The determination of redox states of mantle-derived primary magmas and derivative liquids has been the subject of numerous studies performed over the last decades and is still highly demanding. Thereby, the main information on magma redox state is obtained from the analysis of iron speciation in pillow glasses or in minerals. Amongst all analytical methods available to determine Fe3+/Fe2+ ratios, the EPMA-based flank method provides accurate and high spatial resolution analyses at low cost. However, the routine use of this method is still limited by the need of well-calibrated standards and its lower precision compared to XANES or Mössbauer spectroscopy.

In this contribution, we present further improvements of the flank method approach using a JEOL JXA-iHP200F EPMA. Therefore, we synthesized a set of new standard glasses for calibration ranging from basaltic to andesitic compositions at different fO₂ conditions to cover an Fe2+/∑Fe range from ~0 to 1. We found that the most beam-sensitive glasses in our sample set, namely hydrous andesitic glasses, remain unaltered during the measurement up to an electron beam energy density of 0.23 mJ*µm⁻². To optimize analyses, we recommend measuring basaltic and andesitic glasses under static beam conditions at 80 nA and 20 µm with counting times of 60 s. Additionally, the newly available large TAP diffraction crystals (TAPL) significantly enhance intensity yield, by approximately 2–3 times compared to a standard TAP crystal, resulting in a notable improvement of analytical precision. Ultimately, we applied our optimized flank method approach to natural boninite and fore-arc basaltic pillow glasses from the Izu-Bonin-Mariana arc (collected by IODP Expedition 352) and compare the results with literature data previously obtained by XANES on the same samples and could show that our newly developed approach is working reliably.