodium concentration is
• . ' . 87 •.-••.•••.-,...•
-------
rather high (200-300 ppm).
Trace Elements in the Yenshei-Province Samples
The water samples from two wells (Well I and II) in the Yenshei-Province
of Taiwan were collected in plastic Cubitainers. Two samples each of
unpreserved,water, preserved by addition of 0.1 weight percent of ascorbic
acid, and preserved by acidification to 0.1 M HNOs arrived by air mail at
College Station on July 13, 1980 and were analyzed during the two week
period following arrival.
The total arsenic concentrations and the concentrations of arsenite and
arsenate in these samples are given in Table 23. The total arsenic concen-
trations determined by the hydride generation technique are in the range of
0.84 ppm to 1.1 ppm. Arsenate is the predominant arsenic compound in these
samples. Unpreserved samples and samples preserved by addition of ascorbic
acid had the same arsenite concentrations indicating that the chemical forms
of arsenic in the samples did not change during the time period between
collection and' analysis.
The trace element concentrations' in these water samples were determined
by ICP shortly after: arrival. The results,'however, were unreliable. The
arsenic concentrations in most of,the samples were found to be <60 ppb by
ICP.. When our-;own ICP became operational the water samples were analyzed
again. The arsenic concentrations in the .eight month old samples were in
fair agreement with those found by hydride generation in the freshly
collected, samples. 'The Yenshei:. water samples contained Li, Na, K, Mg,. Ca,
Sr,. Fe, B, Si, P, As and S. The results of these analyses are summarized in
Tables 24 and 25 -for-Well, I and Well II, respectively. The chemical form of
the sulfur in these water samples (0.5 and 1.5 ppm) was not determined.
Sulfate was found in. several well water samples from the'blackfdot disease
area in Taiwan (7la). Additional experiments are needed to determine the
chemical form of .sulfur." 'A-special search should be. made for sulfite •,
which caused decreased DNA synthesis in human lymphocytes and is mutagenic
in microbial test systems (71b) -.•'•• '
Fluorescent Compounds in the Yenshei Water Samples
Blackfoot disease is an affliction peculiar to the south-west coast of
Taiwan. The disease is caused by blockage of the peripheral vascular system
and extreme cases are identified by ulceration and loss of the limbs (72, 73).
Fluorescent pollutants in artesian well waters have been cited as possible
causes of this disease. As a result, several studies have been undertaken
to identify and characterize these fluorescent compounds. Lu et_ al_. (73)
have found that the water produces an intense green fluorescence under
.ultraviolet light. Chloroform extracts of the water were separated into
nine components by thin-layer chromatography. The pH of these waters
gradually increased to 9.0 upon standing in air at room temperature or upon
boiling. Attempts were made to isolate and identify the fluorescent
compounds in the well waters (74). The compounds were adsorbed on an anion
exchange resin and eluted with a 5% ammonium chloride solution. The eluate
88
-------
STABLE 23. RESULTS OF TOTAL ARSENIC ANALYSES AND ARSENITE AND ARSENATE DETERMINATIONS IN THE YENSHEI
'"= WATER SAMPLES* ' ,
00
\o
K '3%
Sample "2 °
o-o
•••' : 33
*^C ^1
^Well I -
+
""•
Well II
+
*Average + average
XI •";'••:'".''.' '•.-'•.'
o° :> TOTAL ARSENIC, ppm
w „ Arsenite, ppm Ar senate, ppm
S| 2 .I,' HG ; ICP ;, by HG by HG
i-H • Cu • • • ;.••---•
. H • ••••'•-.; ,V
+ - 0.84+0.02
0.88+0.01
+ ; 0.85+0.01
+ - 1.0510.08:
- 1.16+0.03;
+ 1.1110.02 ;
6.73+0.01 - -
0.72+0.01 0.024+0.001 0.85+0.01
0.73+0.01 0.022+0.001 0.8310.08
0.95+0.01 -
0.96+0.01 - -
Q.98±0.01 0.02410.01 1.08+0.02
deviation of .at least three determinations per sample :
' • . . .' '••.'•;.. '•':'•.: •':-. . f ' ' '.
-------
TABLE 24. TRACE ELEMENT CONCENTRATIONS IN THE YENSHEI WATER SAMPLES, WELL I, DETERMINED BY ICP
H
V-3
• La
0.007
5.5
Ma
280
42
K
1.0
Rb
Cs
Fr
0.46
Be
34
Mg
8.6
45
Ca
4.3
0.44
Sr
0.08
0.58
Sa
Ra
Concentration of the — •
Element in ppm Found
in the Sample*
*The absence of a yyyy
concentration is at 01
8
Sc
2
Y
La
Ac
16
Ti
-. 4
Zr
Hf
8
V
Nb
Ta
8
Cr
8
Mo
18
w
E
•yyyy
figure indicates that the
r below the detection limit.
