Some remarks on the calculation of the pore size distribution function of activated carbons

Different authors investigated the effects of geometric and energetic heterogeneities on adsorption and on carbon characterization methods. In most theoretical studies carbon structure is modeled as parallel infinite graphite walls that form ideal slit-shaped pores of the fixed widths. In the literature there is the lack of systematic studies showing the influence of pore structural and Lennard-Jones (LJ) potential parameters on the poresize distribution functions. Moreover, the parameters characterizing the properties of the adsorbed phase and the heterogeneity of the adsorbent surface should be taken into account. The Nguyen and Do method with proposed by us ASA algorithm, were utilized for the assessment of the porosity from the series of almost few thousands numerically generated local adsorption isotherms. The values of the mentioned-above parameters are varied over the wide range (ca. ±20%) of the reference ones. Different types of the theoretical and experimental adsorption isotherms (nitrogen at 77 K) were taken into account as the global ones. They were related to the mechanism of the primary, secondary or mixed micropore filling. The variations in some above-mentioned parameters have significant effects only for PSDs (and for average pore widths) corresponding to the primary micropore filling mechanism. On the other hand, for the process of the secondary micropore filling, the influence of these parameters (without the BET coefficient for adsorption on a “flat” surface, cs,B) is rather insignificant. Nevertheless the differences between local and global adsorption isotherms (in the whole range of relative pressures) the absence of micropores having pore half width equal to ca. 1 nm on PSDs was observed for studied adsorbate–adsorbent systems with exceptions of the strictly microporous adsorbents and/or the low values of cs,B. Comparison of the experimental data with the generated theoretical isosteric enthalpy of adsorption indicates that the phenomenal uptake observed from experiment can be explained in terms of the reasonable solid–fluid interaction parameters. Therefore, we varied the heterogeneity of the adsorbent surface via the strength and the range of the solid–fluid potential and the parameter cs,B in order to reproduce the experimental data of enthalpy of adsorption. Note that similar procedure was applied by Wang and Johnson to reproduce some hydrogen adsorption data measured for carbon nanofibres. The analysis of the obtained results shows that the selection of the values of the parameters of the intermolecular interactions and the quantities characterizing the properties of the adsorbed phase and the heterogeneity of the adsorbent walls for molecular simulations should be made with care and the influence of possible errors should be considered. © 2006 Elsevier Inc. All rights reserved