Proceedings of International Conference on Perovskite Thin Film Photovoltaics and Perovskite Photonics and Optoelectronics (NIPHO24)
Publication date: 25th April 2024
Solution-based approach used to fabricate hybrid halide perovskites (HPs) induces polycrystallinity in the layer so that a massive amount of defects (10 16 cm ^-3 ) is generally formed, with a consequent detrimental effect on the open circuit voltage (Voc ) of the respective perovskite solar cells (PSCs). To face this issue, surface passivation of HPs has been improved and large organic salts are ranked among the best candidates to perform it [1]. Unfortunately, the exploration of a plethora of organic salts is driven by the trial-and-error approach, which is time and money-consuming, with limited insight into the organic salts’ features which most affect the effectiveness of the passivation. In this context machine learning (ML) methods may emerge as a valuable tool to guide experimental efforts thanks to a deeper comprehension of the passivation mechanism itself. In this study, we propose a ML approach with Shapley additive explanation to get the organic cation’s features governing the Voc optimization. We found that low halide fraction and hetero atom carbon ratio correlate with increased PSCs Voc. Through optical and morphological characterization of HPs, we concluded that the beneficial role of low halide fraction is dominated by the light Cl- , whose, with its strong binding capacity to positively charged defects in HPs, reduces non-radiative recombination, while the low hetero atom carbon ratio affects the increased flexibility of the molecule, resulting in better coverage of the surface. Finally, we assessed and confirmed the force of the ML algorithm, by using it directly to make predictions about the experimental Voc that would obtain a PSC in which the HP is passivated with a new cation [2].
[1] Sam Teale, Matteo Degani, Bin Chen, Edward Sargent, Giulia Grancini, Nature Energy, accepted
[2] Mattia Ragni, Fabiola Faini, Matteo Degani, Ian Postuma, Giulia Grancini, submitted
The authors acknowledge the “HY-NANO” project that received funding from the European Research Council (ERC) Starting Grant 2018 under the European Union’s Horizon 2020 research and innovation program (Grant Agreement No. 802862) and the “SPIKE” project that received funding from the European Research Council (ERC) Proof of Concept 2022 under the European Union’s Horizon 2020 research and innovation program (Grant Agreement No. 101068936) and the support from the Ministero dell’Università e della Ricerca (MUR) and the University of Pavia through the program “Dipartimenti di Eccellenza 2023-2027".