Effects of potential models on the adsorption of carbon dioxide on graphitized thermal carbon black: GCMC computer simulations

Abstract

In this paper, we evaluate the performance of three intermolecular potential models (TraPPE, HMT and 5-Charge) on the adsorption of carbon dioxide on graphitized thermal carbon black at various temperatures. All models have three dispersive interaction sites, which coincide with the centers of the three atoms of carbon dioxide. To account for the quadrupole, the first two models have three discrete charges, while the other has five discrete charges to reproduce not only the quadrupole moment, but also the higher moments. Simulation results obtained from the GCMC are tested against experimental data of adsorption of carbon dioxide on graphitized thermal carbon black at 193 K [R.A. Beebe, A.V. Kiselev, N. Kovaleva, R. Tyson, J. Holmes, Adsorption and state of carbon dioxide, sulfur hexafluoride and ammonia on a graphitized carbon surface. I. Adsorption isotherms. Russian J. Phys. Chem. 38 (1964) 372–378.] and 273 K [A. Guillot, F. Stoeckli, Reference isotherm for high pressure adsorption of CO2 by carbons at 273 K, Carbon 39 (2001) 2059–2064.]. Very low pressure data for carbon dioxide of Myers and Prausnitz [A. Myers, J. Prausnitz, Hindered rotation in physical adsorption. Trans. Farad. Soc. 61 (1965) 755–764.] at various temperatures are also tested. In general, the adsorption isotherms are well described by the TraPPE model, while the isosteric heat is only reasonably described. The HMT under-predicts the data while the 5-Charge model over-predicts it. In the case of Guillot and Stoeckli data, the agreement between the simulation results and the data is substantially better if the charges of the carbon dioxide molecules are increased by 8% whenever the particles are within the first layer. This may suggest the importance of the surface mediation on the adsorbed molecules. We carried out simulations of adsorption in slit pores of various sizes, using the TraPPE model and the 1C-LJ model, to show the significant implication of the packing effects in confined space.