An Artificial-Intelligence-Assisted Investigation on the Potential of Black Silicon Nanotextures for Silicon Solar Cells

Shaozhou Wang, Tong Xie, Ran Liang, Yu Zhang, Fa‐Jun Ma, D.n. Payne, Giuseppe Scardera, Bram HoexACS 2022 · DOI

Abstract

Black silicon (b-Si) nanotextures are of interest for Si solar cells because of their enhanced light trapping properties. However, the wide range of complex nanotextured b-Si surface morphologies makes a systematic investigation of b-Si solar cells challenging. A comprehensive performance review is necessary to determine the promising b-Si nanotextures for solar cell applications. In this work, we use artificial-intelligence approaches to assist in compiling a systematic and highly refined performance review of b-Si solar cells. We also perform numerical simulations of electrical properties for various nanotextured b-Si morphologies. We find that the weighted average reflectance (WAR) is an effective surface morphology metric for a wide range of surface textures. By correlating solar cell performance parameters to WAR, we show that multicrystalline Si solar cell efficiency can be improved with b-Si nanotexturing, and this is predominately attributed to an increase in short-circuit current density via the blue response improvement. We also show that some b-Si nanotextures can improve the performance of monocrystalline Si solar cells. Device simulations show that the electrical performance of hierarchical (combination of microtexture and nanotexture) and inverted-pyramidal b-Si nanotextures and microtextures can be comparable to or even better than random pyramids. As such, these textures show great potential for monocrystalline Si solar cells.
BibTeX
@inproceedings{wang2022artificial,
  author = {Shaozhou Wang and Tong Xie and Ran Liang and Yu Zhang and Fa‐Jun Ma and D.N. Payne and Giuseppe Scardera and Bram Hoex},
  title = {{An Artificial-Intelligence-Assisted Investigation on the Potential of Black Silicon Nanotextures for Silicon Solar Cells}},
  booktitle = {ACS},
  year = {2022},
  doi = {10.1021/acsanm.2c02619},
}

Loading related content...

Figure