Abstract
Despite the improvement in understanding the structure of the activated carbons, the procedure for developing new carbonaceous materials suitable for the removal of phenolic compounds is still largely based on trial and error. Until now there are no predictive models to assist in the selection or synthesis of these adsorbents. Here, we apply molecular simulation in order to better understand the pore size – adsorption relationship in activated carbons. We simulated a set of phenol isotherms for different carbon pore sizes (8.9, 18.5 and 27.9 Å), named representative pores. The pore size of 8.9 Å is the most efficient in removing diluted phenol in water being effective at concentrations of 1.6 × 10−5 mol/L. The other pores are effective for concentrations of 3 orders of magnitude above this. A predictive approach for phenol removal capacity, based in the representative pore methodology, was proposed and validated for commercial activated carbon. Moreover, we present evidences that this method can be extended to other phenolic compounds.