Microchemical Journal 108 (2013) 16

Contents lists available at SciVerse ScienceDirect

Microchemical Journal
journal homepage: www.elsevier.com/locate/microc

Determination of total arsenic, total inorganic arsenic and inorganic arsenic species in
rice and rice our by electrothermal atomic absorption spectrometry
Ioannis N. Pasias, Nikolaos S. Thomaidis , Efrosini A. Piperaki
Laboratory of Analytical Chemistry, Department of Chemistry, University of Athens, Panepistimiopolis Zografou, 15771 Athens, Greece

a r t i c l e i n f o a b s t r a c t

Article history: The determination of inorganic arsenic species in food has become very important during the last few years,
Received 8 September 2012 since As(III) and As(V) are considered carcinogenic and found at high concentrations in food products, such
Received in revised form 15 November 2012 as rice. The present work describes the development of three different methods for the determination of total
Accepted 16 November 2012
arsenic, total inorganic arsenic, and As(III)As(V) in rice and rice our food products, purchased from the
Available online 23 November 2012
local market by electrothermal atomic absorption spectrometry (ETAAS). The methods were based on the
Keywords:
use of different selective extraction procedures and the optimization of all crucial instrumental and method-
Total inorganic arsenic ological parameters. By these new preparation procedures for the determination of arsenic species, the loss
Arsenic species and transformation of analytes during the extraction and digestion steps was prevented. For the validation
ETAAS of each method, precision, accuracy, and selectivity have been assessed, as performance criteria. All devel-
Chemical modiers oped methods were accurate and precise. The calculated recoveries ranged between 92% and 105% and the
Selective extraction procedures (%) relative standard deviation values, under repeatability or reproducibility conditions, were lower than
Rice and rice our food products 15% for all different concentration levels tested. The validated methods were applied successfully for the de-
termination of total arsenic and total inorganic arsenic in a prociency test organized by the International
Measurement Evaluation Program (IMEP-107). The methods were also applied in rice and rice our samples
purchased from the Greek market and the obtained results indicated that in almost all samples total arsenic
and total inorganic arsenic were detected at ng/g levels.
2012 Elsevier B.V. All rights reserved.

1. Introduction
Arsenic is one of the most toxic trace elements and occurs in both
inorganic and organic forms, which are found in the environment
both from natural sources and from anthropogenic activities. As(III)
and As(V) were deemed group I carcinogen by the International
Agency for Research on Cancer (IARC) [1], and therefore the develop-
ment of a fully-validated method for the determination of both or-
ganic and inorganic arsenic species is necessary.
The World Health Organization (WHO) recommended 15 g of
total inorganic arsenic (t-inAs) per body weight as a provisional toler-
able weekly intake (PTWI). The main source of human exposure to As
is via the food chain. Drinking water, or eating seafood, cereals, and
algae based food products are among the commodities with the
highest levels of arsenic. However, the ratio of total inorganic arsenic
(t-inAs) to total arsenic (t-As) varies among the different kinds of
foodstuffs [25]. According to a recent report published by the Euro-
pean Food Safety Authority (EFSA) Panel on Contaminants in the Food
Chain [6], arsenic exposures from food and water across 19 European
countries, using lower bound and upper bound concentrations, are
estimated to be in the range 0.130.56 g/kg body weight (b.w.) per
Corresponding author. Tel.: +30 210 7274317; fax: +30 210 7274750.
E-mail address: ntho@chem.uoa.gr (N.S. Thomaidis).
day for average consumers. In the same report, it is noted that chil-
dren younger than three years of age are the most exposed to inor-
ganic arsenic (0.502.66 g/kg b.w. per day). Ofcial studies for the
estimated human exposure to inorganic arsenic in Greece are not
available. However, data on the average diet for rice food products
in Greece conducted by a European program called DAFNE (Data
Food Network) showed that citizens consumed about 16 g of those
products per person and per day, putting rice in the list as the top
consumed foods in Greece [7]. Recently, some studies had been
presented in literature concerning the determination of t-inAs in
rice products [811]. In general there are two main approaches for
the arsenic speciation analysis: (a) the separation of arsenic species
after chromatographic separation and detection with specic detec-
tors. These methods usually use different chromatographic tech-
niques, such as gas (GC) or liquid (LC) chromatography with on-line
detection techniques, such as mass spectrometry (MS), inductively
coupled plasma mass spectrometry (ICP-MS), atomic absorption
spectrometry (AAS), or atomic uorescence spectrometry (AFS) and
(b) the separation of arsenic species after the selective extraction(s)
based on the different chemical properties of each of the species.
These methods use mainly spectrometric techniques and chemical
methods for the separation of one species from the others [816].
Electrothermal atomic absorption spectrometry (ETAAS), hydride
generation atomic absorption spectrometry (HG-AAS), inductively

