PGE analyses of MORB have been notoriously challenging and time consuming because of their very low concentrations, large sample sizes required, and dissolution difficulties encountered because of their refractory nature. Most of the existing data has been obtained by neutron activation. However, with the advent of the ICP-MS and improved methods of dissolution, the challenge has been mostly overcome [Shirey and Walker, 1995; O'Neill, 1996; Rehkamper and Halliday, 1997; Rehkamper et al., 1998; Pearson and Woodland, 2000]. It is now possible to analyze basalts and sediments of 1-5 gram size in the ppb-ppt range [e.g Colodner et al., 1992; Fryer and Greenough, 1992; Colodner et al., 1993; O'Neill, 1996; Rehkamper et al., 1998, 1999 a & b]. Investigators at URI are now routinely using an isotopic dilution HR-ICP-MS method for analyzing Ir, Ru, Pt, Pd, Re, Ag, and Cd in basalt glasses. Our new procedure for these elements, modified after that of Pearson and Woodland [2000], is illustrated in Fig. 1 below. Its precision, accuracy and reproducibility has been tested against 6 rock standards with compositions ranging from peridotite to andesite and one in-house mid-ocean ridge basalt (EN026 10D-3).

         Table 1 reports these results and blank levels asserted for 1-2 gram samples. Comparison with recommended or previously reported PGE values is generally excellent (Fig. 2). Note that several PGE recommended, or simply reported, values for these standards are missing. This illustrates the difficulty of analyzing PGE in basalts because of their very low concentrations (part per billion to trillion levels). Reproducibility of our method is in the range of 4-11% (1sd) for the PGE, Cd and Ag and 17% for Re on sample size of 1-2 grams. This is considered good because of the low PGE concentration level present in basalts and the notorious "nugget effect" (inhomogeneous distribution of PGE in rocks due to metal alloy formation). One example of this effect was possibly encountered in our survey of these standards in BCR-2 (note significantly higher variation among various splits).

      A set-up for processing simultaneously over a two-week period a batch of twelve samples (10 unknowns, an in-house standard, and a procedure blank) has been established and is now used routinely. The pre-ICP-MS chemical processing of a batch takes about 10 days on a part time basis, except for a full day for the chromatographic step. The ICP-MS isotope ratio measurements are done over 1.5 days, followed by data reduction. It involves spiking with enriched isotopes and a HF-HNO3 acid dissolution method using screw cap Teflon beakers for basalts, and Parr bombs for picrites or ultramafic rocks. The use of Carius tubes was found unnecessary for basalt glasses as evident from the excellent reproducibility of our in-house basaltic standard shown in Table 1. This step is followed by anion exchange chromatography for separating Cd from the PGE as a group and removing the bulk basaltic matrix, using a three-step aliquot collection for subsequent ICP-MS analysis. As indicated, the method is modified after Pearson and Woodland [2000], with attention to minimizing blanks, and eliminating potential interfering masses during ICP-MS analysis (e.g. 177Hf16O on 193Ir and 108Cd on 108Pd) (a typical elution curve is given in Fig 3. The following Oak Ridge enriched spikes are used: 99Ru, 105Pd, 106Cd, 109Ag, 185Re, 191Ir, and 198Pt. The following isotope ratios are measured in the ICP/MS step: 99Ru/101Ru, 105Pd/108Pd, 185Re/187Re, 191Ir/193Ir, 198Pt/195Pt, 109Ag/107Ag, 106Cd/111Cd. All reagents are purified in-house by redistilling 4 times.

Back to ICPMS Home