Environ Geochem Health
DOI 10.1007/s10653-007-9091-3
ORIGINAL PAPER
An assessment of sampling, preservation, and analytical
procedures for arsenic speciation in potentially
contaminated waters
Youn-Tae Kim Æ Hyeon Yoon Æ Cheolho Yoon Æ
Nam-Chil Woo
Springer Science+Business Media B.V. 2007
Abstract This study was undertaken to ascertain
optimal methods of sampling, preserving, separating,
and analyzing arsenic species in potentially contam-
inated waters. Arsenic species are readily trans-
formed in nature by slight changes in conditions.
Each species has a different toxicity and mobility.
The conventional field sampling method using filters
of 0.45 mm in size could overestimate the dissolved
arsenic concentrations, as passing suspended particles
that can act as a sink or source of arsenic depending
on the site condition. For arsenic species in neutral
pH and iron-poor waters, the precipitation can be
stable for up to 3 days without any treatment, but for
longer periods, a preservative, such as phosphoric
acid, is required. Also, the analytical procedure must
be selected carefully because the levels and hydride
generation efficiencies of arsenic in different species
can vary, even for the same amount of arsenic. For
arsenic speciation in samples that also include
organic species, a hybrid high-performance liquid
chromatography (HPLC) column and inductively
coupled plasma mass spectrometry (ICP-MS) gave
Y.-T. Kim
N.-C. Woo
Department of Earth System Sciences, Yonsei University,
Seoul, Korea
Y.-T. Kim
H. Yoon (&)
C. Yoon
Korea Basic Science Institute, Seoul Center, 136-600
Seongbuk P.O. Box, 12, 88-3, Hawolgok 1-dong,
Seongbuk-gu, Seoul, Korea
e-mail: dunee@kbsi.re.kr
the best resolution and lowest detection limits.
However, the procedure using a solid phase extrac-
tion (SPE) cartridge can be used economically and
conveniently for analyzing samples containing only
inorganic arsenic species, such as groundwater,
especially that related to mine activity.
Keywords Arsenic speciation
Sampling

Preservation
Analytical methods
Sample condition
Introduction
Arsenic is a problematic toxic contaminant, which
has many chemical species, each with a different
toxicity and mobility (Mandal and Suzuki 2002). In
natural water, arsenite [As(III)] and arsenate [As(V)]
predominate (Plant et al. 2005). Organic arsenic
species, such as monomethyl arsonate (MMA) and
dimethyl arsinate (DMA), are only present at low
levels (Bednar et al. 2004). Inorganic arsenic species
include known carcinogens (USEPA, accessed in
June 2006). Of these, As(III) is more toxic than
As(V), while the toxicities of the organic arsenic
species have not been fully evaluated (Thomas et al.
2001; Hughes 2002; Aposhian et al. 2003). As with
their toxicities, the behaviors of arsenic species differ
markedly. The ionic charges of inorganic arsenic
species vary from 0 to 3, depending on the pH and
redox conditions (Raposo et al. 2006). For accurate
risk assessment of arsenic in the environment, the
123

