Low-density silicon allotropes for photovoltaic applications

TL;DR

This study systematically predicts low-density silicon allotropes with promising photovoltaic properties, identifying 11 metastable phases with band-gaps near the optimal range for solar energy conversion.

Contribution

It introduces a modified ab initio minima hopping method to explore low-density silicon phases, revealing new metastable structures with potential for solar cell applications.

Findings

11 silicon allotropes with direct or quasi-direct band-gaps (1.0-1.8 eV)

Enhanced absorption spectra overlap with solar spectrum

Predicted phases are structurally similar to known clathrates

Abstract

Silicon materials play a key role in many technologically relevant fields, ranging from the electronic to the photovoltaic industry. A systematic search for silicon allotropes was performed by employing a modified ab initio minima hopping crystal structure prediction method. The algorithm was optimized to specifically investigate the hitherto barely explored low-density regime of the silicon phase diagram by imitating the guest-host concept of clathrate compounds. In total 44 metastable phases are presented, of which 11 exhibit direct or quasi-direct band-gaps in the range of $\approx$1.0-1.8 eV, close to the optimal Shockley-Queisser limit of $\approx$1.4 eV, with a stronger overlap of the absorption spectra with the solar spectrum compared to conventional diamond silicon. Due to the structural resemblance to known clathrate compounds it is expected that the predicted phases can be synthesized.