How the “Liquid Drop” Approach Could Be Efficiently Applied for Quantitative Structure–Property Relationship Modeling of Nanofluids

The main goal of this paper is the evaluation of the applicability of the geometrical “liquid drop” model (LDM) to describe physicochemical properties of nanofluids in quantitative structure–property relationship (QSPR) modeling. LDM-based descriptors are size-dependent, which allows them to be applied for a series of nanoparticles with the same chemical composition but different sizes. Thermal conductivity of nanofluids as the target property was investigated. Random forest regression as a nonparametric approach was utilized to determine important structural features of nanofluids responsible for enhancing their thermal conductivity.