Polycrystalline silicon is also being studied as a material for producing large quantities of solar cells by automated processing and at low cost. The concept is based on depositing the semiconductor material on a continuous and inexpensive substrate. This technique would eliminate the energy and cost intensive processes which are presently used to fabricate thin solar cells of crystalline silicon. Work on polycrystalline silicon solar cells has only been pursued actively since 1975, and the efforts to date have been devoted to understanding the fundamental operation of the device. However, efficiencies between 4 and 6% have been reported. As expected, a wide variety of problems confront polycrystalline silicon researchers and include high recombination losses, expansion coefficient mismatches, preferential doping at grain boundaries, and backside electrode connections . The recombination loss at crystallite grain boundaries is the most critical parameter of a polycrystalline silicon solar cell. The belief is that if the crystallites can be made large enough, then the major portion of current flow will be vertically rather than horizontally, and efficiency will be preserved. Methods proposed for achieving this include melting and resolidification using electron beams or lasers. The substrate is critical in a polycrystalline silicon solar cell because it affects the structure and composition of the film. Furthermore, depending on how the film is deposited, it must be capable of withstanding high temperatures. Thus far, steel, carbon, and sapphire have been used in experimental cells and aluminum has been proposed. Thin crystalline films of silicon are routinely grown on synthetic sapphire by epitaxial deposition processes for the integrated circuit industry. The technique is successful because the crystal lattice spacing is very close to that of silicon and, therefore, large areas of relatively defect-free films can be formed. A similar process for gallium arsenide is not known to exist. Silicon has also been deposited in the amorphous state which does not have crystalline structure or crystallite boundaries. The expected efficiencies for such cells is as high as 10%. If the size of the crystallites is larger than the thickness of the solar cells, most of the electrical carriers can be collected without encountering grain boundaries. A method has been developed by Solarex which involves casting silicon so that large crystallites on the order of a millimeter are formed. Solarex has announced a cell made by this process with a 10% efficiency. Since solar cell weight is an important factor in the design of the SPS arrays, it is worthwhile to analyze its effects in thin-film solar cells. Data has been shown previously that the minimum thickness for a crystalline silicon solar cell is 0.100 mm. In the case of silicon, the minimum weight has been reached. Even if a polycrystalline silicon
RkJQdWJsaXNoZXIy MTU5NjU0Mg==