Approximately 35 samples were randomly selected from throughout Site 1256 and were prepared at Boston University by both the flux fusion method and the microwave-assisted acid digestion method. All samples were measured by ICP-AES for major elements (Si, Al, Ti, Fe, Mn, Ca, Mg, Na, K, and P) and selected trace elements (Ba, Sr, Cr, Ni, Sc, V, and Zr). As mentioned earlier, both the flux fusion and acid digestion analyses were calibrated with the same standard reference materials.
The analyses from each preparation are in good agreement (Fig. F2; Table T3). For each element, the microwave-assisted acid digestions show a complete digestion as the concentrations fall on or very near to the 1:1 line with the concentrations produced by flux fusions. Importantly, the refractory elements, such as Ti and Zr, also fall on the 1:1 line, signifying that the acid digestions can achieve a complete digestion when using HF in a pressurized microwave for an extensive period of time, which allows for ample conditions to dissolve refractory minerals in these sediments.
Boric acid was added in the microwave-assisted digestion method to prevent the formation of insoluble fluorides and to inhibit the volatilization of Si. A robust 1:1 relationship between flux fusions and microwave-assisted acid digestions for Ca and Al also demonstrates the usefulness of the boric acid and the inhibition of insoluble fluorides. If fluoride precipitation occurred, Ca and Al concentrations are likely to have fallen below the 1:1 line. Moreover, the 1:1 relationship for Si between shipboard flux and shore-based acid digestions is also tight (r2 = 0.982), indicating that the small amount of Si lost by volatilization (even in the presence of the boric acid, and evidenced by the slope of 0.936) is predictable and minor in extent. Overall, the descriptive statistics in Table T4 indicate a strong correlation within the analytical uncertainties.
While most elements show excellent correlation and regression slopes very near unity, Ni and Cr, however, present correlation coefficients better than 0.95 yet yield regression slope values 10%–15% smaller than unity (Table T4). The variability observed with Ni and Cr potentially reflects the variability in the flux fusions, as these elements are less precisely determined by flux methods at these low concentrations than by acid digestion (Table T1).
In summary, the comparison between flux fusions and microwave-assisted acid digestions demonstrates that with the addition of boric acid
These results indicate the usefulness and practicality of using microwave-assisted acid digestions in future IODP laboratories.