0026-265X/$ see front matter 2012 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.microc.2012.11.008
2 I.N. Pasias et al. / Microchemical Journal 108 (2013) 16
coupled plasmamass spectrometry (ICP-MS) and high performance
liquid chromatography coupled with inductively coupled plasma
mass spectrometry (HPLC-ICP-MS) are among the most used tech-
niques. However, the lack of a standard or fully-validated method
for the determination of t-inAs is of great concern. Therefore, the
Directorate General for Health and Consumers (DG SANCO) of the
European Commission (EC) recently requested that the European
Union Reference Laboratory for Heavy Metals in Feed and Food
(EU-RL-HM) evaluate the performance of European laboratories fo-
cused on the determination of total and inorganic As in rice with a
view to future discussions on the need for possible regulatory measures
[12]. The test item used in this exercise was rice purchased in a local su-
permarket which was provided by the University of Aberdeen and
coded as IMEP-107.
The present work describes initially the development and the val-
idation of two different methods for the determination of t-As, t-inAs
by ETAAS. For this purpose, different extraction or digestion proce-
dures and various (permanent and non-permanent) chemical modi-
ers were evaluated and optimized. Furthermore, for the rst time,
Pd-nanoparticles (Pd-NPs) and Pt-nanoparticles (Pt-NPs) were tested
as new potential chemical modiers for the determination of arsenic
in rice and rice our samples. Then, nally, the development of a third
fully-validated method for the determination of As(III) after the reac-
tion of As(III) with ammonium pyrrolidinedithiocarbamate (APDC)
and its extraction with an appropriate organic and inorganic solvent
was accomplished. This method was based on the classical extraction
method proposed for the determination of As(III) in water samples
[17] and was extended for the st time in food matrices. All three
methods were applied for the determination of t-As, t-inAs and
As(III)As(V) in a reference sample used for the prociency test
(IMEP-107) and in rice and rice our samples purchased from the
local market.
2. Materials and methods
2.1. Instrumentation
A Perkin Elmer SIMAA 6000 atomic absorption spectrometer
equipped with a transversely heated graphite atomizer (THGA) and
with longitudinal Zeeman-effect background correction and an
AS-72 autosampler was used. The THGA graphite furnace was
equipped with an integrated L'vov platform. The THGA temperature
program for all methods developed is given in Table 1. The system op-
eration was controlled using Perkin Elmer AA Winlab, version 2.5. A
Perkin Elmer arsenic electrodeless discharge lamp (EDL) at an operat-
ed wavelength of 193.7 nm and an instrument slit width 0.7 nm was
used. A Perkin Elmer EDL System was used to stabilize the lamp cur-
rent at 350 mA. A 20-L volume of the standard or sample solutions
was dispensed in the graphite tube with the AS-72 autosampler.
The rice and the rice our samples were pretreated with a MARS
X-Press (CEM Corporation, NC, USA) microwave oven.
Table 1
Temperature program, under the optimum conditions, for the graphite furnace for the
determination of t-As, t-inAs and As (III) in rice and rice our samples.
Step Temperature Ramp Hold Ar ow rate Read
(C) time (s) time (s) (mL min1)
Drying 1 110 1 15 250
Drying 2 130 15 25 250
Pyrolysis 1100ab 1000c 10a 30bc 20 250
Atomization 2000 0 5 0 ON
Cleaning 2450 1 3 250
a
t-As.
b
t-inAs.
c
As (III).
2.2. Materials
All chemicals were of analytical reagent grade. Ultra pure water
from Millipore Direct-Q-UV water purier (Millipore) was used. The
As(V) stock standard solution (1000 5 mg L 1) (H3AsO4 in
0.5 mol L 1 HNO3) was purchased from Merck. The As(III) stock stan-
dard solution 1000 (mg L1) was prepared by dissolving 0.1320 g of
diarsenic trioxide in 100 mL of 10 M sodium hydroxide, and diluting
to 100 mL with ultra pure water. An aliquot of these solutions was dilut-
ed with water to give the appropriate concentrations before use. To pre-
pare the modier solutions, 100.2 g L 1 of Pd(NO3)2 in 15% HNO3
(Merck), 1000 mg L 1 H2PtCl6 in HCl (2 mol L 1) (CertiPUR, Merck),
1% w/v of Mg(NO3)2 in 1% HNO3 (High-Purity Standards), 1000
5 g mL1 of ZrOCl22H2O in 0.5% HNO3 (High-Purity Standards),
1000 g mL1 of Ir in 2% HCl (High-Purity Standards) were used. The
palladium nanoparticles (Pd-NPs; 0.58 mg Pd + 1 mg MWNTs in 1 mL
(1% v/v) SDS) and platinum nanoparticles (Pt-NPs; 0.42 mg Pt+ 1 mg
CNHs in 1 mL (1% v/v) SDS) were a gift from the National Hellenic Re-
search Foundation [18]. Likewise, HNO3 SupraPUR 65% (Merck) and
HCl 36% (Merck) were used for the extraction or the digestion of the
samples. For the selective determination of t-inAs and As species, an
APDC (99%, Sigma-Aldrich) solution of 1% (w/v) and an EDTA
(99%, Sigma-Aldrich) solution of 5% (w/v) were prepared. For pH
adjusting, HCl 1 M and NaOH 4 M solutions were prepared. 4-Methyl-
2-pentanone spectrophotometric grade, (MIBK, 99%, Sigma-Aldrich)
was also used as an extraction solvent.
For quality assurance and accuracy estimation, the certied refer-
ence material IRMM 804, Rice our (European Commission Joint Re-
search Centre, Institute for Reference Materials and Measurements)
and the IMEP-107 test item for the prociency test were used.
2.3. Experimental procedures
Commercially available rice and rice our samples of various
brands and different varieties were obtained from local super markets
in Lamia, Greece (in the years 2011 and 2012) and these samples
were appropriately coded. The samples were milled and homoge-
nized in a food processor, with a stainless steel cutter, and sometimes
were also ground in a mortar to obtain a more representative
material.
2.3.1. Determination of t-As
0.5 0.1 g of the homogenized rice or rice our sample was
weighted in PFTE vessels and 5 mL HNO3 (65%) were added. The sam-
ples were digested with MARS X-Press (CEM Corporation, NC, USA)
microwave oven after the application of a preselected program
(First stage: Power = 1600 W; Ramp time (min) = 2; Hold Time
(min) = 0 min; Temp (C) = 165 and Second stage: Power =
1600 W; Ramp time (min) = 3; Hold Time (min) = 5 min; Temp
(C) = 175) and then diluted to a nal volume of 20 mL with ultra
pure water.
2.3.2. Determination of t-inAs
0.5 0.1 g of the homogenized rice or rice our sample was
weighted in 50 mL polystyrene centrifuge tubes and 5 mL HNO3
1 M were added. Then the samples were vortexed and ultrasonicated
for 15 min and centrifuged at 4000 rpm for 15 min. Finally, 15 mL
EDTA 0.1% (w/v) were added, vortexed and centrifuged at 4000 rpm
for other 15 min. The supernatant was analyzed by ETAAS.
2.3.3. Determination of s()s(V)
0.5 0.1 g of the homogenized rice or rice our sample was
weighted in PFTE vessels and 10 mL HNO3 1 M were added. The ex-
traction was performed on a microwave oven MARS X-Press (CEM
Corporation) with the following preselected program: First stage:
Power = 800 W; Ramp time (min) = 2; Hold Time (min) = 2 min;
 I.N. Pasias et al. / Microchemical Journal 108 (2013) 16 3

