Analyst Aqgust 1983 Vol. 108 pp. 939-943 939 Spectrophotometric Determination of Arsenic in Biological Tissues and Sediments After Digestion with Nitric Sulphuric and Perchloric Acids and Pre-concentration by Zinc Column Arsine Generation and Trapping W. A. Maher" Department of Oceanography University of Southampton Southampton SO9 5NH A procedure for the determination of total arsenic in environmental extracts is described. Arsenic is converted into arsine using a zinc reductor column, the evolved arsine trapped in a potassium iodide - iodine solution and the arsenic determined spectrophotometrically as an arsenomolybdenum blue complex. The detection limit (based on four times the standard deviation of six blank measurements) is 0.024 p g and the coefficient of variation is 5.1% at the 0.1-pg level.The method is free from interferences by other elements at levels normally found in environmental samples. Keywords Arsenic determination ; hydride generation and trapping ; mole-cular-absorption spectrophotometry ; environmental materials The continuing interest in arsenic in the en~ironmentl-~ has led to a need for methods of improved simplicity and precision for the determination of arsenic in environmental materials. Many methods for the determination of arsenic in environmental extracts use arsine genera-tion after the prior destruction of the matrix to isolate and concentrate arseni~.~,5 Arsine generation methods usually utilise the addition of zinc magnesium aluminium or sodium tetrahydroborate(II1) in the forms of solutions slurries or pellets to acidified solutions.The main problems with using arsine generation separation methods is that severe interferences, affecting both the extent and the rate of hydride generation are caused by the presence of large amounts of other element@-7 and strong oxidising acids.* Interferences in the deter-mination step may also be caused by other hydride forming elements (selenium tellurium, tin antimony and germanium) which are co-generated. Published methods also often require long generation times and suffer from imprecision owing to the variable evolution of hydrogen. The use of a zinc reductor column to generate arsine that can tolerate high concentrations of foreign ions and allow a rapid throughput of prepared samples with little evolution of hydrogen has been reported.9 In this paper the use of a zinc column to generate arsine and the subsequent trapping of the gas have been investigated.This method of separation and concentation has been combined with the determination of arsenic as arsenomolybdenum blue to give a convenient method for determining total arsenic in solutions obtained from biological and sediment samples. Experimental Instrumentation and Apparatus 4-cm microcuvettes (capacity less than 2 ml). was prepared by filing zinc sticks to pass through a 2-mm sieve. 10-20 mesh. Absorbance measurements were made with a Zeiss PMQ 2 spectrophotometer using 1- and The reductor column used to generate and trap the arsine is shown in Fig. 1 . Zinc powder The Drierite (CaSO,) was Reagents All chemicals were of analytical-reagent grade.* Present address Department of Physical and Inorganic Chemistry University of Adelaide Adelaide 5001 Australia 940 MAHER SPECTROPHOTOMETRY OF As IN BIOLOGICAL TISSUES Analyst VoZ. 108 Dissolve 1.3204 g of arsenic(II1) oxide in 1000 ml of 0.1 M sodium hydroxide solution. A reducing solution was prepared by dissolving 15 g of potassium iodide and 15 g of ascorbic acid in 100 ml of distilled water . A trapping solution was prepared by dissolving 1 g of potassium iodide and 0.5 g of iodine in 100 ml of distilled water. A mixed spectrophotometric reagent was prepared by adding 2.2 nil of 18 M sulphuric acid 5 ml of 5% m/V ammonium molybdate solution 1 ml of 0.3% m/V antimony potassium tartrate solution and 12 ml of 15% m/V ascorbic acid solution to 29 ml of distilled water in the order indicated.Arsenic(II1) standard solution 1 000 pg ml-l. Potassium iodide (15% m/V) - ascorbic acid (15% m/V) solution. Potassium iodide (1% m/V) - iodine (0.5% m/V) solution. Mixed reagent. Syringe ll Zinc i 1 ml) - Drierite Pasteur pipette Centrifuge tube Potassium iodide -solution iodine Fig. 1. Apparatus to generate dry and