Food Chemistry 113 (2009) 720726

Contents lists available at ScienceDirect

Food Chemistry
journal homepage: www.elsevier.com/locate/foodchem

Analytical Methods

Sequential extraction combined with HPLCICP-MS for As speciation in dry seafood products
Xuan Cao a, Chunli Hao b, Geng Wang c, Huanghao Yang d, Dengyun Chen e, Xiaoru Wang d,*
a

College of Chemistry and Chemical Engineering, Ocean University of China, Qingdao 266003, PR China Fuzhou University of China, Fuzhou 350000, PR China c Shandong University of China, Jinan 250000, PR China d The First Institute of Oceanographic Institute, State Oceanographic Administration, Qingdao 266061, PR China e Agilent Technologies, USA
b

a r t i c l e

i n f o

a b s t r a c t
A sequential extraction procedure followed by HPLCICP-MS analysis was developed for the speciation and quantization of arsenic species in dry seafood products (DSPs). The extraction process involved three major steps, which produced respectively three As fractions: non-polar (Asnonpolar), polar (Aspolar) and inorganic arsenic species (Asinorganic). The extraction efciency (EE%) is in the range of 87115%. Asnonpolar is 0.610.2% of total arsenic in DSPs, which is served as the index of level of arsenolipid. For Aspolar and Asinorganic, a hyphenated HPLCICP-MS technique was established for the separation and quantication of six arsenic species including arsenobetaine (AsB), arsenocholine (AsC), arsenite (AsIII), arsenate (AsV), monomethylarsonic acid (MMA) and dimethylarsinic acid (DMA). The results indicate that AsB and AsV are the two dominate species in DSPs, while all other species are present in relatively low concentrations. The recovery efciency of 7702.8% could be obtained with spike recovery test in this two-steps extraction. 2008 Elsevier Ltd. All rights reserved.

Article history: Received 19 December 2007 Received in revised form 11 June 2008 Accepted 3 August 2008

Keywords: HPLC ICP-MS Sequential extraction Arsenic Speciation Dry seafood products

1. Introduction Dry seafood products (DSPs) are very popular in China and some Asian countries. The production of DSPs in the region is huge, and their manufacturing involves a widely diverse processes including braise, baking, drying, avoring, etc. Just as fresh seafood, DSPs are also an important source of proteins, polyunsaturated fatty acids, and a wide range of vitamins (B, D, and A) and minerals (calcium, phosphorus, iron, etc.) in this part of the world. (FAO., 1998). In fact, compared with the fresh ones, DSPs are more convenient to store and cook, and are often sold commercially in instant food formulations. Both DSPs and the fresh seafood are sometimes contaminated with trace toxic elements such as arsenic. (Sikorski & Kolakowska, 1994). For safety evaluation, the speciation of these trace elements is just as important as the determination of their total concentration levels since toxicity is critically dependent on their exact chemical forms (Andreae, 1986; Penrose, 1974). For example, the LD50 values (50% lethal dose; mg/kg) of different arsenic species vary widely as: arsine 3, arsenite 14, arsenate 20, monomethylarsonic acid 7001800, dimethylarsinic acid 700 2600, arsenocholine 6500 arsenolipids >8000, and arsenobetaine >10,000 (Nriagu, 1994).
* Corresponding author. Fax: +86 532 88963253. E-mail address: cxfpy@hotmail.com (X. Wang). 0308-8146/$ - see front matter 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.foodchem.2008.08.001

Unfortunately, same as their fresh seafood counterparts, in most countries, there is yet no well-dened, species-specic, legislative control of DSPs (Buchet & Pauwels, 1994; Francesconi & Edmonds, 1994; Phillips, 1994). One of the best cited As standards for arsenic is for drinking water and its value in China is 50 mg/l. The setting of this limit is largely based on toxicity data for inorganic arsenite and arsenate. If this same limit were to applied to DSPs, most of them would probably be banned from the market since the arsenic content in seafood typically exceeds that of the drinking water by a factor of 1000 (Cullen & Reimer, 1989; National Academy of Sciences, 1997). In these cases, it is obvious that the setting of limits based solely on total arsenic concentrations is scientically unfounded, and instead, the identication and quantication of individual arsenic species in the sample would be a more reasonable approach. In this regard, the absence of effective analytical method has been the main obstacles. The subject of arsenic speciation analysis has been extensively reported in the literatures but those related to applications in DSPs are very scarce (Gomez-Ariza & Sanchez-Rodas, 2000; Larsen & Pritzi, 1993; Larsen & Quetel, 1997). In the present work, a sequential solvent extraction procedure has been developed for the fractionation of different arsenic species. These fractions were then analyzed by ICP-MS or hyphenated HPLCICP-MS for total As and arsenic speciation analysis, respectively.

