(234e) Heterogeneous Photocatalytic Degradation of Phenol by Titanium Dioxide Nanoparticles: An Optimization Approach

     Anthropogenic and industrial activities in the past have introduced large quantities of chemicals into the environment causing potential harm to many ecosystems.  Phenol is one such harmful ecotoxins with widespread industrial and commercial application. Phenols are known to possess carcinogenicity, teratogenicity, mutagenicity and endocrine disrupting ability.

     Phenols are routinely used in the manufacture and production of many phenol based products, such as phenolic resins, insulation panels, herbicides and pesticides, and in the formulation of paints, lubricants etc.  Phenol discharges into the environment arises from industries such as petroleum refinery, pulp and paper, metal casting, coal gasification and steel manufacturing.  According to Environment Canada's National Pollutant Release Inventory (NPRI) database, 450 - 500 tonnes of phenol have been released annually into the Canadian environment over past five years. In the United States, annual discharge estimate is over 1000 tonnes.

     Phenol removal from industrial effluents can be accomplished by conventional physical, chemical and biological process depending upon the prevalent phenol concentration.  During biological treatment of phenolic based effluents, microbial processes can become impaired at threshold levels.  The United States Environmental Protection Agency (USEPA) has identified granular activated carbon (GAC) adsorption as best available treatment method (BAT) for phenolic contamination.  Removal of phenol can be also achieved by enzymatic treatment using laccase and horseradish peroxidase enzymes.  Enzymatic treatment has been reported to be more advantageous over biological or other physical/chemical processes.  However, the former process relies on phase transfer or partial polymerization of the contaminant rather than its degradation.  In recent years, heterogeneous photocatalysis using titanium dioxide (TiO₂) have been identified as a potential alternative to existing treatment technologies.  Heterogeneous photocatalysis offer a unique advantage over other alternatives because it provides a green solution; since, toxic organic pollutants are converted into carbon dioxide (CO₂) and water using photonic energy.

     TiO₂ photocatalysis originates from the semiconductor band gap and it is actuated by means of photo-generated electron-hole pair.  Photo-generated electron migrates to the surface of the photocatalyst and initiates the formation of hydroxyl radical (^?OH).  These ^?OH radicals subsequently cause the degradation of organic molecules.  Commercially available TiO₂ particles in the micrometer particle size range lack photocatalytic activity because of a recombination of charge carriers en-route to the catalyst surface.  Augmenting the TiO₂ photocatalytic efficiency is expected to depend on the specific surface area of the catalyst or reducing the diffusion path.  Several nanometer size TiO₂ formulations have been tested for their photocatalytic potential on selected organic compound such as phenol.  For example, Degussa P25 is a commercially available TiO₂ nano-material which has been used to degrade phenol and numerous organic pollutants.

The photocatalytic degradation rate of phenol using TiO₂ is dependent on several factors.  First, the reaction is affected by the number of photons impinging on the reaction surface.  The latter is an inverse function of the wavelength of incident radiation and it is measured directly from the photocurrent produced or it can be quantified using actinometry or radiometry principle.  The second factor affecting the reaction rate is the surface area per TiO₂ particle.  Adsorption onto the photocatalyst at very low concentrations (less than the millimolar range) is known to follow the Langmuir adsorption isotherm.  The third factor is the availability of oxygen in the aqueous phase to generate hydroxyl radicals.  The fourth variable to consider is the amount of catalyst particles available and the crystal structure.  The photocatalytic activity of TiO₂ is affected by the crystal structure, which in turn controls the semiconductor bandgap.  Titanium dioxide exists in four crystalline forms anatase (kinetically stable), rutile (thermodynamically stable), brookite and monoclinic-TiO₂.  In terms of the photocatalytic activity, anatase is more catalytically active than the rutile form.  Most photocatalysts including the Degussa P25 which have been utilized for their excellent photocatalytically activity have the anatase crystal structure.  Although, several reports have already assessed the impact of individual factors on phenolic degradation, further research is required to assess optimizing condition for all the factors using an integrated approach.  Moreover, the photocatalytic rates reported in the literature are difficult to compare due to the different reporting units and experimental conditions.  The purpose of the present study is to address these issues.

