Both shipboard and shore-based samples were measured with Jobin-Yvon (JY) ICP-AES instruments. The JY2000 instrument on the JOIDES Resolution employs a monochrometer, enabling the sequential measurement of individual wavelengths while rapidly scanning through the spectrum. The JY ULTIMA-C instrument at Boston University has the capability to measure both individual and multiple wavelengths simultaneously, as it employs a monochrometer as well as a polychrometer. Thus, the shore-based protocol measured most major elements (Si, Al, Ti, Fe, Mn, Ca, and Mg) and the trace element Sr on the polychrometer with other major and trace elements (Na, K, Ba, Cr, Ni, Sc, V, and Zr) on the monochrometer. The same wavelengths for each element were measured on both instruments. For the monochrometers, the optical focal length on the Boston University instrument (1 m) is also longer than that on the JOIDES Resolution (0.6 m), allowing better resolution of spectra, although this has not shown to be a handicap for the shipboard instrument.

Quality Assurance and Quality Control

The standard reference materials (SRMs) used in calibrating each ICP-AES were slightly different between the shipboard and shore-based protocols. The shipboard SRMs included BCSS-1 (marine mud), BHVO-2 (Hawaiian basalt), JCh-1 (Japanese chert), JLs-1 (Japanese limestone), NIST-1c (argillaceous limestone), and MAG-1 (marine mud) for a six-point calibration curve, with SCo-1 (Cody shale) being measured as an unknown item to check accuracy and consistency between analytical runs (Shipboard Scientific Party, 2003). The shore-based SRMs consisted of MAG-1, SCo-1, JA-2 (Japanese andesite), and BCSS-1 as a four-point calibration curve for the upper terrigenous-rich sequence. For the carbonate-rich samples we used a four-point calibration curve consisting of JLs-1, MAG-1, and two gravimetrically mixed standards of 80% JLs–20% MAG-1, and 60% JLs–40% MAG-1. Constructing these mixed standards enabled a better matrix match and a more appropriate concentration range for the calibration. At times JCh-1 was used in the calibration to also help constrain the silica-rich samples. NIST-1c was used as an accuracy check and was within one standard deviation of the published uncertainty (Table T1). Furthermore, a composite sample from Site 1256 (termed "EEP") was used at Boston University to monitor consistency between runs.

The analytical uncertainties, presented as precision, for the shore-based and shipboard procedures (Table T2) quantify the sum of the uncertainties due to both instrument variability and the sample preparation process. Precision was calculated by percent of variation associated with a random sample that had been prepared and measured separately three times. Comparing shipboard flux fusions with flux fusions performed at Boston University, the precision improved significantly, with Boston University being better than 3% of the respectively measured values for Si, Al, Ti, Fe, Mg, and Ba. Elements that were not reported during the cruise (P and Zr) as a result of poor uncertainty (>30% of the measured values) are both within 4% of the measured value from shore-based analysis. The uncertainty of the alkali elements (Na and K) shipboard is >20% of the measured values, whereas shore-based precisions are better than 3% of the measured values. Some of the high shipboard uncertainty may be caused by the LOI procedure that was performed only at sea, as well as by other reasons such as the physical motion of the ship (see discussion in Quintin et al., 2002). Trace elements such as Cr, Ni, and V are often difficult to measure in carbonates, due to their low concentrations, and thus procedures on board the ship probably approached the procedural detection limit. However, at Boston University, we adjusted the calibration curve to suit carbonate-rich sediments and prepared the samples in a cleaner and more stable environment than aboard the JOIDES Resolution.

Importantly, the difference between flux fusions at Boston University and microwave-assisted acid digestions is minor. The microwave typically yielded precision better than 3% of the measured values for all elements, including Cr, Ni, V, and Zr. As will be discussed below, this agreement between the flux fusion and the acid digestion, even for elements that are traditionally difficult to dissolve, demonstrates the ability of using microwave-assisted digesting systems to prepare samples for ICP analyses.