• Marquardt, Julien

Abstract

Over the last decades the need of electrical energy increased continuously, whereas the percentage of electric energy from renewable sources became larger in the last years. The demand of an energy supply, which is produced by renewable sources completely, is more important than before. Nowadays, solar cells reach power conversion efficiencies of 46% [1] using multi-junction concentrator cells, which are very complicated and expensive in production. The maximum power conversion efficiency for a single junction solar cell is restricted to ~32% [2], named the Shockley Queisser limit. However, the incorporation of transition metals into the wide gap CuGaS2 chalcopyrite type absorber material was proposed to create an intermediate band, which cause two additional absorption ranges and an increase in power conversion efficiency up to 63% [3, 4]. The aim of this study was to determine the solid solubility limits of several transition metals as well as to study their effect on the chalcopyrite type crystal structure and optoelectronic properties. All investigated transition metals were successfully incorporated into the chalcopyrite type structure, by solid state reaction synthesis of pure elements. The lowest solid solubility was obtained from chromium and nickel with 0.003(1)mol% CrS and 0.008(1)mol% NiS in Cu0.5Ga0.5S, which results in no observable changes in the chalcopyrite type crystal structure. A much higher solid solubility limit was observed for manganese with 0.098(1)mol% MnS in Cu0.5Ga0.5S. The pseudo-binary section of Cu0.5(FexGa0.5-x)S was earlier reported [5] to have complete solubility, was showing a phase separation at xi>0.1 into an iron rich and iron poor chalcopyrite type phase, respectively. Two different substitution mechanisms were observed from the iron alloyed chalcopyrite type phases. For those with low initial iron contents (xi<0.1), the trend of chemical composition and lattice parameters indicate a coupled substitution (Cu+Ga↔Fe), whereas higher initial iron contents show a unilateral substitution (Ga↔Fe). The substitution of manganese into the chalcopyrite type structure is rather coupled than unilateral as would be necessary for an intermediate band absorber material. Using the average neutron scattering length analysis method, it was observed that manganese is occupying both cationic sites of the chalcopyrite type structure. From these extrinsic defects (MnGa, MnCu) two optoelectronic active defect states result, located within the band gap of the chalcopyrite type semiconductor, but not corresponding to the proposed intermediate band position. The obtained photoluminescence (PL) spectra was correlated to the predictions given by density functional theory [6]. In addition, correlating the PL spectra with the defect concentration, it can be seen that the dominant PL band at 1.86eV is rather related to the intrinsic GaCu cation anti-site defect than anion vacancies. Based on the obtained anion parameter of the chalcopyrite type phases from X-ray diffraction using the Rietveld method, the effect of copper vacancies on local structural changes was derived and existing calculation models were enhanced to give more precise predictions for overall structural parameter (e.g. tetragonal distortion).

Topics

• density
• impedance spectroscopy
• photoluminescence
• nickel
• chromium
• phase
• x-ray diffraction
• theory
• semiconductor
• chemical composition
• copper
• defect
• density functional theory
• iron
• Manganese
• power conversion efficiency
• neutron scattering