De Novo Design of Molecules with Low Hole Reorganization Energy Based on a Quarter-Million Molecule DFT Screen: Part 2.

Organic semiconductors have many desirable properties including improved manufacturing and flexible mechanical properties. Due to the vastness of chemical space, it is essential to efficiently explore chemical space when designing new materials, including through the use of generative techniques. New generative machine learning methods for molecular design continue to be published in the literature at a significant rate but successfully adapting methods to new chemistry and problem domains remains difficult. These challenges necessitate continual method evaluation to probe method viability for use in alternative applications not covered in the original works. In continuation of our previous work, we evaluate four additional machine-learning-based de novo methods for generating molecules with high predicted hole mobility for use in semiconductor applications. The four generative methods evaluated here are (1) Molecule Deep Q-Networks (MolDQN), which utilizes Deep-Q learning to directly optimize molecular structure graphs for desired properties instead of generating SMILES, (2) Graph-based Genetic Algorithm (GraphGA), which uses a genetic algorithm for optimization where crossovers and mutations are defined in terms of RDKit's reaction SMILES, (3) Generative Tensorial Reinforcement Learning (GENTRL), which is a variational autoencoder (VAE) with a learned prior distribution and optimized using reinforcement learning, and (4) Monte Carlo tree search exploration of chemical space in conjunction with a recurrent neural network (RNN) decoder (ChemTS). The generated molecules were evaluated using density functional theory (DFT) and we discovered better performing molecules with the GraphGA method compared to the other approaches.

De Novo Design of Molecules with Low Hole Reorganization Energy Based on a Quarter-Million Molecule DFT Screen.

Materials exhibiting higher mobilities than conventional organic semiconducting materials such as fullerenes and fused thiophenes are in high demand for applications in printed electronics. To discover new molecules in the heteroacene family that might show improved hole mobility, three de novo design methods were applied. Machine learning (ML) models were generated based on previously calculated hole reorganization energies of a quarter million examples of heteroacenes, where the energies were calculated by applying density functional theory (DFT) and a massive cloud computing environment. The three generative methods applied were (1) the continuous space method, where molecular structures are converted into continuous variables by applying the variational autoencoder/decoder technique; (2) the method based on reinforcement learning of SMILES strings (the REINVENT method); and (3) the junction tree variational autoencoder method that directly generates molecular graphs. Among the three methods, the second and third methods succeeded in obtaining chemical structures whose DFT-calculated hole reorganization energy was lower than the lowest energy in the training dataset. This suggests that an extrapolative materials design protocol can be developed by applying generative modeling to a quantitative structure-property relationship (QSPR) utility function.

Machine-Learning Guided Quantum Chemical and Molecular Dynamics Calculations to Design Novel Hole-Conducting Organic Materials.

Materials exhibiting higher mobilities than conventional organic semiconducting materials such as fullerenes and fused thiophenes are in high demand for applications such as printed electronics, organic solar cells, and image sensors. In order to discover new molecules that might show improved charge mobility, combined density functional theory (DFT) and molecular dynamics (MD) calculations were performed, guided by predictions from machine learning (ML). A ML model was constructed based on 32 values of theoretically calculated hole mobilities for thiophene derivatives, benzodifuran derivatives, a carbazole derivative and a perylene diimide derivative with the maximum value of 10-1.96 cm2/(V s). Sequential learning, also known as active learning, was applied to select compounds on which to perform DFT/MD calculation of hole mobility to simultaneously improve the mobility surrogate model and identify high mobility compounds. By performing 60 cycles of sequential learning with 165 DFT/MD calculations, a molecule having a fused thioacene structure with its calculated hole mobility of 10-1.86 cm2/(V s) was identified. This values is higher than the maximum value of mobility in the initial training data set, showing that an extrapolative discovery could be made with the sequential learning.

Massive Theoretical Screen of Hole Conducting Organic Materials in the Heteroacene Family by Using a Cloud-Computing Environment.

Materials exhibiting higher mobilities than conventional organic semiconducting materials such as fullerenes and fused thiophenes are in high demand for applications in printed electronics. To discover new molecules in the heteroacene family that might show improved charge mobility, a massive theoretical screen of hole conducting properties of molecules was performed by using a cloud-computing environment. Over 7 000 000 structures of fused furans, thiophenes and selenophenes were generated and 250 000 structures were randomly selected to perform density functional theory (DFT) calculations of hole reorganization energies. The lowest hole reorganization energy calculated was 0.0548 eV for a fused thioacene having 8 aromatics rings. Hole mobilities of compounds with the lowest 130 reorganization energy were further processed by applying combined DFT and molecular dynamics (MD) methods. The highest mobility calculated was 1.02 and 9.65 cm2/(V s) based on percolation and disorder theory, respectively, for compounds containing selenium atoms with 8 aromatic rings. These values are about 20 times higher than those for dinaphthothienothiophene (DNTT).