On the prediction of adsorption isotherms of methanol/water vapour mixtures on microporous activated carbon

Abstract

It is well known that the adsorption of organic vapours on activated carbons (AC) can be disturbed in the presence of water vapour. When the adsorbed solutions are miscible, different theories, based on the pioneering Dubinin–Radushkevich (D–R) and Dubinin– Astakhov (D–A) equations are used to describe the experimental data. Among recent works, Linders et al. [1] showed that the isotherm of multicomponents can be modeled by using the ideal adsorbed solution theory (IAST), D–R and D–A equations. This model fits well with the experimental data up to a relative pressure of 0.1 (methanol adsorption at 60% RH) but shows discrepancies beyond these value. No explanation was given by the authors. Taqvi et al. [2] proposed to estimate the binary equilibria using the virial mixture coefficient approach (VMC). The observed trends are explained in terms of molecular interactions and, in particular, water was found to promote the adsorption of methanol. This is explained by the ability of alcohol to form H bond with water. To sum up, it seems that if some points such as the state of the phase adsorbed (homogeneous solution) are well established, other points such as the over-adsorption of methanol or the lack of models for the high methanol relative pressures, need further explanation. In this work, we report our results obtained for the co-adsorption of methanol and water vapours on a commercial activated microporous carbon (AC). The relative humidity (RH) ranges from 25% to 74% and the methanol concentration from 0% to saturation. The obtained results lead us to propose a thermodynamical approach that explains the isotherms of methanol adsorption at constant RH. Furthermore, our approach fits well with several results in the literature. Adsorption isotherms were collected in dynamic flow conditions at 298 K using an intelligent gravimetric analyzer (IGA) supplied by Hiden Analytical Ltd. This apparatus is fully computer controlled and measures gravimetric adsorption on a microbalance. The sample temperature was constantly monitored throughout the experiment, and the variation was found to be ±0.05 K. Further details about the acquisition procedure can be found elsewhere [3]. The AC studied has been obtained by activation of coconut shell in steam at 1123 K. The porosity characterization was determined by benzene adsorption using a procedure described elsewhere [4]. The sample is mainly considered as microporous (Vmic = 0.585 cm 3 /g; Vmes = 0.082 cm 3 /g) with specific surface area close to 1500 m 2 /g and having a poor surface chemistry (acidic groups = 0.06 meq/g and basic groups = 0.5 meq/g as determined by Boehm titration). The isotherms of pure methanol and water adsorption are presented in Fig. 1. The adsorption of methanol, that exhibits type I isotherm, is characterized by strong dispersive interactions of the methyl groups with the carbon pore walls [3]. As a result, a large part of the pore volume is already filled at low relative pressures. The water isotherm is of type V. Then the carbon surface is evaluated to be extremely hydrophobic. The sharp increase of the adsorbed water is attributed to the condensation of water into the micropores. The adsorption isotherms of binary mixtures, presented in Fig. 2, were obtained for a relative humidity of 25%, 50%, 60% and 74% respectively. It should be noted that the number of data for each isotherm decreases with increasing RH due to limitations of our experimental set-up. Moreover, all the isotherms are of type I. The slope at low partial pressure in methanol increases with increasing RH before reaching a plateau that corresponds to the capacity near saturation. The total mass uptake is higher when RH = 25% than for pure methanol, and it then decreases with increasing RH.