What question did this study set out to answer?

This research aims to optimize lead-free CH₃NH₃SnI₃ perovskite solar cells using machine learning and SCAPS-1D modeling.

March 12, 2026Journal of Materials Science Materials in Engineering0 citationsOpen Access

Machine learning–guided SCAPS-1D design of lead-free CH₃NH₃SnI₃ perovskite solar cells with 40.17% simulated PCE

MRMd Arifur RahmanBangladesh University of Engineering and Technology MAMohammad Jahangir AlamBangladesh University of Engineering and Technology

Key Points

This research aims to optimize lead-free CH₃NH₃SnI₃ perovskite solar cells using machine learning and SCAPS-1D modeling.
Conducted large-scale optimization using 35,585 device configurations.
Employed SCAPS-1D simulation coupled with machine learning techniques.
Utilized Random Forest algorithm to predict power conversion efficiency (PCE) accurately.
Achieved a simulated power conversion efficiency (PCE) of 40.17%.
Demonstrated strong predictive capability with R² = 0.96 and MSE = 0.210.
Identified optimal device architecture using WS₂ as an electron transport layer and V₂O₅ as a hole transport layer.

Abstract

Highlights Large-scale SCAPS-1D + ML co-optimization across 35,585 device configurations. CH₃NH₃SnI₃ with WS₂ (ETL) and V₂O₅ (HTL) emerges as the optimal architecture. Random Forest predicts PCE with R 2 = 0.96, MSE = 0.210, enabling rapid design search. 40.17% simulated PCE under 1-sun AM1.5G; sensitivity to series/shunt and interface traps quantified. Actionable design rules: ~ 0.80 µm absorber, low-doping regime, minimized interfacial defects.

Ask AI

Helpful

Bookmark

View Full Paper