Absorption Spectrometry after Microwave Digestion: NMKL1. Collaborative Study. KAARE JULSHAMN. Institute of Nutrition, Directorate of Fisheries, PO Box 185 ...
JULSHAMN ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 83, NO. 6, 2000 1423
RESIDUES AND TRACE ELEMENTS
Determination of Arsenic in Seafood by Electrothermal Atomic 1 Absorption Spectrometry after Microwave Digestion: NMKL Collaborative Study KAARE JULSHAMN Institute of Nutrition, Directorate of Fisheries, PO Box 185, N-5804 Bergen, Norway ARNGRIIMUR THORLACIUS Borregaard Industries Ltd., PO Box 162, N-1701 Sarpsborg, Norway PER LEA Matforsk, Norwegian Food Research Institute, Osloveien 1, N-1430 Ås, Norway Collaborators: Kjetil Barland; Kari Eidem; Karl Olav Gjerstad; Dag Grønningen; Erik H. Larsen; Syverin Lierhagen; Helena Liukkonen-Lilja; Birger Lind
Eight laboratories participated in an interlaboratory method performance (collaborative) study of a method for the determination of arsenic in foodstuffs of marine origin by electrothermal atomic absorption spectrometry after wet digestion using a microwave oven technique. The study was preceded by a practice round of familiarization samples. The method was tested on 8 materials (cod roe, krill, blue mussel, saithe, scampi, cod fillet, shrimp, and cod extract) ranging in As content from 2 to 75 mg/kg. The materials were sent to participants in the study as blind duplicates, and the participants were asked to perform single determinations on each sample. Repeatability relative standard deviations (RSDr) for As ranged from 6.8 to 17.4%. Reproducibility relative standard deviations (RSDR) ranged from 7.6 to 24%. The highest RSDR value was found for the sample with the highest concentration of As.
he levels of arsenic in foods generally reflect normal accumulation from the environment. The chronic and acute toxicity of As and its compounds are well known (1). Chapman (2) reported as early as 1926 that As in marine crustaceans and molluscs was present mainly in a chemical form which was non-toxic to human beings. It has further been suggested that As may be an essential element to humans (3) and it seems clear that As compounds found in seafood have less acute toxicity than inorganic As (4), which can be found in small concentrations in fish fillet (5). In some cases, authorities in countries that import seafood products demand from the
T
producing country documentation of the content of total As in fish and fish products. The Nordic Committee on Food Analysis (NMKL) decided to undertake an interlaboratory study of the determination of total As in seafood products by electrothermal atomic absorption spectrometry (ET AAS) using a wet digestion procedure in a microwave oven, to evaluate performance characteristics (trueness, repeatability, and reproducibility) using judiciously chosen foodstuffs. The present method has been in use for several years at the Institute of Nutrition (Bergen, Norway). Prior to the collaborative study, a preliminary study was performed aimed at studying some critical factors that could affect the trueness and reproducibility in the ET AAS analysis. This study involved comparing different background correction techniques, the use of pyrolytically coated versus uncoated platform graphite tubes, as well as Ni versus Pd/Mg as chemical modifiers. The study was performed using 5 sample solutions of marine Certified Reference Materials (CRM) distributed to 4 participating laboratories. The following results were found: No differences in results were obtained between AAS instruments equipped with Zeeman correction or deuterium arc background correction. Small differences in concentration levels of As in the CRM as well as in characteristic mass were found when chemical modifiers were compared. Pd/Mg was recommended in order to avoid contamination of the graphite furnace with nickel. The characteristic mass was improved by using pyrolytically coated graphite tubes with the L’vov platform as opposed to using uncoated graphite tubes with the platform. In the collaborative study, the standard addition procedure was recommended (6). Collaborative Study
Received December 15, 1999. Accepted by JS March 28, 2000. 1 Nordic Committee on Food Analysis, General Secretariat, c/o National Veterinary Institute, Department of Food and Feed Hygiene, PO Box 8156 Dep. N-0033 Oslo, Norway.
