Published December 2009
Silicon wafer based photovoltaic cells that absorb light photons and convert them to electricity (electrons) appear to be at the edge of commercial cost competitiveness (grid parity). The purity of silicon used to make solar wafers can be considerably lower than electronic grade silicon used to make semiconductors for micro processors, allowing for the use of lower manufacturing cost technologies such as Siemens reactors, fluidized bed reactors, and directional solidification furnaces. Solar cells are made from feedstocks such as upgraded metallurgical grade silicon, polysilicon, thin film amorphous silicon, thin film non-silicon metal complexes (cadmium telluride, cadmium indium gemanium selenide, etc.), and organo-metallics,
Two kinds of companies currently market commercial quantities of polysilicon used to produce solar wafers: 1) on-purpose producers starting with high purity quartz, trichlorosilane (or similar silicon containing gases) that produce polysilicon in high temperature furnaces, and 2) recyclers of electronic grade silicon that has been scrapped from discarded electronic products, or recovered as by-product waste from silicon production. We examine the purity requirements and processing schemes for producing solar grade polysilicon, and report on the corresponding economics.
Until year 2000, photovoltaic cells were produced primarily from electronic grade silicon. More recently, silicon based photovoltaic cells producers have learned how to produce the photovoltaic cells directly from polysilicon, thereby eliminating the costly last step of converting polysilicon first to higher purity single crystal silicon. Also commercialized recently is technology for producing ‘upgraded’ metallurgical silicon that can be blended into higher purity solar grade polysilicon and used to produce photovoltaic cells. There are claims that upgraded metallurgical silicon can be used directly (not blended) to form solar wafers. Directional solidification furnaces operating at high vacuum accomplish this objective.
We present in this report the process engineering technology and corresponding production economics for making at commercial scale ‘solar grade’ polysilicon via: 1) Siemens furnace technology (the market leader), and 2) fluidized bed reactor technology. We also present 3) the process design and corresponding techno-economics for producing upgraded metallurgical silicon ingots in a rotational solidification dimensional electric induction furnace operating at high vacuum.