0.007
Temp (C) = 70 and Second stage: Power = 800 W; Ramp time
(min) = 1; Hold Time (min) = 5 min; Temp (C) = 85 and then the 0.006
samples were diluted to a nal volume of 15 mL with HNO3 1 M. At
0.005
an aliquot of 10 mL of the supernatant, 5 mL EDTA 5% (w/v) were

Peak Area
added and the pH of the resulting solution was adjusted to 4.84.9 0.004
with HCl 1 M and NaOH 4 M. Then, 2 mL of the APDC solution 1%
0.003
(w/v) were added. The mixture was diluted to 30 mL with ultra
pure water and the tube was vortexed for 23 min with 3 mL of an 0.002
organic solvent (MIBK) and centrifuged at 4000 rpm for 15 min. At
an aliquot of 2 mL of the supernatant, 2 mL HNO3 2% (v/v) were 0.001
added and the tube was vortexed for 12 min and centrifuged at 0
4000 rpm for 6 min. An appropriate volume of the bottom layer 0 2 4 6 8 10 12
was used for the determination of As(III) by ETAAS. As(V) was deter- %(w/v) EDTA
mined by the difference of t-inAs content minus the As(III) content.
All the crucial methodological and analytical parameters, con- Fig. 1. The effect of the EDTA concentration on the recovery of As (III).

