Solvent Engineering for High-Performance Two-Dimensional Ruddlesden–Popper CsPbI₃ Solar Cells

Abstract

Two-dimensional (2D) Ruddlesden–Popper (RP) CsPbI3 exhibits enhanced phase stability compared with 3D CsPbI3. However, the issue of the uncontrollable crystallization process limits its photovoltaic performance. Here, the influence of a binary mixed solvent on the film quality and photovoltaic properties of (PEA)2Cs4Pb5I16 (n = 5) is studied in detail. It is demonstrated that the crystallization rate and crystal growth can be controlled by adjusting the amount of dimethyl sulfoxide (DMSO). Optimizing the solvent composition with adding 10% DMSO in pure dimethyl formamide (DMF) leads to perfect coverage, larger flaky 2D grains, reduced grain boundaries, and a better vertical orientation to the substrate due to the formation of a more stable intermediate phase. This can form good interface contact, which is beneficial to charge transport/extraction between TiO2 (electron transport layer, ETL) and perovskite, finally resulting in improved device performance. The enhancement of the power conversion efficiency of the optimized device based on DMF/DMSO (9:1) is 3.57% compared with the reference device based on pure DMF. This work illustrates the role of crystallization kinetics in the RP CsPbI3 film and offers a simple and effective method for high-performance 2D CsPbI3 solar cells.

Accelerated Multisolvent Prediction for Aqueous Stable Halide Perovskite Materials

“…13 10% DMSO is added into DMF to optimize the solvent combination, achieving improved film coverage, reduced boundary area, and stable intermediate phase for perovskite solar cell application. 14 Nevertheless, the virtual design space of multisolvents is highly complex, which is inaccessible by traditional experimental and simulation paradigms. For instance, with a presence of 100 basic solvents in the laboratory, the combination of three random solvents can lead to 10 6 choices, which calls for a more effective and less time-consuming inverse design process, and such multidimensional issue is even more prominent if more solvents are combined together (e.g., singular → binary → ternary → quaternary → quinary) that can be hardly achieved by the traditional trial-and-error methods.…”

“…For example, Jeon and Noh mixed γ-butyrolactone and DMSO solvents to form a homogeneous and densely packed perovskite layer via the formation of a CH 3 NH 3 I–PbI 2 -DMSO intermedia phase . 10% DMSO is added into DMF to optimize the solvent combination, achieving improved film coverage, reduced boundary area, and stable intermediate phase for perovskite solar cell application . Nevertheless, the virtual design space of multisolvents is highly complex, which is inaccessible by traditional experimental and simulation paradigms.…”