Structure-property reduced order model for viscosity prediction in single-component CO₂-binding organic liquids

CO₂ capture from power generation with aqueous solvents remains energy intensive due to the high water content of the current technology, or the high viscosity of non-aqueous alternatives. Quantitative reduced models, connecting molecular structure to bulk properties, are key for developing structure-property relationships that enable molecular design. In this work, we describe such a model that quantitatively predicts viscosities of CO₂ binding organic liquids (CO₂BOLs) based solely on molecular structure and the amount of bound CO₂. The functional form of the model correlates the viscosity with the CO₂ loading and an electrostatic term describing the charge distribution between the CO₂-bearing functional group and the proton-receiving amine. Molecular simulations identify the proton shuttle between these groups within the same molecule to be the critical indicator of low viscosity. The model, developed to allow for quick screening of solvent libraries, paves the way towards the rational design of low viscosity water-lean solvent systems for post-combustion CO₂ capture. Following these theoretical recommendations, synthetic efforts of promising candidates and viscosity measurement provide experimental validation and verification.