cerning the extraction or digestion procedure (extraction time, ap-

propriate solvent, sample mass and pH value) and the instrumental
parameters (appropriate chemical modier, optimum pyrolysis and
atomization temperatures), were investigated and optimized, taking
into account both the sensitivity and the accuracy for control criteria.
Quantication was performed with calibration curves, constructed
from integrated absorbance measurements and prepared by spiking
20 L of different standard solutions of As, which were prepared fol-
lowing the same protocols as described above. Standard additions cal-
ibration curves were also performed for checking matrix interference.
The limit of detection, LOD (g L 1) for each method, was calculated
from the equation LOD = 3.3 SBL/b, where SBL was the standard devi-
ation of ten blank determinations and b the slope of each calibration
curve. Precision and accuracy were also estimated by spiking different
concentrations of the analyte and following the whole pre-treatment
procedure. For quality assurance and accuracy estimation, the certi-
ed reference material IRMM 804, Rice our (European Commission
Joint Research Centre, Institute for Reference Materials and Measure-
ments) with a certied reference value of 49 4 g t-As kg 1 and the
IMEP-107 test item for the prociency test with assigned values for
t-As and t-inAs equal to 172 4 g kg 1 and 107 14 g kg 1, re-
spectively, were also used.
3. Results and discussion
3.1. Optimization of the general extraction procedure
3.1.1. Determination of t-inAs
A preliminary study was conducted on the extraction of t-inAs in
the presence of different extraction media combined with the use of
EDTA as a surface cleaning reagent. In general, several extraction
acid mediums have been proposed in different studies. For example,
Matos Reyes et al. concluded that sonication of freeze-dried vegetable
samples with 1 mol L 1 H3PO4, at room temperature, was an efcient
toxic arsenic extraction procedure that did not modify the original ar-
senic species by hydride generation-atomic uorescence spectrome-
try [19]. On the other hand, Cava-Montesinos et al. demonstrated
that the use of a mixture of 10 mL NO3 3 and Triton XT 114
0.1% (v/v), combined with the use of an DTA solution (0.1% w/v)
was the most appropriate medium for the quantitative extraction of
the method showed both good repeatability (9.4%, n = 3) and accura-
cy [(101 7)%, n = 3]. Under these conditions, extraction time was
also optimized. Extraction time is among the most crucial parameters
investigated in the literature and is directly depended from the ma-
trix of the sample and the technique used. For example, Signes-Pastor
et al. concluded that by leaving the sample overnight in 1% (v/v)
HNO3 combined with the use of a microwave furnace program can
lead to an accurate determination of t-inAs [21]. However, in most
of the studies short extraction procedures have been proposed for
the accurate determination of t-inAs [20,22,23]. In the present
study, it was proved that after a 20 min sonication the extraction
had been accomplished.
3.1.2. Determination of As(III)
The same crucial parameters were also investigated for the speci-
ation of As(III). The use of a microwave program, as described in
Experimental procedures, for the determination of As(III) combined
with the use of 10 mL NO3 1 along with 5 mL EDTA 5% (w/v)
(Fig. 1) was considered as optimum for the accurate determination
of As(III) in a 0.5 g of sample mass. EDTA was used as a masking
agent due to the interferences caused by the metal ions which react
with APDC leading to a less quantitative extraction of As(III) [17]. In
this method, there are two other crucial parameters; pH and APDC
concentration. As previously referred, this method is based on the se-
lective extraction of As(III) in methyl isobutyl ketone (MIBK) and the
back extraction in 1% (v/v) NO3 after the complexation with APDC.
The back extraction step was considered as indispensable due to the
fact that the sensitivity was increased by a factor of 4. The use of a
1% (w/v) APDC solution was considered as optimum for the quantita-
tive reaction with As(III). Furthermore, Fig. 2 shows the effect of pH
value on the enhancement of the peak area of a standard solution
containing 10 g As(III) L 1 in NO3 1 after the complexation
0.02
0.018
0.016
0.014
Peak Area