electro-chemical state of dissolved arsenic must be
retained during sample handling.
The stability of aqueous arsenic species is also
dependent on the presence of other dissolved metal
ions. For example, iron is a common element in
arsenic-contaminated waters, which can precipitate if
oxidation occurs during sampling. Arsenic can co-
precipitate with any Fe-oxyhydroxides or even
change its oxidation state (Daus et al. 2002; Bednar
et al. 2002). Pre-treatment of water samples involving
the complexation of metal ions or acidification of
samples has been proposed to prevent any precipita-
tion of metal oxyhydroxides. Daus et al. (2002) used
nitrilotriacetic acid (NTA), hydrochloric acid (HCl),
phosphoric acid (H3PO4), and acetic acid for this
procedure, and found the best results with
0.01 mol l1 H3PO4. Bednar et al. (2002) used
ethylenedinitrilotetraacetic acid (EDTA), sulfuric
acid (H2SO4), nitric acid (HNO3), and HCl to
preserve inorganic arsenic species in groundwater
and acid mine drainage samples. EDTA has been
found to minimize precipitate formation and redox
reactions. However, both of the above studies focused
on iron-rich waters. Preferred pre-treatment proce-
dures for iron-poor waters remain to be determined.
Various analytical techniques have been devel-
oped to distinguish and reduce the detection limits of
different arsenic species. Different ion-exchange
resins have also been tried (Yu et al. 2003), and
hydride generation procedures have been employed,
in conjunction with atomic absorption spectrometry
(AAS), atomic fluorescence spectrometry (AFS),
inductively coupled plasma-atomic emission spectro-
photometry (ICP-AES) or ICP mass spectrometry
(ICP-MS), not only to reduce the detection limits, but
also to overcome interferences (Hung et al. 2004).
Among these methods, a hybrid high-performance
liquid chromatography (HPLC)/ICP-MS method has
been most widely applied (B’Hymer and Caruso
2004). In the majority of water analyses, most
attention has been paid to the predominant inorganic
arsenic species, which are also recognized carcino-
gens (Thirunavukkarasu et al. 2002). Analytical
methods for the separation and detection of organic
arsenic species have been tested in the laboratory, but
these have only recently been applied to natural
waters (Hung et al. 2004; Bednar et al. 2004; Akter
et al. 2005; Sánchez-Rodas et al. 2005), where
123
Environ Geochem Health
distinguishing the different organic species is gener-
ally difficult due to their low concentrations.
In this study, our primary objective is to identify
an optimal procedure for arsenic speciation in natural
water samples without deforming chemical species.
Sampling and preservation methods commonly used
for arsenic-bearing waters were thoroughly tested
under various conditions. Then, laboratory separation
procedures were compared with those employed in
the field and the limitations of the ICP-AES and ICP-
MS analytical methods for the determination of
arsenic species in natural samples under various
conditions were discussed.
Materials and methods
Water sampling and the application
of the preservation method
Groundwater samples were collected from the aban-
doned mines at Guryong and Ulsan, Korea. The
Guryong mine was an Au-Ag-Cu mine. The tailings
from this mine have been found to contain 7.04–
27.27 wt% Fe2O3, with As2O3 < 0.020 wt% (Kim
2004). Groundwater samples were taken from both
monitoring wells (NX size, 3.500 inside diameter) and
piezometers (100 inside diameter), and surface waters
from a pond in the mine tailings. The Ulsan Fe-W
mine was dug into a skarn that contains a consider-
able amount of arsenopyrite (Choi and Youm 2000).
Groundwater samples were collected from monitor-
ing wells in the tailings and from domestic wells
around the mine. Surface waters were sampled from
creeks and ponds near the Ulsan mine. All water
samples showed high arsenic concentrations, which
were generally in excess of the World Health
Organization (WHO) arsenic guideline of 10 mg l1
(WHO 1993) for drinking water, with the pH ranging
from neutral to weakly alkaline (pH 6.2–8.6).
Along with groundwater sampling for a total
analysis of metal ions, various preservation methods
were applied to the separate samples to investigate
the changes in stability of dissolved ionic species,
especially ‘‘arsenic,’’ by the duration of time after
sampling has been done. Selective filter pore size
and preserving chemicals have been applied to
different samples for these purposes. Groundwater
Environ Geochem Health
samples from the Guryong mine were filtered
through both a 0.2-mm and 0.45-mm pore membrane
syringe filter at the sampling site, and then acidified
with HNO3 to lower the sample pH to below 2. The
sample volume was about 10 ml. This sampling was
designed to see whether the conventional filtering
method in field water sampling, using 0.45-mm pore
filters to differentiate elements in dissolved and
suspended particulate forms, is appropriate for
arsenic assessment.
To test the stability of arsenic species followed by
sampling, water samples from the Ulsan mine were
filtered though a 0.45-mm pore membrane syringe
filter, divided into four sub-samples, and then
preserved in four different ways: the addition of:
1. H3PO4
2. HCl
3. Ethylenedinitrilotetraacetic acid (EDTA) to a
final concentration of 0.01 M
4. No treatment
All samples were stored at 48C. The effect of
suspended particles on the arsenic concentration was
tested by changing the order of the H3PO4 addition,
which is known to be a better preserving reagent for
arsenic, either before or after filtration.
Instrumentation
The arsenic concentrations were primarily deter-
mined by ICP-AES (Ultima 2C; Jobin Yvon, Edison,
NJ, USA), with a radio frequency (RF) power of
1000 W. All arsenic measurements were made at
wavelength 193.7 nm. Arsenic was also analyzed
with an ICP-MS (Elan 6100; Perkin Elmer, Waltham,
MA, USA). The instrumental conditions were opti-
mized before each analytical run, with either the
HPLC or short column (solid phase extraction, SPE)
incorporated to permit arsenic speciation analysis. All
ICP-MS measurements of arsenic were made at m/z
75.
A conventional hydride generation (HG) method
was selectively applied to lower the detection limit.
For metal hydride generation of arsenic species, the
sample was mixed with strong acid (6N HCl) on-line
at a volume ratio of 1:1, then followed by mixing
with 1% NaBH4 (Merck) solution in 0.1% NaOH for
arsine gas (AsH3) generation.
Reagents
A standard stock solution, containing 1,000 mg l1
As(V) in 2–3% HNO3, was purchased from Merck
(Darmstadt, Germany). As(III) was prepared by
dissolving 0.0660 g As2O3 (>>99.5% [RT] Fluka,
Buchs, Switzerland) in 50 ml of 0.3% (w/w) NaOH.
Monomethyl arsonate (MMA) and dimethyl arsinate
(DMA) were made by dissolving 0.1058 g Na2CH3A-
sO3
6H2O (99%, Chem Service, West Chester, PA,
USA) and 0.0782 g Na(CH3)2AsO2 (For Synthesis;
Merck) respectively in de-ionized water. A standard
stock solution of 1031 mg kg1 arsenobetaine (AsB)
(CRM626) was purchased from BCR. Working
standard solutions were prepared by appropriate
dilution at the time of each measurement by diluting
0.1 ml of each stock solution to 100 ml. These
solutions were diluted to specific concentrations with
de-ionized water.
Speciation analysis of arsenic by on-line
hyphenated system analysis
For field separation of the inorganic arsenic species,
As(III) and As(V), SPE cartridges (Accell Plus QMA;
Waters, Milford, MA, USA) were used. Various types
of SPE cartridges can be used for routine arsenic
speciation analysis of water. In this study, a strong
anion exchange SPE cartridge (Accell Plus QMA;
Waters) was used. In the field, 10 ml of the filtered
sample was passed through the cartridge at a rate of
1–2 drop s1. The leachate was acidified with HNO3,
and the cartridge then sealed and cooled. Upon return
to the laboratory, each cartridge sample was extracted
with 10 ml of 0.16 M HNO3.
Laboratory separations were undertaken using a
SPE cartridge (Accell Plus QMA cartridge; Waters)
and an HPLC column (PRP X-100; Hamilton, Reno,
NV, USA). The Accell Plus QMA SPE cartridge
provided a short column, which allowed primary
analysis of the inorganic arsenic species at low
pressure. The eluent used consisted of 20 mM of pH
7.0 phosphate buffer, applied at a flow rate of
1 ml min1 using a peristaltic pump, with a sample
injection volume of 500 ml. The PRP X-100 HPLC
column allowed for analysis of several arsenic
species, including organic species. The eluent used
consisted of a 10 mM, pH 9.0 phosphate buffer for
the Ulsan samples and a 30 mM, pH 6.0 buffer for the
123
 Environ Geochem Health

