Pb-isotope analyses of USGS reference materials

Conventional corrections for thermal ionisation mass spectrometer (TIMS) induced Pb-isotopic fractionation often result in a loss of accuracy because the commonly employed pure Pb reference materials (NIST SRM 981, 982) frequently exhibit markedly different fractionation behaviour to real geological samples. As a result, these SRMs are inappropriate for comparison and/or correction of inter-operator/laboratory biases. A "matrix-matched" reference material would be preferable but, as yet, no systematic study has approached this problem. Here we present high quality Pb-isotope ratio determinations, obtained using a double spike procedure, for six USGS reference materials in an attempt to address this deficiency. Our data suggest that most of these rocks could be used as isotopic reference materials comparable, in terms of uncertainty, to the NIST SRMs. However, significant differences in isotopic composition exist between first (e.g. BCR-1) and second (e.g. BCR-2) generation samples. The cause of these differences remains unclear but has significant implications for their use as trace element reference materials.