inorganic As species using the same technique [20]. In the present 0.012
study, among the media tested, 5 mL of HNO3 1 M combined with 0.01
the use of 15 mL of an EDTA solution (0.1% w/v) was found as opti- 0.008
mum when taking into account both the sensitivity and the calculated 0.006
recovery [(100.9 7.1)%, n = 3] from a spiked sample as control 0.004
criteria. The use of H3PO4 1 M combined with EDTA (0.1% w/v) pro- 0.002
vided lower recoveries and worse precision compared with the mix- 0
ture HNO3 1 M - EDTA (0.1% w/v) [(74 14)% (n = 3)].
-0.002 0 2 4 6 8 10
The solvent-to-sample ratio was also optimized. It was found that, pH
for a xed volume of 5 mL HNO3 and 15 mL EDTA, the optimum sam-
ple mass, spiked with 500 g t-inAs kg 1, was equal to 0.5 g, where Fig. 2. The effect of the pH value on the recovery of As (III).
4 I.N. Pasias et al. / Microchemical Journal 108 (2013) 16
Fig. 3. The pyrolysis atomization curves in both aqueous and matrix solution for the
determination of As when 1.25 g Pd and 200 g20 g ZrIr were tested as chemical
modiers.
with 0.12 mmol APDC at pH values ranged from 1 to 9. The optimum
pH value was dened as 4.8. This conclusion complements the con-
clusion derived by Liang et al. who determined As(III) in water sam-
ples [24].
3.2. Optimization of the analytical and instrumental parameters
The determination of t-As, t-inAs and inorganic arsenic species
was accomplished after the thorough comparison of various chemical
modiers. Pd, Pt in either aqueous solutions or as Pd-nanoparticles
(Pd-NPs) and Pt-nanoparticles (Pt-NPs), Pd + Mg, and the permanent
modier ZrIr were investigated as potential chemical modiers. The
nal choice was based on the calculated characteristic mass (mo) of
standard and matrix solutions. The temperature program for the for-
mation of the coating has been described elsewhere [25].
The sensitivity achieved for As when 1.25 g Pd, as Pd(NO3)2 was
dispensed onto the THGA graphite furnace was proved to be the highest
among all the modiers tested, especially in matrix solutions. Pd-NPs
provided equal sensitivity compared with the use of Pd(NO3)2. The py-
rolysis and atomization temperatures did not appear to be signicantly
effected by the presence of Pd-NPs. Gunduz et al. also concluded that
the use of Au-NPs did not affect the optimum pyrolysis and atomization
temperatures for the determination of As [26]. The addition of Pt in ei-
ther aqueous solutions or as Pt-NPs provided low sensitivity. However,
the use of Pt-NPs resulted in the increase of the atomic absorption signal
( + 47%) compared with the use of Pt in aqueous solution (mo = 54 and
79 pg, respectively). The mixture PdMg was also tested as a potential
chemical modier. The addition of Mg did not affect signicantly the
peak area of As, but had a signicant increase on the background signal
(0.0616 versus 0.0107 when 1.25 g Pd was used).
Results produced with the use of 1.25 g Pd were compared with
those derived when 200 g Zr and 20 g Ir were used as permanent
chemical modier for the determination of t-inAs. The comparison
was based on the calculation of the characteristic mass in both aque-
ous and matrix solutions (Fig. 3). It was proved that the sensitivity
Table 2
Comparison of the calibration curves obtained by aqueous standard solutions, matrix matched solutions and standard additions for the determination of t-As, t-inAs and As(III).
Curves (n = 3, each)
t-As
3 3
Calibration with aqueous y = (2.0 0.04) 10 x 0.00016 (0.0014)
solutions
Standard additions y = (1.9 0.03) 103 x+ 0.0010 (0.0008)
Matrix matched solutions y = (1.9 0.04) 103 x+ 0.0012 (0.0004)
LOD (ng g1) 22.1
LOQ (ng g1) 66.3
achieved when 1 ng As in aqueous solution was dispensed in a THGA
modied with ZrIr was higher than that achieved with the use of
1.25 g Pd (mo = 41 and 49 pg, respectively). On the other hand, the
sensitivity achieved when 0.2 ng As in matrix solution was dispensed
in a THGA modied with ZrIr was lower than that achieved with the
use of 1.25 g Pd (mo = 87 and 53 pg, respectively).
Consequently, 1.25 g Pd was chosen as the optimum chemical
modier for all methods developed. The optimum instrumental con-
ditions are presented in Table 1.
3.3. Analytical gures of merit
Calibration curves for all developed methods were constructed,
following the procedures described in the experimental procedures
session, and applying the temperature program given in Table 1. In
order to investigate the presence of matrix interferences and exploit
the possibility of performing calibration with aqueous standards,
standard additions and matrix-matched calibration were performed.
The results are summarized in Table 2. It was observed that the
slope of the calibration curve from aqueous solutions is statistically
different by those of standard addition and matrix-matched curve
(checked by t-test for a condence level of 95% and at seven degrees
of freedom) only for the determination of t-inAs.
The limits of detection and quantication were calculated and
were found equal to 22.1 g kg 1 to 66.3 g kg 1, 30.1 g kg 1 to
90.3 g kg 1, and 19.0 g kg 1 to 57.0 g kg 1 for the determina-
tion of t-As, t-inAs, and As(III), respectively. These values are slightly
higher than other reported in literature and concerning the determi-
nation of the same analytes by more sensitive techniques, such as
ow injection hydride generation atomic absorption spectrometry
(FI-HG-AAS) [9], high-performance liquid chromatography coupled
to an inductively coupled plasma mass spectrometry (HPLC-ICP-MS)
[11], hydride generation-atomic uorescence spectrometry (HG-AFS)
[19,20,22]. However, in view of the content levels of t-As, t-inAs and
As(III) in rice and rice our samples, it can be concluded that the limits
of detection and quantication presented in this study are t for their
purpose.
Precision experiments were carried out and the intra-day and
inter-day precision values (n = 5), presented here as the mean (%) rel-
ative standard deviation (%RSD) of three different concentration levels,
were equal to 7.4% and 6.1%, 12% and 13%, and 9.5% and 13% for t-As,
t-inAs and As(III), respectively. The HORRATr and HORRATR values
achieved from different concentration levels, measured three times
under repeatability conditions and three times at two different days
under reproducibility conditions, were lower than the crucial value of
two, as described by the European Commission Regulation 333/2007
for the ofcial control of the levels of lead, cadmium, mercury, inorganic
tin, 3-MCPD and benzo(a)pyrene in foodstuffs [2729]. However, this
regulation may be also used when a limited number of fully validated
methods of analysis exist, and a tness-for-purpose approach may be
used to assess the suitability of the analytical method.
Recovery experiments were carried out in order to check the accu-
racy of the proposed method. The (%) calculated recoveries for t-As,
t-inAs, and As(III), expressed as the mean recovery value of three
Equation
t-inAs As(III)
y = (1.6 0.04) 10 x 0.00061 (0.0011) y = (1.58 0.03) 103 x + 0.0038 (0.0029)
y = (1.4 0.08) 103 x 0.0010 (0.00058) y = (1.51 0.03) 103 x + 0.0089 (0.0046)
y = (1.2 0.05) 103 x 0.0039 (0.0016) y = (1.50 0.03) 103 x + 0.0063 (0.0052)
30.1 19
90.3 57
 I.N. Pasias et al. / Microchemical Journal 108 (2013) 16 5

