Many-Body Perturbation Theory Meets Machine Learning to Discover Singlet Fission Materials

PI Noa Marom, Carnegie Mellon University
Marom Aurora ESP

A schematic illustration of the project’s workflow. Image: Noa Marom, Carnegie Mellon University

Project Summary

Supercomputers have been guiding materials discovery for the creation of more efficient organic solar cells. By combining quantum-mechanical simulations with machine learning and data science, this project will harness exascale power to revolutionize the process of photovoltaic design and advance physical understanding of singlet fission, the phenomenon whereby one photogenerated singlet exciton is converted into two triplet excitons—increasing the electricity produced.

Project Description

Singlet fission (SF) is the spontaneous conversion of one photogenerated singlet exciton (opposite-spin electron-hole pair) into two triplet excitons (same-spin electron-hole pairs). Intermolecular SF occurs in crystalline media, where it is mediated by coupling between chromophores in the excited state. Recently, there has been a surge of interest in SF thanks to its potential to significantly increase the efficiency of organic solar cells by harvesting two charge carriers from one photon. However, few materials are presently known to exhibit intermolecular SF with high efficiency, hindering the realization of solid-state SF-based solar cells. The chemical compound space of possible chromophores is infinitely vast and largely unexplored, as most research to date has focused on restricted classes of molecules (mainly acenes and their derivatives). Exploring this configuration space by experimental means alone would be unfeasible. This project takes a data-driven approach to enable computational discovery of SF materials.

Targeting the convergence of simulation, data, and learning by using machine learning (ML) algorithms to analyze data generated by quantum mechanical simulations and to steer simulations for further data acquisition, this work advances the mission of DOE by discovering new intermolecular SF chromophores, which will catalyze the realization of “third generation” solid-state SF-based organic solar cells. Structure-property correlations revealed by machine learning will advance the fundamental understanding of SF by deriving chemical insights and design rules for chromophores and crystal forms with enhanced SF efficiency and mapping the mechanisms by which noncovalent interactions between chromophores lead to intermolecular SF. In addition to SF chromophores, we may discover materials with desirable properties for other applications, such as organic light-emitting diodes, organic transistors, and dyes and pigments.

Quantum mechanical simulations of excited-state properties are combined with machine learning and data science to discover new materials for more efficient organic solar cells, reveal structure-property correlations, and advance the physical understanding of the singlet fission process, whereby one photogenerated singlet exciton is converted into two triplet excitons.

Project Type