Mapping the Carbon Footprint of Organic Chemicals in Commerce

The chemical industry is a major contributor to global CO2 emissions, accounting for ∼10% of total greenhouse gases, underscoring the need for a sustainability transition. However, public carbon footprint data exists for only hundreds of the more than 100,000 chemicals in commerce, limiting effective action. To address this, we collect high-quality primary chemical carbon footprint data and refine a machine learning model to estimate the carbon footprint of 94,697 chemicals in commerce across 26 national and regional chemical inventories. Our analysis reveals clear molecular structure-carbon footprint relationships that may guide future molecular design: carbon footprints are strongly linked to molecular frameworks, the number of ring structures, and types of heteroatoms. Newly commercialized chemicals tend to have higher carbon footprints than older chemicals, emphasizing the urgent need to embed "sustainable-by-design" principles into chemical innovation from the outset. This study provides a database of chemical carbon footprints and actionable insights for the sustainable design and use of chemicals, supporting the environmental decision-making of chemical producers, consumers, and policymakers.