X. Cao et al. / Food Chemistry 113 (2009) 720726

721

2. Experimental 2.1. Standards and reagents Stock solutions of individual arsenic compounds were prepared using the following standard: arsenite triiodide (AsIII), and arsenate oxide (AsV) from Alfa Aesar (Ward Hill, MA), and dimethylarsinic acid (DMA) from Acros Organics (NJ, USA). Arsenobetaine (AsB), arsenocholine (AsC), and monomethylarsinic acid (MMA) were prepared by Tsinghua University (Beijing, China). All the stock solutions were stored in a dark room at 4 C, and the mixed standard solutions containing all arsenic species was prepared from the above single standard solutions. Milli-Q ultra pure deionized (DI) water (Millipore, USA) was used throughout the experiment. The solvents acetone, methanol and hydrochloric acid, used for extraction were of high purity grade purchased from Merck (Darmstadt, Germany), Strong oxidants included 65% nitric acid (Merck) and 30% hydrogen peroxide (Merck) were used for DSP samples digestion. HPLC grade ammonium carbonate from Fischer Scientic (Fairlawn, NJ, USA), was used to prepare the mobile phase used for the chromatographic separations. 2.2. Instrumentation A lyophilizer (Labconco, USA) was used to freeze dry the samples to a constant weight. A variable speed reciprocal shaker (Apparatyus, China) was used for sample homogenization. A rotary evaporator (EYELA, Japan) was used to evaporate the organic solvents and water used in the extraction process. Centrifugation (HERMLE, Germany) was used to separate the extract. The Accelerated solvent extraction (ASE) instrumentation used for sample extraction was an ASE 100 from Dionex Corporation (Sunnyvale, CA, USA). An Agilent 1100 series high performance liquid chromatography (HPLC) system consisting of a binary pump, a vacuum degasser, and a manual-injector with a 20 ll sampling loop was used in HPLC analysis. A PRP-X100 anion exchange column (Hamilton, Reno, NV, USA) was used for the separation of arsenic species. The Arsenic determination was performed by an Agilent 7500a ICP-MS (Agilent Technologies, USA), equipped with a Babington nebulizer, a glass double path spray chamber and a standard quartz torch, and HPLC operating conditions are listed in Table 1. 2.3. Sample and sample treatment Reference material TORT-2 (lobster hepatopancreas) was purchased from National Research Council of Canada (Ottawa, Ontario, Canada). Eight of the more popular dried seafood products in China were selected for real samples in this work. The samples were collected in April 2006 from some large supermarkets in Qingdao. Detail information of the samples was summarized in Table 2. All samples were rinsed three times with deionized water to remove the salt on the surface, and then frozen dry a constant weight. They were grinded into powders and then (particle size: <2 mm) stored in the dark at 70 C before analysis. 3. Determination of total Arsenic Dried seafood (0.2 g) and 5 ml of concentrated HNO3 were added together into a PTFE bomb. After 10 min the bombs were sealed off and put into an oven at 80 C for 4 h. Afterwards the bombs were cooled down and the gas was released to discharge the high pressure in the bombs. The bombs were then placed back in the oven at 170 C for 4 h. After cooling down the bomb, 1 ml H2O2 was added onto the residue inside the bomb to form a clear

Table 1 Instrumental parameters for total As determination and As speciation analysis Total arsenic ICP-MS RF power RF matching Carrier gas ow rate Peristaltic pump ow rate Arsenic speciation HPLC Analytical column Flow rate Injection volume Mobile phase A Mobile phase B Agilent 1100 Series Hamilton PRP-X100 (250 4.1 mm; 10 lm) 1.5 mL/min 25 lL H2O 50 mM (NH4)2CO3 (pH = 9.5) Time (min) A (%) B (%) 0 100 0 15 0 100 30 100 0 Agilent 7500a 1350 W 1.6 V 1.10 L/min 0.5 rps 75 As, 72Ge, 35Cl, 77Se and 83Kr Agilent 7500a 1350 W 1.6 V 1.10 L/min 0.1 rps

