A refined electron probe microanalysis protocol for accurate quantification of F and Cl in mafic silicate glasses

Abstract

Accurate quantification of volatile elements (F and Cl) in mafic silicate glasses is critical for understanding magmatic processes; however, it remains analytically challenging due to spectral interference (e.g., Fe Lα on F Kα) and matrix effects during overlap correction. We develop a refined electron probe microanalysis (EPMA) protocol based on an established single-standard overlapping-peak-correction procedure. This method is applicable across a range of beam currents (60–300 nA), eliminating the need for a multi-standard calibration curve, which typically requires the same beam current as the F measurement in unknown samples. Systematic evaluation using F-free glass standards (G5, G6, and G7) with varying Fe contents demonstrates that calibrated F concentrations in BHVO-2G remain consistent within uncertainty across different beam currents and standards, agreeing with previous studies. Matrix effects from different overlap standards cannot be completely eliminated by the single-standard correction procedure. Consequently, F concentrations determined using different overlap standards clustered into five distinct ranges corresponding to standard types: (1) silicate glasses, (2) garnet, (3) olivine, pyroxene, and spinel, (4) iron oxides and chalcopyrite, and (5) pyrite and pure iron metal. Calibrated values show insignificant variation with the iron valence state in overlap standards. Matrix effects are the dominant factor in the calibration of Fe Lα–F Kα spectral interference. Under optimized analytical conditions (15 kV, 300 nA, 15 μm, and PC0 crystal), detection limits are 46 ppm for F and 12 ppm for Cl. The method's accuracy was validated by comparing the test results of international glass standards with literature data.