In silico study of the solubility of fullerene derivatives in chlorobenzene

Since the seminal discovery of fullerene in 1985, fullerene C60 and its derivatives have emerged as versatile nanostructured materials with applications spanning optics, electronics, cosmetics, and biomedicine. The synthesis of various fullerene derivatives (FDs) involves the addition of functional groups to C60. Primary production costs are intricately tied to the extraction and purification of C60, given its limited solubility in both polar and many organic solvents. A critical physicochemical characteristic defining FDs' industrial utility is their solubility, a factor integral to substance distribution in the environment.

The application of fullerenes, particularly in renewable energy, has driven extensive Quantitative Structure-Property Relationship (QSPR) research into their performance-related properties. This study assesses FD solubility in chlorobenzene. Solubility in organic solvents is essential, as it is linked to bioaccumulation. We have generated linear and nonlinear machine learning-based QSPR models, validated for their robustness and accuracy. These models are based on experimental solubility data obtained for a set of FDs in chlorobenzene. Subsequently, these models were deployed to predict the solubility of a novel set comprising 117 FDs. The examined dataset encompasses diverse groups attached to the C60- and C70-core structures, with side chains strategically linked to the cyclopropane group.

The insights drawn from this study bear significance in understanding how FDs with varying solubility levels can be assessed for distinct applications. By exclusively relying on developed “in silico” models and circumventing extensive testing, these QSPR models offer early evaluative tools, shedding light on FDs with optimal solubility for specific application domains. The outcomes of this research hold promise for advancing our comprehension of fullerene derivatives and streamlining their application in diverse industrial sectors.