Guryong samples. The flow rate was adjusted to

1 ml min1 using a P680 HPLC Pump (Dionex,
Sunnyvale, CA, USA), with a sample injection
volume of 20 ml.

Results and discussion

Sampling and preservation methods

When iron-rich groundwater is exposed to air, iron

oxyhydroxide can readily precipitate, which can
significantly alter the total concentration of arsenic
if the pH is not sufficiently low (Bang et al. 2005).
This is the main reason why researchers must prevent
their samples from coming into contact with air
during sampling. To quantify the effect of precipita-
tion on the arsenic concentration, we tested both the
filter pore size and acidification procedures during
groundwater sampling. The former factor can help
clarify the interaction with suspended submicron
particles, and provide a clue to the stability of the
water sample. The latter factor is pertinent for
suspended particles in rich water samples, and gives
information on the role of suspended particles as a
sink for or source of arsenic.
A total of 22 groundwater samples from the
Guryong mine, with a mean Fe concentration of Fig. 1 Total arsenic concentrations of samples from (a) NX-
208 mg l1 (min 29 mg l1, max 737 mg l1), were size monitoring well, and (b) 1@ diameter piezometer filtered
through 0.2- or 0.45-mm membrane pores
filtered through two different membrane filters, with
pore sizes of 0.2 and 0.45 mm. Each sub-sample was
then acidified and analyzed by HG-ICP-MS. As sampling method using filters of 0.45 mm in size
shown in Fig. 1, the total arsenic concentrations could overestimate the dissolved arsenic concentra-
displayed two different trends, reflecting the diverse tions while suspended submicron-sized particles were
characteristics of the two types of well. Groundwater present.
samples from the monitoring wells show no differ- Interactions between suspended particles and dis-
ence in arsenic concentrations with regard to chang- solved arsenic species eventually lead either to a sink
ing the filter pore size (Fig. 1a). However, in the or to regeneration of arsenic within many natural
samples collected from the piezometers (Fig. 1b), environments. Waters containing iron oxides, or
those filtered through the 0.2-mm pore membrane hydroxides can precipitate during sampling or upon
show roughly half the arsenic concentration of those preservation, and need treating to prevent co-precip-
filtered through the 0.45-mm pore filter. This indicates itation of arsenic (Gallagher et al. 2001). If filtered
that the samples from the piezometers contained water samples having high Fe contents and insuffi-
more arsenic in suspended particles, ranging from 0.2 ciently low pH were not acidified, an abrupt decrease
to 0.45 mm in size, than those from the monitoring in arsenic concentration could occur as a result of the
wells. Suspended particles in natural waters could co-precipitation of arsenic. For example, in a ground-
function as a sink for or source of arsenic depending water sample from the Guryong mine (pH 5.8, Eh
on ambient conditions. The conventional field 235 mV, Fe 154 mg l1, and As 11.0 mg l1), arsenic

