Why Choline Chloride? A Comparative Study of Physicochemical Properties and Machine Learning–Driven Insights Into Melting Point Trends

Theoretically, a wide variety of DESs can be formed through different combinations of components and their interactions. Among these, ChCl is most widely used in DES formulations. What accounts for the prominence of ChCl in DES formation? This study aims to analyze and address this question. A comparison with DESs based on other HBAs highlighted the favorable trade-offs of ChCl in the selected physicochemical properties. Additionally, the study constructed a machine learning model to evaluate the effects of different HBDs on the melting point of ChCl-based DESs. Machine-learning analysis indicated that the intrinsic thermal properties of HBD and the composition of the mixture were the most influential descriptors for melting point prediction. In addition, the local environments of selected functional groups provided complementary structural information associated with melting point variations. DFT calculations provided qualitative molecular-level interpretations consistent with these statistical associations. The presence of polyhydroxy and amide groups tends to lower the melting point, while the presence of polycarboxyl and cyclic structures tends to prevent a decrease in the melting point. These findings highlight the favorable trade-offs of ChCl in the aspects considered, providing useful context for the rational design and optimization of DES systems.