No abstract

We report the results of an extensive study employing numerous methods to characterize carrier transport within copper indium gallium sulfoselenide (CIGSS) photovoltaic devices, whose absorber layers were fabricated by diverse process methods in multiple laboratories. This collection of samples exhibits a wide variation of morphologies, compositions, and solar power conversion efficiencies. An extensive characterization of transport properties is reported here -including those derived from capacitance-voltage, admittance spectroscopy, deep level transient spectroscopy, time-resolved photoluminescence, Auger emission profiling, Hall effect, and drive level capacitance profiling. Data from each technique were examined for correlation with device performance, and those providing indicators of related properties were compared to determine which techniques and interpretations provide credible values for transport properties. Although these transport properties are not sufficient to predict all aspects of current-voltage characteristics, we have identified specific physical and transport characterization methods that can be combined using a model-based analysis algorithm to provide a quantitative prediction of voltage loss within the absorber. The approach has potential as a tool to optimize and understand device performance irrespective of the specific

Thin films of CdTel-xSx with bulk atomic compositions, x=-[S]/([S]+[Te]), ranging from 0 to 0.45 were deposited by vacuum co-evaporation of CdTe and CdS with substrate temperatures of 200 and 250'C. X-ray diffraction analysis revealed that films with x < 0.3 were predominately single phase having the zincblende structure. Films with 0.35 < x < 0.45 contained the wurtzite modification. Lattice parameter determination indicated that each phase exists with compositions well within the miscibility gap shown on published equilibrium phase diagrams. The variation of the optical band gap with x was determined by measuring transmission and reflection of the films. Heat treatment at 415'C in the presence of CdC1 2 caused the films to segregate into two phases consistent with the phase diagram. If the CdC1 2 is assumed to only promote the phase segregation process, then the compositions of the two phases after heat treatment may be taken as measurements of the solubility limits of S in CdTe and Te in CdS respectively. The solubility limit of S in CdTe was thus determined to be 5.8% at 415'C which is the temperature used for the common CdC1 2 treatment of CdTe-based solar cells. An analysis of CdTe/CdS solar cell device structures shows that the atomic composition of alloys created by interdiffusion are consistent with these solubility limits.

Polycrystalline thin films deposited by coevaporation of CdTe and CdS form metastable single-phase CdTe–CdS alloys. Subsequent heating in a kinetic-enhancing ambient segregates CdTe1−xSx and CdS1−yTey phases from the original alloy phase. The equilibrium miscibility gap between the resulting CdTe1−xSx and CdS1−yTey alloy phases is determined for CdTe1−xSx films with initial x ∼ 0.4 treated from 360 to 700 °C. At 625 °C, the equilibrium compositions correspond to published results for mixed crystals. Below 625 °C the miscibility gap widens asymmetrically due to different mixing free energies for S in CdTe and Te in CdS. The solubility thermodynamics are modeled with an excess mixing free energy to account for nonideal mixing behavior.