Experimental constraints on the role of chloride in the origin and evolution of kimberlitic magma

Abstract

graphite tube isolated from a capsule with the sample by the MgO insulator was applied as a heater. The sample powder was loaded in a Au-Pd or Pt capsule and after accurate drying welded by arc welding. The inner walls of Ptcapsules were shielded by Re foil for minimization of iron loss. Each cell contained two capsules: one with the composition of model kimberlite (this study) and another with the composition of natural kimberlite. The results of experiments with the natural composi� tion will be presented in the next paper. The tempera� ture in each run was controlled by a W97Re3-W75Re25 thermocouple located in the center of the heater and isolated by the Al2O3 insulator from it. The tempera� ture gradient in the sample measured using a two� pyroxene thermometer did not exceed 50°С at 1500°С and 5 GPa. The pressure in the sample was maintained by the calibration performed at room tem� perature by change in the electric resistance for the transitions Bi I-II (2.55 GPa), Ba I-II (5.5 GPa), and Bi III-IV (7.7 GPa) and corrected for high tempera� tures by phase transitions graphite-diamond, quartz- coesite, and garnet-perovskite in the CaGeO 3 system. The error in pressure measurement was estimated on the level of 0.2 GPa. The composition of phases was determined on a JEOL Superprobe JXA�8800 micro� probe analyzer at Tohoku University.