Abstract
Kesterite-structured Cu2ZnSnS4 (CZTS) is a semiconductor that is being studied for use as the absorber layer in thin-film solar cells. Currently the power-conversion efficiencies of this technology fall short of the requirements for commercialisation, despite the promising sunlight-matched optical band gap. Disorder in the Cu-Zn sub-lattice has been observed and is proposed as one possible explanation for the shortcomings of CZTS solar cells. Cation site disorder averaged over a macroscopic sample does not provide insights into the microscopic cation distribution that will interact with photogenerated electrons and holes. To provide
atomistic insight into Cu-Zn disorder we have developed a Monte Carlo (MC) model based on pairwise interactions. We utilise two order parameters to relate Cu-Zn disorder to the processing temperature for stoichiometric systems: one based on cation site occupancies (the Q order parameter) and the other based on cation pair-correlation functions. Our model predicts that the order parameters reach a plateau at experimentally relevant low temperatures, indicating that Cu-Zn order in stoichiometric CZTS is thermodynamically limited. Around room temperature, we predict a minimum of 10% disorder in the cation site occupancy
within (001) Cu-Zn planes.