Gradient program ICP-MS RF power RF matching Carrier gas ow rate Peristaltic pump ow rate Monitored signals

solution. This solution was diluted to a nal mass of 25 g with deionized water for ICP-MS analysis. The total arsenic content in the dry seafood samples was measured before and after extraction using inductively coupled plasma mass spectrometry (ICP-MS). The rare earth element 89Y was used as an internal standard. The interference correction equations were chosen from the library for EPA200.8 to correct 75ArCl+ interference. 4. Sequential extraction procedures 4.1. Step I: non-polar arsenic compounds extraction The DSPs were rst extracted by acetone to extract the relatively non-polar arsenic compounds. In the procedure, 0.2 g of powdered samples were accurately weighed into 10 ml centrifugal tubes followed by the addition of 4 ml of acetone. After closing the lids securely, the sample powders and acetone solvent were mixed thoroughly by reciprocal shaker for 5 min. The samples were then sonicated for 20 min, and the resulting mixtures were centrifuged at the rate of 5000 rpm for 10 min at 4 C. The bottom residue was then re-dispersed in the extractant solvent by shaking, and sonicated for another 10 min. The supernatant and the residue were again separated by centrifugation, and this samples extraction procedure was repeated for three times. The three supernatants were combined and ltered through a 0.22 lm membrane lter. The solvent in the ltrate was removed by rotary evaporation, and the nal concentrate was digested by the procedure as described in EPA method 200.3 (Wahlen & McSheehy, 2004). The arsenic concentration in this fraction, denoted as Asnonpolar, was determined by ICP-MS. (McKiernan & Creed, 1999). 4.2. Step II: extraction of polar arsenic compounds The solid residue from step I was dried in an oven at 50 C to drive away the residue acetone. It was then extracted by 4 ml of a mixture of 50% methanol in deionized water (DI). The extraction procedure was the same as those described in step I. After extraction, the solutions were ltered and then placed in an oven at 65 C to remove the solvent mixture. The drying process proceeded slowly to a nal volume of about 35 ml. During the process DI

722

X. Cao et al. / Food Chemistry 113 (2009) 720726

Table 2 Total arsenic concentration in samples by acid digestion and sequential extractions (mg/kg dry weight, n = 3) Samples Source Total arsenica,b Sequential extraction Asnonpolar Sardine Whitebait Squid1 Squid2 Oyster Clam Brown algae Shrimp1 Shrimp2
a b c d

Total extractable arsenic Aspolar 1.9 0.1 4.9 0.5 1.4 0.3 1.1 0.2 6.7 1.1 1.5 0.5 14 1.3 35.3 0.9 6.4 0.3 EE% 87.5 68.3 97.2 81.9 57.0 74.6 74.8 67.6 100.9 Asinorganic 0.05 0.02 2.1 0.4 0.02 0.01 0.03 0.02 5.8 0.7 0.26 0.03 2.7 0.7 9.6 1.2 0.7 0.1 EE% 2.3 29.3 1.4 2.4 49.3 13.2 14.6 18.4 10.4 Asresidue 0.05 0.02 ND 0.01 0.01 ND ND ND 0.04 0.02 0.02 0.01 ND 2.0 7.7 1.42 1.10 13.36 1.91 17.05 45.27 7.30

Total EE%c

EE%

Yellow Yellow Yellow Yellow Yellow Yellow Yellow Bo sea Yellow

sea sea sea sea sea sea sea sea

2.16 0.53 7.17 0.44 1.42 0.28 1.27 0.23 11.7 1.0 1.97 0.01 18.5 1.4 52.3 3.3 6.33 0.99

0.1 0.2 0.7 0.2 NDd 0.03 0.01 0.9 0.1 0.17 0.04 0.5 0.1 0.32 0.05 0.25 0.07

4.6 10.2 2.4 7.9 8.6 2.8 0.6 3.9

93 107 100 87 114 97 92 87 115

The precision is expressed as one standard deviation. Total arsenic concentration were considered as acid digestion arsenic. The extraction efciency (EE%) is dened as the percentage of total acid digestion arsenic. ND represents not detected.

