Activated carbon was used as an adsorbent for the removal of phenol from aqueous solution under various operating conditions. Statistical design using MATLAB® tools was employed to study the effects of three variables on six different response factors. MATLAB® software was employed to solve proposed quadratic model equations and for fitting a quadratic response surface. All the models with a very high adjusted R‐square predicted the experimental data well. The Thomas, the Yoon‐Nelson, the Wolborska, the bed depth service time, and the linear driving force models were used to predict breakthrough curves for experimental data. The results showed that the linear driving force model was suitable for the prediction of the breakthrough curve data.

Granular activated carbon (GAC) was synthesized from Silver berry (Elaeagnus Angustifolia L.) seeds using zinc chloride as the activation agent. To optimize the operating parameters, the effects of the time and temperature of carbonization, impregnation ratio, and heating rate on the iodine number and yield of activated carbon were studied. Optimized parameters were impregnation ratio of 1:1, carbonization temperature of 500 °C, carbonization time of 1 hour, and heating rate of 5 °C/min. The GAC synthesized under optimized conditions was characterized by Nitrogen adsorption-desorption isotherms, SEM, EDX, XRD, FT-IR, Boehm titration, TG-TGA, and TG-IR. It was found that the synthesized GAC has a microporous structure with a BET surface area of 1109 m 2 /g, a micropores volume of 0.317 cm 3 /g, and an average pore diameter of 2.1 nm. The methylene blue (MB) dye was employed as a molecule model to evaluate the porosity and the adsorption capacity of the synthesized GAC. The results showed that the maximum adsorption capacity of MB and the percent portion of the surface area (S MB /S BET ) were 120.48 mg/g and 30.62%, respectively. The experimental results reveal that the synthesized GAC can be used as a low-cost adsorbent for the removal of small and large environmental pollutants.

No abstract

This paper aims at examining the increase of phenol adsorption breakthrough curves spreading caused by the chemical heterogeneity of granular activated carbon fixed beds. The local and the thermodynamic equilibrium assumption, as well as the nonlinear adsorption obeying to Langmuir isotherm, are considered. This study particularly tempts to link the reduced variance of phenol breakthrough curves to a measurable quantifying parameter of the chemical heterogeneity. The investigated artificial heterogeneous media are prepared by alternating layers of two types of granular activated carbon, active and non‐active ones, that have similar physical properties. On the one hand, the chemical heterogeneity is quantified by the active layer relative thickness of the column length, l1/L. On the other hand, it is quantified by the mean value of the probability distribution γ. The latter also represents the mean active grains mass ratio of the total medium mass, hence the medium mean capacity. The obtained results show an increase in the reduced variance and thus the effective global dispersion with the heterogeneity; the increase is as important as the medium capacity decreases. However, the dispersion increase achieves a limit value, even when the heterogeneity increases. The results are statistically modelled using a regression equation function of the capacity variation in terms of γ and the chemical heterogeneity in terms of l1/L. The relationship combining the medium capacity and the chemical heterogeneity is obtained. The relationship implicitly takes into account the effect of the column length.

AbstractThis paper presents a comparison between some numerical methods and techniques for solving the nonlinear advection-dispersion equation, which may be used to describe the adsorption of phenol into a granular activated carbon fixed bed under local equilibrium conditions. The adsorption is described by the Langmuir isotherm, which makes the advection-dispersion equation nonlinear. This equation is solved successively by using the approximation and linearization techniques. For each technique, two types of numerical algorithms are used. Concerning the first one, the Implicit and the Runge Kutta schemes are used. As for the second one, the Modified Picard iteration and the Newton Raphson scheme are applied. Simulation results have been compared to each other and to the experimental data as well. Both of the Implicit and the Runge Kutta algorithms have led to superimposed simulated breakthrough curves. Both of the modified Picard and Newton Raphson schemes have given identical results too. However, comparing to the experimental data, the obtained solution, using the approximation technique, has underestimated the retardation of solute and failed in fitting the experimental breakthrough. The Obtained solution, using the linearization technique, has correctly fitted the experimental results under all the conditions of: feed flow rate, activated carbon bed height and the inlet phenol concentration.