A DOUBLE-OS SPIKE FOR MOLYBDENITE GEOCHRONOLOGY

Over the past several years the molybdenite Re-Os chronometer has become an increasingly useful tool for geochemists.   Because common Os is absent in the mineral, simple isotope dilution determination of the abundances of radiogenic ¹⁸⁷Os and Re permit calculation of sample age.  On the other hand, the mono-isotopic composition of the Os in molybdenite precludes direct application of a fractionation correction to mass-spectrometer measurements of spike-sample mixtures of molybdenite and mono-isotopic Os spikes.  We have developed an osmium spike composed of two isotopes that now allows us to apply a fractionation correction to the Os measurements.

The double Os spike is made from enriched ¹⁸⁸Os and ¹⁹⁰Os in the proportion of about 8:1, as originally proposed by Morgan (1999).  Sample spiking, digestion, equilibration, and chemical separations are carried out in the same manner as for any other spike (mono-isotopic or common Os).  We measure the following four masses of the Os mixture: ¹⁸⁷Os, ¹⁸⁸Os, ¹⁹⁰Os, and ¹⁹²Os.  Radiogenic ¹⁸⁷Os from the sample is determined from the ¹⁸⁷Os/¹⁸⁸Os; the fractionation factor is obtained from the ¹⁹⁰Os/¹⁸⁸Os; and the common Os is determined from the ¹⁹⁰Os/¹⁹²Os.  With the ¹⁹⁰Os peak significantly smaller than the ¹⁸⁸Os peak, a more precise determination of very small amounts of sample common Os is possible.  Correcting for common Os may be important for geologically accurate ages on low-abundance samples.

We have found this spike to be particularly useful for analysis of samples with low ¹⁸⁷Os abundances, that is, samples that are very young or those with very low Re abundances (LLHR samples; Stein et al. 2000).  For such samples we observe age reproducibilities well under one permil.  Although the double-Os spike is also convenient for analysis of higher-abundance samples, replicate analyses of our internal molybdenite standard show only marginal improvement in reproducibility over our results using a single Os spike. This could indicate that mass fractionation is either minimal or constant for higher-abundance samples, for which we can use Faraday cup detectors and relatively low filament temperatures.  In any case, mass fractionation in Re measurements is not addressed by NTIMS, and is probably the source of small residual errors.