Stochastic model for equilibrium adsorption onto activated carbon

Abstract

Most models used in describing adsorption equilibria onto activated carbon of adsorption from organic vapours, or from dilute aqueous solutions. However, these models fail to give accurate predictions if less ideal substances such as dissociated compounds or ionic complexes are present, for example in industrial streams. Empirical equations are usually the only method that can be used to describe adsorption of these systems, but unfortunately no conclusions can be drawn from the parameters of these equations. Therefore, an improved understanding of adsorption is required, so that models can be extrapolated to new combinations of species. To alleviate the shortcomings of existing models, a new model based on statistical thermodynamic principles and particularly the Monte Carlo method is proposed. With the use of this new model, classical non-ideal behaviour such as lateral interactions, heterogeneous carbon surfaces, selective adsorption due to steric interactions as well as irreversible adsorption can be modelled. Many of these influences have been mentioned by other authors but have not usually been included in equilibrium predictions. All the parameters used in the new method have some physical meaning and describe only a single characteristic of the adsorption system. This method was applied successfully to two example systems, i.e. the more classic ideal system containing p-bromophenol and p-nitrophenol, as well as the adsorption of gold and silver cyanide complexes. Some insight into the adsorption reactions taking place in these systems can be gained if the method is used as a diagnostic tool. A sensitivity analysis was also conducted to demonstrate the behaviour of the model under various non-ideal conditions.