123
Environ Geochem Health
was not detected in water if it had not been acidified.
In this case, precipitation can act as a sink for arsenic.
H3PO4 can prevent the As co-precipitation with iron
oxides or hydroxides, and can also extract arsenic
bound to these species (Ko et al. 2005). Therefore,
the test changing the order of the H3PO4 addition,
before or after filtration, can inform us with regard to
the role of suspended particles. Groundwater samples
from the Ulsan mine contained some suspended
particles; only one sample showed a difference in
arsenic concentrations by changing the order of
H3PO4 addition: 4.15 mg l1 in the sub-sample of
H3PO4 addition before filtration, and 0.09 mg l1 in
the sub-sample of the H3PO4 addition after filtration.
This means that these suspended particles containing
absorbed arsenic prior to sample collection can act as
a source of arsenic by condition change.
Analytical methods
Various techniques have been used for the determi-
nation of arsenic speciation. Widely used procedures
have employed exchange resins to separate arsenic
species, coupled with spectroscopic and ICP analyses
to measure the arsenic concentrations (Hung et al.
2004). In our study, both the SPE cartridge and HPLC
column were used for both field and laboratory
separation of arsenic species to examine any prob-
lems that could arise during analysis.
Analytical implications for various arsenic species
The measured intensities of different inorganic
arsenic species yield almost the same values during
ICP-MS analysis (Table 1) if the absolute concentrations
Table 1 Calculated intensities of arsenic species containing the same amount of arsenic, as measured by inductively coupled plasma
mass spectrometry (ICP-MS). All intensities were corrected using an internal standard
Species Intensity for 10 mg l1
of arsenic
As(V) 0.0829
As(III) 0.0843
MMA 0.0878
DMA 0.0836
AsB 0.1195
As(V) arsenate, AS(III) arsenite, MMA monomethyl arsonate, DMA dimethyl arsinate, AsB arsenobetaine
Inorganic species were calibrated using 0, 0.5, 1.0, 10, and 50 mg l1 standards, and organic species using 0, 1.0, 10, 50, and
100 mg l1 standards
of arsenic species are the same. However, different
behavior was observed when with the use of ICP-
AES as the analytical method, the organic arsenic
species in de-ionized water showed only half the
intensities of the inorganic species. This problem can
be resolved by acidifying the samples with nitric acid
(Table 2).
Similarly, different intensities were obtained with
different species during the HG procedure, even
though the same concentrations of each arsenic
species were measured (Maity et al. 2004). The on-
line HG method was tested for both inorganic and
organic arsenic species by applying ICP-AES in order
to achieve the lower detection limit. Figure 2 shows
the hydride generation efficiencies of various arsenic
species in the same arsenic concentrations analyzed
by the on-line HG-ICP-AES. The efficiency of the
on-line HG procedure was best with the reduced
form, As(III), confirming the result of Hung et al.
(2004). According to the results in Fig. 2, the relative
intensity of As(III), As(V), and MMA showed about
50% intensities, and DMA about 10% intensities
compared with 100%, or even. AsB failed to generate
hydride gas.
Until now, a conventional HG method, without
any pre-reduction, has been most frequently used to
measure the total arsenic concentrations in most
applications, largely because organic species and
As(III) are normally rare in natural waters (Maity
et al. 2004). However, where these species are
present at high levels, the use of conventional
methods for the determination of the total arsenic
concentrations of waters would lead to under- or
over-estimations. Consequently, whenever the con-
ventional HG method is used to measure the total
Calibration
y = 0.0082x + 0.0009, R2 = 1.0000
y = 0.0086x  0.0017, R2 = 0.9999
y = 0.0087x + 0.0008, R2 = 1.0000
y = 0.0083x + 0.0006, R2 = 1.0000
y = 0.0119x + 0.0005, R2 = 0.9998
123
 Environ Geochem Health