different spiked samples for each different concentration level, were Table 4
equal to 105.0 1.0, 105.0 3.0, and 92.1 5.6, respectively. Accura- Total arsenic, total inorganic arsenic and arsenic species content in rice and rice our
samples (n = 3).
cy was also tested by analyzing the standard reference material IRMM
804 (Rice our, European Commission Joint Research Centre, Institute Sample code t-As (g/kg) t-inAs (g/kg) As(III) (g/kg) As(V) (g/kg)
for Reference Materials and Measurements), with a certied refer- 1 189 30 53 16 28 11 25 11
ence value of 49 4 g t-As kg 1, and the IMEP-107 test item, with 2 160 11 113 11 80 12 33 12
assigned values for t-As and t-inAs equal to 172 4 g kg 1 and 3 155.1 7.7 111 4.3 92 10 19 10
4 164.3 8.4 86 33 65 13 21 13
107 14 g kg 1, respectively. The determined value for t-As,
5 172 28 86 20 48 11 38 11
when the standard reference material IRMM 804 was analyzed, was 6 110.2 7.4 80 15 54.3 8.5 26.3 8.5
equal to 46.3 5.5 g t-As kg 1 (n = 3). The respective values for 7 197 43 113 9.3 89.8 7.6 23.5 7.6
t-As and t-inAs when the IMEP-107 was analyzed were 167 8 270.8 5.8 130.0 7.6 93 11 37 11
14 g t-inAs kg 1 (n = 3), and 107 10 g t-inAs kg 1 (n = 3), 9 42 25 b30.1 b19 b30.1
10 88 10 b30.1 b19 b30.1
whereas the calculated z-scores were equal to 0.2 and 0.0, respective-
11 189 30 53 16 28 11 25 11
ly. It should be noted that one hundred and three laboratories from 35 12 151 3.7 138 16 70.7 9.1 67.8 9.1
countries registered to the exercise. Ninety-eight laboratories 13 261 15 147 5.8 71.0 9.5 76.2 9.5
reported results for total As, but only 32 participants reported results 14 201 19 132 17 97.1 5.5 34.3 5.5
15 162 15 100 23 63 18 37 18
for inorganic As and among them the z-score value was equal to 0.0
16 b22.1 b30.1 b19 b30.1
only when performing this specic developed method for the deter- 17 34.7 8.2 b30.1 b19 b30.1
mination of t-inAs. Moreover, the IMEP-107 was also analyzed for 18 170 14 107 10 91.2 9.3 15.8 9.3
the determination of As(III), even though there was no assigned
value for this determination. The calculated value was equal to
91.3 g As(III) kg 1. The result was in good agreement with the re- Comparison of rice samples analyzed in the present study showed
spective one found by Huang et al. [11], who also determined As(III) that the lowest mean content of total arsenic and total inorganic arse-
in IMEP-107. nic were found in rice from India. Williams et al. [8] also concluded
The selectivity of t-inAs and As(III) determinations was also tested that the white type, Basmati grain rice samples showed the lowest ar-
by spiking different forms of organic arsenic (DMA and arsenocholine) senic content among all the samples tested. On the contrary,
in rice and rice our samples. The calculated recoveries of organic arse- parboiled type rice sample showed the highest content of t-As and
nic were equal to zero (n = 3, for each species). Moreover, for the selec- t-inAs, followed by the yellow type rice sample. Parboiled rice is the
tivity of As(III) when the As(V) co-exists, a sample was spiked by both one that has been partially boiled in the husk and this procedure
As(III) and As(V) species, and the recovery of As(V) was also equal to might be responsible for this observation.
zero (n = 3). Quality management systems do not seem to have any effect on
the content of total and total inorganic arsenic. For example, samples
3.4. Determination of t-As, t-inAs and As(III)-As(V) in rice and rice our coded 1 to 6 are all white type and long grain rice samples produced
samples under different quality management systems and brands. Among
them, sample 6, which was produced by a local producer near the
The fully-validated methods were applied for the determination of city of Lamia, Greece, for local consumption only, showed the lowest
total and total inorganic arsenic species in rice (n = 15) and rice our content of total and total inorganic arsenic.
samples (n = 3) of different brands and varieties collected from local To assess the risk for consumers, the weekly intake for a 70-kg
super markets in the city of Lamia, Greece. Samples were produced adult was compared with the PTWI (15 g t-inAs/kg body weight/
under different quality management systems and some of them week) [6]. According to the data available through the European pro-
were collected straight from local producers without any industrial gram DAFNE (Data Food Network) based on the average diet per per-
pretreatment (from eld to fork conditions) (Table 3). The samples son, the daily average rice consumption in Greece is equal to 16 g [7].
were analyzed in triplicate and the results are summarized in Table 4. This means that to overcome the PTWI value a 70-kg adult must con-
The content of total arsenic ranged from 42 g/kg to 271 g/kg for the sume 100 kg of rice per week, but without taking in account the mean
rice samples (n = 15) and from b22.1 g/kg to 170 g/kg for the rice consumption of other foods which might contain large amounts of in-
our samples (n = 3). The proportion of total inorganic arsenic was organic arsenic. Children younger than three years old are the most
equal to (64 19)%, whereas the respective percentage of As(III) to exposed to inorganic arsenic. As shown in Table 4, there are samples
total inorganic arsenic was equal to (65 12)%. As(V) was deter- of rice our samples used for baby foods with detectable total inor-
mined by the difference of t-inAs content minus the As(III) content. ganic arsenic. Dietary exposure to inorganic arsenic for children
Similar results were also reported by other researchers [810,21,23]. under three years old is in general estimated to be about 2- to

Table 3
Types and characteristics of the rice and rice our samples analyzed.