Mn
Tc
Re
8
Fe
0.46
Ru
Os
6
Co
Rh
Ir
10
Ni
Pd
25
PI
4
Cu
2
Ag
48
Ail
10
Zn
6
Cd
Hg
9
B
50
A!
Ga
55
In
91
Tl
C
16
Si
3.4
Ge
120
Sn
84
Pb
H
150
P
4.9
57
As
0.73
32
Sb
49
Bi
O
S
0.5
. 84
Se
.103
Te
Po
H
F
Cl
Br
1
At
He
Ne
Ar
Kr
Xe
Rn
Srrs
Eu
Gd
Th
Pa
64
Pu
Cm
Cf
-------
TABLE 25. TRACE ELEMENT CONCENTRATIONS IN THE YENSHEI WATER SAMPLES, WELL II, DETERMINED BY ICP
H
0 4.3
Ll
0.01
5.5
Na
196
42
K
13.6
Rb
Cs
Fr
0.46
Be
34
Mg
23
45
Ca
17.5
0.44
Sr
0.20
0.58
Ba
Ra
Concentration of the--
Element. in ppm Found
in the Sample*
*The absence of a yyyy
concentration Is at o.i
8
Sc
2
Y
La
Ac
16
Ti
4
Zr
HI
8
V
Nb
Ta
8
Cr
;8
Mo
18
m
xxxx-
E
yyyy
Detection Limit in pp
figure indicates that the
rNbelow the detection limit.
Mn
Tc
Re
8
Fe
0.50
Ru
OS
6
Co
Rh
Ir
10
Ni
Pd
25
Pt
4
CIS
2
Ag
48
Au
10
Zn
6
Cd
32
Hg
b
9
B
0.87
50
Al
Ga
55
In
91
Tl
C
16
Si
3.4
Ge
120
Sn
84
Pb
N
150
P
2.3
57
As
0.97
32
Sb
49
Bi
0
S
1.5
84
Se
103
Te
Po
H
•f1
cs
Br
I
At
He
Ne
Ar
Kr
Xe
Rn
Ce
Pr
Nd
Pm
Sm
Eu
Gd
Tb
Dy
Ho
Er
Tm
Yb
Lu
Th
Pa
64
u
Np
Pu
Am
Cm
Bk
Cf
Es
Fm
Md
No
Lr
-------
was evaporated to dryness on a rotary evaporator at 50°. The residue was
extracted with methanol and acetone. The extract produced six blue-green
_fluqrescent bands on aJTLC plate^ One: of^ the compounds was tentatively
Tdeh'tified" as either"TysergTc~acid7 dihydr'olyserg'ic~'a'cid~'o'r'^' related'
compound. Another study (75) has identified ergocalciferol as~i~contaminant
in the drinking water. .In addition to the chemical studies, a number of
publications discuss the toxic (76, 77) and cytotoxic (78, 79) 'effects of
the fluorescent compounds in the drinking water.
f • ' i .
Since ergot alkaloids were suggested to cause the fluorescence in the
artesian drinking water, ion-pair, reversed phase, high performance liquid
chromatography was used in an attempt to identify the compounds. This
method has been recently suggested for ergot alkaloids.
Experimental •
• . "•.•/. ~ •
Liquid chromafeography was performed on an Altex/Beckman Model 312 MP
High Performance Liquid Chromatograph. A Farrand Model 801 Spectrofluoro-
meter equipped with a 10 yL quartz flow cell was used for fluorescence
detection; For preparative liquid chromatography a 15 x 250 mm glass
column was used. A Fisher Model 80 circulating water :bath was employed to
provide elevated'temperature chromatography. Samples were concentrated on a
Brinkman Rotavapor -rotary evaporator.
Amberlite IRA-400C strong anion exchange resins, Partisil 10-SAX
(strong anion exchange column), a Partisil 10-SCX (cation exchange column),
reversed-phase columns (Altex), a 25 cm Spherisorb-C18 and a 15 cm
Ultrasphere-ODS column were used for chromatography. Ergometrine, ergotamine
tartrate, ergocryptine, ergocalciferol and I)-lysergic acid were obtained
from Sigma Chemical Co.
'' •'.''' -t- " • "
Results .:-. *r ..•-..•• .... , '•;
' ' ' *
First, the sample preparation procedure of Lu (74) was•evaluated. The
procedure requires the adsorption of the fluorescent compounds on an
Amberlite anion exchange resin followed by extraction with methanol and then
with acetone. One liter of unpreserved Yenshei Well I water, collected in a
Pyrex bottle, was pumped through a preparative LG•column packed with
Amberlite IRA-400C using a solvent metering HPLC pump with a 2 y fritted
inlet filter. The Farrand fluorometer was used to monitor the column
effluent for the appearance of fluorescent components. After about eight
hours of pumping attempts were made-unsuccessfully toelute the ,
fluorescent components with aqueous 5% NH^Cl, aqueous 10% NHi+Cl and aqueous
10% NHijCl/NHitOH at pH 9. The dark brown band that had formed at the top of
the column did not move. Ethanol, 1 M with respect to HC1, eluted most of
the adsorbed components in the brown band as well as highly fluorescent
resin decomposition^products. The eluate was evaporated to dryness. Dark
brown crystals were noted in the residue. During the extraction of the
residue with MeOH tlje crystals dispersed but did not dissolve. Attempts at
dissolving the brown crystals in water, acetone and cyclohexane failed.
