Article ID Journal Published Year Pages File Type
1510931 Energy Procedia 2014 8 Pages PDF
Abstract

In this paper, we investigate different industrial applicable cleaning sequences on test wafers and PERC solar cells in comparison to a laboratory type RCA clean. The cleaning sequences pSC1, HF/HCl, HF/O3 and HF/O3 show lifetimes between 1 ms and 2 ms which is comparable to a laboratory type RCA clean corresponding to a surface recombination velocity Spass below 15 cm/s. The pSC1, HF/HCl clean achieves lifetimes around 1 ms, whereas the PSG-etch shows poor cleaning quality with lifetimes around 500 μs. Reference PERC cells using a rear protection layer before texturing and diffusion demonstrate efficiencies up to 20.4% for the cleaning sequence pSC1, HF/HCl prior to passivation which is comparable to the RCA clean. The HF/O3 cleans result in lower PERC efficiencies up to 20.0% mainly due to a lower Fill Factor which is likely caused by etching of the emitter and hence increased contact resistance. Investigations of polished test wafers show that the cleaning sequences pSC1, HF/HCl, HF-Dip and pSC1, HF/HCl, HF/O3 are able to sufficiently remove porous silicon from the front side and simultaneously allowing excellent rear surface passivation. A first batch of PERC solar cell results with polished rear surface post texturing and POCl3 diffusion achieves efficiencies of up to 20.7% when applying an RCA clean. However, the pSC1, HF/HCl and pSC1 HF/O3 still exhibit significantly lower efficiencies since in this batch the porous silicon of the emitter was not yet sufficiently removed, which is subject to further optimization.

Related Topics
Physical Sciences and Engineering Energy Energy (General)