An improved cleaning process is developed to remove front surface contamination for single heterojunction solar cells on textured surfaces on ∼25 μm thick exfoliated, flexible mono-crystalline silicon. The process is very effective in cleaning metallic and organic residues, without introducing additional contamination or degrading the supporting back metal used for ultrathin substrate handling. Quantitative analysis of the Auger electron spectra shows significant potassium contamination reduction (∼0.89% atomic) using the new cleaning process. An open-circuit voltage enhancement of 22 mV and an absolute 1.5% increase in conversion efficiency are observed with the new cleaning procedure for the exfoliated thin solar cells.Thin crystalline silicon (c-Si) solar cells are of much interest due to their potential to achieve high efficiency and reduce cost by using less Si material. However, there are significant challenges to commercialize sub-100 μm thin Si substrates as they can easily break or crack with wafer-handling, resulting in low yield in a solar cell manufacturing line. We have introduced in our earlier work, 1 a kerf-less process in which ultra-thin (∼25 μm) and flexible mono-crystalline Si substrates can be obtained through an exfoliation technique from a thicker (>450 μm) parent wafer. These substrates, when exfoliated, have thick (∼50 μm) electroplated nickel (Ni) metal backing, which provides mechanical support to the thin Si and enables ease of processing for semiconductor device fabrication.Previously we have demonstrated single heterojunction (SHJ) solar cells fabricated on this type of substrate exhibiting efficiencies 14.9% on as-exfoliated substrates. 2 However, on textured surfaces efficiency was limited to 11%. We postulated that one of the issues that could be limiting the performance of the cells is unintentional front surface contamination introduced during wet chemical processes before hydrogenated amorphous Si (a-Si:H) deposition of the front surface emitter, which can limit the open-circuit voltage (V OC ) of these solar cells. This could happen due to the presence of potassium ions introduced from potassium hydroxide (KOH) during texturing. For decontamination we could not use SC-2 solution (5:1:1 ratio of H 2 O, H 2 O 2 , HCl at 80 o C) as it reacts rather aggressively with the electroplated Ni back metal. Instead, we used a piranha solution (1:1 ratio of H 2 O 2 , H 2 SO 4 ) for both decontamination from potassium ions and removal of organic contaminants, which did not seem to show corrosion degradation in the back side Ni. The pH level of HCl is slightly lower compared to H 2 SO 4, and SC-2 solution has a stronger effervescent action than piranha solution. This may explain why the Ni is much more affected by the SC-2 clean compared to the piranha clean. Nevertheless, piranha-treatment alone is probably inadequate for metal residues or potassium related contaminant removal after texturing.In this work, we attempted to address the front surface contamination issue by developing an improv...
