On the identification of the sharp spike in the heat curve for argon, nitrogen, and methane adsorption on graphite: Reconciliation between computer simulation and experiments

Experimental data for the adsorption of argon, nitrogen, and methane on a planar graphite surface under subcritical conditions exhibit a very sharp peak in the isosteric heat curve at loadings close to the monolayer concentration. The magnitude of this peak is much greater than the value expected from the extrapolation of the isosteric heat in the submonolayer region to the monolayer loading. This sharp and large peak has been interpreted as the transition of the first layer from a hypercritical fluid phase to a solid phase. Here, we argue that volumetric or calorimetric experiments are in fact carried out in a canonical system, where a dose of known mass of adsorptive is introduced into the system, rather than in an open system exposed to an infinite supply reservoir as assumed in GCMC simulations. We have carried out canonical simulations with the new Mu-CMC scheme proposed recently by Fan et al. (J. Phys. Chem. B2011, 115 (35), 10509-10517) and have reproduced the sharp peak in the heat curve. We show that this sharp spike is observed in a canonical ensemble (and in some cases grand canonical ensemble), whether the surface is unstructured or structured and no matter whether nitrogen is modeled as a 1CLJ or 2CLJ+3q or methane is treated as a 1CLJ or 5CLJ+5q model, suggesting that the heat spike is an intrinsic characteristic of the gas-solid pair. © 2011 American Chemical Society