was slowly added to solvent exchange with methanol. The nal concentrate was then diluted and made up to the volume of 25 ml with the addition of DI. The fraction was denoted as (Aspolar). 4.3. Step III: extraction of inorganic arsenic compounds The solid residue from the step II was dried in oven at 65 C to drive away methanol, then 0.05 mM hydrochloric acid (HCl) was added. The extraction procedure was the same as those of step I, while the extraction solutions were concentrated the procedure as described in step II. The nal solutions from this step were diluted and made up to the volume of 25 ml by DI. The fraction was denoted as Asinorganic. 5. Quality assurance and quality control (QA/QC) The sample extraction and analytical procedures were accompanied by the running of blanks and spiked standards. All samples were analyzed in three independent replicates to ascertain the reproducibility of the analyses. The accuracy of arsenic determination was examined by running Certied Reference Material (CRM), TORT-2 (NRCCNRC, Canada). The arsenic species in TORT-2 were extracted following the method described in Raimund Wahlens reference (Wahlen & McSheehy, 2004), using the separation procedure listed in Table 1. The measured total arsenic in the CRM material is 21.5 1.6 lg/g (n = 3), which agrees well with the certied value of 21.6 1.8 lg/g. Detailed results of As concentrations in individual fractions are listed on the top of Table 3, in which results from this work are compared with those reported by Raimund Wahlen. The two sets of data agree very well, and show the absence of signicant difference between the two sets of data, suggesting the validity of the method developed here. The total arsenic in CRM (TORT-2) was determined by both the sealed bomb and the microwave assistant digestion methods. The latter method followed the procedure reported by Li and wei (2003). The concentrations of arsenic in CRM by the microwave assisted digestion ranged from 90% to 100% of the certied results, and no signicant difference was found between the two methods. Sealed bomb digestion was chosen nally as it could digest more samples at the same time.

2003), the total arsenic concentration (dry weight basis) in DSPs is much higher. Furthermore, marked variation in total arsenic concentrations was observed among the different species of DSPs investigated in this report. Samples including shrimp1, oyster and brown algae containing more arsenicals than those of others and algae, as a kind of arsenic concentrator, indubitably contains the high amount of arsenicals. Interestingly, the concentrations in shrimp1 and oyster are higher or chose to algae. Thus, detailed Arsenic speciation analysis is deserved. 6.2. Optimization of arsenic separation The method of Judith A. Brisbin (Brisbin & Hymer, 2002) was slightly modied to make it adoptable for our experiment. Anion exchange chromatography was used for the separation of the six arsenic species (AsIII, AsV, DMA, MMA, AsB, AsC). Deionized water (A) and 50 mM (NH4)2CO3 (B) were used as the mobile phase. Fig. 1a compares the ion chromatograms observed under different gradient elution conditions. the ow rate and the gradient elution time were optimized, and baseline separations of individual peaks were achieved. The optimization process targeted ve of the arsenical peaks except AsV, which, with its long retention time (15 min), is readily separated from the rests. According to the conditions in Brisbins method, AsIII, AsV, DMA, MMA can be separated completely in an anion exchange column but not AsB and AsC. The two cationic species AsB and AsC co-elute together on an anion column without separation. To solve this problem, gradient elution with low initial concentration of mobile phase coupled with high elution rate were used. Fig. 1a compares the different gradients used in this experiment and their separation efciencies. It is obvious that the best separation was achieved when the ow rate was 1.8 ml/min, and with a linear gradient elution from 100%A to 100%B in 18 min. The results indicated that low (NH4)2CO3 concentration and high elution rate enhance the separation of AsB and AsC. The benecial effects of high ow rate, however, is offset by operational problems in ICP-MS, where high solvent loading resulted in plasma instability. As a compromise, a moderate ow rate (1.5 ml/min) and gradient elution conditions (015 min linear gradient from 100%A to 100%B) were chosen, and the separation achieved under the conditions is illustrated in Fig. 1b. 6.3. Sequential extraction of DSPs A sequential extraction procedure was developed to fractionate the arsenic compounds in DSPs. To optimize the extraction process, the rst step was to remove the fats from the samples. The two aims of this step are: (1) to remove the fats along with other biological matrix interferences before chromatography (McKiernan & Creed, 1999) and (2) to nd out the amount of non-polar

6. Results and discussion 6.1. Total arsenic concentration in DSPs The total arsenic concentrations found in the DSP samples are given in Table 2. Compared with the fresh seafood (Li & Wei,

