Abstract
Quantum chemistry calculations have been very useful in providing many key detailed properties and enhancing our understanding of molecular systems. However, such calculation, especially with ab initio models, can be time-consuming. For example, in the prediction of charge-transfer properties, it is often necessary to work with an ensemble of different thermally populated structures. A possible alternative to such calculations is to use a machine-learning based approach. In this work, we show that the general prediction of electronic coupling, a property that is very sensitive to intermolecular degrees of freedom, can be obtained with artificial neural networks, with improved performance as compared to the popular kernel ridge regression method. We propose strategies for optimizing the learning rate and batch size, improving model performance, and further evaluating models to ensure that the physical signatures of charge-transfer coupling are well reproduced. We also address the effect of feature representation as well as statistical insights obtained from the loss function and the data structure. Our results pave the way for designing a general strategy for training such neural-network models for accurate prediction.
Highlights
In the search for novel molecular materials with desirable properties, it is important to establish reliable theoretical predictions.1 When simple theoretical approaches reach the limit of generating useful prediction, computer simulations, especially those that do not depend on empirical parameters, become essential
We found that percentages of failing the orientation tests are 9.2% (9/98), 4.6% (3/65), and 15.6% (10/64) for models trained by LMAE, LMSE, and LHuber, respectively
We developed a scheme to build an artificial neural networks (ANNs) model
Summary
In the search for novel molecular materials with desirable properties, it is important to establish reliable theoretical predictions. When simple theoretical approaches reach the limit of generating useful prediction, computer simulations, especially those that do not depend on empirical parameters, become essential. When simple theoretical approaches reach the limit of generating useful prediction, computer simulations, especially those that do not depend on empirical parameters, become essential. This is true especially in the case of organic optoelectronic materials. In the prediction of charge transport properties of organic semiconducting materials, conventional Marcus theory for charge-hopping and the band theory are both limited; the intermolecular electronic coupling is often large enough to delocalize the charge, whose dynamics couples with nuclear movements. It is not large enough either, to form a robust polaron band where a standard solid-state theory can apply.. It is not large enough either, to form a robust polaron band where a standard solid-state theory can apply. A direct simulation containing both quantum and classical degrees of freedom of the problem is perhaps the best solution, given the tremendous advancement of computational technologies in this modern era.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