92
-------
A much simpler preparation procedure similar to that of Lu et al. (75),
in which the filtered water is concentrated under vacuum at 50°, was adopted.
The water was filtered through a cellulose filter disk with an average pore
size of 0.5 u. The filtrate was concentrated under vacuum at approximately
50° in a rotary evaporator. Typically 200 mL of well water were concen-
trated to 10.0 mL.
Chromatographic Analyses of Well Water Concentrates
High performance anion exchange chromatography employing a Partisil
10-SAX column and aqueous 0.15 tl NH^Cl at pH 3 or 0.1 M aqueous trisodium
citrate at pH 6 as the mobile phases did not separate the fluorescent
components. Washing the column with water and then methanol revealed that
most of the injected sample had been tightly adsorbed to the column packing.
High performance cation exchange liquid chromatography using a Whatman
Partisil 10-SCX column gave better results. Using gradient elution
(0.001 M -»• .05 M sodium citrate, at pH 6) two highly absorbing and fluorescent
components and an additional, nonfluorescent UV absorbing component were
separated (Fig. 31A). Changing the ionic strength and/or the pH of the
-mobile .phases ;did not improve the, separation. .The frequent appearance of
a shoulder on the strongest fluorescence peak suggested the presence of an
additional compound.. Fractions up to 2.6 mi elution volume (Fig.. 31B) and
fractions with an elution volume >2.6 mL (Fig. 31C) were collected from
four injections, concentrated and rechromatographed. A third fluorescent
component was observed at an elution volume of 2.4 mL (Fig. 31C).
Next the- feasibility of a reversed phase separation of the fluorescent
compounds using ion-pair formation was investigated. Several experiments
with hexanesulfonic acid and heptanesulfonic acid at pH 3 were conducted.
Heptanesulfbnic acid gave the best results. Figures 32 and 33 show
chromatograms of the concentrated water samples and five alkaloid standards
employing a UV-detector at 254 nm and a- fluorescence detector, respectively.
D-Lysergic Acid and ergometrine were not resolved under these conditions.
Calciferol, a rather- insoluble compound, does not: fluoresce and, therefore,
does not produce a signal when a,fluorescence detector is employed (Fig. 33).
The chromatograms contain also peaks,arising" from imparities in the aqueous
solvent which fluoresce and are UV-detectable. Comparing the chromatograms
of the alkaloid standards with the chromatograms of the water concentrates
D-lysergic acid and/or ergometrine and ergotamine could be present in the
water samples. The presence of calciferol cannot, be ruled out. Ergotamine
and a solvent impurity have the same elution volume with the impurity
contributing approximately one third to the total area of the peak. The
chromatograms of the water samples contain several peaks, which do not
correspond to any of the five alkaloid standards.
The water samples,, which had'been concentrated to one twentieth of
.their original volume, deposited a white precipitate, after approximately
36 hours. Such a sample was stored and analyzed after eight weeks to see
whether any changes had taken place. The absorption and fluorescence
chromatograms of the aged water sample are shown in Fig. 34. The
93
-------
uv
Chromatogram of Fraction
with Elution Volume <2;6 ml
Fl uorescence Detector
.Column: Partisil 10-SCX 4.6 mm x 250 mm
Mobile Phase: 0.001 M Trisodium Citrate
at pH 6 (Solvent A)
0.05 M Trisodium Citrate
at pH 6 (Solvent B)
Sample Size: 100 Microliters
Flow Rate: 1 mL/min
Elution Profile: Linear Gradient from 0%
. to 5% Solvent B from 2.5 to 3.0 min.
UV at 254 nm
Fluorescence excitation 358 nm
emission 438 nm
Chromatogram of Fraction
with Elution Volume >2.6 mL
Fluorescence Detector
I I
0 2 5
Elution Volume, mL
2 5
Elution Volume, mL
Figure 31. Cation exchange liquid chromatogram of the concentrated Yenshei
water sample.
94
-------
ALKALOID
STANDARDS
Columns:-Spher1sorb C-18 3.2 nut x 250 ma
connected to U1trasphere-ODS 4.6 ran x
150 mm
Mobile Phase: 0.05 H Heptanesulfonlc Acid. 0.005 M
KH2P04 in Water at pH 3.0 (Solvent A)
Aceton1tr1le (Solvent B)
Sample Size: 100 MicrolUers
Flovt Rite: 1 nt/mln
Elutlon Profile: Linear Gradient from 01 to
1001 Solvent B over 17 Minutes
UV-Oetector at 254 nm
0.64 Absorption Units Full Scale (AUFS)
O —
'II-
ELUTION
10 VOLUME
15
Same Conditions as Above
y
WATER
CONCENTRATE
Figure 32. Ion-pair reversed phase chromatogram of the concentrated Yenshei
water sample and alkaloid standards employing an UV-detector.
