(Invited) Local Electrolyte-Driven Redox Cycling within a Nanoporous Photoanode

Abstract

Nanoporous photoanodes used in dye-sensitized solar energy conversion systems such as solar cells and solar hydrogen generators have been carefully studied to understand factors that influence their performance, as measured by photocurrents and open-circuit photovoltages. We have developed a chemically detailed, multiscale reaction-diffusion model that seeks to connect these observables to the detailed dye redox and electrolyte chemistry for the simpler dye-sensitized solar cell (DSC) system, incorporating dyes bound within a TiO2 nanoparticle-based anode, and an I-/I3 - redox shuttle connecting the anode to the cathode [1]. The model framework permits nanoscale characteristics of the interfacial and solution-phase redox processes, including charge trapping, to be connected to photocurrents and electron densities under various dye excitation scenarios, and is extendable to the more complex interactions occurring in hydrogen generators. In DSCs, it is generally assumed that the redox shuttle needs to make a full round trip between anode and cathode to convert I3 - to 3 I- at the cathode after dye reduction reactions in the anode. It is also thought that electron trapping by electrolyte components, in particular I2, is deleterious, and degrades performance. Detailed simulations assess these mechanistic elements. The results show that during operation, I- is deeply depleted in the photoanode pores, leading to significant populations of oxidized dyes and a shift in the I-/I3 - equilibrium with formation of I2. Electron trapping by I2, which forms I2 - at the photoanode-electrolyte interface, does not invariably affect photocurrent or photoanode charging. This is because the increased rate of I2 - disproportionation in the electrolyte enables local regeneration of I- within the pores and efficient dye reduction after photoexcitation, without requiring I3 - diffusion to the cathode. The possibility of local cycling, which results in photoanode redox activity without net current flow, has not been reported previously and suggest that causes of less-than-ideal performance may lie elsewhere. The simulations provide nanoscopic predictions for interfacial processes that control dye cycling. The timing of electrolyte anion diffusion toward the anode-dye interface within the pores, relative to charge injection and to dye neutralization, will be discussed. [1] F. A. Houle, J. Phys. Chem. C 123, 14459-14467 (2019).