Deep Learning Approaches for Predicting the Surface Tension of Ionic Liquids

Area of Science:

• Materials Science
• Computational Chemistry
• Chemical Engineering

Background:

• Ionic liquids (ILs) are versatile solvents with tunable properties crucial for applications like electrolytes and heat transfer fluids.
• Experimental determination of IL properties, such as surface tension, is challenging due to the vast number of possible combinations and high costs.
• Computational methods are essential for accelerating the discovery and design of ILs with desired characteristics.

Purpose of the Study:

• To develop accurate deep learning (DL) models for predicting the surface tension of ionic liquids (ILs).
• To utilize simplified molecular input line entry system (SMILES) representations for feature extraction in DL models.
• To provide a computational tool for rapid screening and rational design of ILs with specific surface tension values.

Main Methods:

• Development of two deep learning models using SMILES representations of ionic liquids.
• Training and validation of the models using a broad range of experimental surface tension data across various temperatures.
• Evaluation of model performance using R-squared and root-mean-square error metrics.

Main Results:

• Both DL models achieved high accuracy in predicting IL surface tension, with an R-squared value of 0.990.
• The models demonstrated a low root-mean-square error of 0.792 mN/m, indicating excellent agreement with experimental data.
• The models effectively captured the relationship between IL molecular structure and surface tension.

Conclusions:

• Deep learning models based on SMILES representations offer a powerful and efficient approach for predicting ionic liquid surface tension.
• These predictive models can significantly reduce the time and cost associated with experimental characterization.
• The developed models facilitate the rational design and discovery of novel ionic liquids with tailored surface tension properties for diverse industrial applications.