95
-------
ALKALOID STANDARDS
Sane Conditions as Below
Sample Size: 100 Mcrollters Containing
0.035 mg 0-Lyserg1c Add
0.031 mg Fr-gome trine
0.014 mg Ergotamlne Tartrate
0.031 mg a-Ergocrypt1ne
0.010 mg Calciferol
Fluorescence Detector Attenuation 0.03
1 t 1
' 1
5
i i
ELUTON •
1 1 1
10
1 1
VOLU« .
i i i
15
i i i
i i i
Columns: SpheHsorfc C-18 3.2 nui x 250 run connected to
Ultrasphere-OOS 4.6 ran x 150 m
Mobile Phase: 0.05 M heptanesulfonlc Acid. 0.005 H
KH,PO. in ffater at pH 3.0 (Solvent A)
AcitoAttrlle (Solvent 8)
Sample Size: 100 Micro]Hers
Flow Rate: 1 mL/m
-------
Sax Conditions as In Figure 33
Fluorescence Detector •
Attenuation. 0.01
VOLIW. it
Sane Conditions as 1n-Figure 32
UV-Oetector-.
ELUTIOJI -10 VOLUK. rt,
20
Figure 34. Ion-pair reversed phase chromatograms of the aged, concentrated
Yenshei water sample employing a UV-detector at 254 nm and a
fluorescence detector.
97
-------
chromatograms of the aged sample are markedly different in the elution
volume region from 5 to 7 mL and from 15 to 17 mL when compared to the
chromatograms of the fresh samples (Fig. 32, 33).
Unconcentrated artesian well water was also analyzed by rapid scanning
f luorometry as described previously (81) . The emission-excitation matrix
for this samples is shown in Figure 35. The intensity values represent the
sum of four single frame scans electronically added to comprise a single
integrated spectrum. The data indicate a blue-green fluorescence with the
emission spread over more than 100 mn from ^350 nm to 500 nm. Rapid
scanning fluorometry could be used for routine and quick characterization of
well waters containing fluorescent compounds .
,.11753
276
Scattered Light Spectrum
Figure 35. Rapid scanning fluorescence analysis of unconcentrated
' ""••'• Yenshei well water.
Chromatography and analysis of the fluorescence of the Yenshei water
sample showed that several alkaloids and other unknown fluorescent compounds
were present. A sufficient amount of sample was not available to separate
these compounds chromatographically and analyze them by mass spectrometry.
Additional work in this area is urgently needed.
WATER SAMPLES FROM NOVA SCOTIA
The initial set of Nova Scotia samples was received in linear-
polyethylene bottles. The second set was collected in one gallon Cubitainers
at the same time (April 21, 1977). Sample 1 was taken from a dug well
suspected of causing chronic arsenic poisoning and anemia. It contained a
98
-------
TABLE 26. ANALYSIS RESULTS FOR THE NOVA SCOTIA WATER SAMPLES
Element
Arsenite, ppm As
6 weeks after collection*
at time of sample collection^"*"
24 hrs. after collection''"''
Ar senate, ppm As
6 weeks after collection*
at time of sample collection ' '
24 hrs. after collection ^
Total Inorganic Arsenic (Arsenite and Arsenate)
6 weeks after collection*
at time of sample collection''"''
24 hrs. after- collection"*^
Total Arsenic (6 weeks after collection)
AAS**
. . .';. '---• ••..,.••• -.. NAAtt .. •
Sodium, ppm .
Copper, ppm
Manganese, .ppm
Zinc, ppm - '
Iron, ppm
Ammonia, ppm'"''
Oxygen, ppmft
pHtt. .... . . ....
Sample 1
John Hartlin
0.68
4.5 ,
. -
4.48
3.5
"
5.16
8.0
6.8
8.2
60
0.08
2.4
: «0.1 •'""-'
«0.1
15
<1.0
6.6
Sample 2
H. Sullivan
0.05
0.31
0.21
0.23
0.32
0.18
-0.28
0.63
0.39
0.52
1.1
50
0.21
<0.1
•0.9
«0.1
. -•• •
3.2
8.0
* Determined by Hydride Generation.
Neutron Activation Analysis.
** Graphite Furnace Atomic Absorption Spectrometry.
Data provided by C. E. Tupper, Nova Scotia Department of Public Health.
99
-------
substantial quantity of large particulate matter. Sample 2 was collected
from a drilled well with no known harmful physiological effects.
