Real-Time Photodynamics of Squaraine-Based Dye-Sensitized Solar Cells with Iodide and Cobalt Electrolytes

Abstract

A series of dye-sensitized solar cells (DSSCs) has been prepared by using indole-based or quinoline-based squaraines (SQs) as the sensitizer and containing the commonly used I3–/I– redox pair or the lately employed cobalt complexes, [Co(dimethylbipyridine)3]3+/2+, [Co(bipyridine)3]3+/2+, and [Co(phenanthroline)3]3+/2+ redox electrolytes. The photodynamics of the different electron transfer reactions have been investigated by means of the femto- to millisecond pump–probe techniques. In the femtosecond transient absorption experiments, the electron injection rate constants and efficiencies, kei and φei, were determined for each cell. Larger values of kei and φei for the indole-based (SQ 8) compared to the quinoline-based (SQ 12) squaraines were obtained (13.2 × 1010 s–1 and 0.95 × 1010 vs 6.9 × 1010 s–1 and 0.81 for SQ 8 or SQ 12 with the I3–/I– pair, respectively), despite the similar values of the electron injection driving forces (−ΔG0ei = 0.75 vs 0.76 eV). This is due to the lower electron density in the lowest unoccupied molecular orbital at the anchoring group (−COOH) in SQ 12 compared to SQ 8. However, the type of electrolyte did not affect the kinetics of the electron injection processes. In the flash photolysis experiments, the kinetic parameters of the electron recombination via dye or electrolyte and the cation regeneration were calculated from the decays of the transient absorption signals of the electrons (1550 nm) or the SQ cation (570 nm). It was found that the electron recombination with the oxidized redox species is faster with the Co-based compared to the I3–/I– electrolytes for both SQs, τrec = 3 versus ∼0.5–1 ms. This proves that the steric hindrance in these SQs is not sufficient to avoid the approach of the Co3+ species to the surface of the TiO2 nanoparticle. Moreover, the regeneration rate constants and efficiencies, kreg and φreg, are considerably smaller for the cells with the different Co-based electrolytes compared to those with the I3–/I– pair (i.e., kreg = 30 × 104 vs 8 × 104 M–1s–1 and φreg = 0.96 vs 0.75 with the [Co(dmb)3]3+/2+ for SQ 8). This is explained by the lower regeneration driving force, −ΔGreg, in the Co-based electrolytes (0.3–0.1 eV). Thus, the use of Co-based electrolytes in these two SQs is detrimental to the overall efficiency of the cell, since −ΔGreg values below 0.4 eV do not give complete regeneration efficiency. Finally, we have compared the measured photocurrent with the calculated electron injection and regeneration efficiencies, and we found a good correlation between both parameters.