Sample code Rice type Grain Region Mainly used Quality system

1 white Long Greece For pilaf and salads GAPa

25 white Long Greece For pilaf and salads First quality
6 white Long Greece For pilaf and salads Local producer, from eld to fork sample
7 white Medium Greece For risotto and salads GAPa
8 white Medium Greece For risotto and salads First quality
910 white Basmati long India First quality
11 Whole brown Long Greece For natural diet GAPa
12 White Long but with no gluten Greece First quality
13 parboiled Long Greece For pilaf GAPa and rst quality
14 yellow Long Greece For various dishes First quality
15 White Thai long Thailand For various dishes First quality
1618 Rice our Greece For baby foods
a
GAP: Good Agricultural Practice.
6 I.N. Pasias et al. / Microchemical Journal 108 (2013) 16
3-fold that of adults [6]. For this reason, a more appropriate estima-
tion of the PTWI value according to the more recent results reported
in the literature should be undertaken.
4. Conclusions
This study proposes three fully validated methods for the determi-
nation of total arsenic, total inorganic arsenic and inorganic arsenic spe-
cies by ETAAS. Different extraction or digestion procedures and various
chemical modiers (permanent and non-permanent) were evaluated
and optimized. A new method for the determination of As(III) after
the reaction of As(III) with ammonium pyrrolidinedithiocarbamate
(APDC) and its extraction with an appropriate organic and inorganic
solvent was developed for the rst time and validated for non-
aqueous matrices. The methods were also applied for the determination
of t-As and t-inAs in a test sample used for the prociency test IMEP-107
and the z-score values achieved proved that the methods were accurate
and precise. The methods were applied for the determination of t-As,
t-inAs and As(III)As(V) in rice and rice our samples purchased from
the Greek market due to the lack of data focusing on the content of
those analytes in rice and rice our coming from Greece. It was proved
that t-As and t-inAs were detected in almost all rice and rice our
samples. For adults, the average weekly intake is lower that the PTWI
value, but due to the fact that exposure of young children is in general
estimated to be about 2 to 3-fold that of adults, this value should be
reconsidered and focused on the more recent results reported in the
literature.
References
[1] International Agency for Research on Cancer, Some Drinking-Water Disinfectants
and Contaminants, Including Arsenic, in: IARC Monographs on the Evaluation of
Carcinogenic Risks to Humans, vol. 84, ARC (International Agency for Research
on Cancer), Lyon, France, 2004.
[2] S. Kapaj, H. Peterson, K. Liber, P. Bhattacharya, Human health effects from chronic
arsenic poisoning a review, J. Environ. Sci. Health. Part A 41 (2006) 23992428.
[3] Y.J. Zavala, J.M. Duxbury, Arsenic in rice: I. Estimating normal levels of total arse-
nic in rice grain, Environ. Sci. Technol. 42 (2008) 38563860.
[4] J.M. Duxbury, A.B. Mayer, J.G. Lauren, N. Hassan, Food chain aspects of arsenic
contamination in Bangladesh: effects on quality and productivity of rice, J. Envi-
ron. Sci. Health A 38 (2003) 6169.
[5] Y.-G. Zhu, P.N. Williams, A.A. Meharg, Exposure to inorganic arsenic from rice: a
global health issue? Environ. Pollut. 154 (2008) 169171.
[6] European Food Safety Authority (EFSA), Scientic opinion on arsenic in food
EFSA panel on contaminants in the food chain, EFSA J. 7 (2009) 13511355.
[7] DAFNE-ANEMOS, Data Food Networking, http://www.hhf-greece.gr/dafnesoftweb/
Main.aspx?type=multi, (last access 10-06-2012).
[8] P.N. Williams, A.H. Price, A. Raab, S.A. Hossain, J. Feldmann, A.A. Meharg, Variation
in arsenic speciation and concentration in paddy rice related to dietary exposure,
Environ. Sci. Technol. 39 (2005) 55315540.
[9] S. Torres-Escribano, M. Leal, D. Vlez, R. Montoro, Total and inorganic arsenic con-
centrations in rice sold in Spain, effect of cooking and risk assessments, Environ.
Sci. Technol. 42 (2008) 38673872.
[10] Y.J. Zavala, R. Gerads, H. Grleyk, J.M. Duxbury, Arsenic in rice: II. Arsenic speci-