The results of the arsenic speciation studies performed during June 1977
and total arsenic, sodium, copper, manganese, zinc, iron, ammonia, dissolved
oxygen, and pH determinations are shown, in Table 26. The total arsenic
concentration in sample 1 is approximately 6 ppm. The arsenite and arsenate
analyses carried out in the field by the.Nova Scotia Department of Public
Health indicated comparable concentrations for arsenite (4.5 ppm) and
arsenate (3.5 ppm> in.sample 1. The dissolved oxygen content, for sample 1
was very low at less than 1 ppm'. Six weeks, after the collection the sample
contained 0.68 ppm arsenite and 4.48 ppm arsenate. Since a considerable
period of time had elapsed between sampling and analysis, aerial oxidation
of the arsenite to arsenate could have occurred. The sum of the arsenite
and arsenate levels is less than the total arsenic concentration determined
by GFAA. The differences for samples 1 and 2 are 1.6 and 0.24 ppm,
respectively. It is not certain whether these differences reflect changes
in the total arsenic content as a function of time (decreasing with
increasing time),, are indicative of undetermined arsenic species or were
caused by interferences ..specific for GFAA. The dissolved arsenic concen-
tration in sample 2 dropped sharply after 24 hours from 0.63 ppm to 0.39 ppm
with the remainder present as particulate arsenic (0.17 ppm).
An unusually high manganese content (2.4 ppm) was found for sample 1.
Copper in the range 0.1 to 0.2 was also detected in both Nova Scotia
samples. Zinc was found in sample 2 at the 0.1 ppm level. The sodium
levels were quite moderate at ^50 ppm.
DISCUSSION
The water samples analyzed as part of this project had total arsenic
concentrations in the range 18 ppb to 8 ppm. The arsenite/arsenate ratios
are in the range 0.007 to 3.4 (Table 27). No indications of the presence of
methylated arsenic compounds,, which are reducible to methylarsine or
dimethylarsine, have been found. Experiments with the high pressure liquid
chromatograph/graphite furnace atomic absorption spectrometer system, which
would provide information about the presence of organic arsenic compounds
not reducible to methylarsines, detected only arsenite and arsenate.
Comparison of total arsenic concentrations with the sum of the arsenite and
arsenate concentrations place an upper' limit on the concentrations of any
other arsenic compounds which might be present. .These, upper limits are in
most cases in the low ppb range.
The other trace elements found in these water samples by ICP are listed
in Table 27. There are no particularly offensive elements in these water
supplies with exception of beryllium (7.4 ppm) in the untreated Antofagasta
samples.
The various physiological effects observed in populations exposed to
these arsenic-containing drinking water supplies (Table 28) could have been
caused by the presence of varying amounts of arsenite and arsenate
100
-------
TABLE 27. SUMMARY OF TOTAL ARSENIC, ARSENITE, ARSENATE AND TRACE ELEMENT CONCENTRATIONS IN DRINKING WATER SAMPLES*
Total As
Arsenite
Arsenate
Arsenite/
Arsenate Ratio
Inckley
PQ
0.18
0.010
0.18
0.06
CO
4J
01
0
0.02
0.010
0.010
VI. 0
. 8
MH
0)
H
m
3.1
2.4
0.7
3.4
Mtt
01
1
.- '" »
• 4.5-6.0
0.35-4.6
0.1-4.3
ntofagasta
ritreated
< P
0.75
0.016
0.74
0.02
ntofagasta
reated
< H
0.41**
0.003
0.41
•
0.007
eiishei I
w ., '
0.85
0.023
6.84
'0.03
M
M
•H .
01
1
01 .
1.1
0.024
1.08
i
j 0.02
ova Sco'tia 1
2
8.0
4.5
3.5
1-3
ova Scotia 2||
a .
0.63
0.31
0.32
1.0
Al
B
Ba
Be
Ca
Cu
Fe
K
LI
Mg
Mn
Na
P
Pb
S
SI
Sr
Ti
V
Zn
. _ _
1.2
0.026
.
3.24
0.08
,
-
-
1.83
233
-
.
*
13.4
0.10
_
.
•
_
0.01
0.045
-
15.4
0.26
0.007
-
-
6.65
-
65
-
1.17
*
15.3
0.52
-
_
-
0.06-0.50
"
0.17
. -
180-291
- ; •
29-53
12
-
52-113
0.42
37-47
0.27
_
28
11.0
0.47
0.008
0.15
0.14
.0.06-0.54
•'
0.29
- •
200-309
• "• ..'-''••
25
13-118
: 0.007
55-139
0.56-0.76
35-50
• 'f-
.
17.5
11.0
0.60
0.009
0.17
0,30
0.08
2.8
0.008
7.4
20.0
0.003
0.11
1313
0.62
7.4
0.002
102
0.34
-
*
38.7
0.28
..
-
0.10.
3.1
2.8
0.008
-
20.3
0.007
0.30
13.2
0.64
7.5
0.006
103
0.26
-
*
36.3
0.29
- " ,:.
_
0.10
> ' :
,T
'- ' '.
• - ':'
• ^3
6.46
1.0
0.007
: !3.6 ;
•: ^'- ' "
280
4.9
'.
