Impact of sensitizer kinetics on the small-signal characteristics of dye-sensitized solar cells

Recent research on dye-sensitized solar cells (DSCs) has pointed out that dye kinetics (involving recombination, regeneration and injection mechanisms) may significantly affect the achievable power conversion efficiency. Dye-kinetic limitations could be at the root of a few experimental evidences, such as the decrease of Incident Photon-to-Current Efficiency (IPCE) under forward bias, the impossibility to reconstruct the measured current voltage (J-V) characteristic under light by the sum of dark J-V and short circuit current density (Jsc), and the degradation of lifetime and diffusion length (Ln) under illumination with respect to dark. These facts suggest the existence of at least two parallel recombination mechanisms for the photogenerated electrons involving electrolyte species and oxidized dye molecules, and a possible dependence of dye regeneration/injection on the local electron density. In this work we investigate through numerical simulations and experimental characterizations the static and small-signal behavior of DSCs with clear dye-kinetic limitations. From the modeling standpoint, this is done by including a detailed description of electron recombination rates with electrolyte and oxidized dyes, and of dye regeneration/injection kinetics. The results allow for an in-depth analysis of the influence of dye-kinetics not only on the photovoltaic behavior but also on the transport parameters and figures of merit (such as Ln) usually derived from Electrochemical Impedance Spectroscopy (EIS). An example is shown in Fig.1. The mismatch between J-V under illumination and the J-V predicted by the superposition of the dark J-V and Jsc (Fig.1-A) is correlated to the decrease of the dye regeneration rate and the associated increase of the dye-electron recombination rate as forward bias increases, see Fig.1-B. Dye kinetics has a significant influence on the (calculated) local Ln (Fig.1-C) across the film and the small-signal Ln (Fig.1-D) extracted from EIS under different operating conditions. For the sake of comparison, we show in the inset of Fig.1-C the value of the Ln estimated from the IPCE ratio technique. The outcome of these results is twofold: they give a robust support to the use of EIS for evaluating Ln, even in presence of complex and possible nonlinear recombination mechanisms and trap-limited kinetics; they highlight the several different behaviors that can arise from EIS measurements carried out in different operating conditions, due to dye-kinetic limitations. The analysis also suggests the possible impact of dye-kinetics on optical small-perturbation techniques which will be addressed in the full work.