Theoretical insights into cobalt redox shuttles for third-generation dye-sensitized solar cells: effect of π-spacer alteration on photovoltaic properties.

Understanding how the electronic structure of redox shuttles and sensitizers governs dye regeneration and photovoltaic output is important for the development of efficient dye-sensitized solar cells (DSSCs). In this work, density functional theory (DFT) and time-dependent density functional theory (TDDFT) were employed to examine a series of ligand-tuned cobalt redox shuttles, [Co(PY5Im)]2+/3+, [Co(PY5ImDMA1)]2+/3+, and [Co(PY5ImDMA2)]2+/3+, in combination with a reference dye D35 (Dyenamo orange; Dye 1) and five modelled new dyes (Dye 2-6) that were adjusted by altering the fused heteroatoms π-spacers in the reference dye with the aim of improving its power conversion efficiency. The computed redox potential energy levels for RS1 (-4.88 eV), RS2 (-4.78 eV) and RS3 (-4.65 eV) show that the cobalt redox shuttles offer broader and more tunable redox properties than the conventional iodide/triiodide shuttle. The computed photovoltaic parameters show that all the dyes display negative ΔGinj and ΔGreg values, confirming the thermodynamically favourable electron injection and dye regeneration with all three cobalt redox shuttles. The modelled dyes show different light-harvesting efficiencies (LHE) and photovoltaic responses, highlighting the vital role of altering the fused heteroatoms π-spacers in tuning the performance of the device. Among the series, Dye 4 emerges as the most potent sensitizer, combining high LHE (81.6%) and the highest predicted short-circuit current density (7.70 mA cm-2) and leading to the highest predicted power conversion efficiency of 5.30%. Experimentally, in comparison to the LEG4 system with these cobalt RS, an efficiency in the range of 1.9% to 2.3% was determined. This shows that combined tuning of redox shuttles and dyes offers an effective approach for improving the performance of DSSCs. The strong alignment between the theoretical and experimental results supports the reliability of the computational approach.