X. Cao et al. / Food Chemistry 113 (2009) 720726 Table 3 Speciation of arsenic compound in sample of CRMs, step II and step El (mg/kg dry weight, n = 3) Sample CRM TORT-2b TORT-2 Sample Step II Sardine Whitebait Squid1 Squid2 Oyster Clam Brown algae Shrimp1 Shrimp2 Step IIIg Sardine Whitebait Squid1 Squid2 Oyster Clam Brown algae Shrimp1 Shrimp2
a b c d e f g f

723

AsC

AsB

AsIII

DMA

MMA

As V

Total extractable arsenica 21.6 1.8 21.5 1.6 Totald 1.8 4.2 1.0 0.7 5.7 1.4 11.9 34.3 6.1 0.04 1.8 0.01 0.02 5.3 0.2 2.3 6.6 0.61 Astotale 1.9 0.1 4.9 0.5 1.4 0.3 1.1 0.2 6.7 1.1 1.5 0.5 14 1.3 35.3 0.9 6.4 0.3 0.05 0.02 2.1 0.4 0.02 0.01 0.030.02 5.8 0.7 0.26 0.03 2.7 0.7 9.61.2 0.7 0.1

0.024 0.010 ND AsC 0.17 0.01 0.16 0.07 ND ND ND 0.17 0.01 ND 0.61 0.03 0.17 0.01 ND ND ND ND ND ND ND ND ND PE%c 8.5 2.1 8.9 1.3 2.3

14.25 1.08 14.0 1.1 AsB 1.2 0.4 3.0 0.9 0.57 0.04 0.39 0.08 3.6 1.8 0.28 0.04 11.2 1.7 27.8 1.3 5.4 1.5 PE% 60.0 39.0 40.1 35.5 26.9 14.7 65.7 61.4 74.0 ND 0.25 0.01 ND 0.01 0.01 2.7 0.5 ND 0.35 0.03 1.5 0.7 0.01 0.01

0.0928 0.0371 ND AsIII ND 0.2 0.1 ND ND ND ND ND ND ND 3.2 0.9 20.2 2.1 3.3 0.1 PE% 2.6 ND ND ND ND 0.7 0.1 5.2 ND 0.3 0.1 1.8 ND ND

0.84 0.10 0.8 0.2 DMA ND ND ND ND 1.01 0.03 0.1 0.1 0.45 0.04 0.20 0.04 0.14 0.06 ND ND ND ND ND ND ND ND ND PE% 7.6 5.2 2.6 0.4 1.9

0.093 0.069 0.09 0.02 MMA ND 0.2 0.1 ND ND 0.31 0.05 0.6 0.3 ND 0.41 0.03 0.2 0.1 ND ND ND ND ND ND 0.058 0.005 0.06 0.01 ND PE% 2.6 2.3 31.4 0.9 2.7 0.3 0.1

0.0928 0.0371 0.09 0.01 AsV 0.4 0.1 0.6 0.1 0.42 0.05 0.3 0.1 0.9 0.2 1.0 0.1 0.2 0.1 5.3 1.4 0.12 0.09 0.04 0.03 1.5 0.3 0.01 0.01 0.01 0.01 1.9 0.3 0.2 0.1 1.58 0.03 5.05 0.01 0.6 0.1 PE% 20.0 7.8 29.6 27.3 6.7 52.4 1.2 11.7 1.6 2.0 19.5 0.7 0.9 14.2 10.5 9.3 11.2 8.2

Total extractable arsenic determined by ICP-MS with standard addition calibration in aqueous extracts (sonication). These were reference value given by Raimund Wahlen (14). PE% is dened as the percentage of total extractable arsenic. Sum of arsenic speciation. The direct determination of arsenic concentration in extraction solution of step II and step III using ICP-MS. Extracted by 1:1 MEOH/water. Extracted by 0.05 M HCl.