trap arsenic hydride. Procedure Biological tissues and sediments were freeze-dried and ground (to less than 200 pm) before analysis. A weighed sample (less than 0.5 g) was placed in a 30-ml Pyrex centrifuge tube 5 ml of concentrated nitric acid were added and the mixture was allowed to stand for at least 12 h at room temperature to ensure complete dissolution (this avoids foaming on heating).The tube was then placed in an aluminium heating block and refluxed until the evolution of brown fumes ceased. After cooling 5 ml of a nitric - sulpliuric - perchloric acid mixture (5 + 1 + 3 V / V ) were added and heating continued until dense fumes of sulphur trioxide appeared. The digest was diluted with 5 nil of 1.5 M hydrochloric acid 1 ml of reducing solution was added and the solution allowed t o stand for 40 min to reduce all of the inorganic arsenic to the trivalent form. The solution was then made up to 25 ml in a calibrated flask with 1.5 M hydrochloric acid. The apparatus was assembled as in Fig. 1 with 1.5 ml of the potassium iodide - iodine solution in the centrifuge tube. The nitrogen gas flow-rate was adjusted to 150 ml min-l, 1 ml of sample was injected on to the zinc column and the evolved arsine collected for 2 min.Between each sample 1 ml of 1.5 M hydrochloric acid was injected to prevent any accumulation of potentially interfering material on the column. After the arsine had been trapped the centrifuge tube was removed and 0.5 ml of the mixed spectrophotometric reagent added. The solution was mixed by means of a vortex mixer and allowed to stand for 30 min to develop the arsenomolybdenum blue. Calibration graphs of absorbance veyszts amount of arsenic (0-1 and 0-5 pg) were prepared by using arsenic standards carried through the entire analytical procedure. The absorbance was measured at 866 nm August 1983 AND SEDIMENTS AFTER ACID DIGESTION AND PRE-CONCENTRATION Results and Discussion Optimisation of Hydride Generation and Trapping 90 80 70 941 ---To optimise the conditions for hydride generation and trapping the effects of acid concentra-tion gas flow-rate and trapping solution composition on the evolution and trapping of arsine were investigated.Sample solutions containing 5 pg ml-l of arsenic (1 ml injections) and 1.5 ml of 1 % m/V potassium iodide - iodine trapping solution were used for the optimisation of acid concentration and the gas flow-rate. The effects of flow-rate and acid concentration are shown in Figs. 2 and 3. It was desirable to use the lowest acid concentration possible as an excess of acid necessitated the frequent replacement of zinc and also caused excessive entrain-ment of water and acid vapour which resulted in the rapid clogging of the absorbent.Hydro-chloric acid was preferable to sulphuric acid which consumed larger amounts of zinc during hydride evolution. A hydrochloric acid concentration of 1.5 M and a gas flow-rate of 150 mlmin-l were chosen for all further work. Under these conditions the collection of arsine was complete within 2 min. A trapping solutioii containing 1% nz/V potassium iodide and 0.5% m/V iodine quantitatively trapped up to 5 pg of arsenic (from a 1-ml injection) as arsine; no arsenic was found in a second bubbler trap in series with the first. The iodine concentration was not critical as long as an excess was present. The equilibrium I + I- + 13- produces 13- which traps the arsine. Recoveries of arsenic added to the zinc column were all greater than 97% (97 & 1 97.6 &- 0.5 and 98.2 & 0.8% for 0.1 2.0 and 5.0 pg ml-l of arsenic respectively using the proposed conditions).The use of smaller capillary tubing for trapping did not increase the trapping efficiency. 100 I s! -. 