Prediction of the Appropriate Temperature and Pressure for Polymer Dissolution Using Machine Learning Models | JoVE Visualize

Area of Science:

Chemical Engineering
Materials Science
Computational Chemistry

Background:

Polymer solutions are vital in the chemical industry, but determining optimal dissolution conditions is challenging.
Traditional experimental methods for estimating polymer dissolution parameters are time-consuming and expensive.
Predicting dissolution parameters requires understanding complex interactions between polymer molecular weight, chemical structure, solvent properties, and weight percent.

Purpose of the Study:

To develop a novel machine learning-based approach for predicting the minimum temperature and pressure required for polymer dissolution.
To establish correlations between polymer and solvent molecular characteristics and their dissolution parameters.
To offer a more efficient alternative to traditional experimental methods for determining polymer processing conditions.

Main Methods:

Compiled a dataset from existing literature on polymer dissolution, including molecular weight, chemical structure, solvent properties, and weight percent.
Extracted chemical bond information from the molecular structures of polymer-solvent systems.
Trained and evaluated six machine learning algorithms: linear regression, k-nearest neighbors, regression trees, random forests, multilayer perceptron neural networks, and support vector regression.

Main Results:

The Random Forest model demonstrated the highest predictive accuracy for both temperature (R² = 0.931) and pressure (R² = 0.942).
The models successfully correlated molecular weight, chemical structure, and solvent properties with dissolution parameters.
The machine learning approach significantly reduced the need for iterative experimental testing.

Conclusions:

Machine learning offers an efficient and accurate method for predicting polymer dissolution conditions.
This approach can accelerate the development and optimization of processes involving polymer solutions.
The developed models provide a valuable tool for chemical engineers and materials scientists working with polymers.