Case Study: Grace achieves rapid thermodynamic property estimation with Citrine Informatics

Grace estimates thermodynamic properties with Citrine Informatics using 127+ molecular features

The specialty chemicals company Grace needed to optimize its manufacturing processes but faced a challenge: key thermodynamic properties for many molecules could not be measured experimentally due to their constantly changing forms. To address this, Grace worked with Citrine Informatics and used the Citrine Platform to develop a new method for estimating these essential properties.

Citrine Informatics implemented a solution using machine learning models on their platform to predict critical properties based solely on a molecule's SMILES string. The resulting models achieved high accuracy, which was successfully validated on a challenging test molecule, trimethyl aluminum dimer. This new capability allows Grace to rapidly estimate properties without experimental measurement, accelerating their process optimization and saving significant time.

Grace