METHODS

Hardground crusts were cut with a rock saw to obtain thick slides and thin sections. Iron and Mn elemental maps were obtained using an electron microprobe, and ⁹Be, Nd, Co, Ni, Mn, Fe, and Cu concentrations were measured by inductively coupled plasma–mass spectrometry (ICP-MS) at the Institute for Inorganic Chemistry of ETH (Switzerland). ¹⁰Be concentrations were measured at the accelerator mass spectrometry (AMS) facility of the Paul Scherrer Institute and the Institute for Particle Physics of ETH following standard procedures. Chemical preparation of samples for the ¹⁰Be AMS measurements closely followed a previously described method (Henken-Mellies et al., 1990). ¹⁰Be concentrations were normalized to internal standard S555 with a nominal ¹⁰Be/⁹Be ratio of 95.5 x 10^–12. Age calculations were carried out by applying a ¹⁰Be radioactive half-life of 1.5 m.y. Nd isotope compositions were determined using a Nu Plasma multicollector ICP-MS following the chemical separation procedure of Cohen et al. (1988). ¹⁴³Nd/¹⁴⁴Nd was normalized to ¹⁴⁶Nd/¹⁴⁴Nd = 0.7219 to correct for instrumental mass bias. Age-corrected _Nd(T) values were calculated using ¹⁴⁷Sm/¹⁴⁴Nd = 0.115. The errors shown on the figures are 2 external reproducibilities based on repeated measurements of the JMC-Nd standard. The 2 external precision for the Nd isotope measurements of different sessions varied between 18 and 41 ppm. All ratios presented are normalized to given standard values: for JMC-Nd a given ¹⁴³Nd/¹⁴⁴Nd ratio of 0.511833 cross-calibrated to the La Jolla standard (0.511858) was used. Run precision for each sample was better than the external reproducibility. Nd isotope ratios are expressed as _Nd values, which denote the deviation of the measured ¹⁴³Nd/¹⁴⁴Nd ratio from the chondritic uniform reservoir CHUR (0.512638) multiplied by 10,000.