Preparation of Furfuryl Alcohol-Derived Activated Carbon Monolith for Liquid Adsorption
Darmadi, (2009) Preparation of Furfuryl Alcohol-Derived Activated Carbon Monolith for Liquid Adsorption. PhD thesis, Universiti Putra Malaysia.
The preparation and optimization of carbon coated monolith is reported. The aim is to produce mesoporous activated carbon monolith for liquid adsorption using the dipcoating method. The materials required are a carbon source (furfuryl alcohol), a pore former agent (poly ethylene glycol), a binder (pyrrole), and polymerization catalyst (nitric acid). Furfuryl alcohol (FA) is first polymerized, followed by the impregnation of monolithic structure, carbonization, and activation. The effect of poly ethylene glycol (PEG) on the structure of carbon monolith is first investigated. The carbon coated monoliths are characterized by thermo gravimetrical analysis (TGA), elemental analysis, scanning electron microscopy (SEM), Fourier transform infrared (FTIR) and textural analysis. The carbon monolith prepared without the addition of pore former agent (only FA) exhibits adsorption Type I which is a characteristic of microporous material, whilst the carbon monolith prepared with the addition of pore former agent (FA + PEG) is of Type IV indicating mesoporous material. Brunauer, Emmett, and Teller (BET) surface areas measured by N2 adsorption are 264 and 431 m2 g-1 for sample FA and sample (FA + PEG), respectively. Total pore volume of the samples FA and FA + PEG are 0.13 and 0.38 cm3 g-1 , respectively. The PEG is completely decomposed during carbonization to create new mesoporosity. The optimization of pore volume and surface area of carbon coated monolith is studied using the response surface methodology (RSM) based on the Box- Behnken design. The carbonization temperature, concentration of PEG, and molecular weight of PEG are identified as the dominant parameters in controlling the pore size distribution, pore volume, and surface area. The maximum pore volume found from the RSM is 173 mm3 g-1 at carbonization temperature of 680 oC and concentration of PEG of 38% vol. with molecular weight of PEG of 1000 g mol-1, whilst maximum surface area is 585 m2 g-1 at carbonization temperature of 660 oC and concentration of PEG of 31% vol. with molecular weight of PEG of 1000 g mol-1. To confirm these results, synthesis of carbon coated monolith is performed. Experimental results obtained are pore volume of 161 mm3 g-1 and surface area of 553 m2 g-1, which are very close to the prediction by RSM. The performance of activated carbon monolith is evaluated using the methylene blue (MB) adsorption. Equilibrium adsorption data are predicted by three isotherms, i.e. the Langmuir, the Freundlich, and the Redlich-Peterson isotherms. The best fit to the data is obtained with the Redlich-Peterson and the Langmuir isotherms with correlation coefficient (R2) of 0.997 and 0.998, respectively. The maximum monolayer adsorption capacity is 191 mg g-1. The Langmuir isotherm is used for modeling and simulation as it is a two parameter model and has similar accuracy in describing the isotherm data in this work. The dimensionless equilibrium parameter (RL) is calculated as 0.1, indicating that the adsorption is favorable. The kinetics of adsorption of MB is studied in terms of pseudo first order and second order mechanism for chemical adsorption as well as an intraparticle diffusion mechanism by applying the linear driving force (LDF) approximation in batch system. Kinetic parameters and correlation coefficients are determined. It is shown that the pseudo first order kinetic equation fits well to describe the adsorption kinetics with rate constants 6.6 × 10-3 min-1, 4.0 × 10-3 min-1, and 1.5 ×10-3 min-1 for initial concentrations 20, 50, and 100 mg L-1, respectively. The LDF model for a monolithic system is developed. The pseudo first order rate constant is used as initial guess to estimate the LDF mass transfer coefficient ( LDF k ) by matching the simulation with experimental data. The comparison of the results calculated using the LDF model and experimental data is in good agreement with LDF k values 8.25 × 10-3, 5.20 × 10-3, and 2.50 × 10-3 min-1 for initial concentrations 20, 50, and 100 mg/L, respectively. A predictive dispersed plug flow model, with adsorption rate described by the LDF model, is developed to predict the breakthrough curve of a monolith column. The model parameters are from batch adsorption experiments. The result of simulation is found to agree excellently with the experimental data. The effect of LDF mass transfer coefficients is investigated numerically. The LDF mass transfer coefficients are found to significantly affect the shape of the breakthrough curve.
Repository Staff Only: Edit item detail