Improving performance of copper(I)-based dye sensitized solar cells through I₃ ^-/I^- electrolyte manipulation

The performances of dye-sensitized solar cells (DSSCs) containing a heteroleptic bis(diimine)copper(I) dye and an I₃ ^-/I^- liquid electrolyte have been investigated as a function of the electrolyte composition. Starting from a standard electrolyte with LiI (0.1 M), I₂ (0.05 M), 1-butyl-3-methylimidazolium (BMII) ionic liquid (0.6 M), 1-methylimidazole additive (0.5 M) in 3-methoxypropionitrile (MPN), a series of ten electrolytes was initially screened; the solvent, additives, concentration of I₂, and sources (LiI and/or BMII) and concentrations of I^- were varied. The highest short-circuit current densities (J_SC = 7.85 and 7.60 mA cm^-2) and photoconversion efficiencies (η = 2.64 and 2.70% relative to 6.56 and 6.11% for two N719 references) were observed for DSSCs with an electrolyte comprising I₂ (0.03 M), BMII (0.6 M), 4-tert-butylpyridine (TBP, 0.4 M) and guanidinium thiocyanate (GNCS, 0.1 M) in MPN solvent; no LiI was added. This composition was further optimized by varying the amounts of TBP and GNCS. A low concentration of I₂ (0.015 M) is optimal. The presence or absence of GNCS has little effect on overall DSSC performance, but increasing the concentration of TBP is detrimental to DSSC performance, and an absence of TBP is beneficial. Open-circuit voltage decay (OCVD) measurements confirm that the addition of GNCS and/or TBP reduces recombination rates at the TiO₂-dye interface. The best DSC parameters obtained for the copper(I)-based dye were J_SC = 7.80 mA cm^-2, open-circuit voltage (V_OC) = 501 mV and η = 2.76% (relative to 6.72% for N719).