arsenicals reside in this extract. This portion of the arsenicals is often neglected by other researchers because of their low concentrations. In this experiment, the efciencies of acetone (McKiernan & Creed, 1999) and N-hexane (Schmeisser & Goessler, 2005) were rst evaluated and compared. Table 4 compares the amounts of arsenicals extracted by the two solvents used alone or in sequence. Acetone gave the best performance and was therefore chosen as the solvent in further experiments. As for the water-soluble arsenicals, methanol or its water solution was the best choice because of its high extraction efciency (EE%), universal solvency power and water miscibility. The MeOH, however, has low EE% towards inorganic arsenicals species. Thus, in the nal step, the samples were further extracted by dilute hydrochloric (0.05 mM HCl) acid because of the its reported high EE% on inorganic arsenicals extraction (Mir & Rutter, 2007). The extraction yield and efciency (EE%) of different extractants towards samples TORT-2 and shimp1 are shown in Table 4. 1:1(V:V) MeOH/water and 0.05 mM hydrochloric (HCl) acid were nally xed as the extractants in threesteps, respectively. The oyster, brown algae and whitebait samples in general contain high amounts of Asnonpolar, (Table 2). The highest value occurred in oyster (0.9 0.1 mg/kg), while brown algae (0.5 0.1 mg/kg) also contains high concentration of Asnonpolar. It has been reported that oyster and algae contains high concentration of arsenolipid (Schmeisser & Goessler, 2005), which can be extracted by acetone. Therefore, the Asnonpolar can serve as the index of the level of arsenolipid in DSPs. In our DSP samples, acetone extraction accounted for 0.610.2% of the total arsenicals, and the value for fresh sh falls in the same range as the DSPs. (McKiernan & Creed, 1999), suggesting the absence of apparent differences between fresh seafood and DSPs. For the extractable arsenicals by polar solvent 1:1 MeOH/Water, the fractional yield of arsenicals ranged from

57100.9%. The results of extraction efciency indicate the 1:1 MeOH/Water are suitable for DSPs extraction except for Oyster (57.0%). As a method for food safety determination, 57% extraction efciency would be unacceptable. So dilute hydrochloric (HCl) acid is needed to enhance EE%. The combined yield after these threestep extraction range from 87 to 115%, indicating good recovery and material balance of the sequential extraction procedure total EE% was obtained. The extraction efciency, which is above 100%, may result from the incompletely evaporation of MeOH in this sample. It is reported that carbon-containing compounds, such as MeOH, would enhance the signals in ICP-MS because of the carbon effect (Allain & Jaunault, 1991). The arsenic in the residues of extraction were also determined by acid digestion, and relatively low arsenic concentration was observed. The results indicate that this sequential extraction procedure could extract quantitatively the arsenicals in DSPs. 6.4. Arsenic speciation in DSPs Speciation analysis of arsenicals was performed on fractions obtained from step II (Aspolar) and step III (As inorganic) by hyphenated HPLCICP-MS technique (Table 3). AsB and AsV were detected in all DSP samples studied, and they represented the most abundant arsenic species among total extractable arsenicals in the two fractions. In contrast to those of other seafood, AsB (arsenobetaine) represents the vast majority of total arsenic in marine organisms. It has been proposed that AsB represents the end point of the arsenic cycle in the marine ecosystem (Borak & Hosgood, 2007), because AsB is not transformed further in humans and other mammals. AsB could be excreted essentially unchanged, it is considered as a harmless arsenical (Cannon & Edmonds, 1981). Thus, even though the total arsenic concentration in seafood is high,

724

X. Cao et al. / Food Chemistry 113 (2009) 720726

a
Aboudance (cps)

48000 40000 32000 24000 16000 8000 0 0 100 200 300 Time (sec) 400 500 600 1 2 5 (I) (II) (III) 3 4

b
Abundance (cps)

3500 3000 2500 2000 1500 1000 500 0 0

AsB

AsC

DMA MMA As(III) As(V)

100

200

300

400

500

600

700

800

900

Time (sec)

c
Abandance (cps)

2400 2000 1600 1200 800 400 0 0 1

Agilent 7500a Agilent 7500ce

100

200

300

400 Time (sec)

500

600

700

800

Fig. 1. Gradient anion exchange ICP-MS chromatograms of As III, DMA, MMA, AsB and AsC.

the overall toxicity is low since AsB accounts for majority of the As species in these samples. The highest AsB concentration was found in the sample of shimp1 (step II + step III): 29.3 mg/kg (64.7% of total extractable arsenic), followed by brown algae: 11.6 mg/kg (67.8% of total extractable arsenic). arsenobetaine also represented the most abundant arsenicals in the samples. Interestingly, another dominating arsenicals found in DSPs was AsV, whose amount exceeds 50% of total extractable arsenic in clam. There is little report on this phenomenon, and inorganic arsenic was customarily considered as a kind of trace arsenicals in marine animals compared with organic arsenic. To eliminate the possibility of 75ArCl+ interference for arsenic in ICP-MS detection, an advanced type of ICP-MS (Agilent7000ce) with a reaction/collision cell was used to eliminate polyatomic interference. For the same sample, there is

