Deep Learning-Driven Molecular Generation and Electrochemical Property Prediction for Optimal Electrolyte Additive Design

Recently, generative models have rapidly advanced and are being applied to various domains beyond vision and large language models (LLMs). In the field of chemistry and molecular generation, deep learning-based models are increasingly utilized to reduce experimental exploration and research costs. I...

Full description

Saved in:
Bibliographic Details
Main Authors: Dongryun Yoon, Jaekyu Lee, Sangyub Lee
Format: Article
Language:English
Published: MDPI AG 2025-03-01
Series:Applied Sciences
Subjects:
Online Access:https://www.mdpi.com/2076-3417/15/7/3640
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850212598797041664
author Dongryun Yoon
Jaekyu Lee
Sangyub Lee
author_facet Dongryun Yoon
Jaekyu Lee
Sangyub Lee
author_sort Dongryun Yoon
collection DOAJ
description Recently, generative models have rapidly advanced and are being applied to various domains beyond vision and large language models (LLMs). In the field of chemistry and molecular generation, deep learning-based models are increasingly utilized to reduce experimental exploration and research costs. In this study, we conducted research on Variational Autoencoder-based molecular generation and property prediction to screen for optimal molecules in the design of electrolyte additives for lithium-ion batteries. Using a dataset composed of promising electrolyte additive candidate molecules, we generated new molecules and predicted HOMO and LUMO values, which are key factors in electrolyte additive design. For approximately 1000 newly generated electrolyte additive candidate molecules, we performed DFT calculations to obtain HOMO and LUMO values and calculated the mean absolute error (MAE) between the predicted values from the trained model and the DFT-calculated values. As a result, the model demonstrated exceptionally low errors of approximately 0.04996 eV (HOMO) and 0.06895 eV (LUMO), respectively. This means that battery experts can receive recommendations for new molecules, refer to their predicted HOMO and LUMO values, and select potential electrolyte additives for further validation through experiments. By replacing the traditional electrolyte additive development process with deep learning models, this method has the potential to significantly reduce the overall development time and improve efficiency.
format Article
id doaj-art-182d907a89bf455d91cc96c557fc0f78
institution OA Journals
issn 2076-3417
language English
publishDate 2025-03-01
publisher MDPI AG
record_format Article
series Applied Sciences
spelling doaj-art-182d907a89bf455d91cc96c557fc0f782025-08-20T02:09:18ZengMDPI AGApplied Sciences2076-34172025-03-01157364010.3390/app15073640Deep Learning-Driven Molecular Generation and Electrochemical Property Prediction for Optimal Electrolyte Additive DesignDongryun Yoon0Jaekyu Lee1Sangyub Lee2Energy IT Convergence Research Center, Korea Electronics Technology Institute, #25, Saenariro, Bundang-gu, Seongnam-si 13509, Gyeonggi-do, Republic of KoreaEnergy IT Convergence Research Center, Korea Electronics Technology Institute, #25, Saenariro, Bundang-gu, Seongnam-si 13509, Gyeonggi-do, Republic of KoreaEnergy IT Convergence Research Center, Korea Electronics Technology Institute, #25, Saenariro, Bundang-gu, Seongnam-si 13509, Gyeonggi-do, Republic of KoreaRecently, generative models have rapidly advanced and are being applied to various domains beyond vision and large language models (LLMs). In the field of chemistry and molecular generation, deep learning-based models are increasingly utilized to reduce experimental exploration and research costs. In this study, we conducted research on Variational Autoencoder-based molecular generation and property prediction to screen for optimal molecules in the design of electrolyte additives for lithium-ion batteries. Using a dataset composed of promising electrolyte additive candidate molecules, we generated new molecules and predicted HOMO and LUMO values, which are key factors in electrolyte additive design. For approximately 1000 newly generated electrolyte additive candidate molecules, we performed DFT calculations to obtain HOMO and LUMO values and calculated the mean absolute error (MAE) between the predicted values from the trained model and the DFT-calculated values. As a result, the model demonstrated exceptionally low errors of approximately 0.04996 eV (HOMO) and 0.06895 eV (LUMO), respectively. This means that battery experts can receive recommendations for new molecules, refer to their predicted HOMO and LUMO values, and select potential electrolyte additives for further validation through experiments. By replacing the traditional electrolyte additive development process with deep learning models, this method has the potential to significantly reduce the overall development time and improve efficiency.https://www.mdpi.com/2076-3417/15/7/36403D molecule generationelectrolyte additive designmolecular property regression
spellingShingle Dongryun Yoon
Jaekyu Lee
Sangyub Lee
Deep Learning-Driven Molecular Generation and Electrochemical Property Prediction for Optimal Electrolyte Additive Design
Applied Sciences
3D molecule generation
electrolyte additive design
molecular property regression
title Deep Learning-Driven Molecular Generation and Electrochemical Property Prediction for Optimal Electrolyte Additive Design
title_full Deep Learning-Driven Molecular Generation and Electrochemical Property Prediction for Optimal Electrolyte Additive Design
title_fullStr Deep Learning-Driven Molecular Generation and Electrochemical Property Prediction for Optimal Electrolyte Additive Design
title_full_unstemmed Deep Learning-Driven Molecular Generation and Electrochemical Property Prediction for Optimal Electrolyte Additive Design
title_short Deep Learning-Driven Molecular Generation and Electrochemical Property Prediction for Optimal Electrolyte Additive Design
title_sort deep learning driven molecular generation and electrochemical property prediction for optimal electrolyte additive design
topic 3D molecule generation
electrolyte additive design
molecular property regression
url https://www.mdpi.com/2076-3417/15/7/3640
work_keys_str_mv AT dongryunyoon deeplearningdrivenmoleculargenerationandelectrochemicalpropertypredictionforoptimalelectrolyteadditivedesign
AT jaekyulee deeplearningdrivenmoleculargenerationandelectrochemicalpropertypredictionforoptimalelectrolyteadditivedesign
AT sangyublee deeplearningdrivenmoleculargenerationandelectrochemicalpropertypredictionforoptimalelectrolyteadditivedesign