ICP-MS and Te Co-precipitation after nickel sulphide fire-assay collection for pd determination in roadside soils

Abstract

The analytical determination of palladium derived from car emissions and deposited on topsoil has close resemblance to the determination of palladium in geological samples in general. On Earths crust Pd occurs associated with other precious metals (Pt, Rh, Ir, Ru and Os, which are collectively known as platinum group elements, PGE). Their abundance is very low (just few ppb or even less) and mineralised rocks may contain just few parts per million of PGE, as discrete platinum-group minerals and in solid solution in sulfides, arsenides and sulfarsenides (Cabri 1992). Palladium is also sometimes found associated with gold (Cabral et al. 2002). The random and inhomogeneous distribution of PGEs phases implies that when precious metals are to be determined in rock, sediment and soil samples, large sample mass should be taken (at least 5 g, but it can be as large as 50 g, Potts 1992). The commonest analytical techniques require that sample should be taken in solution, to separate the precious metals from the matrix elements and an easy choice could be acid dissolution. Among the precious elements used in catalytic converters, palladium has the larger mobility (Wood and Middlesworth 2004), which is associated to the higher solubility of this element, which was attributed to reactions with nitrogen oxides (Amosse and Delbos 2002). As in minerals, in catalytic converters Pd is alloyed with other precious elements and total dissolution requires strong oxidizing reagents, like aqua regia. The solubility of PGE containing phases in aqua regia is substantial, but not always complete (Gowing and Potts 1991) and it can be quantitative if pressure is applied, in apparatus like Carius tubes or a high pressure asher (Pearson and Woodland 2000, Pennebaker and Denton 2001, Meisel et al. 2003, Fritsche and Meisel 2004). The use of closed vessels has constrains regarding sample mass and alternatives are fusion methods. One successful procedure is the sintering or fusion with sodium peroxide (Enzweiler et al. 1995; Morcelli and Figueiredo 2000) which has been successfully applied to large samples mass (Amosse 1998; Amosse and Delbos 2002). Another fusion method is the so-called fire assay (also known as docimasy or FA). In this method, a large sample is fused in alkaline fluxes (usually mixtures of sodium carbonate and sodium tetraborate) and metallic compounds which will collect precious metals. The most frequent collectors are lead (added to the initial sample+flux as lead oxide or carbonate) or nickel sulphide (added as metallic Ni and elemental S). Lead is used for Au collection and Pd can be also quantitatively recovered. When other PGEs are of interest, nickel and sulphur are used (NiS). This is also the choice when analysis of PGE associated with catalytic converters deposited on soils is of interest. During sample fusion, the nickel sulphide phases separate from the slag, sinking to the bottom of the clay crucible and the small droplets coalesce, forming a button which size depends on the initial amounts of Ni and S. Precious metals are collected by the NiS, due to their high affinity (partition coefficient) to the sulphide phase. The coprecipitation with Te after the HCl digestion of the NiS button is reported to improve the PGE recovery (Shazali et al. 1987). Here we describe the analytical method of NiS-FA and Te coprecipitation followed by ICP-MS as applied to the determination of Pd originated from catalytic converters in roadside soils collected from a high traffic road in Sao Paulo, Brazil. Method accuracy was evaluated by the analysis of geological reference materials with Pd certified values.