Multiscale investigation of CO₂ solubility behavior in oil-water mixtures: Experiments, molecular dynamics simulations, and predictive model optimization

This study systematically investigates the dissolution behavior and regulatory mechanisms of carbon dioxide (CO₂) in oil-water mixtures through a multiscale approach integrating experiments and molecular dynamics (MD) simulations. Experimental results demonstrate that CO₂ solubility exhibits significant pressure-increasing and temperature-decreasing trends, reaching a peak value of 1.422 mol·L⁻¹ at an oil-to-water ratio of 2:1 (5.2 MPa), which is 37 % higher than the 1:2 ratio. CO₂ solubility in crude oil is 2–3 times higher than in the aqueous phase. MD simulations reveal the enrichment effect of CO₂ at oil-water interfaces (density 1.9 times higher than the bulk phase) and the mass transfer mechanism dominated by interfacial tension reduction ( γ decreases from 57.9 to 46.2 mN·m⁻¹). Based on experimental data, the Taylor model (for the oil phase, MRE (Mean relative error) = 1.75 %) and Duan model (for the aqueous phase, MRE = 2.17 %) were optimized. A composite predictive model for oil-water mixtures was developed, achieving an overall MRE of 3.54 %, significantly outperforming traditional models (error reduction by 85 %). This research provides a high-precision theoretical framework for CO₂-enhanced oil recovery (EOR) and carbon sequestration, elucidating the critical role of multiphase interfacial behavior in dissolution kinetics. • A multiscale framework is constructed integrating experiments, simulations, and modeling. • CO₂ solubility variations are quantified under diverse conditions via experiments. • CO₂ interfacial enrichment mechanisms are uncovered via molecular dynamics simulations. • A composite solubility prediction model is optimized for heterogeneous reservoirs.