A novel smart framework for sustainable nanocomposite electrolytes based on ionic liquids of dye-sensitized solar cells by a covalently multifunctional graphene oxide-vinyl imidazole/4-tert-butylpyridine cobalt complex

In this study, covalently multifunctional graphene oxide (GO) with vinyl imidazole/4-tert-butylpyridine cobalt complex (GO-VI/TBP cobalt complex) was synthesized through a novel approach. To this end, GO was functionalized with 1-vinyl imidazole (VI), followed by establishing a reaction between GO-VI, 4-tert-butylpyridine (TBP), and anhydrous CoCl2. Then, in dye-sensitized solar cells (DSSCs), this synthesized compound was used as an effective additive in sustainable nanocomposite electrolytes based on 1-butyl-3-methylimidazolium iodide (BMII) and 1-ethyl-3-methylimidazolium iodide (EMII) ionic liquids (ILs). Adding 0.6 wt% of optimal GO-VI/TBP cobalt complex to electrolyte increased the conversion efficiency of DSSCs significantly up to 7.359 % compared to 4.130 % in the initial standard DSSCs. This 78.18 % efficiency increase demonstrates how the GO-VI/TBP cobalt complex as a molecular bridge affects the conductivity and electron transport in the electrolyte based on ionic liquids. By enhancing the I-/I3- diffusion coefficient, cobalt complexes with the TBP ligand and nitrogen-containing heterocyclic compounds in GO-VI/TBP cobalt complex compounds accelerated electron transfers and ion conductivity in the electrolyte. As a result, the short circuit current density (Jsc) increased from 8.131 to 14.301 mA cm−2, and the open-circuit voltage (VOC) rose from 0.725 to 0.754 V. Density functional theory (DFT) studies showed an increase in the conduction band of the TiO2 electrode after the adsorption of the electrolyte additives on its surface. This upward shift resulted in the quick injection of electrons from the dye’s lowest unoccupied molecular orbital (LUMO) to the TiO2 electrode’s conduction band. Finally, a soft computing system was designed to predict experimental features (VOC and JSC) based on effective factors. Based on the results, Random Tree, Random Forest, and Multilayer Perceptron (MLP) methods with correlation coefficients greater than 0.92 have the highest efficiency for creating the soft sensor. Overall, this work can be employed as a novel strategy for advancing the usage of graphene derivatives in this sector to boost the performance of electrolytes.