Table 2 Relative spectral intensities produced by the same Moreover, in natural waters, organic arsenic species,
amount of arsenic for different species within different matri- such as MMA and DMA, are present as anions.
ces, as measured by inductively coupled plasma-atomic emis-
Consequently, As(III) will not be trapped in an SPE
sion spectrophotometry (ICP-AES)
cartridge, while As(V), MMA and some of the
Species Intensity for 50 mg l1 of arsenic organic species present as anions will (Yu et al.
De-ionized water 2% HNO3 2003), and can be extracted in the laboratory prior to
analysis. In the results of Yu et al. (2003), the
As(V) 1,817.64 1,881.95 percentages of As(III), As(V), DMA, MMA, and AsB
As(III) 1,597.83 1,772.39 retained on 2 g of strong anion exchange sorbent are
MMA 700.67 1,835.86 0, 100 ± 3, 41 ± 5, 100 ± 2, and 6.5 ± 3%. By
DMA 550.98 1,689.41 application of this method, Bednar et al. (2004) found
AsB 777.77 2,173.96 a recovery of 99% for 25 mg l1 As(V) after 47 days.
However, this technique can only be applied to
separate inorganic species if there are, as two or more
arsenic species, present as anions, cannot be sepa-
rated within the SPE cartridge. Further, it is useful to
know the rough arsenic concentrations involved due
to the limited cartridge capacity. For the analysis of
extractants with elevated anion contents, it is better to
avoid ICP-MS analysis, as arsenic when measured at
m/z 75 suffers from 40Ar35Cl+ interference.

Laboratory separation

Laboratory separation procedures were conducted

Fig. 2 Hydride generation (HG) efficiencies for different
arsenic species, as measured by on-line hydride generation
inductively coupled plasma-atomic emission spectrophotome-
try (HG-ICP-AES). As(V) arsenate, AS(III) arsenite, MMA
monomethyl arsonate, DMA dimethyl arsinate, AsB arsenobe-
taine
arsenic concentration in water, the arsenic species
present must be considered, and pre-reduction of
arsenic species to As(III) should be carried out prior
to applying the HG procedure.
Field separation
Employing field separation procedures using SPE
cartridges can avoid the changes in the chemical state
of arsenic that might result from sample storage. The
ionic charges of inorganic arsenic species in solution
can readily occur with changes in pH (Raposo et al.
2006). As(III) has a zero charge under neutral pH
conditions, whereas the change for As(V) is negative.
using a SPE cartridge (Accell Plus QMA cartridge;
Waters), as a short column and an HPLC column
(PRP X-100; Hamilton). The results obtained are
shown in Fig. 3. A 100 mg l1 standard solution,
including As(III), As(V), MMA, and DMA, was
analyzed by SPE-HG-ICP-AES (Fig. 3a). The DMA
peak is not shown, and the MMA peak overlaps with
that of As(III). This result indicates that short
columns are only capable of separating inorganic
species. If ICP-MS is used, the detection limit might
be reduced. Figure 3b shows the peaks for a 5 mg l1
standard solution, containing both As(III) and As(V),
analyzed by SPE-ICP-MS, while Fig. 3c shows the
results for a 5 mg l1 standard solution, which
included AsB, As(III), As(V), MMA, and DMA,
analyzed by HPLC-ICP-MS.
Consequently, when arsenic speciation has a
significant impact on researchers’ projects such as
toxicity or risk assessment of potentially As-contam-
inated waters, the HPLC-ICP-MS method can be
applied because of its capacity to separate more
arsenic species present in the solution and to detect to
low concentrations. However, since the HPLC col-
umn will take more time and more effort to optimize

