Prediction of solid vapor pressures for organic and inorganic compounds using a neural network

A method to estimate solid vapor pressures (PS) for organic and inorganic compounds using an artificial neural network (ANN) is presented. The proposal consists of training an ANN with PS data of a defined group of substances as a function of temperature, including as learning variable five physicochemical properties to discriminate among the different substances. The following properties were considered: molecular mass, dipole moment, temperature and pressure in the triple point (upper limit of the sublimation curve), and the limiting value PS → 0 as T → 0 (lower limit of the sublimation curve). 152 substances (1520 data points) have been used to train the network. Then, the solid vapor pressures of 60 other solids (600 data points) have been predicted and results compared to experimental data from the literature. The study shows that the proposed method represents an excellent alternative for the estimation of solid vapor pressures and can be used with confidence for any substances.