no obvious change between them except for some differences on peak area (Fig. 1c), which were result from 75ArCl+ interference. Since DSPs was made from fresh seafood, we hypothesized that some arsenicals (AsIII or organic arsenicals) were transformed to inorganic forms during these process (braise, baking, drying, avoring, etc). Further research would be needed to substantiate this claim. Other arsenicals (DMA, MMA, AsIII, AsC) were also found in some DSPs. However, compared with AsB and AsVI, they were not the main component in amount. We did not take into account the species of arsenosugar in DSPs because of standard shortage. This could explain the small differences observed between the sum of six species of arsenicals and the total extractable arsenic in each extraction step. Recovery experiments were conducted using shimp2. After step I extraction, mixed spike arsenicals were added

X. Cao et al. / Food Chemistry 113 (2009) 720726 Table 4 Total arsenic in samples by sonication in different extraction solution(mg/kg dry weight, n = 3) Sample Asnonpolarb Acetone Shrimp1 EE%a TORT-2 EE%
a b c d

725

N-hexane acetonned 100% MeH N-hexane ND 0.03 0.04 AcetoneN-hexanec ND ND ND 0.7 0.1 0.9 0.1 64 12.28 0.01 57

1:1 MeOH/Water 1:2 MeOH/Water Water

0.1 M HCl

0.05 M HCl

0.02 M HCl

ND 0.84 0.07

0.95 0.03 67 14.27 0.02 67

0.83 0.01 59 10.69 0.01 50

0.88 0.01 62 8.77 0.09 41

0.86 0.05 60 12.53 0.02 59

1.19 0.03 84 13.67 0.02 65

0.78 0.01 54 10.03 0.04 47

The extraction efciency (EE%) is dened as the percentage of total acid digestion arsenic. The precision is expressed as one standard deviation. This value is obtained by N-hexane extraction after acetone extracted. This value is obtained by acetone extraction after N-hexane extracted.

Table 5 Result of spike recovery experiment (mg/kg dry weight, n = 3) Extraction procedure Spikes Spiked conc. Recovered spike conc.a AsC Aspolar + AsC AsB AsIII DMA MMA As VI 1 4 1 2 2 1 0.9 0.3 ND ND ND ND ND 77.0 AsC AsB AsIII DMA MMA As VI 1.5 6 1.5 3 3 1.5 1.39 0.09 ND ND ND ND ND 83.2 AsB ND 9.8 1.3 ND ND ND ND 104.1 ND 10.8 1.2 ND ND ND ND 94.7 AsIII ND ND 0.8 0.1 ND ND ND ND 80.0 ND ND 1.5 0. ND ND ND 100 DMA ND ND ND 2.2 0.7 ND ND 102.8 ND ND 5 ND 3.2 0.6 ND ND 101.9 MMA ND ND ND ND 2.1 0.4 ND 95.4 ND ND ND ND 2.9 0.8 ND 90.6 AsVI ND ND ND ND 1.8 0.7 104.7 ND ND ND ND ND 2.0 0.3 90.1

Asinorganic

Spike Recovey % Aspolar +

Asinorganic

Spike Recovery %
a

Spiked and recovered spike concentration (average 1 S.D. from three independent determinations) are reported in mg/kg (dry weight).

into the sample, extracted as steps II and III. The results are listed in Table 5. 77102.8% recoveries were obtained using these twostep extraction processes, indicating the validity of the method. 7. Conclusion Sequential extraction is an effective method to extract the arsenicals in DSPs, compared to extraction by single solvent extraction for which the efciency is much lower. High extraction efciency and high recovery (77102.8%) were obtained using this method. An optimized HPLC procedure was developed, in which baseline separation of AsB and AsC. The method has been applied to the screening and speciation of As species in DSPs. AsB was observed as the most component in the DSPs. It reaches 14.7% (clam) to 74.1% (shimp2) of total extractable arsenic. AsV was also observed in a high concentration in some DSPs, such as shrimp1, oyster and clam, suggesting the possible occurrence of As transformation processes during the manufacturing processes. Acknowledgments We gratefully acknowledge the nancial support by the National Natural Science Foundation of China (No: 20675021). References
National Academy of Sciences, (1997), Arsenic. National Research Council, Washington, DC, US. Allain, P., & Jaunault, L. (1991). Signal enhancement of elements due to the presence of carbon-containing compounds in inductively coupled plasma mass spectrometry. Analytical Chemistry, 63(14), 14971498.

