Solar silicon via improved and expanded metallurgical silicon technology. Quarterly report, No. 3

Abstract

The goal of this program is to produce solar-cell-grade silicon at 3000 Mt/y for less than $10/kg by 1986.The approach is to improve the quality of the 200,000-Mt/y process for producing $1/kg, 98-percent pure metallurgical silicon. High-purity sources of quartz and quartzite have been identified for smelting with charcoal in an upgraded arc furnace. Typical quartz analyses indicate Al = 30 ppm, B <1 ppM, Fe = 30 ppM, and P <10 ppM. Charcoal purification by halogen treatment at 1500 - 2500/sup 0/C reduces impurity levels to less than, or near, 10 ppM. Trade-off studies between charcoal smeltability and impurity contribution to silicon must be determined. It was demonstrated that impurity segregation during Czochralski crystal growth is just as effective when pulling a polycrystalline ingot as when pulling a single-crystalline ingot, provided that constitutional supercooling is prevented. This permits improved silicon yield in the overall process. A Czochralski ingot pulled from upgraded arc-furnace silicon was analyzed at the 65-percent-of-the-melt-pulled point and found to have transition metal impurity concentrations below 0.003 - 0.01 ppMa limits of detection. Estimates of actual transition metal concentrations indicate that the material should meet or exceed anticipated definitions of solar-grade silicon. Segregation of impurities has beenmore » demonstrated using the potentially high-capacity Bridgman technique of crystal growth.« less