Understanding the origin of disorder in kesterite-type chalcogenides A2ZnBQ4 (A = Cu, Ag; B = Sn, Ge; Q = S, Se): the influence of inter-layer interactions

Semiconducting quaternary chalcogenides with A2ZnBQ4 stoichiometry, where A and B are monovalent and tetravalent metal ions and Q is a chalcogen (e.g. Cu2ZnSnS4 or CZTS) have recently attracted attention as potential solar-cell absorbers made from abundant and non-toxic elements. Unfortunately, they exhibit relatively poor sunlight conversion efficiencies, which has been linked to site disorder within the tetrahedral cation sub-lattice. In order to gain a better understanding of the factors controlling cation disorder in these chalcogenides, we have used powder neutron diffraction, coupled with Density Functional Theory (DFT) simulations, to investigate the detailed structure of A2ZnBQ4 phases, with A = Cu, Ag; B = Sn, Ge; and Q = S, Se. Both DFT calculations and powder neutron diffraction data demonstrate that the kesterite structure (space group: I4[combining macron]) is adopted in preference to the higher-energy stannite structure (space group: I4[combining macron]2m). The contrast between the constituent cations afforded by neutron diffraction reveals that copper and zinc cations are only partially ordered in the kesterites Cu2ZnBQ4 (B = Sn, Ge), whereas the silver-containing phases are fully ordered. The degree of cation order in the copper-containing phases shows a greater sensitivity to the identity of the B-cation than to the chalcogenide anion. DFT indicates that cation ordering minimises inter-planar Zn2+Zn2+ electrostatic interactions, while there is an additional intra-planar energy contribution associated with size mismatch. The complete Ag/Zn order in Ag2ZnBQ4 (B = Sn, Ge) phases can thus be related to the anisotropic expansion of the unit cell on replacing Cu with Ag.