A description of the method and an invitation to participate in the study were sent to 11 laboratories in Denmark, Finland,
1424 JULSHAMN ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 83, NO. 6, 2000
Iceland, Norway, and Sweden. Eight laboratories accepted and agreed to follow the method procedure and the time schedule of the study. The design, conduct, and interpretation of the study followed guidelines recommended by AOAC INTERNATIONAL (7) and NMKL (8). Participants represented industry, commercial laboratories, universities, and government laboratories.
Study Materials Each participant received 8 test materials (all materials were obtained from Rieber & Søn ASA, Bergen Norway): hard cod roe powder (hard cod roe cooked, mixed with antioxidant, vacuum dried, and milled into fine powder); krill powder (Euphausia superba; meat of Antartic krill without shell that has been cooked, dried, and milled into fine powder); mussel powder (Mytilus edulis; meat from fresh, frozen mussel cooked, minced, dried, and milled into fine powder); saithe powder (Pollacius virens; deboned, fresh, frozen fish cooked, minced, dried, and milled into fine powder); scampi powder (Nephros norvegicus; scampi powder made of tails from fresh, frozen Norwegian lobster; the tails were cooked, minced, dried, and milled into fine powder); cod muscle powder (Gadus morhua; cod muscle powder made of fresh, frozen muscle of cod, cooked, dried, and milled to fine powder); shrimp powder (Pandalus borealis; shrimp powder produced of fresh, cooked, and frozen shrimp that have been dried and milled to fine powder); cod extract powder (Trisopterus esmarkii; cooked cod, fluid separated by filtration, concentrated by vacuum drying, and milled to fine powder). The expected values of As in the materials used in the study were obtained by analyzing 10 replicates per material (Table 1). The homogeneity of the test materials was investigated by estimating within-container and between-container variations of the 8 study materials. The homogeneity test was performed by taking 5 subsamples of 5 g from each food sample. Two replicates of 0.25 g (according to the method) from each subsample were digested. In total, 10 replicates of each food sample were analyzed for As. Results were analyzed by 1-way analysis of variance (ANOVA). Values of p > 0.05 were obtained for all test materials, indicating that they were homogeneous.
Protocol Before the full collaborative study, participants were given the opportunity to become familiar with the method in a pre-trial test. The materials included 2 certified reference materials: blue mussel and oyster tissue, both purchased from the U.S. National Institute of Standards and Technology (NIST; Gaithersburg, MD). The blue mussel contained a certified concentration of 6 mg/kg As and the oyster 14 mg/kg As. Six laboratories reported results for the certified reference materials. The results show that some laboratories had problems in analyzing the samples with acceptable trueness and repeatability, probably because of insufficient experience with the present method. Hopefully, the experience gained through the pretrial tests was sufficient to perform adequately in the main study. The 8 test materials of the collaborative study were all dry and packed in small plastic containers. They were presented to participants as blind duplicates, that is, as 16 randomly coded materials. The fact that blind duplicates were included in the set of materials was not disclosed to the participants. Participants were asked to perform single determinations of As concentration of the materials according to the method described below and to report results in mg/kg on a dry weight basis. Participants were also asked to give information on the microwave oven and the temperature program used, as well as to report name and model of the AAS instrument, type of graphite tube, background corrector, and acid purity used. All test samples were dried, with a residual moisture of 2 to 8%. The participants were asked to perform dry matter determinations on the test materials and report their values in mg/kg dry weight. METHOD
Fields of Application The method is applicable to quantitative determination of As in various types of seafood products. The method has been tested primarily on dry products but may be used for fresh samples as well.
Principle Table 1. Types of materials included in the study and their expected As concentrations (average of 10 independent decompositions per sample) Material
Expected values, mg/kg (dry weight)
Hard cod roe powder
2.5 ± 0.2
Krill powder
3.5 ± 0.2
Mussel powder
8.5 ± 0.4
Saithe powder
13.5 ± 0.5
Scampi powder
20.0 ± 0.8
Cod muscle powder
25 ± 1
Shrimp powder
40 ± 1
Cod extract powder
80 ± 2
Concentrated nitric acid and hydrogen peroxide are added to samples, which are then digested in a microwave oven. Any commercially available laboratory microwave oven may be used. As concentrations are determined by ET AAS using a standard addition procedure or external standardization.