123
Environ Geochem Health
Fig. 3 Spectra of standard solutions analyzed using different
procedures and various analytical conditions. (a) 100 mg l1
standard of As(III), As(V), MMA, and DMA analyzed by solid
phase extraction HG-ICP-AES. (b) 5 mg l1 standard of As(III)
and As(V) measured by SPE inductively coupled plasma mass
spectrometry (SPE-ICP-MS). (c) 5 mg l1 standard of AsB,
As(III), As(V), MMA, and DMA determined by high-
performance liquid chromatography ICP-MS
analytical conditions than the short column, the
procedure using the short column can be used
economically in samples containing only inorganic
arsenic species.
Comparison of preservatives in iron-poor water
Preservation methods are introduced to sampling
procedures to maintain both the amounts and species
of the elements of interest (i.e., arsenic in this study)
from undergoing chemical changes due to metal
oxyhydroxide precipitation, photochemical oxidation
or redox reactions. Several reagents have been tested
in the laboratory for the preservation of iron-rich
water samples containing arsenic (Daus et al. 2002;
Bednar et al. 2002).
Herein, EDTA, H3PO4 and HCl were tested as
preservatives for groundwater from the Ulsan mine
(Fig. 4a) and surface water (Fig. 4b), both of which
had a neutral pH and low iron contents. These
samples were analyzed using laboratory separation
procedures employing HPLC-ICP-MS after storage
periods of within 18, 72, 120, 192, and 360 hours.
The concentrations of other metals, such as Al, Mn,
Cd, Cr, Cu, Co, Ni, and Pb, were all below
0.1 mg l1, but the total arsenic concentrations were
73.3 mg l1 and 23.7 mg l1 respectively.
The results from the arsenic speciation analysis are
shown in Fig. 4. Concentrations of arsenic species are
converted into the compositional ratio in percentage
of total arsenic concentrations measured with ICP-
MS. Both samples have only inorganic arsenic
species, and As(V) revealed as the dominant arsenic
species, with a minor As(III) content. The dashed
lines show the results of the field separations, and the
four point symbols represent the As(III) (bottom) and
As(V) (top) concentrations in the samples treated
with each of the preservatives.
In the groundwater sample (Fig. 4a), As(III) is
present at a very low concentration (1.17 mg l1), and
thus it is difficult to detect via laboratory separation
using HPLC-ICP-MS analysis. In the surface water
sample, As(III) at a concentration of (4.77 mg l1)
appears to be relatively stable for up to 360 h
(15 days) of storage with no significant effects of the
preservatives on the concentration. On the contrary,
As(V) continuously decreases over storage time, in
all sub-samples treated with different methods and in
both groundwater and surface water. The decrease
rate of As(V) concentration appears to become stable
after 192 h (8 days). Regarding the efficiencies of the
individual preservatives, H3PO4 was proved to be the
best.
This result was very different with the previous
tests on the As(III)/As(V) equilibria. In the artificial
samples spiked with Fe2+ 2 mg l1 and As(III)
200 mg l1 without any preservatives, As(III) con-
centration was dramatically decreased: 190 mg l1 at
fresh, 69 mg l1 after 24 h, then < 1 mg l1 after
9 days (Daus et al. 2002). In reagent water initially
containing an equal distribution of As(III) and As(V)
at 10 mg l1, the distribution of the arsenic species in
samples without preservatives changes within 2 days
with or without light exposure: As(III) was reduced to
As(V) (Bednar et al. 2002).
Because As(III) concentration was maintained,
As(V) was inferred not to be transformed, but to be
removed from water. A possible assumption is an
adsorption of negative charged As(V) onto the
surface of the bottle or invisible settled particles. In
this case, As(III), having a zero charge, is not
123
 Environ Geochem Health

Fig. 4 The stability of arsenic species with different preser- 23.7 mg l1 and Fe 0.03 mg l1). Dashed lines were
vatives over time compared using laboratory separation concentrations of As(III) (bottom) and As(V) (top) using field
methods for (a) groundwater (pH 7.8, AsT 73.3 mg l1 and separation methods
Fe < 0.01 mg l1), and (b) surface water (pH 7.1, AsT

adsorbed. Therefore, neutral and iron-poor water

samples for the arsenic analysis, even if for total
arsenic concentrations, should be measured within a
short time. To reduce bias, it would be better if
neutral and iron-poor water were not treated if
samples could be analyzed within 3 days (±10%
error compared with field separation); however, if
samples could not be analyzed within 3 days,
preservatives become necessary. The efficiencies of
the individual preservatives were similar, but H3PO4
proved to be the best.

