Properly combining highly siderophile element (HSE: Re, Pd, Pt, Ru, Ir, Os) abundance data, obtained by isotope dilution, with corresponding 187Os/188Os and 186Os/188Os measurements of rocks requires efficient digestion of finely‐ground powders and complete spike‐sample equilibration. Yet, because of the nature of commonly used methods for separating Os from a rock matrix, hydrofluoric acid (HF) is typically not used in such digestions. Consequently, some silicates are not completely dissolved, and HSE residing within these silicates may not be fully accessed. Consistent with this, some recent studies of basaltic reference materials (RMs) have concluded that an HF‐desilicification procedure is required to fully access the HSE (Ishikawa et al. (2014) Chemical Geology, 384, 27–46; Li et al. (2015) Geostandards and Geoanalytical Research, 39, 17–30). Highly siderophile element abundance and Os isotope studies of intraplate basalts typically target samples with a range of MgO contents (< 8 to > 18% m/m, or as mass fractions, < 8 to > 18 g per 100 g), in contrast to the lower MgO mass fractions (< 10 g per 100 g) of basalt and diabase RMs (i.e., BIR‐1, BHVO‐2, TDB‐1). To investigate the effect of HF‐desilicification on intraplate basalts, experiments were performed on finely ground Azores basalts (8.1–17 g per 100 g MgO) using a ‘standard acid digestion’ (2:1 mixture of concentrated HNO3 and HCl), and a standard acid digestion, followed by HF‐desilicification. No systematic trends in HSE abundances were observed between data obtained by standard acid digestion and HF‐desilicification. Desilicification procedures using HF do not improve liberation of the HSE from Azores basalts, or some RMs (e.g., WPR‐1). We conclude that HF‐desilicification procedures are useful for obtaining total HSE contents of some young lavas, but this type of procedure is not recommended for studies where Re‐Pt‐Os chronological information is desired. The collateral effect of a standard acid digestion to liberate Os, followed by HF‐desilicification to obtain Re and Pt abundances in samples, is that the measured Re/Os and Pt/Os may not correspond with measured 187Os/188Os or 186Os/188Os.
Read full abstract