United States
Environmental Protection
Agency
Health Effects Research
Laboratory
Research Triangle Park NC 27711
xvEPA
Research and Development
EPA-600/S1-82-010 Nov 1982
Project Summary
Speciation of Arsenic
Compounds in Water Supplies
Kurt J. Irgolic
The objectives of this project were
to develop and testanalytical methods
that would allow the chemical 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 attri-
buted to arsenic. Analytical techniques
were developed for the determination
of arsenate (differential pulse polaro-
graphy), for inorganic and organic
arsenic compounds (high pressure
liquid chromatography with graphite
furnace atomic absorption spectro-
metry as element-specific detector)
and for the detection of arsenocho-
line, arsenobetaine, and iodoarsines
(mass spectrometry). These tech-
niques, inductively coupled argon
plasma emission spectrometry, and
hydride generation/DC-helium arc
emission were used for the character-
ization of water samples from Utah,
Alaska, Antofagasta, Taiwan and
Nova Scotia. The total arsenic con-
centration ranged from 18 ppb to 8
ppm with arsenite/arsenate ratios
between 0.007 and 3.4. No organic
arsenic compounds were detected in
any of the water samples. The trace
elements Al, 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 ex-
posed to the arsenic-containing water
supplies could not be caused by
arsenic compounds other than ar-
senite or arsenate. Other trace ele-
ments acting in concert with arsenite
and/or arsenate might produce these
symptoms. However, sufficient data
are not yet available to evaluate these
hypotheses.
This Project Summary was devel-
oped by EPA's Health Effects Re-
search Laboratory. Research Triangle
Park. NC, to announce key findings of
the research project that is fully
documented in a separate report of the
same title (see Project Report ordering
information at back).
Introduction
Arsenic is an element possessing a
rich chemistry Inorganic and organic
arsenic compounds may contain tri-
valent arsenic The trivalent arsenic
compounds are generally more toxic
than the pentavalent derivatives. Many
inorganic and organic arsenic com-
pounds 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 supplied by weathering of
arsenic-containing rocks, human use,
and disposal of various arsenic com-
pounds
-------
Since arsenic is ubiquitous, man
consumes small amounts of arsenic
compounds with the food he eats and
the water he drinks. Life developed in
the presence of arsenic. Therefore,
organisms are expected to tolerate a
certain, not yet clearly defined, dosage
of arsenic. Certain geographically
limited groups oT people have taken
arsenic compounds into their systems
over extended periods of time. These
arsenic compounds are present in their
drinking water supplies The most
famous localities where arsemc-con-
tainmg waters have been consumed are
certain regions in Taiwan and the city of
Antofagasta in Chile Hyperpigmenta-
tion, skin cancer, vascular problems and
other ailments have been attributed to
the arsenic present in the drinking
water Other groups, such as the people
in Fallen, Nevada, have been exposed to
similar arsenic levels in their drinking
water without any ill effects 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.
Thus far, arsenite, arsenate, methy-
larsmic acid, dimethylarsinic acid,
methylarsines, arsenobetame, trimeth-
ylarsoniolactic acid and arsenic-contain-
ing lipids have been identified in
environmental samples.
Preservation of Arsenic
Compounds in Aqueous
Solutions
Water samples generally cannot be
analyzed for trace elements immediate-
ly after collection. Several hours or even
several days elapse between collection
and analysis. During this time the
chemical nature of a trace element,
such as arsenic, can change. Trace
elements can be lost by volatization
and/or can be adsorbed on container
walls. The absorption of many metal
ions and of phosphate ions has been
studied, but arsenic was rarely included.
Disagreements exist in the literature
as to the extent of loss of arsenic from
solutions stored in various containers.
Even less certainty exists about the
conditions under which various arsenic
compounds can be preserved Experi-
ments have shown that arsenite,
arsenate and dimethylarsinic acid are
not adsorbed from 1 ppm solution on the
walls of Cubitainers* (soft polyethylene
containers manufactured by Kimberly)
Walls of Pyrex containers removed
approximately one percent of the
arsenic from the solutions. Ascorbic
acid at a concentration of 1 mg/mL has
been found to prevent the oxidation of
arsenite to arsenate in distilled water
solutions at room temperature On the
basis of these results most water
samples were collected and stored in
Cubitainers and some of the samples
were preserved by adding ascorbic acid
Development of Analytical
Techniques for the
Determination of Arsenic
Compounds
Whereas adequate methods for the
determination of total arsenic concen-
trations do exist, the choice of tech-
niques for the estimation of arsenic
compounds is limited 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 arsenite, arsenate,
methylarsonic acid, dimethylarsinic
acid, trimethylarsine oxide, ethyl-,
propyl- and butylarsonic acid, and the
arsmes obtainable from these com-
pounds. Non-volatile arsenic com-
pounds and arsenic compounds not
reducible to volatile arsmes could not be
determined with the existing methods.
