Inverted random nanopyramids patterning for crystalline silicon photovoltaics

We demonstrate a nanopatterning technique for silicon photovoltaics, which optically outperforms conventional micron-scale random pyramids, while decreasing by a factor of ten the quantity of silicon lost during the texturing process. We combine hole-mask colloidal lithography, a bottom-up nanolithography technique, with reactive ion etching to define nanopyramids at the surface of a silicon wafer. Thanks to the self-organised aspect of the technique, the beads are randomly distributed, however keeping a interbead distance of the order of their diameter. We tune the nanopattern feature size to maximize the absorption in the crystalline silicon by exploiting both anti-reflection and light trapping. When optimized, the nanopyramids lead to a higher absorption in the crystalline silicon than the conventional micron-scale random pyramids in the visible and near the band edge, with a superior robustness to variations of the angle of the incident light. As the nanopatterning technique presented here is simple, we expect that it could be readily integrated into the crystalline silicon solar cell fabrication processing.