1. Introduction to the Ionic Liquid Property Prediction Models
Using XGBoost, separate predictive models were developed for five different properties of ionic liquids: room temperature conductivity (σ), electrochemical window (ECW), melting point (Tm), as well as the pre-exponential factor (lnA) and activation energy (Ea) in the Arrhenius equation. The datasets used for training were sourced from the ILthermo ionic liquid database. The specific training processes for these models follow the methodology described in the unpublished paper ‘IoLiGen: A Data-Driven Discovery of Ionic Liquids for High-Performance Lithium-Ion Batteries Based on Small Datasets,’ which includes detailed training procedures and model performance.
2. Notes
• Users must use the provided Excel file to input relevant parameters. • Ensure the Name column does not contain illegal characters (e.g., Chinese characters, commas, periods, slashes, etc.). • Ensure the SMILES notation is correct (RDKit can be used to validate SMILES), otherwise the program will not function correctly.
3. !!!!! Important Considerations !!!!!: • After the user uploads the Ionic_liquid_list.xlsx file and clicks the Start button, calculations will begin. • Once the Start button is clicked, the user should not exit or refresh the current page, as this may result in the loss of calculation results. Please do not close or refresh the browser until the computation is complete. • By utilizing a pre-trained model for property prediction, large-scale ionic liquid property predictions can be conducted in a short amount of time.
4. Download EXCEL FILE
Download ionic liquid list EXCEL FILE