Arsenic speciation in water samples

Figure 5 shows the analytical results for the real

water samples collected from various environments,
using different sampling methods and with specific
preservation procedures, that were then analyzed by
various arsenic species separation methods. A
groundwater sample from the tailings dam of the
Guryong mine (Fig. 5a) was initially treated with
H3PO4 and then analyzed by SPE-ICP-MS. Arsenic
was mainly present in the As(III) form, with a
concentration of 1.96 mg l1. A similar groundwater
sample from the tailings dam at the Ulsan mine
(Fig. 5b) was also treated with H3PO4, but this
sample was analyzed by HPLC-ICP-MS. This sample
contained both As(III) and As(V) at concentrations of
11.1 and 10.3 mg l1 respectively. A sample from a Fig. 5 Spectra of (a) iron-rich groundwater from the Guryong
mine analyzed by SPE-ICP-MS, (b) neutral groundwater from
pond in the Ulsan area was not treated, but was the Ulsan mine, (c) surface water from the Ulsan pond, (d)
analyzed within 24 h of sampling, which showed surface water from the Ulsan creek. Samples (b), (c), and (d)
As(V) to be the dominant species (26.0 mg l1), with were analyzed by HPLC-ICP-MS

123
Environ Geochem Health
small amounts of AsB, As(III) (1.9 mg l1) and DMA
(3.4 mg l1; Fig. 5c). AsB is the predominant species
in marine organisms, but is rare in freshwater
(Dembitsky and Rezanka 2003). A creek water
sample from the Ulsan area (Fig. 5d) showed high
arsenic concentrations. Although the dominant ar-
senic species in most surface waters is As(V), the
As(III) concentration in this sample was 25.6 mg l1,
notably higher than the 11.0 mg l1 for As(V).
Construction involving the disturbance of surface
sediments was being undertaken at this creek during
sampling, and the suspended surface sediments could
have influenced the arsenic species.
Conclusion
In order to evaluate the toxicity, mobility, and risk of
arsenic in waters, accurate speciation is required. The
following conclusions can be drawn with regard to
the sampling and preservation required to enable
accurate analyses of arsenic species:
1. The conventional field sampling method using
0.45-mmfilters could over-estimate the dissolved
arsenic concentrations.
2. Various arsenic species show different responses
in de-ionized water analyzed with ICP-AES, and
different hydride generation efficiencies. There-
fore, samples containing various arsenic species
should be acidified or pre-reduced to As(III) prior
to analyzing total arsenic concentration.
3. For detailed assessment of arsenic speciation in
water samples, the HPLC-ICP-MS procedure
should be applied due to its capacity to separate
more arsenic species, which are present in the
solution, and to detect low concentrations. How-
ever, the procedure using SPE cartridges can be
used economically and conveniently for samples
containing only inorganic arsenic species.
4. Water samples with a neutral pH and low iron
concentration are stable for up to 3 days, but
require preservative treatment if they are to be
stored for longer periods. Phosphoric acid
(H3PO4) proved to have the greatest preservative
efficiency in this study.
5. Groundwater, especially that related to mine
activity, contained only inorganic arsenic spe-
cies. However, surface waters also showed small
amounts of organic arsenic species. Hence, the
SPE cartridge, used as a short column, is
appropriate for arsenic speciation in groundwater
samples, but an HPLC column is required to
speciate arsenic in surface water samples.
In conclusion, researchers should select their field
and laboratory procedures with caution and pay
attention to the procedures of sample collection,
preservation, and analysis as well as consideration of
site characteristics.
Acknowledgement This research is a part of the Ph.D.
dissertation work of Youn-Tae Kim, Yonsei University.
Financial support was provided by the research fund from
the Korea Basic Science Institute (N26051) and the Korea
Research Foundation (KRF-2005-C00081).
References
Akter, K. F., Chen, Z., Smith, L., Davey, D., & Naidu, R.
(2005). Speciation of arsenic in ground water samples: A
comparative study of CE-UV, HG-AAS and LC-ICP-MS.
Talanta, 68, 406–415.
Aposhian, H. V., Zakharyan, R. A., Avram, M. D., Kopplin, M.
J., & Wollenberg, M.L. (2003). Oxidation and detoxifi-
cation of trivalent arsenic species. Toxicology and Applied
Pharmacology, 193, 1–8.
Bang, S., Johnson, M. D., Korfiatis, G. P., & Meng, X. (2005).
Chemical reactions between arsenic and zero-valent iron
in water. Water Research, 39, 763–770.
Bednar, A. J., Garbarino, J. R., Ranville, J. F., & Wildeman, T.
R. (2002). Preserving the distribution of inorganic arsenic
species in groundwater and acid mine drainage samples.
Environmental Science and Technology, 36, 2213–2218.
Bednar, A. J., Garbarino, J. R., Burkhardt, M. R., Ranville, J.
F., & Wildeman, T.R. (2004). Field and laboratory arsenic
speciation methods and their application to natural-water
analysis. Water Research, 38, 355–364.
B’Hymer, C., & Caruso, J. A. (2004). Arsenic and its specia-
tion analysis using high-performance liquid chromatog-
raphy and inductively coupled plasma mass spectrometry.
Journal of Chromatography A, 1045, 1–13.
Choi, S.-G., & Youm, S.-J. (2000). Compositional variation of
arsenopyrite and fluid evolution at the Ulsan deposit,
Southeastern Korea: A low-sulfidation porphyry system.
The Canadian Mineralogist, 38, 567–583.
Daus, B., Mattusch, J., Wennrich, R., & Weiss, H. (2002).
Investigation on stability and preservation of arsenic
species in iron-rich water samples. Talanta, 58, 57–65.
Dembitsky, V. M., & Rezanka, T. (2003). Natural occurrence
of arseno compounds in plants, lichens, fungi, algal spe-
cies, and microorganisms. Plant Science, 165, 1177–1192.
Gallagher, P. A., Schwegel, C. A., Wei, X., & Creed, J. T.
(2001). Speciation and preservation of inorganic arsenic
in drinking water sources using EDTA with IC separation
and ICP-MS detection. Journal of Environmental Moni-
toring, 3, 371–376.
123

