Improving Cu(In,Ga)Se2 solar cell absorbers based on atomic-level modeling

Cu(In, Ga)Se2 (CIGS)-based solar cells are among the most promising candidates to replace crystalline silicon solar cells, thanks to their high efficiencies and low costs. The defect microstructure of the CIGS light absorber layer influences the optical and electronic properties of CIGS solar cells. The recent progress in their efficiency (up to 23.35 %) is mainly due to the incorporation of different alkali metal atoms into the absorber layer. As efficiency increases towards the Shockley-Queisser limit (31 % for a cell with a band gap of 1.3 eV), it becomes more difficult to improve. Thus, knowledge about native point defects and impurities, as well as about the formation of secondary phases in the CIGS absorber layer, is among essential information for optimizing CIGS solar cell performance. In this thesis, the choice of computational methods and their details strongly affects even the qualitative features of the obtained results. Therefore, the effects of the most important computational parameters are studied carefully in the thesis. Native point defects, native point defect complexes, and alkali metal impurities in the CIGS absorber layer are investigated using first-principles calculations within density functional theory (DFT) in order to understand their effect on the electronic structure. Moreover, based on DFT calculations, a mechanism for secondary phase (e.g. alkali indium selenide) formation is suggested. Calculating defects in CIGS is not straightforward. It is impossible to model defects directly in CIGS because In and Ga randomly occupy the same sites. Therefore all the calculations presented in this thesis are performed for CuInSe2 and CuGaSe2. In this thesis, calculations of native point defect formation energies in CuInSe2 provide information about the abundances of acceptors and donors for materials of different Cu concentrations. Moreover, it is shown that light alkali metal atoms prefer to accumulate on the Cu sublattice as impurities, and incorporation of heavy alkali metal atoms contributes mostly by phase separation. The formation of alkali indium/gallium selenide secondary phases during the post-deposition treatment is predicted by considering possible reactions between CuInSe2/CuGaSe2 and different alkali metal compounds by calculating their formation enthalpies. Interfaces between the secondary phases and the CuInSe2 absorber layer are studied in terms of band offsets. Finally, comparisons between the hard X-ray photoelectron spectroscopy (HAXPES) data and the density of states calculations with potassium post-deposition treatment (PDT) as a case study reveal the formation of the KInSe2 phase on the CIGS absorber surface after heavy potassium post-deposition treatment. In summary, the results in this thesis give information about the energetic characteristics of native point defects, native point defect complexes, alkali metal impurities, and alkali metal secondary phases. The results help to analyse already existing experimental observations of the abundances of point defects, migration mechanisms, and the formation of secondary phases.