A serious gap in the field of nanofluids' modeling is disregarding the effect of molecular structure that must be highlighted. For the first time, the Monte Carlo method was utilized to model isobaric heat capacity and density of nanofluids. A creditable data set was selected contains nitride-based nanofluids (aluminum nitride, titanium nitride, and silicon nitride all dispersed in ethylene glycol). Quasi-simplified molecular-input line-entry system (quasi-SMILES) was applied to represent the structure of nanofluids, successfully. This format made possible incorporating molecular structure besides experimental conditions into the modeling process. The developed models were evaluated precisely; it was found that the statistical qualities were good and their performance was superior to the classical equation. Also, results revealed that some molecular features of nanofluids such as double and triple bond affects isobaric heat capacity and density, while the size of nanoparticles did not impressive affect these properties. It is remarkable to point out that the proposed models introduce a new trend to estimate the thermophysical properties of nanofluids. The utilized approach could be useful for a more reliable and accurate prediction of the other nanofluids' properties.
