Enhanced deep-learning model for carbon footprints of chemicals

Millions of chemicals have been designed and registered to contribute to the global efforts toward accelerating the transition to sustainable chemistry; however, their product carbon footprints (PCF) are often unknown, leaving questions on their sustainability. This general lack of PCF data is because the data needed for comprehensive environmental analyses are generally not available at early design stages. Several predictive tools have been developed to estimate the PCF of chemicals, which are only applicable to a narrow range of common chemicals and have limited predictive ability. Here we propose FineChem 2, which is based on a Transformer framework and first-hand industry data, for accurately predicting the PCF of chemicals. Compared to previous tools, FineChem 2 demonstrates significantly better predictive power, and its applicability domains are improved by ~75% on high production volume chemicals, daily chemicals, and chemical additives in food and plastics. The better interpretability from the attention mechanism enables FineChem 2 to successfully identify PCF-intensive substructures and critical raw materials of chemicals, providing insights into the design of sustainable molecules and processes. Therefore, we expect wide application of FineChem 2 for chemical PCF estimations, leading to advancements in sustainable chemistry.

This repository contains code for producing figures used in the FineChem2 paper.