There is an increased need to create inherently flame-retardant polymers to reduce the use of additives, coatings, loss of mechanical structure, etc. Many of the application areas for these polymers will be in environments that will experience a high temperature excursion, and hence will need a high heat distortion temperature. In amorphous polymers, the heat distortion temperature is characterized by the glass transition or Tg of the polymer. This property is typically governed by the rigidity of the polymer backbone. From the point of view of non-flammability, it is most desirable to have polymers that char since this will continue to preserve structural integrity, insulate the underlying polymer from the heat of the flame and prevent further burning. In this work we leverage existing experimental data and build statistical models using chemical fingerprints. Using machine learning techniques we present the strategy to optimize polymers with the desired charability and Tg.

Design and optimization of polymers with high charability and high Tg using machine learning

253rd National Meeting (2017)

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.