We report experiments designed to help optimize accelerator mass spectrometry (AMS) for 10Be. In many challenging geochronologic applications, the precision of AMS is restricted by Poisson-distributed counting statistics for 10Be, which scale inversely with the square root of the total number of collected ions. For samples with low cosmogenic nuclide abundance, AMS precision is inherently low, even with long counting times. At LLNL CAMS, BeO is mixed with a Nb powder matrix, which enhances ion beam currents. This method has improved sensitivity at CAMS for applications such as measuring the erosion rate of very unstable eroding terrain and burial dating of shielded samples.
We systematically investigated the matrix-induced enhancement by monitoring ion beam currents as a function of matrix composition, matrix concentration, and sample position in the ion source. By testing ion beam currents for BeO prepared in a variety of matrices, we found that (1) the effectiveness of Nb diminishes at higher Nb:BeO mole ratios, (2) other matrices, such as V, W, and Mo, behave comparably to Nb, (3) and Ta provides the best counting efficiency (ions detected per Be atom loaded) in dilute samples. In addition, we observed an inverse linear correlation between matrix electron affinity and AMS integrated (300 s) anion beam current per mass of analyteAMS counting efficiency. This correlation suggests that the propensity of the matrix to competitively attach electrons significantly impacts ion beam current.
In order to explore practical implications of this work, we have studied the effect of common oxide impurities, serving as potential electron sinks, which may exist in BeO samples separated from quartz digests (e.g. TiO2 and Fe2O3). We find that TiO2 (e.a. = 1.59 eV) does not significantly affect ion beam currents, possibly on account of its low electron affinity compared to BeO (e.a. = 2.10 eV). Conversely, Fe2O3 (e.a. = 3.16) exhibits a retarding effect on beam currents, presumably due to its relatively high electron affinity. These observations are substantiated by additional experiments, in which we have measured the effect of an artificial, electron absorbing impurity, Au (e.a. = 2.31), which significantly decreases ion beam currents, albeit to a lesser extent than Fe2O3. The process inferences which these data permit are important for improving analytical sensitivity in geochronologic applications for which sample sizes are limited or 10Be activity is very low.
See more of Poster Session I
See more of The 10th International Conference on Accelerator Mass Spectrometry (September 5-10, 2005)