It recently has been reported that in-situ oxidation during POCl3 diffusion allows for very low emitter recombination current values (J0E)[1]. Homogeneous distribution of the resulting layer thickness of silicon oxide – or phosphorus silicate glass (PSG) is required throughout the whole diffusion boat for optimal optical outcome of the resulting cells. Ideal tools for such diffusion process a state-of-the-art high throughput low pressure diffusion furnaces. With defined passivating PSG layer the PSG etching step is no longer required. In consequence the process chain can be simplified and the costs per solar cells can be reduced. In this work we demonstrate the suitability of in-situ oxidation / passivating PSG formation during phosphorus diffusion for industrial solar cell processing. Results on the homogeneity of the resulting emitter structures are presented. The superior passivation due to the SiO2 – Si interface and absence of a “dead layer” of the emitter profiles is shown on Cz-Si wafers. However, transferring this process to mc-wafer based cell production (in this case: ISC Konstanz’ pilot line and SoliTek’s industrial scale production line) is not so straight forward. First, no sacrificial dead layer for the gettering [2] of metal contamination is obtained by the process and, second, the oxide growth requires relatively high temperatures which are expected to be problematic for mc silicon wafers. Yet, recent times have seen drastic improvement in mc material quality and in consequence state-of-the-art material may be suitable for such emitter formation processes. Therefore, we studied the effects of this diffusion process on mc wafers: The suitability of different material quality was evaluated and the impact of key diffusion parameters on material of different ingot positions, lifetime, and defect densities is presented.
Conference Programme