Understanding Cu2ZnSnS4 as a Photovoltaic Absorber for the Future of Solar Electricity

The world needs solar electricity to replace a large fraction of traditional, fossil-fuel-generated electricity over the coming decades if it is to avoid the worst effects of climate change and continue to meet the needs of an increasingly energy-dependent society. This transition is currently well underway. The installed generating capacity of solar electricity continues to grow exponentially, having reached 307 GW in 2016 (2 % of average global electricity demand), which means that replacing a large majority of fossil fuel use, requiring several terawatts of capacity, in the coming decades is entirely realistic. Cu2ZnSnS4 (CZTS) is a potential material for the absorber layer in photovoltaic solar cells. It has the advantages over silicon, which currently provides 95 % of the solar electricity market, of lower processing costs and a direct band gap, which means much less material is required. Most other alternative absorber materials will ultimately be limited by high material costs, low elemental abundances, or toxicity, but CZTS has none of these problems, making it a very promising material indeed. However, its record photovoltaic efficiency (11.0 %) is well below those of some other materials (>20 %) because of low open-circuit voltage. The outstanding areas of current CZTS research are the absorber-buffer interface, band gap fluctuations caused by point defects, and secondary phases. This thesis presents work investigating the latter two, primarily using bulk samples fabricated by solid-state reaction. Firstly, compositional, structural, and optoelectronic analysis techniques were used to study the effect of composition on material properties. It was found that the quasi-ternary phase diagram commonly used for CZTS is incorrect; and that no common analysis technique can quantify cation disorder in CZTS, despite Raman spectroscopy commonly being used to do so. Secondly, neutron diffraction was used to study the order-disorder phase transition at around 550 K. It was found that the transition temperature is dependent on elemental composition; and that Cu-Zn disorder is present on all cation lattice sites, not merely the 2c and 2d sites of the kesterite crystal structure as has previously been assumed. Thirdly, anomalous X-ray diffraction was used to study cation disorder further. It was found that two distinct phases of CZTS can be present in the same sample, with different elemental compositions resulting from the prevalence of different point defect complexes; two new such types of CZTS were identified; and a mechanism of phase formation was proposed. Finally, a fabrication route for thin-film CZTS by sputtering and sulphurisation annealing was established