Therefore, an analytical system with an
element-specific detector had to be
developed that was capable of separat-
ing volatile and non-volatile arsenic
compounds in complex matrices.
The development efforts produced a
high pressure liquid chromatography-
Hitachi Zeeman graphite furnace ato-
mic 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, or-
ganyl lodoarsmes, arsenocholine and
arsenobetaine.
Hitachi Zeeman Graphite
Furnace Atomic Absorption
Spectrometer as an Element-
Specific Detector for High
Pressure Liquid
Chroma tograph y
Liquid chromatography and specially
high pressure liquid chromatography
(HPLC) with the great resolving power of
its microparticulate columns are poten-
'Mention of tradenames or commercial products
does not constitute endorsement or recommen-
dations for use
tially the best techniques for the
simultaneous detection and determina-
tion of arsenic compounds A water
sample may contain many substances
in addition to arsenic compounds. The
common detectors 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 A graphite furnace atomic
absorption spectrometer (GFAA) com-
bines the advantage of element-speci-
ficity with high sensitivity for many
elements An HPLC-GFAA analytical
system was developed employing a
Hitachi Zeeman GFAA with a sample
valve, an injector, and associated
electronics to control the analysis
sequence
The HPLC-GFAA system has func-
tioned almost flawlessly during the past
three years Aliquots of the column
effluent are automatically transferred
into the graphite cup of the GFAA for
analysis The time interval between
consecutive analyses can be made
within 30 seconds. 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 injection.
Upon migration through the chromato-
graphic column the arsenic compounds
are separated and spread out into
bands. Aliquots of 40 fjL are withdrawn
from the effluent. The 40 ^L aliquots
taken from the center of the band 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.
Conditions have been found which
allow the separation of inorganic
arsenic (arsenite and arsenate), arsen-
ocholine and arsenobetaine employing
a C-18 reverse phase column, organ-
ylsulfonates as countenons and mix-
tures of water/acetonitrile/acetic acid
as the mobile phase. Arsenite, arsenate,
methylarsonic acid and dimethylarsinic
acid were similarly separated using
water/methanol mixtures saturated
with tetraheptylammonium nitrate as
the mobile phase. The HPLC-GFAA
system, of course, is neither limited to
the analyses of the arsenic compounds
listed above nor to compounds contain-
ing only arsenic.
-------
Differential Pulse
Polarographic (DPP)
Determination of Arsenate
and Arse nit e
Arsenite is reducible at the dropping
mercury electrode and can be deter-
mined polarographically at concentra-
tions as low as 0 3 ppb. Arsenate is
polarographically inactive under these
conditions. Addition of polyhydroxy
compounds to an acidic solution of
arsenate makes arsenate reducible
Among 11 polyhydroxy compounds, D-
mannitol at 0.5 M concentration in 2 0
M aqueous perchloric acid produced the
largest reduction peak for arsenate The
DPP curve of arsenate under these
conditions is characterized by maxima
at -0 55 V and -075 V 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 into the current maxi-
mum and becomes a shoulder at
As(arsenate) concentrations of approx-
imately 5 ppm The rather low intensity
peak between-0 Wand-0.8 V might be
obscured at low arsenate concentra-
tions by the solvent breakdown and
at high arsenate concentrations by the
current maximum The arsenite reduc-
tion wave in 2 0 M perchloric acid
solution shifts from -0.425 V to -0 34V
upon addition of mannitol
When arsenite and arsenate are
present in solution, the arsenate
reduction peak at -0 55 V can be used
for the determination of arsenate with
some confidence only when the con-
centration of As(arsemte) is between
100 ppb and approximately 500 ppb, the
current is not lower than 2 /uA and the
arsenate concentration is equal to or
higher than the arsenite concentration
If these conditions are not fulfilled,
arsenite must be oxidized to arsenate by
cenumflV) ammonium nitrate. Excess
cenum(IV) must be reduced with hy-
droxylamme hydrochlonde Arsenate is
then determined in the presence of
mannitol using the peak at -0 55 V.
