To gauge our accuracy and precision for the analytical method, at least one sample of an "in-house standard" was included in each run. This in-house standard, "ODP Site 1098 Consistency Standard," is a composite sample of Site 1098, a finely-ground, diatom-rich sediment from the Palmer Deep (provided by Linda Anderson at the University of California, Santa Cruz). The lithology of Site 1098 ranges from diatom ooze to diatom-bearing silts and clays (Barker, Camerlenghi, Acton, et al., 1999). Core descriptions (Barker, Camerlenghi, Acton, et al., 1999) indicate that Site 1098 sediments are essentially a two-component mixture of diatoms and siliciclastics (primarily quartz and clays). The volcanic glass component, based on 100 smear slide analyses, is zero (see "Results for Site Survey Samples"). Summary statistics for repeated runs of the in-house standard are given in Table T1 and show no significant difference in the SiO2 weight percent content of the Consistency Standard for the two extractions (Na2CO3 and KOH). From these results, it is suggested that the harsher KOH treatment, relative to the Na2CO3 digestion, does not dissolve significant amounts of SiO2 from the nonbiogenous sources, which are primarily clay minerals and quartz.
A second question regarding the analytical method is whether or not digestion proceeds to completion. Several parameters may affect the efficiency of digestion, including digestion time, bath temperature, molar strength of reagent, and the weighed sample mass to reagent volume ratio. One indicator of incomplete digestion, aside from visual proof, would be a positive correlation between the amount of digested SiO2 as a function of sample mass:reagent volume. First, no visual proof of undigested fossils was found in the recovered residue after either of the alkaline extractions. Second, there is no correlation between the amount of digested silica as a function of the weighed sample mass to reagent volume ratio (Fig. F2). Digestion times of 8.5-10 hr were determined based on the offsets in the amount of SiO2 leached from the Consistency Standard when digestion times were <8 hr. The average SiO2 values increased for both extractions and corresponded to a lower standard error when digestion times were increased to 8.5-10 hr. Although this digestion time at 85°C should produce good results (complete extraction) for samples containing up to 31 wt% SiO2, several factors, as discussed below, will affect the outcome of unknown samples.