Optimization of cosmogenic 10Be and 26Al extraction for precise AMS measurements of low concentrations

Abstract

Both burial dating and recently introduced isochron-burial dating require accelerator mass spectrometry (AMS) analysis of low cosmogenic nuclide concentrations with low backgrounds and low uncertainties for more precise ages. Therefore, the aim of this study is to optimize the extraction of 10Be and 26Al from quartz for the AMS measurements in the range of ca. 5000 10Be at.g−1 and 50000 26Al at.g−1 in 50 g of purified quartz. The latter is more feasible when total Al concentrations are ca. 20 ppm or less, with <10% uncertainty. To do this, we modified our sample preparation protocol. In order to reduce total Al concentrations, we added a treatment step with orthophosphoric acid. ICP-OES analysis of aliquots from different leaching steps yielded a dramatic decrease in total Al concentrations, e.g. from ca. 3000 down to ca. 10 ppm. To increase 27Al currents, Al was mixed with FeO, Cu, Nb, and Ti in two different molar mixing ratios (1:5 and 1:10). At the ETH Tandy AMS, the low energy current of two cathodes for each mixing ratio and metal was measured for >2000s. The best performance was found for Al plus Cu at a molar mixing ratio of 1:5. 10Be/9Be and 26Al/27Al measurements of several low concentration samples yielded ratios in the mid 10−14 for Beryllium and low 10−13 for Aluminum within less than 10% uncertainties, thus, 10Be and 26Al concentrations of ca. 5000 ± 500 at.g−1 and ca. 50000 ± 5000 at.g−1, respectively. Additional tests are currently being carried out to further improve the performance of BeO samples containing very low amounts of 9Be carrier. The implications of our study are : (1) very young surface exposures, on the order of few hundred years, can be dated with few tens of years of uncertainty ; and (2) and isochron-burial dating of ca 100 ka old sediments is well possible.