Determination of arsenic in arsenic compounds and marine biological tissues using low volume microwave digestion and electrothermal atomic absorption spectrometry

Arsenic in marine biological tissues was measured by a low volume microwave dissolution technique and electrothermal atomic absorption spectrometry using a palladium and magnesium nitrate matrix modifier. Freeze dried samples (0.1 g dry wt) and 1 mL of nitric acid were placed into 7 mL screw top Teflon vessels and treated using a programmed three stage digestion. The effect of power and time parameters on arsenic recovery was examined. Preliminary studies showed that a short, maximum power first stage was beneficial for effective digestion, followed by cooling and an additional longer lower power stage. The first two microwave programme stages were kept consistent (600 W, 2 min; 0 W, 2 min) and the third stage was varied from 150 to 450 W over 15 to 60 min. Optimised microwave power and time conditions for arsenic recovery in marine reference materials were found to be 600 W, 2 min; 0 W, 2 min; 450 W, 45 min. Complete recovery of arsenic from a range of arsenic compounds and reference marine biological materials [National Institute of Standards and Technology (NIST) SRM 1566a Oyster Tissue; National Research Council Canada (NRCC) DORM-1 Dogfish Muscle; NRCC TORT-1 Lobster Hepatopancreas; NIST RM 50 Albacore Tuna; International Atomic Energy Agency (IAEA) MA-A-2 Fish Tissue; and AGAL-2 Shark Tissue] was achieved. The effects of single modifier and modifier combinations of magnesium, copper, nickel and palladium nitrates on the arsenic atomic absorption signal magnitude were examined for the inorganic species of arsenic [As(III) as trichloride and As(V) as sodium arsenate], a phosphate spiked arsenite standard, organic arsenic compounds (arsenobetaine, arsenocholine, monomethylarsonic acid, dimethylarsinic acid and trimethylarsine oxide) and digests of reference marine biological materials. All digests were prepared using optimised microwave conditions and diluted to 50 mL to bring the concentration of As in the samples to within the normal working range of the standards and reduce the observed phosphate chemical interferences for full recovery of the analyte. Three dimensional peak area surface response diagrams for Pd/Mg and Ni/Mg showed the following trends: (1) Mg alone recorded the highest signal magnitude (on average up to 17% higher than the optimum concentrations of Pd/Mg or Ni/Mg) for all arsenic compounds; (2) Mg alone exhibited poor modifying action for the digested marine tissues; (3) relative to the signal magnitude for Pd or Ni alone, the addition of Mg to the Pd or Ni modifier for arsenic compounds, in general, did not give a significant increase in signal magnitude; (4) in the case of digested marine tissues, the addition of Mg to Pd or Ni for analysis was essential for quantitative recovery of arsenic, providing the highest signal magnitude for the reference materials. The addition of phosphate to arsenic standard compounds at concentrations that result in approximately 0.052 mu mol P (1.5 mu g P) on the platform caused a significant suppression of the As signal. This concentration of phosphate also significantly reduced the ability of Mg2+ to act as a modifying agent, Mg no longer providing the highest signal magnitude for arsenic in arsenic compounds. The thermal stabilisation of As in the furnace was also reduced from 1200 to 900 degrees C. Pd and Ni were shown to provide higher signal magnitudes for phosphate spiked As standards, with the addition of Mg only increasing the signal magnitude by about 10% relative to Pd or Ni alone. The use of Cu, Ni and Pd was also found to increase significantly the thermal stabilisation of As in the furnace (1300 to 1400 degrees C). Higher atomisation temperatures were also required (up to 2200 degrees C). Examination of the Ni/Mg and Pd/Mg three dimensional peak area diagrams for digested marine tissues and the phosphate spiked As standard showed that both sample types exhibited similar optimum working ranges, with a smaller quantity of Pd being required to achieve the same modification as Ni. Cu/Mg did not appear to afford any advantage over Ni/Mg or Pd/Mg. In general, Pd/Mg gave a signal magnitude comparable to the equivalent Ni/Mg modifier for As, however, on the surface response diagrams, significantly larger optimal working regions for the marine tissue digests were recorded and Pd/Mg gave significantly higher signal magnitudes for the phosphate spiked standard. Pd is more effective in stabilising As in the presence of phosphate or in complex sample matrices and was therefore chosen as the preferred modifier to Ni. Solutions containing phosphate concentrations resulting in less than 0.0052 mu mol P (0.15 mu g P) on the platform did not exhibit any phosphorus interference. Digested tissues (0.1 g) needed to be diluted to 50 mL to lower the phosphate concentrations to these levels. The optimum signal response for arsenic in biological tissues using As(III) as an inorganic standard was obtained with Pd/Mg at a level of 0.15 mu mol Pd+0.4 mu mol Mg (15.9 mu g Pd+9.7 mu g Mg) on the platform. This value was selected as a compromise between the optimum amount of modifier needed for the stabilisation of arsenic compounds in aqueous standards and that required by the digested marine tissues.