A linear models approach to optimize carbazole-based dyes for solar cell applications

Engineering notes

Engineering notes will be added by the Power for Solar editorial team.

Chinese explanation / 中文解读

中文解读待补充：本站会优先为光伏效率、钙钛矿太阳能电池、储能技术、太阳能热利用、BIPV、并网技术等高价值论文补充中文说明。

Original abstract

Abstract The excellent characteristics of carbazole-based dyes have resulted in DSSCs with power conversion efficiencies (PCEs) that are among the highest for metal-free dyes. However, the rational modification of these structures remains a challenge. In this work, 126 carbazole-based dyes were used in the development of models to predict the efficiency of DSSCs and to describe the relation between dye structure and cell efficiency. Using variable selection techniques (genetic algorithm and best subsets) on a pool of 2393 molecular descriptors (2D, 3D, quantum, and spectroscopic), three multiple linear regression models were obtained that satisfied well-established validation criteria (e.g., R 2 Pred > 0.7, Q 2 LOO > 0.6, r 2 m (LOO) > 0.5 and, ∆r 2 m (LOO) < 0.2). The best model indicated important molecular characteristics: π-bridge length, increased molecular branching, ionization potential, and presence of electronegative atoms. Based on these features, 36 novel sensitizers were designed, three of which achieved a predicted PCE approaching 9%, representing a potential improvement of almost two times over the reference sensitizers (LY‑S, LY‑P, and LY‑F). When we evaluated a small sample of the modified structures using common DFT and TD-DFT calculations for assessing new sensitizers, a favorable tuning of the electronic and spectroscopic properties was observed, which corroborates the predictions made by the models. Therefore, the best model presented can be employed for the rapid screening of carbazole-based sensitizers.

Links and sources

• PDF from original source

Comments