Development of a scaled equation and an Artificial Neural Network for the surface tension of alcohols through a statistical and factor analysis approach

This work presents a new formula to calculate the surface tension of alcohols. As a first step, an analysis of the available data of the surface tension for alcohols was made. A total of 1643 data were collected for n-alcohols, aromatic alcohols, cycloaliphatic alcohols, 2-alkanols, and methyl alkanols. The data were then analyzed through a factor analysis approach. The factor analysis, combined with physical considerations, allowed to regress the experimental data with a scaled equation that is very simple to employ. The same physical parameters considered in the scaled equation were also adopted as input parameters in a multi-layer perceptron neural network, to predict the surface tension of alcohols. The proposed multilayer perceptron has one hidden layer with 30 neurons, determined according to the constructive approach. The proposed model was trained, validated, and tested for the set of collected data, showing that the accuracy of the neural network model is very high.