     Photocatalytic experiments were performed in a photocatalytic reactor (25 mm ID x 250 mm length) which was fabricated using GE-214 clear fused quartz silica (Technical Glass Products Inc., Painesvile, Ohio).  Sealed reactors containing a phenol solution and TiO₂ photocatalyst were placed in a modified Rayonet RPR-100 UV photocatalytic reactor (Southern New England Ultraviolet Co., Connecticut).  The reactor was equipped with sixteen phosphor-coated low-pressure mercury lamps on the outer perimeter and a centrally located rotating inner carousel.  Six fused quartz reaction tubes were placed on the inner rotating carousel.  Radiation emitted from the lamps (300 nm monochromatic UV light) had an average irradiance of 9 mW/cm² (measured using a calibrated UV-X radiometer).  To minimize variation in irradiance among the UV lamps, control experiments were performed to optimize the rotational speed of the inner carousel.  All experiments were performed in Milli-Q water (18 Mohm) and at a temperature of 37^oC. Over the duration of each experiment, a fixed amount of aqueous solution was withdrawn at specific time intervals and stored in capped aluminum foil wrapped tubes for further analysis.  Phenol degradation was monitored using a high performance liquid chromatograph (Dionex Ultimate 3000, Sunnyvale, CA) which was equipped with a UV-visible photodiode array detector set at 215 nm wavelength.  The instrument was configured with an Acclaim C18-3 mm-2.1 mm (ID) x 100 mm (length) column (Dionex, Sunnyvale, CA) and an oven temperature set at 45^oC.  The eluent (acetonitrile-water mixture (1:4)) flow rate was set at 0.4 ml/min. Headspace carbon dioxide was analyzed using a Varian CP 3800 gas chromatograph (Palo Alto, CA). 

     Response surface methodology was adopted for optimization of photocatalytic degradation of phenol.  A Box-Benkhen design (triplicate set) with three central points was used for optimization study. The experimental design factors used in the study are shown in Table 1.  The factors and the experimental levels for each factor were selected based on literature, available resources and results from preliminary experiments. The three TiO₂ nanoparticles with different particle size (surface area) had identical properties and crystal structure (confirmed by X-ray diffraction). The experimental data were analyzed using Minitab 15 statistical software (Minitab Inc. State College, PA).

     Photocatalytic degradation of phenol was performed at the selected levels of each factor.  The experimental response was expressed as apparent degradation rate.  Analysis of variance (ANOVA) was performed with the process responses for a full quadratic model involving the experimental factors under study.  A backward elimination method was applied and statistically insignificant terms (p > 0.05) were removed to achieve the final response surface model. The quantum efficiency was compared for the different catalyst particle sizes.  The quantum efficiency is defined as the ratio of phenol molecules degraded to the number of incident photon on the reaction surface.

     A typical data set showing the apparent degradation rate (quantum efficiency) versus specific surface area of TiO₂ is shown in Figure 4.  An optimum photocatalytic degradation rate as well as apparent quantum yield was observed with 10 nm (210m²/g) TiO₂ particles.  This observation could be attributed to the quantum size effect, the phenomenon of confinement of charge carriers (electron or hole) and the increase in bandgap as the particle size approaches the order of de-Broglie wavelength of the charged carrier.

     A practical constraint in terms of the applicability of TiO₂ nanoparticles was the poor settling tendency of suspended particles.  Studies to immobilize the TiO₂ nanoparticles onto a solid support medium are underway as a means to eliminate the latter problem.  The bottleneck of the immobilized catalyst system was reported as the potential aggregation tendency of the tiny catalyst particles on the support surface during the thermal stabilization step.  Future work is underway to develop an efficient TiO₂ immobilized system with optimum photocatalytic activity.