Andreae, M. O. (1986). Organoarsenic compounds in environment. In P. J. Craig (Ed.), Organoarsenic Compound in Environment, Principles and Reactions (pp. 198228). New York: Wiley. Borak, J., & Hosgood, H. D. (2007). Seafood arsenic: Implications for human risk assessment. Regulatory Toxicology and Pharmacology, 47(2), 204212. Brisbin, J. A., & Hymer, C. B. (2002). A gradient anion exchange chromatographic method for the speciation of arsenic in lobster tissue extracts. Talanta, 58(1), 133145. Buchet, J. P., & Pauwels, J. (1994). Assessment of exposure to inorganic arsenic following ingestion of marine organisms by volunteers. Environmental Research, 66(1), 4451. Cannon, J. R., & Edmonds, J. S. (1981). Isolation, crystal structure and synthesis of arsenobetaine, a constituent of the western rock lobster, the dusky shark, and some samples of human urine. Australian Journal of Chemistry, 34(4), 787798. Cullen, W. R., & Reimer, K. J. (1989). Arsenic speciation in the environment. Chemical Review, 89(4), 713764. FAO. (1998), Composicion qumica. In El Pescado Fresco: su Calidady Cambios de su Calidad; Huss, H. H., Ed., FAO: Rome, 2034. Francesconi, K. A., & Edmonds, J. S. (1994). Biotransformation of arsenic in the marine environment. In J. O. Nriagu (Ed.), Arsenic in the Environment, Part 1: Cycling and Characterization (pp. 221263). New York: Wiley. Gomez-Ariza, J. L., & Sanchez-Rodas, D. (2000). A comparison between ICP-MS and AFS detection for arsenic speciation in environmental samples. Talanta, 51(2), 257268. Larsen, E. H., & Pritzi, G. (1993). Arsenic speciation in seafood samples with emphasis on minor constituents: an investigation using high-performance liquid chromatography with detection by inductively coupled plasma mass spectrometry. Journal of Analytical Atomic Spectrometry, 8(8), 10751084. Larsen, E. H., & Quetel, C. R. (1997). Arsenic speciation in shrimp and mussel from the Mid-Atlantic hydrothermal vents. Marine Chemistry, 57(34), 341346. Li, Weihua, & Wei, Chao (2003). A survey of arsenic species in Chinese seafood. Food and Chemical Toxicology, 41(8), 11031110. McKiernan, J. W., & Creed, J. T. (1999). A comparison of automated and traditional methods for the extraction of arsenicals from sh. Journal of Analytical Atomic Spectrometry, 14(4), 607613. Mir, K. A., & Rutter, A. (2007). Extraction and speciation of arsenic in plants grown on arsenic contaminated soils. Talanta, 72(4), 15071518. Nriagu, J. O. (1994). Arsenic in the environment: Human health and ecosystem effects. Advances in Environmental Science and Technology (Vol. 27). New York: John Wiley & Sons (chap. 12).

726

X. Cao et al. / Food Chemistry 113 (2009) 720726 Sikorski, Z. E, & Kolakowska, A. (1994). Tecnologa de los Productos del Mar Recursos, Composicio n Nutritivay Conservacin. In Z. E. Sikorski (Ed.), Composicin nutritiva de los organismos alimenticios marinos (pp. 4172). Zaragoza, Spain: Acriba. Wahlen, R., & McSheehy, S. (2004). Arsenic speciation in marine certied reference materials Part 2. The quantication of water-soluble arsenic species by highperformance liquid chromatography-inductively coupled plasma mass spectrometry. Journal of Analytical Atomic Spectrometry, 19(7), 876882.

Penrose, W. R. (1974). Arsenic in the marine and aquatic environment. Analysis, occurrence, and signicance. CRC Critical Review Environmental Control(4), 465482. Phillips, D. J. H. (1994). The chemical forms of arsenic in aquatic organisms and their Interrelationships. In J. O. Nrigau (Ed.), Arsenic in the Environment, Part 1: Cycling and Characterization (pp. 263288). New York: Wiley. Schmeisser, E., & Goessler, W. (2005). Direct measurement of lipid-soluble arsenic species in biological samples with HPLC-ICP-MS. The Analyst, 130(7), 948955.