PERFORMANCE ENHANCEMENT OF FLEXIBLE THIN-FILM ORGANIC SOLAR CELLS

Abstract

This thesis is set within the current context of the energy transition, where the optimization of lightweight, flexible, and low-cost photovoltaic technologies represents a major challenge. It explores organic solar cells (OSCs) through an approach that combines numerical modeling, algorithmic optimization, artificial intelligence, and experimental validation. First, a state-of-the-art review positioned OSCs among other photovoltaic technologies and identified their main challenges in terms of efficiency and stability. Then, using the SCAPS-1D simulator, different optimization strategies were investigated. For tandem cells, we focused on the PM6:PY-IT (bottom cell) and PM7:PIDT (top cell) systems. Adjusting the thickness and defect density of the active layers, as well as engineering the electron transport layers (ETLs), enabled us to achieve remarkable efficiencies (PCE = 24.5% and 22.83%). Furthermore, we studied the engineering of transport layers in a single-junction cell based on the PM6:PY-IT absorber, where several ETL combinations were tested. Among them, the PCBM/TiO₂ configuration showed a significant improvement in charge extraction (PCE = 19.35%, VOC = 1.14 V, FF = 73.76%). Machine Learning models, particularly XGBoost, were then employed to predict the compatibility and performance of materials with high accuracy (R² = 0.9957). Finally, a practical demonstration was carried out by integrating an OPV module into a smartphone charging system, illustrating the potential of autonomous organic photovoltaic solutions for portable electronics and embedded systems. The results highlight the relevance of a multidisciplinary approach and open up new perspectives, particularly in improving long-term stability, exploring new materials, and integrating OSCs into smart systems and large-scale applications