Change in peak temperature at low pressure vs. peak temperature at 1000 mbar for the minerals listed in Table
(Ep = epsomite, Mo = montmorillonite, Go = goethite, Ki = kieserite, Ka = kaolinite, Al = alunite, Ma = magnesite, Ca (1) = <150 μm calcite, Ca (2) = 3 mm calcite). For example, the peak temperature of magnesite at 1000 mbar (x axis) is 629°C. The change in peak temperature (or ΔT, plotted on the y axis) is -13, -23, and -34°C at 150, 30, and 12 mbar, respectively. The difference in the Ca(1) and Ca(2) decomposition temperatures is due to particle size effects, discussed in detail in the main text. Ep, Ki, Mo, and Al appear multiple times in the plot because they decompose in multiple gas-releasing steps (e.g. epsomite releases water at low temperature and SO2 from sulfate decomposition at higher temperature). This figure shows that, although there might be a small trend in increasing ΔT with peak temperature at 1000 mbar, the scatter in the data is large, making prediction based on these data difficult. Furthermore, the effects of sample particle size are as large as the effects of pressure, showing that a pressure-only model is not sufficient to predict the change in decomposition temperatures. Samples were run using a 3 sccm He carrier gas flow and a ramp rate of 35°C/minute.