Samples analyzed in this study were splits taken from XRD and XRF samples that were analyzed on board during Leg 178. The initial XRD and XRF samples, originally ~10 g, were crushed to a fine gravel, and ~1 g of this gravel was set aside for possible INAA analyses. The remainder was powdered in an agate mortar, washed in nanopure water to remove salts, sonicated for 30 to 40 min, and centrifuged. The centrifuged sediments were dried for at least 12 hr at 120°C and powdered in a tungsten carbide mixer mill. An ~1-g aliquot was ignited for 10 hr at 950°C to determine loss on ignition (LOI). This sample was retained for onshore major element analyses. Thus, the INAA samples are distinct from the major element samples and were only minimally processed on board to ensure that trace element chemistry was not affected.
Once ashore, the INAA samples were ground to a fine powder in a high-purity alumina mortar. In a series of three irradiations, 39 samples were analyzed along with U.S. Geological Survey (USGS) geochemical standards BHVO-1 and G-2 (Govindaraju, 1989). Powders weighing ~200 mg were irradiated in polyethylene vials with standard flux monitors at the University of California, Irvine, Research Reactor for 4 hr at a neutron flux of 1.8 × 1012 n/cm2/s. Samples were counted on high-purity, intrinsic Ge detectors at ~1 week and 4 weeks following irradiation, and gamma-ray spectra were used to determine the concentrations of Sc, Cr, Fe, Co, Zn, As, Rb, Sr, Sb, Cs, Ba, La, Ce, Sm, Eu, Tb, Yb, Lu, Hf, Ta, and Th.
The powders taken from shipboard XRF and LOI were ground again in a high-purity alumina mortar. About 20 mg of sample was fused into a glass bead at temperatures of ~1700°C, using a Mo strip heater (Brown, 1977) under an Ar atmosphere. Glass beads were mounted in epoxy, polished, and analyzed by electron microprobe. Analyses were performed at 11-nA beam current and 15-keV accelerating voltage on a CAMEBAX electron microprobe. Standardization was obtained by using a set of Smithsonian Institution mineral standards (Jarosewich et al., 1980). In some samples, fusion was not complete and quartz crystals were observed in backscatter electron images. These samples were ground further and fused again until complete melting was achieved. Analyses were performed by scanning the electron beam across a box 20 µm in diameter. Typically, eight analyses were performed on each sample and these were spaced out over much of the exposed surface of the glass bead. The multiple analyses were averaged to obtain the bulk chemical composition.