Solvent effects on CO₂ capture by simple amino acids: an integrated density functional theory - machine learning...

Area of Science:

• Computational Chemistry
• Materials Science
• Environmental Science

Background:

• Amino acids are explored for carbon capture technologies.
• Understanding molecular structure and solvent effects on CO2 capture mechanisms is crucial.
• Existing methods require optimization for efficiency and scalability.

Purpose of the Study:

• To investigate CO2 capture by glycine, alanine, and serine anions in various solvents.
• To elucidate the reaction mechanisms and identify optimal conditions for CO2 capture.
• To correlate molecular structure and solvent properties with capture efficiency.

Main Methods:

• Density functional theory with implicit solvation was employed.
• Investigated CO2 capture by amino acid anions in gas phase, water, DMSO, glycerol, and lactic acid.
• Utilized correlation analysis and machine learning (gradient boosting regression) for parameter identification.

Main Results:

• Glycerol proved to be the optimal solvent, yielding highly exothermic reactions and stabilized transition states.
• Glycine exhibited favorable gas-phase thermodynamics and low reaction barriers.
• Alanine and serine showed complex behaviors influenced by steric hindrance and solvent interactions, respectively; CO2 loading capacity negatively correlates with amino acid hydrogen bond donors.

Conclusions:

• Glycerol-based formulations with glycine or alanine are superior for industrial CO2 capture.
• Solvent hydrogen bond-donating capability, not dielectric constant, critically governs capture efficiency.
• Strategic solvent selection and understanding amino acid structure are key for designing tunable carbon capture systems.