A fluidized-bed reactor for the photocatalytic mineralization of phenol on TiO(2)-coated silica gel

TiO(2) photocatalysis represents a promising class of oxidation techniques that are intended to be both supplementary and complementary to the conventional approaches for the removal of refractory and trace organic contaminants in water and air. Powdered TiO(2) dispersion systems employed in most studies require an additional separation step to recover the catalyst from the effluent water, which represents a major drawback for large scale applications. The optimization of photocatalytic treatment systems involves merging the benefits of catalyst immobilization on a retainable support, thus eliminating the need for downstream catalyst separation, maximization of photon-exposed catalyst area, and continuous operation. Aiming to integrate such conditions into a single system, a bench-scale annular photo-reactor with concentric UV-C lamp was built to study the photocatalytic mineralization of phenol on fluidized silica gel beads coated with sol-gel-synthetized TiO(2). Reactor efficiency was investigated for different silica particle diameters (224, 357 and 461 μm), fluidized-bed concentrations in the bulk liquid (5, 10, 20 and 30 g L(−1)), initial phenol concentrations in the aqueous solution (0.25 mmol L(−1) to 4.0 mmol L(−1)), and single and multiple sol-gel depositions. Then, the resulting optimum reactor configuration was compared to that of the same process on suspended Degussa P25 TiO(2) nanoparticles under similar experimental conditions. The latter is expected to be more efficient, but post-treatment catalyst recovery, being an energy intensive process, represents a major limitation for large scale applications. Process efficiency was measured as a function of the accumulated energy necessary for the mineralization of 50% of the initial dissolved chemical oxygen demand (COD), or, Q(0.5). Results showed that for any given mass of fluidized bed material, photo-oxidation efficiency increases with decreasing particle size (even for bed concentrations with similar equivalent surface area), decreasing initial phenol concentrations, and increasing number of sol-gel coatings. It was found that, for any given particle size and contaminant mass, there is an optimum bed concentration of 20 g L(−1) for which Q(0.5) reaches a minimum. Finally, under the optimum configuration, the fluidized-bed reactor efficiency is only 30% lower than that of photocatalysis on suspended TiO(2) nanopowder, thus making the proposed fluidized system a viable alternative to slurry-TiO(2) reactors.