Multisite-occupancy adsorption is described by using a new formalism based on the occupation balance approximation (Roma, F.; Ramirez-Pastor, A. J.; Riccardo, J. L. J. Chem. Phys. 2001, 114, 10932). In this framework, the adsorption isotherm is characterized by a correction function C, which relates to the conditional probability of finding the ith empty site to a lattice with i - 1 already vacant sites. A simple semiempirical adsorption isotherm is proposed by approximating C as a combination of the correction functions corresponding to exact 1-D calculations and the Guggenheim-DiMarzio approximation, with adequate weights. Results are compared with corresponding ones from Monte Carlo simulations.

Experimental adsorption isotherms of five n-paraffins (ethane, propane, butane, pentane, and hexane) in 5A zeolite were described by means of a statistical thermodynamics model for linear adsorbates (MLA) developed by Ramirez-Pastor et al. (1999) and compared with the well-known multisite Langmuir model (MSL) of Nitta et al. (1984). The experimental data, obtained by different authors in a wide range of temperatures and pressures, were correlated by using an algorithm of multiple fitting. Two main conclusions were drawn from the analysis of experimental data: (i) for small molecules (ethane, propane), MLA is the more accurate model, validating the hypothesis of the linear rigid character of the adsorbate and reinforcing previous results obtained from the analysis of computational experiments developed for dimers and linear trimers; (ii) for large molecules (n-butane, n-pentane, n-hexane), the better performance of the MSL model suggests that the admolecules adsorb in a nonlinear structure. The isosteric heat of adsorption dependence on the number of carbons obtained from our study, ranging between 23.84 kJ/mol for ethane and 59.26 kJ/mol for hexane, showed a very good agreement with previous results reported in the literature, confirming the consistency of our analysis.

A generalization of the classical monomer site-bond percolation problem is studied in which linear k-uples of nearest neighbor sites (site k-mers) and linear k-uples of nearest neighbor bonds (bond k-mers) are independently occupied at random on a square lattice. We called this model the site-bond percolation of polyatomic species or k-mer site-bond percolation. Motivated by considerations of cluster connectivity, we have used two distinct schemes (denoted as S intersection of B and S union of B) for k-mer site-bond percolation. In S intersection of B(S union of B), two points are said to be connected if a sequence of occupied sites and (or) bonds joins them. By using Monte Carlo simulations and finite-size scaling theory, data from S intersection of B and S union of B are analyzed in order to determine the critical curves separating the percolating and nonpercolating regions.