Low-cost machine learning prediction of excited state properties of iridium-centered phosphors
Prediction of the excited state properties of photoactive iridium complexes challenges ab initio methods such as time-dependent density functional theory (TDDFT) both from the perspective of accuracy and of computational cost, complicating high-throughput virtual screening (HTVS). We instead leverag...
| Main Authors: | , , , |
|---|---|
| Outros Autores: | |
| Formato: | Artigo |
| Idioma: | English |
| Publicado em: |
Royal Society of Chemistry
2024
|
| Acesso em linha: | https://hdl.handle.net/1721.1/156915 |
| _version_ | 1826217428147765248 |
|---|---|
| author | Terrones, Gianmarco G Duan, Chenru Nandy, Aditya Kulik, Heather J |
| author2 | Massachusetts Institute of Technology. Department of Chemistry |
| author_facet | Massachusetts Institute of Technology. Department of Chemistry Terrones, Gianmarco G Duan, Chenru Nandy, Aditya Kulik, Heather J |
| author_sort | Terrones, Gianmarco G |
| collection | MIT |
| description | Prediction of the excited state properties of photoactive iridium complexes challenges ab initio methods such as time-dependent density functional theory (TDDFT) both from the perspective of accuracy and of computational cost, complicating high-throughput virtual screening (HTVS). We instead leverage low-cost machine learning (ML) models and experimental data for 1380 iridium complexes to perform these prediction tasks. We find the best-performing and most transferable models to be those trained on electronic structure features from low-cost density functional tight binding calculations. Using artificial neural network (ANN) models, we predict the mean emission energy of phosphorescence, the excited state lifetime, and the emission spectral integral for iridium complexes with accuracy competitive with or superseding that of TDDFT. We conduct feature importance analysis to determine that high cyclometalating ligand ionization potential correlates to high mean emission energy, while high ancillary ligand ionization potential correlates to low lifetime and low spectral integral. As a demonstration of how our ML models can be used for HTVS and the acceleration of chemical discovery, we curate a set of novel hypothetical iridium complexes and use uncertainty-controlled predictions to identify promising ligands for the design of new phosphors while retaining confidence in the quality of the ANN predictions. |
| first_indexed | 2024-09-23T17:03:29Z |
| format | Article |
| id | mit-1721.1/156915 |
| institution | Massachusetts Institute of Technology |
| language | English |
| last_indexed | 2025-02-19T04:26:23Z |
| publishDate | 2024 |
| publisher | Royal Society of Chemistry |
| record_format | dspace |
| spelling | mit-1721.1/1569152025-02-14T15:53:17Z Low-cost machine learning prediction of excited state properties of iridium-centered phosphors Terrones, Gianmarco G Duan, Chenru Nandy, Aditya Kulik, Heather J Massachusetts Institute of Technology. Department of Chemistry Massachusetts Institute of Technology. Department of Chemical Engineering Prediction of the excited state properties of photoactive iridium complexes challenges ab initio methods such as time-dependent density functional theory (TDDFT) both from the perspective of accuracy and of computational cost, complicating high-throughput virtual screening (HTVS). We instead leverage low-cost machine learning (ML) models and experimental data for 1380 iridium complexes to perform these prediction tasks. We find the best-performing and most transferable models to be those trained on electronic structure features from low-cost density functional tight binding calculations. Using artificial neural network (ANN) models, we predict the mean emission energy of phosphorescence, the excited state lifetime, and the emission spectral integral for iridium complexes with accuracy competitive with or superseding that of TDDFT. We conduct feature importance analysis to determine that high cyclometalating ligand ionization potential correlates to high mean emission energy, while high ancillary ligand ionization potential correlates to low lifetime and low spectral integral. As a demonstration of how our ML models can be used for HTVS and the acceleration of chemical discovery, we curate a set of novel hypothetical iridium complexes and use uncertainty-controlled predictions to identify promising ligands for the design of new phosphors while retaining confidence in the quality of the ANN predictions. 2024-09-20T15:31:56Z 2024-09-20T15:31:56Z 2023 2024-09-20T15:25:21Z Article http://purl.org/eprint/type/JournalArticle https://hdl.handle.net/1721.1/156915 Chem. Sci., 2023,14, 1419-1433 en 10.1039/d2sc06150c Chemical Science Creative Commons Attribution https://creativecommons.org/licenses/by/3.0/ application/pdf Royal Society of Chemistry Royal Society of Chemistry |
| spellingShingle | Terrones, Gianmarco G Duan, Chenru Nandy, Aditya Kulik, Heather J Low-cost machine learning prediction of excited state properties of iridium-centered phosphors |
| title | Low-cost machine learning prediction of excited state properties of iridium-centered phosphors |
| title_full | Low-cost machine learning prediction of excited state properties of iridium-centered phosphors |
| title_fullStr | Low-cost machine learning prediction of excited state properties of iridium-centered phosphors |
| title_full_unstemmed | Low-cost machine learning prediction of excited state properties of iridium-centered phosphors |
| title_short | Low-cost machine learning prediction of excited state properties of iridium-centered phosphors |
| title_sort | low cost machine learning prediction of excited state properties of iridium centered phosphors |
| url | https://hdl.handle.net/1721.1/156915 |
| work_keys_str_mv | AT terronesgianmarcog lowcostmachinelearningpredictionofexcitedstatepropertiesofiridiumcenteredphosphors AT duanchenru lowcostmachinelearningpredictionofexcitedstatepropertiesofiridiumcenteredphosphors AT nandyaditya lowcostmachinelearningpredictionofexcitedstatepropertiesofiridiumcenteredphosphors AT kulikheatherj lowcostmachinelearningpredictionofexcitedstatepropertiesofiridiumcenteredphosphors |