Polymer Informatics Method for Fast and Accurate Prediction of the Glass Transition Temperature from Chemical...

Area of Science:

• Polymer Science
• Materials Informatics
• Computational Chemistry

Background:

• Predicting the glass transition temperature (Tg) of polymers is crucial for material design and application.
• Traditional methods like group additive properties (GAP) and quantitative structure-property relationships (QSPR) have limitations when used independently.
• A unified approach is needed to leverage the strengths of both GAP and QSPR for improved Tg prediction.

Purpose of the Study:

• To develop and validate a novel polymer informatics framework integrating GAP and QSPR methodologies.
• To accurately predict the glass transition temperature (Tg) of polymers based on their chemical structures.
• To identify key molecular descriptors influencing Tg predictions in polymers.

Main Methods:

• Development of a hybrid QSPR-GAP framework combining additive group contributions and molecular descriptors.
• Application of the framework to a dataset of 146 linear homo- and copolymers from the poly-(aryl ether ketone) (PAEK) family.
• Utilizing a genetic algorithm to identify dominant molecular descriptors driving Tg predictions.

Main Results:

• The QSPR-GAP framework achieved a median root mean square error of 8 K for Tg prediction in the PAEK dataset.
• This represents a significant improvement compared to standalone QSPR or GAP models.
• Two molecular descriptors were identified as primary drivers of Tg predictions.

Conclusions:

• The integrated QSPR-GAP framework offers a powerful and accurate method for predicting polymer Tg.
• The framework demonstrates versatility and can be adapted for predicting other physical properties and activities (QSAR).
• This approach is transferable to diverse polymer families, including conjugated and biopolymers.