Evaluation of slurry preparation procedures for the determination of mercury by axial view inductively coupled plasma optical emission spectrometry using on-line cold vapor generation

Abstract

Five different slurry preparation procedures were tested, after grinding the solid samples to a particle size ≤53 μm: (1) using aqua regia plus HF, 30 min of sonication, standing time of 24 h followed by another 30 min of sonication; (2) same as the previous one, except that the standing time and the second ultrasound treatment were omitted; (3) same as the previous one, except that HF was not used; (4) same as the previous one, except that the aqua regia was replaced by nitric acid; (5) same as the previous one, except that the acid nitric was replaced by tetramethylammonium hydroxide (TMAH). The Hg vapor was generated on-line, and the emission signal intensity measured at 253.652 nm by axial view inductively coupled plasma optical emission spectrometry. Initially, four experimental conditions were optimized using a multivariate factorial analysis: the concentrations of HCl and of the reducing agent, NaBH 4, used in the cold vapor generation, and two instrumental parameters, the plasma radiofrequency power and the carrier gas flow rate. The radiofrequency power was statistically significant, but limited to 1.2 kW for practical reasons. The procedures were applied to 11 biological and environmental materials. Both, the slurries and the filtrates were analyzed, using calibration solutions in the same medium as in the slurries. The first three procedures produced results in agreement with the certified values. The two last procedures, using nitric acid or TMHA could not be used for quantitative analysis. For practical reasons, Procedure 3, with a detection limit (3s, n=10) of 0.06 μg g −1 for a sample mass of 20 mg in a final volume of 15 mL is recommended. The relative standard deviations for mercury in the investigated materials, using the recommended procedure, were lower than 12.5%, indicating a good precision for slurry sampling. The recommended procedure is simple, rapid and robust.