Multiscale Modelling and Simulation of Organic Photovoltaic Materials

SNIC 2022/1-23


SNIC Large Compute

Principal Investigator:

Carlos Moyses Graca Araujo


Karlstads universitet

Start Date:


End Date:


Primary Classification:

10304: Condensed Matter Physics

Secondary Classification:

10302: Atom and Molecular Physics and Optics

Tertiary Classification:

10402: Physical Chemistry



Non-Fullerene Acceptors (NFA) are receiving a great deal of attention due to recent breakthroughs in the power conversion efficiency of organic photovoltaic devices obtained with this type of materials. However, a fundamental understanding of the underlying photophysics of the polymer:NFA blends is still lacking. NFAs are able of absorbing light in the visible range opening a new channel for charge photogeneration, which involves a hole-transfer process from the acceptor to the donor. Here, many fundamental questions arise, including: (i) the exciton-state energy transfer at the interfaces; (ii) the low non-radiative voltage losses and high photocurrent generation quantum efficiency in low-gap systems and (iii) exciton dissociation at interfaces with low ionization potential and electron affinity offsets. In this project, we will develop and employed multiscale modelling and simulation of polymer:NFA complexes within the framework of density functional theory (DFT) and time-dependent-DFT (TDDFT), combined with molecular dynamics simulations. Some of the most representative NFAs (e.g. Y5, Y6, perylene and indacenodithiophene derivatives) and efficient polymer-donors (e.g. PBDB-T) will be investigated. The MD simulations will cover two time and length scales, through the all atom molecular dynamics (AAMD) and coarse-grained molecular dynamics (CGMD) simulations. These simulations will be used to understand and predict the intermolecular interactions of oligomers and polymers in the blend structures, and even include the effect of the solvents. The MD and DFT coupling will be achieved in two-ways: MD will provide input to DFT in the form of states from the configurations space, while DFT will provide information (e.g. charge states) to further refine the force fields. Observables – suitable averaged quantities – will be computed so that our computational approach can be verified against experiment. Here, we will focus on the electronic structure, optical and transport properties. The TDDFT will be used to describe the optical transitions and excited states. Although similar methodology has been extensively employed to study solvated molecules (in liquid environment), it has been applied for polymeric systems only in a limited number of studies. Therefore, the development of such methodology for the OPVs is a novelty aspect of our proposal. Additionally, genetic algorithms interplayed with DFT will also be implemented to predict the crystal structures of the organic compounds. Finally, neural network (NN) models will be trained on the DFT data to predict polymer:NFA complexes by passing the time demanding first-principles calculations. The NN models will allow the investigation of a large materials library facilitating the prediction of novel OPVs. This is an interdisciplinary project that will involve the departments of Physics, Chemistry and Mathematics at Karlstad University. Furthermore, it will be carried out in close in-house collaboration with the experimentalists.