High precision calcium isotope analysis using 42Ca-48Ca double-spike TIMS technique

Abstract:

Double spike techniques are widely used for determining calcium isotopic compositions of natural samples. The most important factor controlling precision of the double spike technique is the choice of appropriate spike isotope pair, the composition of double spikes and the ratio of spike to sample（C_Sp/C_N）. We propose an optimal ⁴²Ca-⁴⁸Ca double spike protocol which yields the best internal precision for calcium isotopic composition determinations among all kinds of spike pairs and various spike compositions and ratios of spike to sample, as predicted by linear error propagation method. It is suggested to use spike composition of ⁴²Ca/(⁴²Ca+⁴⁸Ca) = 0.44 mol/mol and C_Sp/(C_N+ C_Sp)= 0.12mol/mol because it takes both advantages of the largest mass dispersion between ⁴²Ca and ⁴⁸Ca (14%) and lowest spike cost. Spiked samples were purified by pass through homemade micro-column filled with Ca special resin. K, Ti and other interference elements were completely separated, while 100% calcium was recovered with negligible blank. Data collection includes integration time, idle time, focus and peakcenter frequency, which were all carefully designed for the highest internal precision and lowest analysis time. All beams were automatically measured in a sequence by Triton TIMS so as to eliminate difference of analytical conditions between samples and standards, and also to increase the analytical throughputs. The typical internal precision of 100 duty cycles for one beam is 0.012‒0.015 ‰ (2δ_SEM), which agrees well with the predicted internal precision of 0.0124 ‰ (2δ_SEM). Our methods improve internal precisions by a factor of 2‒10 compared to previous methods of determination of calcium isotopic compositions by double spike TIMS.

We analyzed NIST SRM 915a, NIST SRM 915b and Pacific Seawater as well as interspersed geological samples during two months. The obtained average δ^44/40Ca (all relative to NIST SRM 915a) is 0.02 ± 0.02 ‰ (n=28), 0.72±0.04 ‰ (n=10) and 1.93±0.03 ‰ (n=21) for NIST SRM 915a, NIST SRM 915b and Pacific Seawater, respectively. The long-term reproducibility is 0.10‰ (2 δ_SD), which is comparable to the best external precision of 0.04 ‰ (2 δ_SD) of previous methods, but our sample throughputs are doubled with significant reduction in amount of spike used for single samples.