Cu2ZnSnS4 (CZTS) has recently attracted attention as a light absorber for photovoltaic applications because of its band gap (εg≈1.4 eV), the relative abundance and low cost of its constituent elements (which permits large-scale low-cost module production), and its demonstrated solar-to-electricity power conversion efficiencies over 8% 1. While many materials have been synthesized by atomic layer deposition (ALD), the application of this method to technologically important metal sulfides is underexplored, and homogeneous quaternary metal sulfides are absent from the literature. We outline a first-to-date ALD process to synthesize CZTS 2, in which a trilayer stack of binary metal sulfides (i.e., Cu2
S, SnS2 and ZnS) was deposited and mixed by thermal annealing. Since this ALD route relies on the facile solid state diffusion of chalcogenides for mixing we investigate (i) intermixing in the initial stack |substrate/Cu2 S/SnS2/ZnS| and (ii) effect of the annealing temperature. The first case is of fundamental interest, while the second one provides parameters needed to fully mix the binaries into a homogenous CZTS alloy. The composition profiles were measured by time-of-flight secondary ion mass spectrometry in high resolution dual-beam regime (gentleDB TOF SIMS) 3 for both as-deposited trilayer stacks (at 135 C) and after their annealing at elevated temperatures (275, 350 and 425 C) in argon ambient for 60 min. Diffuse interfaces between layers were found in the as-deposited case, indicating that binaries greatly premixed at the synthesis temperature of 135 C. By using high-resolution SIMS depth profiles we were able to estimate roughly the diffusion coefficients between adjunct layers in the trilayer stack.
Under annealing, gentleDB SIMS results suggest that mixing of the metal-sulfide binaries into CZTS obeys the following predicted mechanism 4: (1) Cu2 S + SnS2 → Cu2 SnS3 + ZnS; (2) Cu2 SnS3 + ZnS → Cu2 ZnSnS4, Step (2) limits the overall solid-state reaction. Particularly, it explains why CZTS forms only at elevated temperatures in excess of 400 C (Cu and Sn concentrations homogenize at much lower temperatures) as was previously evidenced by Raman spectroscopy and X-ray diffraction 2. Temperature of 425 C was found to be sufficient to fully mix the binaries. One can consider it as a starting point for further device performance-driven process optimization in this ALD route to the CZTS system.
1 B. Shin et al. Prog. Photovoltaics: Res. Appl. DOI: 10.1002/pip.1174 (2011) 2 E. Thimsen et al. Chem. Mater. DOI: 10.1021/cm3015463, article ASAP (2012) 3 S.V. Baryshev et al. Rapid Commun. Mass Spectrom. 26, 2224 (2012) 4 F. Hergert and R. Hock. Thin Solid Films 515, 5953 (2007)
Sergey Baryshev graduated in 2004 from St. Petersburg State Polytechnical University (St. Petersburg, Russia) with MS in Applied Physics and received his PhD in Condensed Matter Physics from Ioffe Physico-Technical Institute (St. Petersburg, Russia) in 2008. Dr. Sergey Baryshev has been a Postdoctoral Fellow in Surface Chemistry Group of Materials Science Division at Argonne National Laboratory. His research interests are in the field of interaction of charged particles and dense laser irradiation with solids and atoms, time-of-flight mass spectrometry, the development of trace analysis and advanced 3D chemical/topographical imaging with nanometer lateral and depth resolution. Additional interests are physics of HTSC (transport, microwave and noise properties); sub-micrometer field- and photoemission electronics; collective phenomena in mesoscale systems; methods of electron microscopy such as X-ray microanalysis, EBIV/EBIC, lithography, cathodoluminescence and their applications to materials important for accelerator physics and technology.
Sergey V. Baryshev, Elijah Thimsen, Shannon C. Riha, Alex B.F. Martinson, Jeffrey W. Elam, Michael J. Pellin, Igor V. Veryovkin
Cite this work
Researchers should cite this work as follows:
University of Illinois Urbana-Champaign, Urbana, IL