Measuring wafer quality

Silicon wafers are by far the most expensive consumable for silicon wafer solar cell production. In addition, imperfections can have severe consequences regarding performance and yield of the production line. Consequently, both wafer suppliers as well as solar cell manufacturers pay close attention to the incoming wafer quality and measure a whole range of relevant properties. It is for example quite easy to do a 2D geometric inspection of the wafer to check for chips or large cracks where generally all imperfect wafers will be discarded. The detection of microcracks is also critical as they can develop further and eventually result in a cracked wafer and it is thus of crucial importance to take them out before spending further resources on it. Often companies will also check for saw marks as they can affect all the etching process. Also the bulk resistivity of the wafers can very quickly be checked by non-contact eddy current techniques as well as the thickness and total thickness variation of the wafer by capacitance techniques.

The advent of photoluminescence imaging has been the biggest revolution in incoming wafer inspection. Photoluminescence imaging allows the detection of various efficiency limiting defects. This is particularly relevant for multicrystalline silicon wafers as they can have a wide range of structural and impurity related defects. Figure 1 shows a photoluminescence image of incoming silicon wafer with an overlay indicating low effective minority carrier lifetime regions related to impurities (yellow) and dislocations (blue). Impurities can typically be either gettered or passivated in the solar cell fabrication process, while dislocations are more challenging to neutralize electronically.

as-cut overlay 2.jpg
Figure 1: Photoluminescence (PL) image of an unprocessed multicrystalline silicon wafer. The wafer has various low effective minority carrier lifetime regions which are characterised by a low PL intensity. Via image processing, the nature of the low lifetime regions can be attributed to impurities (yellow) or dislocations (blue).