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

Abstract

In this work, we develop a machine learning (ML) strategy to map the molecular structure to condensed phase charge-transfer (CT) properties including CT rate constants, energy levels, electronic couplings, energy gaps, reorganization energies, and reaction free energies which are called CT fingerprints. The CT fingerprints of selected landmark structures covering the conformation space of an organic photovoltaic molecule dissolved in an explicit solvent are computed and used to train ML models using kernel ridge regression. The ML models show high predictive power with R 2 > 0.97 and both mean absolute error and root-mean-square error within chemical accuracy. The CT landscape for millions of molecular dynamics sampled structures is thus constructed, which allows for instant prediction of CT rate properties, given any conformation of the molecule. We demonstrate some immediate utilities of the CT landscape such as calculating the ensemble-averaged CT rate constant and interpreting the effects of molecular structural features on the CT rate. The unprecedented CT landscape will be useful for investigating real-time CT dynamics in nanoscale-and mesoscale-condensed phase systems and for the optimal fabrication design for homogeneous and heterogeneous optoelectronic devices.

All-Atom Nonadiabatic Semiclassical Mapping Dynamics for Photoinduced Charge Transfer of Organic Photovoltaic Molecules in Explicit Solvents

“…In the photoinduced CT processes in the solution phase, the triad is initially prepared on the ground (G) state, CPC 60 , in thermal equilibrium with the solvent, and then suddenly gets vertically photoexcited to the P-localized excitonic ππ* state, CP*C 60 . Then, there could occur nonradiative transitions to the excited P-to-C 60 CT state, CP + C 60 – , which is referred to as CT1, or to the excited C-to-C 60 charge-separated state, C + PC 60 – , which is referred to as CT2. , We previously found that the Marcus CT rate constant for ππ* → CT1 transition (picosecond) is larger than that for ππ* → CT2 transition (sub-microsecond), and the CT rate constants vary significantly with the triad conformations . Furthermore, LSC nonequilibrium Fermi’s golden rule (NE-FGR) ,, and its classical limit, that is, the instantaneous Marcus theory (IMT) , calculations, reveal that the triad has a significant nonequilibrium effect caused by the nonequilibrium initial nuclear state, which is manifested in the transient CT rate coefficient that can be enhanced to be about 40 times larger than the CT rate constant, dramatically changing the CT dynamics. , However, in previous studies, only a specific electronic transition can be simulated using LSC NE-FGR or IMT, that is, either pathway or pathway could be simulated on the all-atom level, where the population exchange with the third excited state is not allowed. , A general strategy for nonadiabatic dynamics on multiple electronic states with atomistic details is hence desirable.…”

Learning Conductance: Gaussian Process Regression for Molecular Electronics

“…Recently, machine learning (ML) approaches have become an alternative to traditional quantum chemical calculations, which can bring down computational cost by an order of magnitude or more. − Examples involve the generation of force fields for MD simulations, − prediction of charge transfer integrals, , or even trying to directly solve the Schrödinger equation . While artificial neural networks are attractive in situations where large data sets are available for training [On the flipside, this implies that artifical neural networks typically also require more data points for training.…”

Charge-Separated States Determined Photoinduced Electron Transfer Efficiency in a D-D-A System in an External Electric Field

“…In recent years, many researchers have also proposed many strategies to improve the charge-transfer rate. Sun and Brian found that when the nuclear degree of freedom starts from a nonequilibrium state, the transition rate of the charge-transfer process from porphyrin to C 60 can be significantly increased …”

“…Sun and Brian found that when the nuclear degree of freedom starts from a nonequilibrium state, the transition rate of the charge-transfer process from porphyrin to C 60 can be significantly increased. 23 In order to establish a better donor−acceptor system to generate long-lived charge-separated states, the photoinduced charge transfer of organic heterojunctions formed by the combination of two donors and an acceptor under the action of an external electric field is simulated. The heterojunction formed by the donor−acceptor combination was TPA-TT-BODIPY-C 60 , where triphenylamine (TPA)-terthiophene (TT) was the first donor and BF 2 -boron dipyrromethene (BODIPY) was the second donor.…”