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

Abstract

The process of CO₂ capture by amino acids offers a promising approach for carbon capture technologies, yet the influence of the molecular structure and solvent environment on the reaction mechanisms remains to be understood. The present study investigates the CO₂ capture by glycine, alanine, and serine anions across five environments, namely, gas phase, water, DMSO, glycerol and lactic acid, using density functional theory with implicit solvation. The reaction proceeds via a barrierless nucleophilic attack forming a zwitterionic intermediate followed by the rate-determining intramolecular proton transfer. Glycerol emerges as the optimal medium, exhibiting highly exothermic reaction enthalpies (−50.8 to −53.7 kcal mol⁻¹) and stabilized transition states below the reactant energy levels due to its extensive hydrogen bonding network. Structural variations reveal a kinetic-thermodynamic trade-off in which glycine shows the most favorable gas-phase thermodynamics (−21.4 kcal mol⁻¹) and the lowest barriers (+19.4 kcal mol⁻¹), while the methyl group of alanine introduces steric hindrance and the hydroxymethyl substituent of serine creates a complex solvent-dependent behavior, including an endothermic reaction in DMSO (+0.4 kcal mol⁻¹), due to over-stabilization of the serine–DMSO complex. A correlation analysis of the key parameters reveals that the CO₂ loading capacity negatively correlates with amino acid hydrogen bond donors (r = −0.59), explaining the serine-suppressed aqueous activity. Machine learning analysis (gradient boosting regression, R² = 0.85) identifies a molecular weight threshold (∼105 g mol⁻¹), where the side-chain complexity dominates the reactivity, and demonstrates that the solvent hydrogen bond-donating capability rather than the dielectric constant critically governs the capture efficiency. These findings establish glycerol-based formulations with glycine or alanine as superior candidates for industrial CO₂ capture (ΔG₂₉₈ = −39 to −43 kcal mol⁻¹), highlighting strategic solvent selection for designing tunable amino acid-based carbon capture.