Molecular design and performance improvement in organic solar cells guided by high‐throughput screening and machine learning

Abstract

Over past two decades, organic photovoltaics (OPVs) with unique advantages of low cost and flexibility meet significant development opportunities and the official world record for the power conversion efficiency (PCE) of organic solar cells (OSCs) has reached to 17.3%. Traditionally, efficiency breakthrough need the constant input of intensive labor and time. The artificial intelligence, as a rising interdisciplinary, brings certainly a revolution in research methods. In this review, we introduce a state‐of‐art theoretical methodology of the synergy of high‐throughput screening and machine learning (ML) in accelerating the discovery of high‐efficient OSC materials. We present key details, rules and experience in database construction, selection of molecular features, fast‐screening calculations, models training and their predication capabilities. Meanwhile, three typical ML frameworks are concluded to reveal the structure‐property‐efficiency relationship, suggesting that this theoretical methodology can train powerful models with just molecular configurations and theoretical calculations for molecular design and efficiency improvements.

Molecular Modeling in Anion Exchange Membrane Research: A Brief Review of Recent Applications

“…In recent years, many researchers studied the development and applications of machine learning (ML) models in molecular design and performance improvement of energy materials and devices [ 154 , 155 , 156 , 157 , 158 , 159 ]. ML trained, based on QM/MM, may allow MD simulations at an accurate level close to the electronic-structure method chosen to generate a training set [ 160 , 161 ].…”

Charge-Transfer Landscape Manifesting the Structure–Rate Relationship in the Condensed Phase Via Machine Learning

“…Machine learning (ML) has recently emerged as a powerful technique for obtaining a highly accurate prediction about physical properties with significantly reduced computational cost. − In the OPV field alone, many fruitful works have been reported for the prediction of molecular properties such as electronic coupling, ,,− reorganization energy, , and energy gap, − as well as device-level properties such as power conversion efficiency, open-circuit voltage, short-circuit current density, and fill factor. ,, However, an ML model for directly predicting the CT rate in the condensed phase based on bottom-up atomistic description is still missing.…”

Simultaneous Optimization of Donor/Acceptor Pairs and Device Specifications for Nonfullerene Organic Solar Cells Using a QSPR Model with Morphological Descriptors

“…8 In that context, reliable quantitative structure-property relationship (QSPR) between photoactive materials and device efficiencies is a shortcut for further improving PCE of OSCs 9 , and data-driven machine learning (ML) is one of the promising technologies in this scheme. 10 The impact of D and A molecules on PCE of OSCs is non-additive, and in particular, synergies have been observed for some D/A pairs. 11,12 However, most theoretical studies in this field were focused on optimizing either D or A separately.…”

“…8 Optimal molecular structures for D/A pairs and bulkheterojunction characteristics can reduce exciton recapture. In that context, reliable quantitative structure−property relationship (QSPR) models between the descriptors of photoactive materials and device efficiencies have the potential to improve the PCE of OSCs prior to any experiment, 9 and data-driven machine learning (ML) 10 is an efficient and flexible method for developing QSPR models for such complex systems.…”