No 182W excess in the Ontong Java Plateau source

Small-scale W isotope variations in ancient and modern terrestrial rocks provide insights into Earth's accretion and early differentiation history as well as the long-term evolution of the Earth's mantle. Tungsten isotope studies on such rocks have exploited advances in mass spectrometry, both NTIMS and MC-ICPMS, which now permit the determination of W isotope ratios at unprecedented precision. While W isotope studies performed in different labs by MC-ICPMS and NTIMS generally exhibit excellent agreement, obtaining accurate W isotope data at this level of precision remains analytically challenging. For example, a recent NTIMS study reported a relatively large, +24 ppm excess in 182W/184W for a Phanerozoic sample from the Ontong Java Plateau (OJP), but no such 182W/184W anomaly was found in another study by MC-ICPMS. The present study aims to resolve the discrepancy between these two previous studies, and more generally to evaluate the agreement between different recent W isotope studies by NTIMS and MC-ICPMS. To this end, we report new W isotope data for OJP drill core samples obtained by MC-ICPMS. The OJP samples analyzed here exhibit no resolvable 182W/184W excess relative to the standards and most terrestrial rocks. Moreover, the OJP samples as well as the terrestrial rock standards analyzed here exhibit small but variable W isotope variations for ratios involving 183W, producing coupled variations in both 'radiogenic' (i.e., 182W/183W) and 'non-radiogenic' (i.e., 183W/184W) ratios. These W isotope variations are analytical in origin, induced during sample preparation, and very likely caused by a nuclear field shift isotope fractionation affecting primarily the odd isotope (183W). The recently reported 182W excess for an OJP sample may result from this nuclear field shift effect, as the NTIMS analyses had to rely on a double normalization involving the 183W/184W ratio. More generally, these results demonstrate that using 183W data from any MC-ICPMS or NTIMS study requires a careful quantification of any potential analytical 183W effect. Nevertheless, once such effects are taken into account, then both 182W/184W and 183W/184W can accurately be determined to a very high level of precision.