Abstract
Surface phenomena such as corrosion, crystal growth, catalysis, adsorption and oxidation cannot be adequately comprehended without the full knowledge of surface energy of the concerned material. Despite these significances of surface energy, they are difficult to obtain experimentally and the few available ones are subjected to certain degree of inaccuracies due to extrapolation of surface tension to 0 K. In order to cater for these difficulties, we have developed a model using computational intelligence technique on the platform of support vector regression (SVR) to establish a database of surface energies of hexagonal close packed metals (HCP). The SVR based-model was developed through training and testing SVR using fourteen experimental data of periodic metals. The developed model shows accuracy of 99.08% and 100% during training and testing phase, respectively, using test-set cross validation technique. The developed model was further used to obtain surface energies of HCP metals. The surface energies obtained from SVR-based model are closer to the experimental values than the results of the well-known existing theoretical models. The outstanding performance of this developed model in estimating surface energies of HCP metals with high degree of accuracy, in the presence of few experimental data, is a great achievement in the field of surface science because of its potential to circumvent experimental difficulties in determining surface energies of materials. (C) 2015 Elsevier B.V. All rights reserved.