0.5 ,
3.4 •
•- 0;08
'-.-.' :-
•;
*' '/-••
_
0.87
"-
-
17.5
- '
0.50
13.6
0.01
23
* '-
196
2.3
1.5
3.4
0.20
-
- '
•
*
*
*
*
*
0.008
<0.1
*
*
*
2.4
60
* •
*
*
*
*
<0.1
* '
*
*
*
*
*
*
0.21
<0.1
*
*
*
<0.1
50
*
*
*
*
*
<0.9
*
*
* Not determined. ** As In treated water Is normally less than 100 ppb. See text for discussion.
t The concentrations are given In ppm. ft Results from two different samples collected one year apart.
-------
TABLE 28. ARSENIC-CONTAINING WATER SUPPLIES AND THEIR PHYSIOLOGICAL
. MANIFESTATIONS IN MAN
Sampling Location
Type of Water and Range
of the Total Arsenic
Concentration....
Symptoms Observed in
the Population
Taiwan (37 villages)
Chile (Antofagasta)
Bakersfield, California
Fallen, Nevada,
artesian well waters used
for 45 years; arsenic
leached from geologic
deposits;- 0.017-1.097 ppm;
median 0.5 ppm. .
drinking water supply
since 1958; 0.8 ppm before
water treatment,0.1
after-water treatment.
community drinking water
supply; 0.3-0.7 ppm.
drinking water, 0.1 ppm.
melanois, keratosis,
skin cancer; 15% pre-
valence among: males
over age 60, normal
incidence 2-3%.
melanosis, hyperkera-
tosis; vascular
manifestations:
myocardialischemia,
hemipleglia with
occlusion of the
carotid artery,
. mesenteric arterial
thrbmbos is, pneumonia.
no adverse effects
reported.*
no known adverse
physiological effects*
*In a study (82) of the arsenic exposure of populations in Bakersfield,
California, and Fallen, Nevada, by a questionnaire .designed to elicit
information about arsenic related, symptoms and diseases very few symptoms
were reported. The incidence of these symptoms was not significantly
different from control populations not: exposed to arsenic.
(Table 27). It is also conceivable that one or more of the trace elements
present in the water supplies act in concert with arsenic to cause the
observed effects. Many more samples need to be analyzed and the results
of these analyses correlated with epidemiological studies before a definite
statement can be made about the interactions of trace elements with arsenite
or arsenate.
The chromatographic work on the fluorescent compounds in the Taiwan well
waters strongly suggests the presence of alkaloids such as D_-lysergic acid,
ergometrine and calciferol. Additional experiments (preparative
chromatography, mass spectrometry) could not be carried out because of
insufficient sample.
102
-------
REFERENCES
1. National Academy of Sciences,. Committee on Medical and Biological Effects
of Environmental Pollutants. Arsenic. Washington, D.C., 1977.
2. Doak, G. 0., and L. D. Freedman. Organometallic Compounds of Arsenic,
Antimony and Bismuth. Wiley Interscience, New York, 1970.
3. Reference 1, pp. 11, 80; and references cited there.
4. Challenger, F., ACS Symp. Series, 812. (1978) 1.
5. Cox, D. P., ACS Symp. Series, 7 (1975) 81.
6. Braman, R. S., ACS Symp. Series, _7 (1975) 108.
..'• . • -. " ••;..- • • • • • •••-'','••• '.>•' •••• ••/."•
7. Woolson, E. A., Environ. Health Perspectives, i9. (1977) 73.
8. Irgolic, K. J.., E. A. Woolson, R. A. Stockton, R. D. Newman,
N. R. Bottino, R. A. Zingarp, P. C. Kearney, R. A. Pyles, S. Maeda,
W. J. McShane, and E* R. Cox, Environ. Health Perspectives, j.9_ (1977)
61, and references cited there.
9. Bottino, N. R., E. R. Cox, K. J. Irgolic, S. Maeda, W. J. McShane,
R. A. Stockton, and R. A. Zingaro, ACS Symp. Series, J52_ (1978) 116.
10. Edmonds, J. S., K. A. Francesconi, J. R. Cannon, C. L. Raston,
B. W. Skeltpn, and A. H. White, Tetrahedron Lett., (1977) 1543.
11. Cooney, R. V., R. 0. Mumma, and A. A. Benson, Proc. Natl. Acad. Sci. USA,
_75 (1978) 4262.
12. Francesconi. K. A.. Nature. 289 (1981) 602.
13. Reference 1, p. 33.
14. Reference 1, pp. 4-15.
15. Zief, M., and J. W. Mitchell. Contamination Control in Trace Element
Analysis. John Wiley & Sons, New York, 1976, pp. 27-33.
16. Robertson, D. E., Ana1. Chem. Acta.42 (1968) 533.
17. Ryden, J. C., J. K. Syers* and R. F. Harris, Analyst, 97 (1972) 903.
103
-------
18. Guimont, J. , M. Bichette, and N. Rheaume, At. Absorption Newsletter,
16 (1977) 53.