Arsenite is determined in another
aliquot of the sample in the absence or
presence of mannitol. The arsenate
concentration in the sample is obtained
as the difference between the total
arsenic concentration and the Asfarse-
nite) concentration The detection limits
for arsenate under these conditions are
6 ppb at the 95 percent confidence level
Mass Spectrometry of
Organylarsonic Acids,
Diorgan ylarsinic A cids,
Organyliodoarsmes,
Arsenocholine and
Arsenobetaine
Organic arsenic compounds could
perhaps be determined by mass spectro-
metry in the residues obtained by
evaporation of the water samples
Therefore, the mass spectral behavior of
several organic arsenic compounds was
studied
At probe temperatures between
110°C and 250°C required to obtain
satisfactory mass spectra, organylar-
sonic acids, RAs03H2, and diorganylar-
smic acids, R2AsOOH, formed anhy-
drides and decomposed The products of
these thermal reactions were then
ionized and fragmented yielding com-
plicated mass spectra with many peaks
at m/e values higher than those
expected for the molecular ions A
detailed investigation of the spectra of
five arsonic acids and nine arsinic acids
indicated that mass spectrometry was
of little value for the identification of
arsonic acids, but can be used to
establish the presence of diorganylar-
smic acids in the residues from water
samples Exact mass measurements by
high resolution mass spectrometry
might be necessary to distinguish
arsenic-containing from arsenic-free
ions
Organyliodoarsmes, RnAsU-n (n=1,2),
are much more volatile than arsinic or
arsonic acids and are easily prepared by
treating these acids with hydnodic acid
All of the 14 organyliodoarsmes investi-
gated gave intense molecular ion peaks
Fragmentation proceeded by loss of
alkyl groups, iodine and hydrogen
abstraction Organyliodoarsmes are
well suited for the mass spectrometric
identification of organic arsenic com-
pounds which can be converted to
iodoarsmes
Arsenocholine chloride, [(CH3)3AsCH2
CHgOHJCI, and arsenobetame chloride,
[(CH3)3AsCH2COOH] Cl, produce rich
mass spectra which do not contain
molecular ion peaks The highest mass
peaks correspond to (CH3);3AsCH2CH20
and (CH3)2AsCH2COOH, which were
formed by thermal cleavage of HCI and
CHsCI from the arsonmm salts In spite
of the absence of molecular ions, mass
spectrometry can provide an indication
of the presence of arsenocholme
and/or arsenobetaine in a sample.
Analysis of Water Samples
Samples of arsenic-containing drink-
ing water supplies selected by the EPA
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 Hmckley, Utah, Delta, Utah,
Barefoot Site, Alaska, Mauer Site,
Alaska, Antofagasta, Chile, Yenshei,
Taiwan; Hartlm Site, Nova Scotia; and
Sullivan Site, Nova Scotia were invest-
igated
Graphite furnace atomic absorption
spectrometry, differential pulse polaro-
graphy, high pressure liquid chromato-
graphy with a GFAA as an element-
specific detector, the hydride genera-
tion technique with a DC-helium arc
detector, and inductively coupled argon
plasma emission spectrometry were
employed for the determination of
concentrations of total arsenic, and
trace elements The samples were
collected and stored in Cubitamers or
Pyrex glass containers Unpreserved
samples and samples preserved with
ascorbic acid or nitric acid were
analyzed The analyses were carried out
as soon as possible after receipt of the
samples
The water samples had total arsenic
concentrations in the range of 18 ppb to
8 ppm The arsenite/arsenate ratios
were in the range of 0 007 to 3 4 (Table
1) No indications of the presence of
methylated arsenic compounds, which
are reducible to methylarsme or di-
methylarsme, have been found Exper-
iments 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 methylarsmes, detected
only arsenite and arsenate Comparison
of total arsenic concentrations with the
sum of the arsenite and arsenate
concentrations placed an upper limit on
the concentrations of any other arsenic
compounds which might be present
These upper limits were m most cases
in the low ppb range
The other trace elements found in
these water samples by ICP are also
listed m Table 1. There were no signifi-
cant concentrations of these elements
-------
Table 1.
