Artificial Neural Network Approach to Predict the Solubility of C₆₀ in Various Solvents

Abstract

A multiparameter artificial neural network (ANN) approach was successfully utilized to predict the solubility of C60 in different solvents. Molar volume, polarizability parameter, LUMO energy, saturated surface, and average polarizability molecular properties were chosen to be the most important factors determining the solubilities. The results show that in a large number of solvents (126) the solubility decreases with increasing molar volumes of the solvents and increases with their polarizability and saturated surface areas. A method is suggested to the approximate determination of experimentally not easily measurable solubility related thermodynamic parameters, e.g., the Hildebrand parameter, based on reliable solubility measurements.

Multiparameter solubility model of fullerene C60

“…Both Marcus et al and Ruoff et al concluded that the solubility increased when the molar volume of solvents increased. However, Kiss et al [33] used several computed and experimental properties of solvents and multiparameter artificial neural network method to correlate solubility, and reported the solubility decreased with increasing molar volume. Fig.…”

Effect of Solvents on the Electronic Properties of Fullerene Based Systems: Molecular Modelling

“…There are many different approaches to calculate and predict C60 solubility in organic solvents. Some of them are fully mechanistic [5][6][7][8][9], developed from the thermodynamical point of view; others are statistically based, with good correlation coefficients, but not transparent and complicated in interpretation [10][11][12][13][14][15][16].…”

Application of neural network molecular modeling for correlating and predicting Henry's law constants of gases in [bmim][PF6] at low pressures

“…Neural network has been previously used for gas solubility and phase equilibrium modeling [29,[35][36][37][38][39][40][41][42][43][44][45][46][47][48]. Nowadays, models developed based on the neural network approach are considered as valuable models, only if thermodynamic models show some drawbacks in prediction of properties and/or phase behavior of a system (or when there is no available method).…”