Performance and Matching Band Structure Analysis of Tandem Organic Solar Cells Using Machine Learning Approaches

Abstract

Organic solar cells (OSCs) based on tandem configuration have been drastically studied for boosting power conversion efficiency (PCE) in the the past decade and are a potential improvement to current commercial photovoltaic solar cell technology. Although a series of promising efficiency achievements on tandem OSCs have been reported, the crucial issue is how to estimate tandem OSCs performance based on known physical properties of different photoactive materials. Herein, a set of electronic features of photovoltaic materials is trained using machine‐learning algorithms for accurate efficiency predictions of tandem OSCs. The well‐trained machine‐learning model presented herein aims at 1) designing the matching band structures of active blends in each sub‐cell and 2) identifying the characteristics of the materials to be used to achieve high PCE values. The machine‐learning approach provides an effective strategy for suggesting the ideal properties of photovoltaic materials in terms of highest occupied molecular orbital (HOMO), lowest unoccupied molecular orbital (LUMO), and bandgap, which is useful for the rational design of high‐efficiency tandem OSCs.

Effect of Increasing the Descriptor Set on Machine Learning Prediction of Small Molecule-Based Organic Solar Cells

“…Note that one could also use learning curves to guide hyperparameter optimizations or perform further optimization of the hyperpameter in the region of greatest interest as proposed, for example, in ref . Considering our best results in Table , we have improved the correlation r with respect to our previous work and the accuracy of our results are comparable to alternative ML methods proposed by others. − …”

Machine learning and molecular dynamics simulation-assisted evolutionary design and discovery pipeline to screen efficient small molecule acceptors for PTB7-Th-based organic solar cells with over 15% efficiency

“…No doubt, several studies about the machine learning analysis of organic solar cells have been reported. [24][25][26][27][28][29][30][31][32][33][34][35] So, the question is: how is the present study better than previously reported studies? In this regard, several aspects make our study unique.…”

“…In a few studies, just the energy levels were used to train the models and other properties were ignored. [33][34][35] In the present study, we have calculated a large number of molecular descriptors to consider the maximum structural and chemical features of the acceptor materials. More importantly, we have also predicted energy levels on the basis of experimental data.…”

Recent Applications of Carbon Nanotubes in Organic Solar Cells

“…In this regard, solar energy, a low-cost, renewable, naturally abundant and clean energy source, has attracted enormous research effort as a promising alternative to traditional energy sources ( Subramanyam et al, 2019 ; Zhong et al, 2020 ; Rabaia et al, 2021 ). Among the new-generation photovoltaic devices that convert solar energy into electricity, organic solar cells (OSCs) are being widely investigated owing to their low production cost, facile fabrication procedure, abundance of raw materials, easy scalability, lightweight, excellent flexibility and environmentally friendly nature ( Gusain et al, 2019 ; Lee, 2019 ; Li et al, 2020a ; Shah et al, 2020 ; Shoyiga et al, 2020 ; Liu et al, 2021 ).…”