6533b829fe1ef96bd12897cc

RESEARCH PRODUCT

Investigation of matrix effects in 193nm laser ablation-inductively coupled plasma-mass spectrometry analysis using reference glasses of different transparencies

subject

description

Abstract The degree of transparency of glasses, which depends on the Fe content, may influence the ablation behavior during laser ablation-inductively coupled plasma-mass spectrometry (LA-ICP-MS) analysis. To test possible matrix effects when using a 193 nm Nd:YAG laser, we have analyzed transparent and opaque NIST, BAM and USGS reference glasses. These reference materials are ideal for such investigations, because they are well characterized, most elements are homogeneously distributed at the micrometer scale, and their Fe content varies over a very large range, from 16 to 130,000 μg g− 1. Our measurements show that the fractionation factors of refractory and volatile lithophile elements, such as Sr, Ba, and Rb, are 1.00 ± 0.03 and independent of the degree of transparency. However, for volatile chalcophile/siderophile elements (e.g., Zn and Pb) the fractionation factors vary significantly between 0.7 and 1, depending on the spot sizes and the transparency of the material. Mass-load-induced matrix effects may also influence the accuracy of LA-ICP-MS analysis. They are less than 2% for the lithophile and up to 10% for volatile chalcophile/siderophile elements when the mass load varies by a factor 2.4. Relative sensitivity factors used for calibration of lithophile elements agree within uncertainty limits for transparent and opaque glasses when using a 193 nm laser. Even for volatile/chalcophile elements they differ only by 5–10%. The reliability of the LA-ICP-MS analyses is demonstrated by presenting concentration data of 27 trace elements in the NIST, BAM and USGS reference glasses using NIST SRM 612 for calibration, where highly accurate reference values are available. For trace element concentrations in the range between 1 and 500 μg g− 1, the reproducibility and the uncertainties at the 95% confidence level of the measurements vary between 1–4%, and 7–10%, respectively.