19. Ray, B. J. , and D. L. Johnson, Anal. Chim. Acta,. 62. (1972) 196.
20. Talmi, Y., and D. T. Bostick, J. Chromatogr. Sci., 13 (19?5) 231.
21. Whitnack, G. ~C., and R. G. Brophy, Anal. Chim. Acta, 48 (1969) 123.
22. Al-Sibaai, A. -A., and A. G. Fogg, Analyst. 9£ (1973) 732.
23. Portmann, J. E. , and J. P. Rlley, Anal.; Chim. Acta,. 31 (1964) 509.
24. Wagner, R. , Z. Anal. Chem. . 282 (1976) 315.
25. Reference 15, p.. 27.
26. Harrison, S. H. ; P. D. LaFleur, and W. H. Zoller, Anal. Chem. , 4_7
(1975) 1685.
27. Andreae, M. 0. , Anal. Chem., 49 (1977) 820. .
28.; Andreae, M. 0. , Deep-Sea Res. , 25_ (1978) 391.
29. Andreae, M. 0., Limnol. Oceanogr., ^4 (1979) 440.
30. Owens, J. W. , and E. S. Gladney, At. Absorption .Newsletter, 15_ (1976) 47.
31. Feldman, C. , Anal. Chem., 51 (1979) 664.
32. Taqui Khan, M. M. , and A. E. Kartell, J. Am. Chem. Soc. , 89_ (1967) 4176.
33. Reference 1, p . 255 . .
34. Talmi, Y. , and C. Feldman, ACS Symp. Series, 7 (1975) 13.
35. Peoples, S. A. , J> Lakso, and T. Lais, Proc. West. Pharmacol. Soc. , L4
v .'. ' . (1971) " ' ' • '''' -''
36. Johnson, D. L. , and M. E. Q. Pilson, Anal. Chim. Acta, _58_ (1972) 289.
37. Myers, D. J., and J. Osteryoung, Anal. Chem., 45 (1973) 267.
38. Elton, R. , and W. E. Geiger, Jr., Anal. Lett., £ (1976) 665.
39. Bess, R. C., K. J. Ir'golic, J. E. Flannery, and T. H. Ridgway,
Anal. Lett., J9 (1976) 1091.
40. Bess, R. C. , K. J. Irgolic, J. E. Flannery, and T. H. Ridgway,
Anal. Lett., 10 (1977) 415.
104
-------
41. Braman, R. S., and C. C. Foreback, Science. 182 (1973) 1247.
42. Braman, R. S., D. L. Johnson,. C. C. Foreback, J. M. Ammons, and
J. L. Bricker, Anal. Chem., 49^ (1977) 621.
43. Talmi, Y., and D. T. Bostick, Anal. Chem.. 41 (1975) 2145.
44. Edmonds, J. S., and K. A. Francesconi, Anal. Chem., 48 (1976) 2019.
45. Crecelius, E. A., Anal. Chem., 5£ (1978) 826.
46. Stringer, C. E., and M. Attrep, Jr., Anal. Chem., 5_1 (1979) 731.
47. Talmi, Y., and V. E. Norvell, Anal. Chem., 47 (1975) 1510.
48. Cox, D. P., and M. Alexander, Bull. Environ. Contain. Toxicol., 9. (1973)
••....,. 84., '.:.-,• . ' . . ... ;,., ... .
49. Salmi, E- J., K.. Merivuori, and E. Laaksonen, Suomen Kemistilehti, 19B
(1946) 102. '••'••""• ' •'• •• , /^"^ ' •"':"""- •'-' ; ' """';- •"•'".'.'.,.-"
50. Soderquist, C. J., D. G. Crosby, and J. B. Bowers, Anal. Chem., 46
: ,.(1974)155,.- . -......,.,,,.,,/;. :,.,'-.,.:.:-.. ...,.-•... ,>.." ':,-'.•••
51. Lodmell, Ji P.. Piss. Abstr. Int. B, 34 (1974) 5359.
52. Johnson, L. D.,, K. L.^ Gerhardt, and W. A.. Aiie, Sci. Total Environ., I
(1972) 108.
53. Parris, G.. E.; W. R. Blair, and F.. E. Brinckman, Anal. Chem., 49_ (1977)
378. ' '"" " __.. •. '.-'':'.' •"'•'"' ;,/ .''.':' "•' . ' :• • • ' ' ' •
54. Fernandez, F. J.', At. Absorption Newsletter, 16 (1977) 33.
55. Cantillo, A.. Y., and D. A. Segar, Proc. Int. Conf. Heavy Metals in
Environment, Toronto, Canada, 1976, p.
