Phase Stability of the Earth-Abundant Tin Sulfides SnS, SnS2, and Sn2S3

The various phases of tin sulfide have been studied as semiconductors since the 1960s and are now being investigated as potential earth-abundant photovoltaic and photocatalytic materials. Of particular note is the recent isolation of zincblende SnS in particles and thin-films. Herein, first-principles calculations are employed to better understand this novel geometry and its place within the tin sulfide multiphasic system. We report the enthalpies of formation for the known phases of SnS, SnS2, and Sn2S3, with good agreement between theory and experiment for the groundstate structures of each. While theoretical X-ray diffraction patterns do agree with the assignment of the zincblende phase demonstrated in the literature, the structure is not stable close to the lattice parameters observed experimentally, exhibiting an unfeasibly large pressure and a formation enthalpy much higher than any other phase. Ab initio molecular dynamics simulations reveal spontaneous degradation to an amorphous phase much lower in energy, as Sn(II) is inherently unstable in a regular tetrahedral environment. We conclude that the known rocksalt phase of SnS has been mis-assigned as zincblende in the recent literature. ■ INTRODUCTION Photovoltaic (PV) devices are of growing importance due to increasing population and diminishing reserves. Today, PV technology predominantly uses silicon as an absorber material but because of the low optical absorption coefficient, up to 500 μm thick films are needed to absorb significant fractions of visible light. More optimal absorber materials need less than 5 μm thickness, giving rise to so-called thin-film technologies that require less material and much cheaper processing conditions than silicon, indeed the lowest among commercial PV technologies. Successful examples include the commercially available cadmium telluride (CdTe) and copper− indium−gallium−selenide (CIGS) cells that have achieved record efficiencies close to 20%. Unfortunately, tellurium, indium, and gallium are rare and expensive; alternatives must be sought if PV is ever to scale up to the level of energy generation provided by nonrenewable methods: tera-watt production. Quaternary blends of more common elements can circumvent the issue of precursor availability and cost; where properties are tailored to PV applications by varying the stoichiometry of individual components. Most notable among these is Cu2ZnSnS4 (CZTS), which has achieved efficiencies of greater than 10%. As an alloy of Cu2S, ZnS, and SnS2, element availability is not a concern, but controlling the component ratios can be difficult. It has been shown that the desirable phase of CZTS occupies just a small fraction of the overall phase space for the system and has little or no thermodynamic barrier to phase separation. Herein, we consider tin sulfide, which is one of the components of CZTS and is itself attractive for PV applications because it is abundant, environmentally benign, and inexpensive. For example, tin extraction and importation to the European Union has an associated carbon footprint of less than one tenth of that of gallium (data obtained from “tin at regional storage” system process and “gallium, semiconductor grade, at regional storage” system process of the ecoinvent database within SimaPro7 software) and has an occurrence of 2 ppm on the earth’s crust. Tin sulfide single crystals have been grown by the Bridgman method and chemical vapor transport; and thin-films can be formed by chemical vapor deposition, chemical bath deposition, atomic layer deposition, electrodeposition, sulfurization of tin films, solid-state multilayer synthesis, and successive ionic layer adsorption and reaction. Nanostructures reported to date include, but are not limited to, nanoporous SnS frameworks by templated synthesis, nanodisks by electrodeposition, nanosheets by pyrolysis, nanoflowers by hydrothermal synthesis, nanobelts by a molten salt solvent method, and fullerene-like nanoparticles by laser ablation. Significantly, it has been claimed that zincblende (ZB) tin monosulfide microparticles have been synthesized and deposited as thin-films. This would allow for increased compatibility with existing technologies based on II−VI and III−V tetrahedral semiconductors. For example, the current generation of thin-film solar cells relies on a clean interface between the absorber material and the zincblende structured cadmium sulfide window layer. ZB structures also tend to exhibit a direct fundamental bandgap and large optical Received: September 14, 2012 Revised: October 24, 2012 Published: November 5, 2012 Article