Summary of Total Arsenic, Arsenite, Arsenate and Trace Element Concentrations in Drinking Water Samples^
Hinckley Delta
Barefoot]1^
Antofagasta Antofagasta
/Wauertt Untreated Treated Yenshei I Yenshet II Nova Scotia 1 Nova Scotia 2
Total As
Arsenite
Arsenate
Arsenite/
Arsenate Ratio
Al
B
Ba
Be
Ca
Cu
Fe
K
Li
Mg
Mn
Na
P
Pb
S
Si
Sr
Ti
V
Zn
0 18
0010
0 18
006
12
0026
—
324
008
—
—
—
7 83
—
233
—
—
*
734
0 10
—
—
—
002
0010
0010
-7 0
007
0045
—
154
026
0007
—
—
665
—
65
—
7 77
*
753
052
—
—
—
3 7
24
07
34
006-050
—
0 77
—
180-291
—
29-53
12
—
52-113
042
37-47
027
—
28
11 0
047
0008
0 15
0 14
45-60
0 35-4 6
01-43
0 06-0 54
—
023
—
200-309
—
25
13-118
0007
55-139
056-0 76
35-50
—
~
775
77 0
060
0009
0 77
030
0 75
0016
0 74
002
008
28
0008
74
200
0003
0 11
133
0 62
74
0002
102
034
—
«
38 7
028
—
—
0 70
047"
0003
047
0007
3 7
2,5
0008
—
203
0007
030
732
064
75
0006
703
026
—
•
363
029
—
—
0 JO
0 85
0023
084
003
—
—
—
43
—
046
7 0
0007
86
—
250
49
—
05
34
005
—
—
—
7 7
0024
1 08
002
057
—
—
775
—
0.50
736
007
23
—
796
23
—
75
34
020
—
—
—
SO
45
35
7 3
»
*
*
»
»
0008
<0. 1
»
*
»
24
60
*
*
*
*
*
<0 7
»
*
063
037
032
1 0
»
*
»
*
*
027
<0 7
»
*
*
<0 7
50
*
+
*
*
*
<09
*
*
"Not determined
'**As in treated water is normally less than 100 ppb See text for discussion
'[The concentrations are given in ppm
^Results from two different samples collected one year apart
in these water supplies with the
exception of beryllium found in the
untreated Antofagasta sample.
The various physiological effects
observed in populations exposed to
these arsenic-containing drinking water
supplies (Table 2) 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 m the water
supplies acted in concert with arsenicto
cause the observed effects 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, ergometrme and calciferol
Additional experiments (preparative
chromatography, mass spectrometry)
could not be carried out because of
insufficient samples
Table 2.
Sampling
Location
Arsenic-Containing Water Supplies and Their Physiological
Manifestations in Man
Taiwan
(37 villages)
Chile
(Antofagasta)
Bskersfield, CA.
Fallon. NV
Type of Water and Range of
the Total Arsenic
Concentration
Symptoms Observed in the Pop-
ulation
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.
Melanosis, keratosis, skin cancel
15% prevalence among males
over age 60, normal incidence
2-3%
Melanosis, hyperkeratosis,
vascular manifestations:
myocardialischemia, hemipleglia
with occlusion of the carotid
artery, mesenteric arterial
thrombosis, pneumonia
No adverse effects reported *
No known adverse physiological
effects. *
*(n a study of the arsenic exposure of populations in Bakersfield, Ca , and Fallon, Nv.,
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
4
-------
Recommendations
The use of methods such as the
hydride generation technique; differential
pulse polarography, high pressure
liquid chromatography with sensitive,
element-specific detectors and colori-
metnc methods for determining con-
centrations of total arsenic and 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. Additional
arsenic-containing water samples must
be analyzed in support of or in prepara-
tion for epidemiological studies. Exper-
ience has shown that analysis by one
method is not sufficient to produce
reliable results. At least two independent
techniques should be used for the
determination of arsenic compounds.
Kurt J. Irgolic is with Texas A&M University, College Station, TX 77843.
Frederick C. Kopfler is the EPA Project Officer (see below).
The complete report, entitled "Speciation of Arsenic Compounds in Water
Supplies, "(Order No. PB 82 -257 817; Cost: $ 12.00, subject to change) will be
available only from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Health Effects Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
•&U. S. GOVERNMENT PRINTING OFFICE: 1982/659-095/553
-------
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
Postage and
Fees Paid
Environmental
Protection
Agency
EPA 335
Official Business
Penalty for Private Use $300
S
Kt S
230 $ DEARBOKN SlRtET
CHICAGO it
------- |