56. Brinckman, F. E., W. R. Blair, K. L. Jewett, and W. P. Iverson,
J. Chromatogr. Sci. ,V 15 (1977) 493. • : ,. • ' '
57. Stockton, R. A., and K. J. Irgolic, Intern. J. Environ. Anal. Chem., 6^
. . ' (1979) 313. ''••"... " ' . • • , :;;;- ."' ' :: " " ': . • ' :'- .;- "." -
58. Brinckman, F. E., K. L. Jewett, W. P. Iverson, K. J. Irgolic,
K. C. Ehrhardt, and R. A. Stockton, J. Chromatogr., 191 (1980) 31.
59. Jung,. W. G. 1C Timer Cookbook. Howard W. Sams and Co., Inc.,
4300 W. 62nd Street, Indianapolis, Indiana, 46268; pp. 244-247.
105
-------
60. Pierce, F. D., and H. R.. Brown, Anal. Chem., 49 (1977) 1417.
61. Woolson, E. A., and N. Aharonson, J. Assoc. Off. Anal. Chem., 6i3_ (1980)
523.
62. Ricci, G. R., L. S. Shepard, G. Colovos, and N. E. Hester, Anal. Chem.,
53. (1981) 610.
63. Arnold, J. P., and R. M. Johnson, Talanta,. 16 (1969) 1191.
64. Meites, L., J. Am. Chem. Soc.. 7$. (1954) 5927.
65. White, M. C., and A. J. Bard, Anal. Chem., _38 (1966) 62.
66. Princeton Applied Research, Application Brief A-117.
67. Smith, M. R., K;. J. Irgolic, and R. A. Zingaro, Thermochim. Acta, 4_
(1972) 1.:
68. Irgolic, K. J., R. A. Zingaro, and L. J. Edmonsqn, Jr., Phosphorus, 5_
(1975) 183. ; " . • . : .".''/.. .
69. Borgono, J. M.., P. Vicent, H. Venturing, and A. Infante, Environ.
Health Perspectives, 19_ (1977) 103.
70. Communication from the personnel at the Antofagasta water treatment
plant. •'.-'.
70a. Song, T. J., and G.. S. Logsdon, J. Am. Water Works Assoc., 70 (1978)
379.
70b. Groth, D. H., L. Stettler, and G. Mackay, in "Effects and Dose-
Response Relationships of Toxic Metals," G. F. Nordberg, Ed.,
~~ETsevie"r~ Scientific Publishing Company, Amsterdam, 1976, p. 527.
70c. Sunderman, F. W., Jr., Biol. Trace Element Research, 1 (1979) 63.
70d. Casto, B. C., W. J. Pieczynski, R. L. Nelson, and J. A. DiPaolo,
Proc. Amer. Assoc. Cancer Res.. 1_7_ (1976) 12; and Casto, B. C.,
J.Meyers, and J. A. DiPaolo. ibid, 18 (1977) 155.
71. .Tseng, W. F.,~¥7~K'. Chu, S. H. How,~T."~M7~Fong , C. S. Lin, and S". Yeh,~
:' J .
71a. Shen, Y. S., and C. S. Chen, Adv. Water Poll. Res.. 1 (1964) 173.
71b. U.S. Environmental Protection Agency, "Health Effects of Sulfur
Oxides" External Review Draft, 1980.
72. Chen, K. P., and H. Y. Wu, J. Formosan Med. Assoc.. 61 (1962) 611.
106
-------
73. Lu, F.-J., C.-K. Yang, and K.-H. Ling, J. Formosan Med. Assoc., 74
(1975) 596.
74. Lu, F.-J., C.-K. Yang, and K.-H.. Ling, J. Formosan Med. Assoc., 76
(1977) 209.
75. Luv F.-J., M.-H. Tsai, and K.-H. Ling, J. Formosan Med. Assoc.. 77_
(1976) 68.
76. Lu, F.-J.,, M.-H- Tsai, and K.-H. Ling, J. Formosan Med. Assoc., 76
(1977) 58.
77. Lu, F.-J., and K.-H. Ling. J. Formosan Med. Assoc.. 79 (1979) 314.
78. Chen, J.-Y., F.-J. Lu, and. K.-H Lin, Chinese J. of Microbiology, 10_
(1977) 115. •;••••.: ,-••:•"••
79. Jam, I.-R., J~.-Y. Chen, F.^-J.Lu, and K.-H. Ling, J. Formosan Med.
••.-'• Assoe., 77 (1978) 361. >,. '•:,.r-.-- ••••••:...•:.:•• •...••••'. ••-••••'•• ''•-'
•80. O'sborne,. B.: G.,:;Altex CHfomatogram:,::3 (1980) 5. . ; :
81. Warner, I. M. ,.M. P.. Fogarty, and D.. C. Shelly,. Anal. Chim. Acta. 109
"" "• '- -(1979) 361. - - .•.••:-:'" '' "":''" ' ' -•:;>••"- •'-'••"•'•.-•••^•-. • " • : •
82. Valentine, J. L., School of Public Health,, University of California,
Los Angeles, private communication. :
107
------- |