Electrical Resistivity and Phase Separation in Dilute Liquid Alloys of Sodium in Selenium

The electrical resistivity of pure liquid selenium has been measured by many investigators [ 1 — 7 ] ; yet only a few of them describe the way in which their samples were prepared. Abdullaev et al. [3] and later Gobrecht et al. [4] and Koningsberger [5] have found that small amounts of oxygen have an enormous influence on the electrical resistivity of the liquid. Furthermore it was found [3, 5] that the resistivity of deoxidized selenium is a smooth function of temperature at the melting point Tm, whereas the resistivity of ordinary selenium of high metallic purity exhibits a jump of about three orders of magnitude at Tm. Koningsberger [5] performed both ESR and resistivity measurements on deoxidized as well as ordinary liquid selenium. Only from samples exhibiting a negligible jump in resistivity at Tm, reproducible ESR results were obtained. This confirmed the result of Abdullaev et al. [3] that the jump in the resistivity at Tm is a measure of the oxygen content of the material. The deoxidation method we used is based upon the method described by Koningsberger [5]. A layer of about 20 grammes of selenium shot (0^3mm; Alfa Ventron; met. imp. < 10 ppm) is placed in a carefully cleaned quartz vessel. The vessel is brought into a stainless steel tube, which through a liquid nitrogen trap is connected with a high vacuum diffusion pump. The tube is heated slowly (30 K/hr) under high vacuum (5 X 10~ Torr) up to a temperature of about 160 °C and kept at that temperature for about four hours. Thereafter, the tube is filled with helium, disconnected from the vacuum system and opened in a high purity, helium filled glove box (oxygen < 1 ppm; water < 1 ppm). To test the deoxidation method, a