Predicting thermodiffusion in an arbitrary binary liquid hydrocarbon mixtures using artificial neural networks
New thermal diffusion coefficients of binary mixtures are measured for n-decane-n-alkanes and 1-methylnaphthalene-n-alkanes with 25 and 75 wt % at 25 degrees C and 1 atm using the thermogravitational column technique. The alkanes range from n-pentane to n-eicosane. The new results confirm the recently observed nonmonotonic behavior of thermal diffusion coefficients with molecular weight for binary mixtures of n-decane- n-alkanes at the compositions studied. In this work, the mobility and disparity effects on thermal diffusion coefficients are quantified for binary mixtures. We also show for the binary mixtures studied that the thermal diffusion coefficients and mixture viscosity, both nonequilibrium properties, are closely related.