Chemicals and Reagents (a) Concentrated nitric acid (HNO3).—65% Suprapur. (b) Nitric acid, 0.65% (w/v).—Dilute 10 mL concentrated HNO3, (a), to 1000 mL with water. (c) Hydrogen peroxide (H2O2).—30%, analytical grade. (d) Deionized and possibly filtered water.—Specific resistance, >18 MΩ/cm. (e) As standard.—1000 mg/L (commercial standard solution).
JULSHAMN ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 83, NO. 6, 2000 1425
(f) As standard solutions.—Dilute As standard, (e), to appropriate volume with 0.65% HNO3, (b). Prepare fresh solutions daily. (g) Palladium-matrix modifier (As concentration < 0.01 mg/L), 10 mg Pd/mL.—Based on Pd(NO3)2. (h) Palladium-matrix modifier solution, 1.5 mg Pd/mL.—Dilute palladium–matrix modifier, (g), to 10 mL in a volumetric flask with 0.65% HNO3, (b). (i) Magnesium-matrix modifier (As concentration < 0.01 mg/L), 10 mg Mg/mL.—Based on Mg(NO3)2⋅6H2O. (j) Magnesium-matrix modifier solution, 1 mg Mg/mL.—Dilute magnesium-matrix modifier, (i), to 10 mL in a volumetric flask with 0.65% HNO3, (b). (k) Mixed matrix modifier.—(h) + (j), 1 + 1. Ten µL contains 7.5 µg Pd and 5.0 µg Mg.
Apparatus, Equipment, and Gases (a) Analytical balance. (b) Laboratory microwave oven.—With 100 mL digestion containers and withstanding a pressure of 30 bar. (c) Dispenser. (d) Glassware.—25, 50, 100, and 250 mL volumetric flasks. (e) Polyethylene tubes or glass tubes with screw caps.—Fifteen and 50 mL. (f) Polyethylene tubes with screw caps.—Ten mL. (g) Automatic pipets with tips. (h) Atomic absorption spectrometer.—With background corrector (Zeeman, deuterium lamp, or high current HCl pulsing). (i) Electrodeless discharge lamp, hollow cathode lamp or a boosted hollow cathode lamp for As. (j) Autosampler. (k) Autosampler cup, 2 mL. (l) Graphite furnace. (m) Pyrolytically coated graphite tubes with pyrolytically coated L’vov platform. (n) Argon with purity of 99.998% or better.
Preparation of Test Samples The procedure for determining dry matter content may involve freeze-drying and thermal drying at 105°C for 12 h until constant weight is obtained. Weigh into the digestion container an amount of homogeneous sample corresponding to 0.20–0.25 g dry material. Each digestion series must contain 2 reagent blanks, that is, HNO3 (Reagents, a) and H2O2 (Reagents, c) only without sample materials. If possible, include certified reference materials containing As in amounts corresponding to those found in samples to reveal systematic or random errors. Note: The following 2 sections should be regarded as examples. Digestion programs and amounts of acids will vary with different digestion systems.
Digestion Add 2 mL concentrated HNO3, and 0.5 mL H2O2, to each container. Seal the containers in the capping station. Place carousel with the digestion containers in the microwave oven and start the program: S1, 250 W, 1.00 min; S2, 0 W, 1.00 min; S3, 250 W, 5.00 min; S4, 400 W, 5.00 min; and S5, 650 W, 5.00 min. This program is suggested for 6 containers. For a lesser or greater number of containers, the applied power for each step should be altered proportionally, i.e., for 12 containers the power for each step should be doubled. Open the cooled containers and rinse with water any condensed water in the cap and on the walls. Quantitatively transfer the sample solution to a 25 mL volumetric flask and dilute to the mark with water. Transfer the sample solution to a polyethylene tube.
Calibration of the Microwave Power Transfer 1 kg water to a microwave-transparent beaker (made from Teflon, polyethylene, or polypropylene) and measure the temperature of the water. Place the beaker in the microwave oven and heat at full power for 2 min. Stir the water thoroughly for ca 30 s and remeasure the temperature. The absorbed power, P, is calculated with the expression: P = 34.4 · )T, where T is the temperature increase measured in centigrade. The absorbed power should depart
