Trace element analysis of olivine is challenging but vital, since trace elements in olivine can potentially record crystallization conditions and, correspondingly, conditions of magma generation, evolution, and transport dynamics. Electron microprobe analysis (EPMA) is one of the most widely used methods for trace element analysis in olivine for many reasons including but not limited to high-spatial resolution and high precision. EPMA protocols with high-beam currents and long-counting times provide accurate trace element concentrations with low-detection limits. However, these protocols neglect the effect of secondary fluorescence (SF) across phase boundaries, which may lead to fictitious analytical results when analyses are performed near other phases. In these cases, obtaining accurate EPMA data requires correction of SF effects from the adjacent or surrounding phase. In this study, the contents of olivine trace elements Ca, Ni, Mn, Ti, Cr, and Al have been measured in run products of experimental high-temperature studies on mafic magmatic systems consisting of olivine grains embedded in basaltic glass. Olivine grain fragments with known composition, were embedded in basaltic powder, and heated to 1250 °C in 1-atm vertical gas-mixing furnace for a few minutes and quenched. EPMA analyses of the olivine grains before and after the experiments were used to quantify the effect of SF, which was found to be most significant for Ca and Ti, and to a lesser extend for Al, and to depend on the distance from the interface, olivine grain geometry, accelerating voltage, and olivine and glass compositions. By combining our experimental results with systematic SF calculations performed with the Monte Carlo simulation program PENEPMA, analytical expressions that give the apparent Ca, Ti, and Al concentrations due to SF in equant olivine grains embedded in basaltic glass were derived. These equations, which allow for a rapid correction of SF effects, are used to assess the average overestimation due to SF of olivine analyses included in current experimental databases, which might be the source of major discrepancies observed in petrological tools utilizing trace element contents in olivine.
Read full abstract