SOLVENT EXTRACTION/ANION EXCHANGE SEPARATION AND DETERMINATION OF PGES (OS, IR, PT, PD, RU) AND RE-OS ISOTOPES IN GEOLOGICAL SAMPLES BY ISOTOPE DILUTION ICP-MS

Abstract

A method is described that allows the separation of Re, Os, Pt, Pd, Ir and Ru for determination by isotope dilution via inductively coupled mass spectrometry (ICP-MS) at pg levels, from the same sample aliquot. Samples are digested in Carius tubes, enabling effective attack of PGEs in a wide range of geological materials, including pure chromite separates. Os is isolated by solvent extraction so that it may be precisely analysed for isotopic ratios by ICP-MS or negative-ion thermal mass spectrometry. When Os is analysed by ICP-MS, sample memory is minimised by using a direct injection nebuliser. Following Os extraction, Re, Pt, Pd, Ir and Ru are separated from their sample matrix by anion exchange chromatography. The method thus enables PGE abundances and Os isotopic ratios to be determined on the same sample dissolution, thereby avoiding problems of sample inhomogeneity when comparing data. Furthermore, analysis of Re for the same sample dissolution makes geochronological studies possible. Total procedural blanks are routinely less than 10 pg/g for all elements except Pt (15 to 25 pg/g). Blanks of less than 1 pg/g can regularly be obtained for Ir and Os. We estimate procedural detection limits (3*STD deviation) of c. 3 pg/g for Os and Ir, 5 pg/g for Re, Ru and Pd, and 15 pg/g for Pt. Such detection limits permit analysis of most bulk rock geological samples. The method is evaluated by replicate dissolution of a komatiite reference material, WITS-1 [Tredoux, M., McDonald, I., 1996. Komatiite WITS-1, low concentration noble metal standard for the analysis of non-mineralised samples. Geostand. Newsl. 20, 267-276] together with unknowns. For WITS-1, concentrations of Os, Ir and Ru reproduce well, within a factor of 2 of that expected based on sampling theory. Potential heterogeneity for Pt, Pd and Re is identified at the 1 g sampling level used, that may be related to alteration. Replicate analyses of picritic basalts in the 1 to 10 ppb range show average 2σ reproducibilities of better than 5% for Re, Os, Ir, Pt and Pd, with Ru being better than 10%. This compares very favorably with other analytical techniques. We present data for andesites with Ir and Os at the sub 10 ppt level. As expected, reproducibility decreases as sample concentrations decrease but are sufficient at the 10 ppt level to confidently identify inter-element PGE fractionations. The analytical method works well for a wide variety of rocks types.