Benchmarking admittance spectroscopy methodologies for solar cell characterization

Impedance spectroscopy has been extensively applied for the diagnosis and evaluation of various types of solar cells. The method applies a sinusoidal electrical stimulus and subsequently measures the impedance (conductance and capacitance) across the terminals of the device for a range of different frequencies and/or temperatures. Based on this measurement, it is possible to extract the energy resolved trap distribution using analytical formulas that make use of the derivative of capacitance vs frequency (Walter et al. [1]), or the derivative of capacitance vs temperature (Li and Levi [2]). Assuming Arrhenius trap activation, the two approaches become dual to each other [2]. It is unclear, however, how the extracted trap densities compare to the actual trap densities. It is also not clear how different bias conditions, trap distributions, temperatures, and device dimensions affect the measurements. In this paper, we use “exact” numerical simulations to benchmark two impedance spectroscopy methodologies against each other, and against actual trap densities input to the simulations. The results help clarify the limitations of existing impedance spectroscopy analysis methodologies.