Abstract
A study was conducted to determine how variables during metallothermic purification of chloride salts affect the oxygen content of the final product. 100 g batches of ternary MgCl(2)-NaCl-KCl mixture (50-20-30 wt %) were purified for 3 h with flowing sparge gas and 0.25 wt % Mg addition as a reducing agent. After slag removal and remelting, oxygen content was measured via inert gas fusion analysis (IGFA). The as-received, melted, and baseline-purified salts had oxygen concentrations of 3466 ± 513, 889 ± 135, and 436 ± 151 ppm, respectively, with purification showing statistical significance (>99% confidence). A parametric study with 12 trials identified processing time as the most influential factor, as extending it to 6 h further reduced oxygen concentration to 143 ppm (>99% confidence). Variations in sparge and flush gas had no significant effect, and all tests used excess Mg. Uncertainty analysis of the IGFA technique estimated a 9% measurement uncertainty, primarily due to calibration standard variability.