Nano-photocatalytic mineralization and disinfection for water reclamation: from catalyst engineering to process optimization and modelling.

Abstract

In this thesis, a feasible photocatalytic technology for water treatment was explored using a bottom-up approach in four separate research and developmental stages. These include (1) the synthesis, characterizations and photocatalytic activity (PCI) evaluation of a thin nanocrystals layer of titanium dioxide (TiO₂) immobilised onto modified mesoporous kaolin clay; (2) optimization and kinetics study of the photocatalytic reaction with recalcitrant organic dye Congo red (CR); (3) optimization and kinetics study of the photocatalytic disinfection with surrogate Escherichia coli and (4) assessment of a sequential batch reactor (SBR) mode for semi-continuous removal of dissolved pharmaceutical organic matters from secondary municipal wastewater. A modified two step sol-gel approach was used to immobilise layered of TiO₂ nanocrystals onto structurally stable kaolin (TiO₂-K) particles for enhanced physical properties. The TiO₂-K demonstrated a superior settling ability, adsorption capacity, PCI and stability compared to other conventional TiO₂ particles. Microscopic characterizations revealed that the modified kaolin provides a delaminated sandwich silica structure which minimizes chemical intercalation and further promotes high external surface area for heterocoagulation with microporous TiO₂ nanocrystals. Thermal regeneration cycles for the photocatalysts lifespan study examines that the PCI of TiO₂-K was improved over six treatment cycles, as a result of the change in average TiO₂ nanocrystals size and porosity. The real operational performances of the TiO₂-K photocatalysts for the organic degradation in water were investigated in a self-designed laboratory scale annular slurry photoreactor (ASP) system. The effect of key operational factors for the ASP system, such as TiO₂-K loading, pH, aeration rate and CR concentration were investigated. Results show that pH was the most significant factor that affects the adsorption and photocatalytic reactions in the ASP system. The point of zero charge (PZC) for the TiO₂-K particles was found to shift towards more basic extent of 9.5, resulting in a detrimental PCI when the ASP was operated at pH > PZC (TiO₂-K). The optimum operating conditions of the ASP was found to be 6.0 g L⁻¹ TiO₂-K loading, pH 7.0, 7.5 L min⁻¹ aeration rate and 40 mg L⁻¹ CR concentration. Under these optimum ASP conditions, the CR was completely photo-degraded in 4 h along with 80% reduction in chemical oxygen demand (COD) of the LC/MS-identified intermediates species. Owing to the lack of understanding and predictive in the singular optimum ASP condition, the Taguchi method was employed to collectively optimise the operating factor; determine the synergistic factor interactions and key influential factors, and further develop an empirical response surface model for the PCI prediction. From the Taguchi experiments, the photo-oxidation kinetics of CR exhibited saturation kinetics and thus, the Langmuir-Hinshelwood (L-H) model was applied. The Taguchi method predicted the optimum L-H apparent first order rate constant of 3.46 x 10⁻² min⁻¹ under the ASP operating conditions of 8.0 g L⁻¹ TiO₂-K, pH 5.0, 7.5 L min⁻¹ aeration rate and 40 mg L⁻¹ CR. Analysis of variance revealed that the CR concentration is the most significant factor, while pH appears to be the least significant one. The aeration rate in ASP was determined to have a significant synergistic effect on the TiO₂-K loading from the 3-D response surface plot. When the photo-oxidation of CR was referenced to the anatase titanate nanofiber (TNC) with PZC of 4.6, it was observed that the ASP operation was constrained by a narrow functional pH range. The optimum ASP operating conditions with slurry TNC were 4.0 g L⁻¹ TNC loading, pH 3.0, 5 L min⁻¹ aeration rate and 60 mg L⁻¹ CR, resulting in a degradation rate of 3.47 x 10⁻² mol L⁻¹ min⁻¹. Subsequent Taguchi analysis found that the low PZC (TNC) has a profound effect on the synergistic interaction with its loading concentrations. The 3D response plots showed that the low PZC (TNC) could be compensated with a high TNC loading at pH >PZC (TNC) and low aeration rate for optimal conditions. The Taguchi method predicted that the optimum ASP operating conditions were 6.0 g L⁻¹ TNC, pH 9, 5.0 L min⁻¹ aeration rate and 20 mg L⁻¹ CR. Analysis of variance shows that the pH, aeration rate and CR concentration were the significant factors, while TNC loading is the least significant one. The PCI for both photocatalysts were also tested against the photo-disinfection of Escherichia coli (ATCC 11775) as the surrogate indicator in the batch ASP system. In both investigations, the photocatalytic inactivation kinetics was found to exhibit non-linearity in the enumerated bacteria against irradiation time. The modified Hom model was used to fit a sigmoidal-shape bacterial survivor curve with strong shoulder and tailing characteristics. Using the TNC, the dissolved oxygen level in the open ASP system was found to be constantly replenished and further affects the photocatalytic inactivation kinetics. An L-H mechanistic model was proposed to determine oxygen transfer limitation in the photocatalytic disinfection process at different TNC loadings. A Fe²⁺ up to 1.0 mg L⁻¹ could initiate the residual disinfecting effect (RDE) of the photocatalytic reaction in the ASP, with constant dissociation of hydrogen peroxide (H₂O₂) to hydroxyl radicals (OH•). The RDE was diminished with increasing COD values. To effectively suppress the bacterial regrowth, the dissolved organic carbon should be well suppressed below 16 mg COD L⁻¹. Finally, the ASP system was operated as a SBR mode to allow semi-continuous treatment of the secondary municipal wastewater spiked with pharmaceutical Carbamazepine (CBZ) compound. A microfiltration module of 0.2 μm in porosity was fitted parallel to the reactor light source. The effects of key factors such as SBR cycles, nitrate (NO₃⁻), phosphate (PO₄³⁻), COD and photocatalysts on the effective photocatalytic CBZ removal were investigated. When the CBZ was degraded in the presence of high molecular weight effluent organic matter (EOM) in wastewater, the photocatalytic reaction appeared to have a preferential attack on the EOM that subsequently lower the degradation efficiency of CBZ. Other than this, the PO₄³⁻ in the wastewater showed a detrimental effect of photocatalyst fouling and deactivation on both TiO₂ catalysts used, resulting in a strong catalyst deactivation, and fouling of the catalysts. The deactivation and fouling are site specific and do not completely retard the photoactivity of the catalysts used. The sequential batch-annular slurry photoreactor (SB-ASP) was effectively operated up to two SBR cycles at catalyst loading (1 gL⁻¹) without any photocatalyst replacement, as the semi-continuous operation allows simultaneous water discharge and UV photocatalyst reactivation phase.