Hughes, D. F. (2002). Arsenic toxicity and potential mecha-
nisms of action. Toxicology Letters, 133, 1–16.
Hung, D. Q., Nekrassova, O., & Compton, R. G. (2004).
Analytical methods for inorganic arsenic in water: A re-
view. Talanta, 64, 269–277.
Kim, J. Y. (2004). Mineralogical and physico-chemical prop-
erties of the tailings in the Guryoung Mine Area (Master
Thesis, Yonsie University).
Ko, I., Chang, Y.-Y., Lee, C.-H., & Kim, K.-W. (2005).
Assessment of pilot-scale acid washing of soil contami-
nated with As, Zn and Ni using the BCR three-step
sequential extraction. Journal of Hazardous Materials A,
127, 1–13.
Maity, S., Chakravarty, S., Thakur, P., Gupta, K. K., Bhatta-
charjee, S., & Roy, B.C. (2004). Evaluation and stan-
dardization of a simple HG-AAS method for rapid
speciation of As(III) and As(V) in some contaminated
groundwater samples of West Bengal, India. Chemo-
sphere, 54, 1199–1206.
Mandal, B. K., & Suzuki, K. T. (2002). Arsenic round the
world: a review. Talanta, 58, 201–235.
Plant, J. A., Kinniburgh, D. G., Smedley, P. L., Fordyce, F. M.,
& Klinck, B. A. (2005). Arsenic and selenium. In B. S.
Lollar (Ed.), Environmental geochemistry (33 pp). Ox-
ford: Elsevier-Pergamon.
123
Environ Geochem Health
Raposo, J. C., Zuloaga, O., Sanz, J., Villanueva, U., Crea, P.,
Etxebarria, N., Olazabal, M. A., & Madariaga, J. M.
(2006). Analytical and thermodynamical approach to
understand the mobility/retention of arsenic species from
the river to the estuary. The Bilbao case study. Marine
Chemistry, 99, 42–51.
Sánchez-Rodas, D., Gómez-Ariza, J. L., Giráldez, I., Velasco,
A., & Morales, E. (2005). Arsenic speciation in river and
estuarine waters from southwest Spain. Science of the
Total Environment, 345, 207–217.
Thirunavukkarasu, O. S., Viraraghavan, T., Subramanian, K.
S., & Tanjore, S. (2002). Organic arsenic removal from
drinking water. Urban Water, 4, 425–421.
Thomas, D. J., Styblo, M., & Lin, S. (2001). The cellular
metabolism and systemic toxicity of arsenic. Toxicology
and Applied Pharmacology, 176, 127–144.
USEPA, IRIS, CASRN 7440-38-2, accessed in June 2006.
WHO (1993). Guidelines for drinking water quality, Vol. 1.
Geneva: World Health Organization.
Yu, C., Cai, Q., Guo, Z.-X., Yang, Z., & Khoo, S. B. (2003).
Inductively coupled plasma mass spectrometry study of
the retention behavior of arsenic species on various solid
phase extraction cartridges and its application in arsenic
speciation. Spectrochimica Acta Part B, 58, 1335–1349.