Analysis of geological glasses by electron probe microanalysis under low beam current density conditions

Abstract

Precise and accurate determination of major components in alkali silicate glasses provides critical information concerning element distribution, petrogenesis and magmatic evolution. However, as one of the most important in situ microanalytical technologies, the application of electron probe microanalysis (EPMA) was limited by the time-dependent loss of X-ray intensity of alkali ions (especially Na) during electron beam bombardment on alkali silicate glasses. In this study, an optimized analytical method was developed using a JEOL JXA 8100 EPMA, aiming to get a long enough incubation duration suitable for accurate analysis of major element mass fractions. The optimal analytical conditions were determined as follows: 15 kV accelerating voltage, 2 nA beam current, 20 μm beam diameter, 40 s peak counting time, Na and K analyzed with top priority, time-saving mode with no additional detector movement and no peak search functioned. The availability of this low beam current density condition (6.37 × 10⁻³ nA μm⁻²) to 24 anhydrous reference glasses, with their compositions covering the entire spectrum from the ultramafic to highly siliceous range, was verified by the good agreement of the EPMA measured data with their GeoReM preferred values. Similarly, the methodology presented here is also suitable for analysis of hydrous glasses with H₂O contents up to approximately 9 wt%, accompanied by no obvious Na loss found and no complicated mathematical corrections needed, and therefore can be used for the estimation of H₂O contents in H₂O-rich (>1 wt%) geological glasses by the difference method. Despite its analytical accuracy and precision (>0.5 wt%) not better than those of SIMS and FTIR, EPMA can be used as a substitute technique for H₂O determination with its main advantages of excellent spatial resolution (down to 5–10 μm) and analytical convenience, especially for small melt inclusions or tiny experimental quench products, and the case with no appropriate calibration standards or no reported extinction coefficients available for those compositions of interest. In daily application, our optimized method is low cost and easy to implement and can achieve comparable precision and detection limits with acceptable efficiency (∼4–5 min/point for ten major elements) relative to routine analysis conditions.