RESUMO
Time-resolved photoluminescence (TRPL) is applied to determine an effective lifetime of minority charge carriers in semiconductors. Such effective lifetimes include recombination channels in the bulk as well as at the surfaces and interfaces of the device. In the case of Cu(In,Ga)Se2 absorbers used for solar cell applications, trapping of minority carriers has also been reported to impact the effective minority carrier lifetime. Trapping can be indicated by an increased temperature dependence of the experimentally determined photoluminescence decay time when compared to the temperature dependence of Shockley-Read-Hall (SRH) recombination alone and can lead to an overestimation of the minority carrier lifetime. Here, it is shown by technology computer-aided design (TCAD) simulations and by experiment that the intentional double-graded bandgap profile of high efficiency Cu(In,Ga)Se2 absorbers causes a temperature dependence of the PL decay time similar to trapping in case of a recombinative front surface. It is demonstrated that a passivated front surface results in a temperature dependence of the decay time that can be explained without minority carrier trapping and thus enables the assessment of the absorber quality by means of the minority carrier lifetime. Comparison with the absolute PL yield and the quasi-Fermi-level splitting (QFLS) corroborate the conclusion that the measured decay time corresponds to the bulk minority carrier lifetime of 250 ns for the double-graded CIGS absorber under investigation.
RESUMO
Recently recorded efficiencies of Cu(In,Ga)Se2 based solar cells were mainly achieved by surface treatment of the absorber that modifies the buffer-absorber interface region. However, only little is known about the electronic properties within this region. In this manuscript voltage dependent admittance spectroscopy is applied to low temperature grown Cu(In,Ga)Se2 based solar cells to detect near interface defect states in the absorber. Under non-equilibrium conditions even defect states close to the interface may cross the Fermi level and hence are detectable using capacitance based measurement methods, in contrast to the case of zero bias conditions. Such defects are of potential importance for understanding device limitations and hence, adequate characterization is necessary. A SCAPS model is developed including a near interface deep acceptor state, which explains the frequency and voltage dependence of the capacitance. Using the same model, also the experimental apparent doping density is explained.
