The influence of size on freezing temperature of metallic and semi-metallic nanoparticles is investigated according to the classical nucleation theory by considering size and temperature effect on solid-liquid interface energy. The advantage of this improvised model is the ability to predict the freezing temperature for low dimensional materials, literally free from any arbitrarily adjustable parameters. It predicts that freezing temperature decreases almost exponentially with decreasing size due to the increasing surface/volume ratio. In addition, the critical radius at which the difference between the melting and freezing temperature disappears has been predicted. The model predictions agree well with available experimental results of Pb and Bi nanoparticles, which supply an easy way to understand freezing behavior of nanomaterials. (C) 2014 Elsevier Ltd. All rights reserved.