ANALYST, APRIL 1989, VOL. 114 527 Determination of Selenium in Fish Flesh by Hydride Generation Atomic Absorption Spectrometry Lena Hansson, Jean Pettersson and Ake Olin Department of Analytical Chemistry, University of Uppsala, P.O. Box 531, S-751 21 Uppsala, Sweden Selenium was determined in freeze-dried fish flesh from perch, pike-perch and the fish flesh reference materials MA-A-2 No. 1174 and MA-6-3 No. 151 (both from the International Atomic Energy Agency) by hyd ride generation atomic absorption spectrometry. Fou r different decom position procedures were tested. They yielded consistent results for the four specimens, but the certified concentration level of selenium in the reference material MA-A-2 was not reached. This indicated losses or the presence of very stable selenium-containing compounds in this fish flesh.Neutron activation analysis of the reference material, however, was in agreement with those results obtained. Keywords: Decomposition procedures; selenium; h ydride generation atomic absorption spectrometry; fish flesh; reference material The determination of selenium in biological specimens of aquatic origin can present problems. Reports indicate that low recoveries of the element are sometimes obtained. 1.2 It is possible that organic compounds containing selenium that are resistant to oxidative attack are present in the samples. Recently, we were asked to determine selenium in fish muscle in a project initiated by the Swedish National Environmental Protection Board.3 Therefore, the perform- ance on fish flesh samples of our procedure developed previously for the determination of selenium in biological materials has been investigated.The procedure uses hydride generation atomic absorption spectrometry (HG-AAS) for the final determination of selenium and is combined usually with a magnesium nitrate - nitric acid - hydrochloric acid digestion step to break up the sample. This method has proved reliable for a number of biological materials.”s Freeze-dried flesh from perch, pike-perch and fish flesh reference materials were subjected to four different decomposition procedures prior to HG-AAS. Experimental Reagents and Materials All the reagents and chemicals were of pro analysi quality. The water used for dilution and washing was de-ionised, distilled and filtered through a Milli-Q system.The reference materials were MA-A-2 No. 1174 and MA-B-3 No. 151 fish homogenate (International Atomic Energy Agency. Monaco). The filter-paper used in the closed flask combustion procedure had a low ash content (OOM class 1, Munktells, Grycksbo, Sweden). Apparatus The instrumentation for HG-AAS and the equipment for sample work-up have been described previously.4 Procedure Decomposition Three digestion methods were applied to the samples: (i) the “magnesium” method, (ii) the nitric acid - perchloric acid - sulphuric acid method and (iii) digestion with nitric acid in a closed bomb. These methods are described in reference 4. In addition, method (iv), closed flask combustion in an oxygen atmosphere, was tested and is described below. Weigh about 0.05 g of the sample accurately on to a filter- paper cut out as described in reference 6.Fold the paper to form a small package containing the sample and place the package in a platinum gauze fastened to the stopper of a 1-1 Erlenmeyer flask by a platinum wire (gauze, 20 x 20 mm; wire, 50 mm). Allow a wick to protrude from the paper. Add 15 ml of 4~ hydrochloric acid to the flask as an absorbing medium. Fill the flask with oxygen from a gas tube by flushing for about 10 s, then ignite the wick and introduce the sample package quickly into the flask. Hold the stopper firmly in place and turn the flask upside-down so that the hydrochloric acid can provide a liquid seal for the gases formed. After a few seconds the combustion is complete. Allow the combustion products to be absorbed by the hydrochloric acid for 30 min.Swirl the flask occasionally to facilitate the absorption. Remove the stopper and boil the contents for 15 min to remove the dissolved gases and to reduce any selenate to selenite. After cooling, dilute the solution to 50 ml. Table 1. Concentrations (ug g-1) of selenium in fish samples obtained using four decomposition methods. SD = standard deviation, n = number of deter 111 in at io ns (i) (ii) (iii) (iv) Sample Mcan SD n Mean SD n Mean SD rz Mean SD n Perch . . , . . . . . , . . . 1.29 0.06 7 1.18 0.06 5 1.1s 0.06 4 1.32 0.07 4 MA-A-2No.1174’ . . . . . . 1.15 0.06 12 1.11 0.07 6 1.13 0.06 3 1.29 0.07 4 MA-B-~No. 151t . . . . . . 1.34 0.03 6 1.44 0.03 3 1.41 0.02 3 1.43 0.08 2 Certified concentration, 1.7 & 0.3 pg g-1.Pike-perch . , . . . . . , . . 0.51 0.02 7 0.48 0.03 3 0.48 0.02 5 0.54 0.03 4 NAA result: direct measurement, 1.10 k 10 pg g 1 ; after chemical separation, 1.12 * 0.06 pg g-I (Studsvik AB, Nykoping, Sweden). i Certified concentration, 1.46 pg g-1 (90°/, confidence interval, 1.35-1.70 pg g-I).528 ANALYST, APRIL 1989, VOL. 114 Evaluation The concentration of selenium was evaluated from a calibra- tion graph obtained from standards prepared by adding known amounts of selenite to combusted blanks. Application of the standard additions method to the samples did not change the results significantly. Results and Discussion Concentrations of selenium found in the fish specimens are shown in Table 1. They were calculated on a dry mass basis.The result from the analysis of one of the reference materials using neutron activation analysis (NAA) is also included. In general, the results are consistent both between the digestion procedures and between HG-AAS and NAA. The certified mean value for MA-A-2 of 1.7 k 0.3 pg g-* of selenium, however, was not obtained. This could indicate a systematic error in our procedures of about -30% (over-all mean value was 1.17 pg g-1). but in that instance NAA suffers from the same error (over-all mean value, 1.11 pg g-1). Neutron activation analysis was performed both directly on the irradiated sample and after chemical separation. A further explanation for the discrepancy may be that the MA-A-2 is certified incorrectly for selenium. This is supported by the fact that the result obtained for MA-B-3 was consistent with the certified value References 1. Welz, B., and Melcher, M., Anal. Chem., 1985, 57, 427. 2. Bye, R., and Lund, W., 2. Anal. Chem., 1985, 321, 483. 3. H$kansson, L., "Liming and Mercury," Swedish Environment Protection Board, Stockholm. 4 Pettersson, J . , Hansson, L., and Olin, A,. Talanra. 1986, 33, 249. 5 Hansson, L., Pettersson, J . , and Olin, A., Tafanta, 1987, 34, 829. 6. Gorsuch, T. T., "The Destruction of Organic Matter," Pergamon Press, Oxford, 1970. p, 32. Paper 8f04535F Received November 14th, 1988 Accepted December 12th, I988