Metal matrices to optimize ion beam currents for accelerator mass spectrometry

• Published in: Nuclear instruments & methods in physics research. Section B, Beam interactions with materials and atoms Vol. 243; no. 1; pp. 216 - 222
• Main Authors: Hunt, A.L., Petrucci, G.A., Bierman, P.R., Finkel, R.C.
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01.01.2006
• Subjects: 10Be; Accelerator mass spectrometry; Ion source; Accelerator mass spectrometry; 07.77.K; 10Be; Ion source
• ISSN: 0168-583X; 1872-9584
• DOI: 10.1016/j.nimb.2005.07.220

We report experiments designed to help optimize accelerator mass spectrometry (AMS) of 10Be (in the form of BeO) for geochronologic and geomorphologic applications. In many applications, the precision of AMS is restricted by counting statistics for 10Be, which are in turn limited by the intensity of BeO− beam currents. We show that ion beam currents depend strongly on the metal matrix in which BeO is dispersed, on the matrix:BeO ratio, and for some metals, such as Ag, on the depth to which the sample is packed in the AMS cathode. Typical instantaneous Be3+ currents (μA) produced by the LLNL CAMS Cs sputter ion source and measured in a Faraday cup after the accelerator are 7.6 for samples in Ag, 19.9 in Ta, 20.9 in Mo, 21.1 in W, 25.5 in Nb and 27.5 in V. The AMS counting efficiency (ions detected per Be atom loaded) for a routine analysis time (300s) for equimolar mixtures of BeO and matrix is in the range 2×10−4–6×10−4 in the order V>Ta>Mo>W>Nb>Ag. Additionally, an inverse linear correlation between electron affinity of the matrix and beam current is observed, suggesting that the propensity of the metal matrix to attach electrons impacts the ion signal. These data and the process inferences they allow will provide for significant improvements in detection limits for a widely applied geological dating and tracing technique.