Harnessing COSMO-RS for Sustainable Innovation: From High-Throughput Solvent Screening to Smart Adsorbent Design

Researchers at the University of Delaware have utilized AMS COSMO-RS in two complementary studies that together showcase a predictive, data-driven framework for reducing experimental burden and accelerating innovation in materials and process development.

In the first study, Gupta et al. [1] applied COSMO-RS to streamline the extraction of valuable chemicals from food waste, focusing on potato peels (PPW)—a rich but underutilized source of antioxidants and bioactive molecules. The team conducted high-throughput in silico screening of over 2,400 solvents to evaluate their suitability for dissolving target phenolic compounds. COSMO-RS identified more than 100 solvents with superior solubility compared to conventional solvents like ethanol and methanol.

Among these, dimethylformamide (DMF) emerged as the top-performing solvent due to its strong hydrogen bond-accepting ability and polarity, as confirmed by COSMO σ-profile analysis. To validate these predictions, the researchers performed solubility tests on five key solutes and conducted extraction trials using actual PPW samples. The results showed strong agreement between COSMO-RS predictions and experimental outcomes, confirming the model’s reliability in complex, multicomponent systems. The study also uncovered how molecular interactions—such as polarity and hydrogen bonding—govern extraction efficiency, offering insights that are difficult to obtain through experimentation alone.

In a follow-up study, Gupta et al. [2] extended the use of COSMO-RS to the design of bio-based adsorbents for purifying the extracted phenolics. COSMO-RS was employed to analyze σ-profiles of 28 candidate monomers, assessing their polarity and hydrogen bonding potential with chlorogenic acid, the target molecule. These insights were critical for selecting functional monomers for molecularly imprinted polymers (MIPs), which rely on precise monomer-template interactions to achieve high selectivity.

COSMO-RS identified itaconic acid, a bio-based monomer, as the most promising candidate, exhibiting the highest hydrogen bond acceptor charge density and favorable polarity for interacting with chlorogenic acid. Experimental validation confirmed that itaconic acid-based MIPs outperformed traditional monomers like acrylamide in both selectivity and adsorption capacity. The resulting adsorbent could purify up to 92% of chlorogenic acid from real food waste extracts in a single cycle.

By integrating COSMO-RS into both solvent and monomer selection workflows, the researchers established a predictive, end-to-end strategy for sustainable chemical separations—minimizing trial-and-error experimentation and enabling circular economy solutions.