100 F 8 90 L I I Gas flow-rate/mI min-’ Fig. 2. Effect of gas flow-rate on hydride generation and trapping. Hydrochloric acid concentration 2 M. I I I I I 1 2 3 4 Acid concentrationlhn Fig. 3. Effect of acid concentration on hydride generation and trapping. A, Hydrochloric acid media ; B sulphuric acid media. Spectrophotometric Determination The conditions for the determination of arsenic as arsenomolybdenum blue were investigated by Portmann and RileylO and the optimum concentrations that they recommended have been used in this study.It was thought possible that the absorption spectrum of arsenomolybdenum blue might be modified in the presence of an excess of potassium iodide. The spectrum was plotted and no shift in the absorption maximum was found; thus all measurements were made at 866 nm. The calibration graph of absorbance versus arsenic injected on to the zinc column was linear in the range 0-5 pg for 1-cm cells and 0-1.25 pg for the 4-cm cells. Interferences Possible interference by other elements was investigated by measuring the arsine generated and trapped in the presence of elevated concentrations of other elements. The concentrations at which certain elements interfere are shown in Table I. Various other elements [Al(III), B(III) Ca(II) Cd(II) Co(II) Cr(VI) Fe(III) K(I) Li(I) Mg(II) Mn(II) Na(I) Ni(II) 942 MAHER SPECTROPHOTOMETRY OF As IN BIOLOGICAL TISSUES Analyst VoZ.108 TABLE I EFECT OF INORGANIC IONS ON THE GENERATION AND TRAPPING OF ARSINE All tests used 0.1 p g of As(II1) in 1 ml of 1.5 M hydrochloric acid. generation and trapping conditions were used. Optimised hydride 22 Species . . . . . . Cu(I1) Hg(I1) Mo(V1) Sb(II1) Se(1V) Si(1V) Interference level/pg . . . . 50 2.5 400 25 0.1 50 Pb(II) S(VI) Sn(I1) and Zn(II)] showed no significant interference at the 500-pg level. Only low selenium concentrations in extracts can be tolerated. However few environmental samples contain appreciable amounts of selenium. As selenium is not reduced to hydrogen selenide on the column selenium will not interfere in the final determination step but probably suppresses either arsenic reduction or arsine formation.Selenium appears to suppress arsine generation at high arsenic concentrations but causes a slight enhancement at low arsenic concentrations (around 0.1 pg) which could not be traced to arsenic impurities in the selenium standard used. Accuracy Precision and Detection Limit The accuracy of the method was assessed by recovery experiments and the analysis of standard reference materials (orchard leaves NBS SRM 1571 and oyster tissue NBS SRM 1566). As shown in Table 11 complete recovery of added arsenic was obtained within experi mental error for selected biological tissues and a sediment. The arsenic concentration obtained by replicate analysis of the orchard leaves (9.7 & 0.3 pg g-l) and oyster tissue (13.2 0.4 pg g1) were in agreement with the certified values of 10 & 2 and 13.4 & 1.9 pg gl, respectively.The precision was determined from replicate analyses of arsenic standards carried through the entire procedure. The relative standard deviation at the lowest concentration examined (0.1 pg of arsenic) was 5.1%; for 2.0 and 5.0 pg ml-1 of arsenic the relative standard deviations were 3.7 and 2.4% respectively. The standard deviation of the blank (6 determinations) corresponded to 0.006 pg of arsenic. TABLE I1 RECOVERY OF ARSENIC ADDED TO SELECTED BIOLOGICAL TISSUES AND A SEDIMENT Arsenic added as inorganic arsenic. Arseniclpg 1 Sample kdded Found Recovery % Macroalgae Ecklonia radiata (0.25 g) .. 0 21.3 f 0.9 10 30.9 f 0.9 96 20 51 f 1 99 5 21.8 f 0.6 100 10 26.6 f 0.4 98 5 7.9 f 0.3 96 10 12.9 f 0.6 98 Crayfish J a w s novae hollandiae (0.5 g ) . . . . 0 16.8 f 0.6 Sediment (0.5 g) . . 0 3.1 -+ 0.2 Conclusion The experimental method described in this paper allows the determination of arsenic down to 0.1 pg with a relative standard deviation of 5.1%. The advantages of this method are the inexpensive equipment required and the high concentration of foreign ions that can be present without causing an interference. References 1. 2. Penrose W. R. CRC Crit. Rev. Environ. Control 1914 4 465. Lunde G. Environ. Health Perspect. 1977 19 47 August 1983 AND SEDIMENTS AFTER ACID DIGESTION AND PRE-CONCENTRATION 943 3. 4 5. 6 . 7. 8. 9 . 10. Benson A. A. and Summons R. E Science 1981 211 482. Talmi Y and Bostick D. T. J . Chromatogr. Sci. 1975 13 231. Yamsmoto Y . Kumamaru T. Hayashi Y. and Kamada T. Bull. Clzem. SOC. Jpn. 1973 46 2604. Smith A. E. Analyst 1975 100 300. Guimont J. Pichette M. and Rhekume N. A t . Absorpt. Newsl. 1977 16 53. Kang H. K. and Valentine S. L. Anal. Chem. 1977 49 1829. Maher W. A. Talanta 1982 29 532. Portmann J. E. and Riley J. P. Anal. Chim. Acta 1964 31 509. Received November 8th 1982 Accepted March loth 198