Computational Analysis of Performance in Perovskite and Tandem Solar Cells

Abstract

This thesis explores the advancement of perovskite solar cells (PSCs) through interdisciplinary methodologies, combining device physics simulations, material science, and machine learning (ML). Perovskite solar cells have emerged as a transformative technology in photovoltaic research due to their exceptional light absorption, defect tolerance, and tunable properties. However, challenges such as surface passivation, material selection, and efficient fabrication methods remain critical barriers to realizing their full potential. The work is organized into several focal areas. First, it investigates the structural and bandgap design of PSCs to optimize light absorption and energy conversion. Advanced computational simulations assess the geometrical and optical characteristics of perovskite films, ensuring precise modeling for device optimization. Emphasis is placed on surface passivation strategies, particularly the use of alkyl ammonium iodides, to enhance device stability and efficiency by mitigating surface defects and recombination losses. A data-driven framework employing machine learning was developed to analyze and predict the performance of PSCs. By integrating material properties and synthesis parameters into interpretable ML models, the study identifies key factors influencing device performance. This approach enables the optimization of PSCs through targeted experimentation informed by computational insights. Furthermore, this thesis explores tandem solar cell configurations, focusing on light management, current matching, and efficiency measurement advancements, with particular attention to the challenges in monolithic perovskite/silicon tandem technologies. Strategies to reduce measurement times for mass production and mitigate metastability issues are proposed. Lastly, the integration of AI into autonomous experimental design and device simulations is presented as a future direction, highlighting the potential for self-driven laboratories in accelerating solar cell innovation. This comprehensive study aims to bridge the gap between theoretical modeling, experimental validation, and data-centric analysis, providing a pathway for the development of next-generation perovskite solar cells with enhanced efficiency, stability, and scalability.