Validation of LA-ICP-MS trace element analysis of geological glasses using a new solid-state 193 nm Nd:YAG laser and matrix-matched calibration

We have investigated the performance of a combined single-collector sector-field ICP-MS and a new solid-state 193 nm Nd:YAG laser ablation system for trace element analysis of geological glasses. Experiments were made with certified MPI-DING, USGS and NIST reference glasses. Two different measurement protocols were used: the conventional three-spot analysis ( spot sizes: 50 - 100 mm) for the determination of 26 trace elements using a combination of magnetic ( Bscan) and electrical scan ( Escan) modes and the major element Ca as internal standard ( protocol 1), and a high-precision technique for the determination of three elements using the fast Escan at a constant magnetic field and a trace element as internal standard ( protocol 2). We obtained a precision ( 1 RSD) of element ratio measurements of about 0.8% ( protocol 1) and 0.2% ( protocol 2), which is better than that obtained using a 213 nm Nd: YAG laser ( about 1.1% and 0.6%, respectively). The mean reproducibility ( 1 RSD) of trace element data obtained over a period of three months was about 2% ( 100 mm spot size) and 3% ( 50 mm spot size). The detection efficiencies of the mass spectrometer for elements with Z = 37 ( Rb) - 92 ( U) range between 0.2 x 10(-3) and 2 x 10(-3) detected ions/ablated atoms. They are about 40% higher than those using a 213 nm laser because of lower sample consumption of the 193 nm laser ablation system ( ca. 0.05 mm ablation per shot). Detection limits decrease with increasing atomic number; they range between 0.1 mu g g(-1) ( Rb) and 0.0003 mg g(-1) ( U). We also investigated element fractionation and possible matrix effects, which may bias the analytical results. Element fractionation for the 193 nm laser is lower than for the 213 nm laser. Low ( percent range), but significantly different fractionation factors of some elements ( e. g., Rb, Y, Pb) for glasses having different major element compositions are observed. Highly accurate LA-ICP-MS analyses therefore require suitable reference materials of similar matrices. The eight MPI-DING glasses having different major element compositions seem to be such samples for geoanalytical work. The basaltic KL2-G glass and the komatiitic GOR132-G glass were used for calibration of six basaltic and one komatiitic reference glasses, respectively. Most LA-ICP-MS data agree with published compilation values and isotope dilution data within 95% confidence limits.