Surface Tension Estimation of Binary Mixtures of Organic Compounds Using Artificial Neural Networks

Abstract

The surface tension of binary mixtures at different temperatures and compositions are required in much scientific and technological research. Therefore, having an exact correlation between surface tension and easily accessible physical properties is essential.In this work, the sensitivity of the surface tension to some physical properties was studied by using artificial neural networks to find the most effective ones. Furthermore, the artificial neural networks were used to estimate the surface tension of binary systems as a function of the most effective physical properties including critical pressure, reduced temperature, acentric factor, and molar density. The experimental data, collected for training and verifying the networks, includes various materials such as alkanes, alkenes, aromatics, alcohols, organic acids, as well as chlorine, iodine, sulfur, nitrogen, and fluorine-containing compounds in the composition ranges from zero to 100 mole percent, and temperatures between 116 K and 393 K. The average absolute relative deviation of the most accurate network, obtained for all 2038 data points regarding 83 binary mixtures, is 1.75%.

Surface Tension of Liquid Organic Acids: An Artificial Neural Network Model

“…This may be due to the very different behaviours of the acids considered [ 23 ] and the large amount of available data. It is clear that more complex methods are needed for obtaining significantly better results, and the ANN models presently seem to be an excellent alternative [ 16 , 21 , 30 , 31 , 32 , 33 , 34 , 35 , 36 , 37 , 38 ].…”

“…As an alternative to the previous methods, some authors have used artificial neural networks (ANNs) for the study of the surface tension and other thermophysical properties [ 16 , 30 , 31 , 32 , 33 , 34 , 35 , 36 , 37 , 38 ]. ANNs take their inspiration from biological neural networks, with the mathematical model of a single neuron and the way in which neurons are interconnected (i.e., the “architecture”), leading to a framework for different machine learning algorithms.…”

“…Various ANN models have already been used for the surface tension of pure liquids of different kinds [ 16 , 21 , 30 , 31 , 34 , 35 , 36 , 37 ] and mixtures [ 31 , 39 , 40 ]. In particular, the proposals for pure liquids, published from 2008 to 2017, were summarized by Mulero et al [ 34 ].…”