ation in USA grain and implications for human health, Environ. Sci. Technol. 42
(2008) 38613866.
[11] J.-H. Huang, P. Fecher, G. Ilgen, K.-N. Hu, J. Yang, Speciation of arsenite and arse-
nate in rice grain verication of nitric acid based extraction method and mass
sample survey, Food Chem. 130 (2012) 453459.
[12] M.B. de la Calle, H. Emteborg, T.P.J. Linsinger, R. Montoro, J.J. Sloth, R. Rubio, M.J.
Baxter, J. Feldmann, P. Vermaercke, G. Raber, Does the determination of inorganic
arsenic in rice depend on the method? Trends Anal. Chem. 30 (2011) 641651.
[13] D.J. Butcher, Environmental applications of arsenic speciation using atomic spec-
trometry detection, Appl. Spectrosc. Rev. 42 (2007) 122.
[14] A.S. Stasinakis, N.S. Thomaidis, Fate and biotransformation of metal and metalloid
species in biological wastewater treatment processes, Crit. Rev. Environ. Sci.
Technol. 40 (2010) 307364.
[15] S.-H. Nam, H.-J. Oh, H.-S. Min, J.-H. Lee, A study on the extraction and quantitation
of total arsenic and arsenic species in seafood by HPLCICP-MS, Microchem. J. 95
(2010) 2024.
[16] J. Koh, Y. Kwon, Y.-N. Pak, Separation and sensitive determination of arsenic spe-
cies (As3+/As5+) using the yeast-immobilized column and hydride generation in
ICPAES, Microchem. J. 80 (2005) 195199.
[17] T. Kamada, Selective determination of arsenic (III) and arsenic (V) with ammoni-
um pyrrolidinedithiocarbamate, sodium diethyldithiocarbamate and dithizone
by means of ameless atomic-absorption spectrophotometry with a
carbon-tube atomizer, Talanta 23 (1976) 835839.
[18] N. Karousis, G.-E. Tsotsou, F. Evangelista, P. Rudolf, N. Ragoussis, N. Tagmatarchis,
Carbon nanotubes decorated with palladium nanoparticles: synthesis, character-
ization, and catalytic activity, J. Phys. Chem. C 112 (2008) 1346313469.
[19] M.N. Matos Reyes, M.L. Cervera, R.C. Campos, M. de la Guardia, Non-chromato-
graphic speciation of toxic arsenic in vegetables by hydride generation-atomic
uorescence spectrometry after ultrasound-assisted extraction, Talanta 75
(2008) 811816.
[20] P. Cavas-Montesinos, K. Nilles, M. Luisa Cervera, M. de la Guardia, Non-chromato-
graphic speciation of toxic arsenic in sh, Talanta 66 (2005) 895901.
[21] .J. Signes-Pastor, C. Deacon, R.O. Jenkins, P.I. Haris, A.A. Carbonell-Barrachina,
A.A. Meharg, Arsenic speciation in Japanese rice drinks and condiments, J. Envi-
ron. Monitor. 11 (2009) 19301934.
[22] . Gonzlves, A. LIorens, M.L. Cervera, S. Armenta, M. de la Guardia,
Non-chromatographic speciation of inorganic arsenic in mushrooms by hydride
generation atomic uorescence spectrometry, Food Chem. 115 (2009) 360364.
[23] N.M. Smith, R. Lee, D.T. Heitkemper, K.D. Cafferky, A. Haque, A.K. Henderson,
Inorganic arsenic in cooked rice and vegetables from Bangladeshi households,
Sci. Total. Environ. 370 (2006) 294301.
[24] P. Liang, L. Peng, P. Yan, Speciation of As(III) and As(V) in water samples by dis-
persive liquidliquid microextraction separation and determination by graphite
furnace atomic absorption spectrometry, Microchim. Acta 166 (2009) 4752.
[25] V.P. Kalantzis, N.I. Rousis, I.N. Pasias, N.S. Thomaidis, E.A. Piperaki, Evaluation of
different modiers for the determination of arsenic in leachate samples from san-
itary landlls by electrothermal atomic absorption spectrometry, Anal. Lett. 45
(2012) 592602.
[26] S. Gunduz, S. Akman, A. Baysal, M. Culha, The use of gold nanoparticles as an ef-
fective modier for the determination of arsenic and antimony by electrothermal
atomic absorption spectrometry, Microchim. Acta 172 (2011) 403407.
[27] W. Horwitz, L.R. Kamps, K.W. Boyer, Quality assurance in the analysis of foods and
trace constituents, J. Assoc. Off. Anal. Chem. 63 (1980) 13441354.
[28] M. Thompson, Recent trends in inter-laboratory precision at ppb and sub-ppb
concentrations in relation to tness for purpose criteria in prociency testing,
Analyst 125 (2000) 385386.
[29] Commission regulation (EC) 8 March 2007, Laying down the methods of sam-
pling and analysis for the ofcial control of the levels of lead, cadmium, mercury,
inorganic tin, 3-MCPD and benzo(a)pyrene in foodstuffs 333/2007(L 88/29).