First-Principles Studies of Electronic, Optical and Defect Properties of Photovoltaic Materials

The development of the technology depends heavily on the development of materials. However, how to select the best materials for a specific purpose — i.e. materials selection, is a tricky problem in academia, industry, and our daily lives. Recently, because of the rapid development of computers, ab initio theoretical calculations can be used to aid in materials selection. However, since many approximations in the theoretical calculations exist, choosing appropriate approximations to obtain accurate and predictable materials properties is still difficult. This is the main focus of this thesis. More specifically, we will focus on the materials selection for photovoltaics, which plays a significant role in the energy field today. While modern commercial thin-film PV cells, e.g., based on metal chalcogenide zinc-blende-type materials (Cu(In,Ga)(S,Se)2 (CIGSSe), CdTe) suffer from problems like relying on elements that are either toxic or rare in the earth’s crust, a recent alternative candidate based on kesterite Cu2ZnSn(S,Se)4 (CZTS) peaked at relatively low efficiencies (12.6%) due to the limited open circuit voltage (Voc) caused by the prevalent anti-site structure disorder (e.g. Cu on Zn, Zn on Cu). A possible path forward to reduce this antisite disorder is to pursue materials in which the Cu/Zn combination is replaced by elements that are chemically less similar but that retain the same valence. Recently, Cu2 BaSnS4 ́x Sex (CBTSSe) materials with a trigonal structure (space group P31 ) and composed of only earth abundant metals have been proposed and demonstrated as emerging PV absorbers to address the above issues of CZTSSe. Results obtained as part of this thesis elucidated the band structure and electronic properties of the CBTSSe alloys. A recent device prepared from the Cu2BaSnS4 ́xSex (x « 3) has now been demonstrated with power conversion efficiency (PCE) exceeding 5%. Starting from this early prototype, many avenues remain to optimize the materials, including the underlying chemical positions, the electronic, optical and defect properties of specific compounds. In this thesis, we expand on the CBTSSe paradigm by exploring 16 related compounds, denoted I2-II-IV-VI4 (I=Cu,Ag; II=Sr,Ba; IV=Ge,Sn; VI=S,Se), and some of their alloys for their possible utility as thin-film PV absorbers.

A main methodological result of this thesis concerns the appropriate approximations we can use to obtain accurate and predictable structure, electronic, optical and defect properties for photovoltaic materials. Specifically, structure optimization using computationally expensive hybrid density functional theory is more appropriate than the normally used (semi)local functional (PBE, LDA) and can lead to reasonable and predictable structure and electronic properties. Furthermore, a detailed approach to obtain accurate carrier effective masses is pursued. For the optical properties, the effect of different broadening functions on the onset of absorption coefficients is discussed, and the correct onset behavior can be obtained using Gaussian broadening. At last, a validation of the infinite-size limit of charged defect formation energies calculated by supercell approach is given based on a benchmark study for the gallium vacancy (within charge state q = 0, -1, -2, -3) in GaAs. In general, the bare supercell approach, a supercell approach developed earlier by Freysoldt and coworkers, and a cluster approach can lead to the same infinite-size limit for the charged defect formation energies. Then, based on the appropriate approximations mentioned above, a study of materials properties is described in the I2-II-IV-VI4 (I=Cu,Ag; II=Sr,Ba; IV=Ge,Sn; VI=S,Se) 16 compound systems based on the theoretical structure, electronic and optical properties. Four compounds (Cu2BaSnS4, Cu2BaSnSe4, Cu2BaGeSe4, Cu2SrSnSe4) are identified as potential PV candidates based on their appropriate electronic, and optical properties. Then, two further re- finements are pursued for the Cu2BaSnS4 and Cu2BaGeSe4 compounds. The specific alloys Cu2BaGe1 ́xSnxSe4 (x « 3/4) and Cu2BaSnS4 ́xSex (x « 3) prove to be the best candidates for photovoltaics absorbers among the alloys of these two compounds.