Grain Boundary Metallization on Silicon Solar Cells

Research project P 14432 Grain boundary metallization on silicon solar cells Johann SUMMHAMMER 08.05.2000 Solar cells from multicrystalline silicon are a cheaper alternative to cells made from single crystalline material. However, their conversion efficiency of sun light into electrical power is also lower. In this project we investigate whether the output power of multicrystalline silicon solar cells can be increased by laying the fine metal grid, through which electrons are collected from the front side of the cell, onto the grain boundaries between individual crystallites, wherever this is possible. Theoretical simulations suggest such an increase, and so do preliminary experimental studies on small samples, in which the metal grid had been drawn manually. In the present project the grain boundaries on the 100 x 100 mm2 silicon wafers shall be detected automatically, using pattern recognition methods. The actual metallisation shall also be performed in a fully automated mode. This eliminates the subjective influence on the choice of grain boundaries to be covered with a metal line. A detailed characterisation of the cells produced in this manner shall be performed, from which a clear answer about the usefulness of grain boundary metallisation shall be deducible. Because of the automation steps in grain boundary metallisation the results will also be of significance for industrial manufacturing.

Both in 2001 and 2002 the output of the world wide solar cell industry grew at a yearly rate of 39%. The combined electric power of the solar cells of these two years amounts to over 900 MW. At around noon on a sunny day this would be enough to replace a large nuclear power plant. Some 50% of these solar cells were made from multicrystalline silicon. This has made this material the most widely used base material for solar cells. Presumably this will remain so for the next decade. At the Atominstitut of the Vienna University of Technology research on this base material has been going on for several years. The aim of this project was to equip the solar cells with a new kind of contacting structure for the collection of current on the front face of the solar cells, which should increase the power output. Each silicon wafer, which is to be processed into a solar cell, is subjected to a high resolution photographic scan. Image processing is used to identify the macroscopic boundaries between the individual crystallites, which make up the wafer. These boundary lines then serve as input for the calculation of a metallization grid, whose lines try to follow these boundaries as much as possible. In a further step, which is automated, this grid is then applied onto the silicon wafer. Two different methods have been developed for this process. Both methods show, that the new rid for current collection on the solar cells extracts between 2.5 and 8 % more electrical power than is possible with the metallization grid used in the industry so far. The biggest improvements were observed with multicrystalline silicon wafers of lesser quality. This should make it possible to use cheaper silicon and yet produce solar cells with higher electrical output.