PhD. (Department of Chemical Engineering, Faculty of Engineering and Technology), Vaal University of Technology. / The use of solar photocatalysis for the treatment of aromatic chemicals like phenol in wastewater has attracted significant attention due to the low cost of sunlight. However, sunlight is unreliable since its intensity fluctuates during the day. This drawback can be addressed by supplementing sunlight with artificial UV lamps when the solar intensity reduces. In this work, such a hybrid solar/UV lamp reactor, internally illuminated by the UV lamp and externally by sunlight, was designed. Phenol was used as the model pollutant and the nanophase Aeroxide P25 TiO2 was employed as the photocatalyst and fluidized by compressed air. The catalyst and bubble distribution in the reactor was analysed using computational fluid dynamics (CFD) while the Monte Carlo (MC) method was used to model the light distribution and reaction kinetics. Finally, a lamp controller was designed to specify the required UV lamp output as a function of the solar intensity. The CFD simulation using ANSYS CFX 17 showed that a fairly homogeneous distribution of the catalyst was achieved in the reactor. Consequently, accurate simulations of the light distribution could be achieved without considering the hydrodynamics. The MC models revealed that bubbles did not significantly influence light absorption at the optimum catalyst loading. This showed that air was a good medium for fluidization as it could provide good mixing and oxygen electron acceptor without negatively affecting light absorption. The forward scattering behaviour of the P25 TiO2 and the increase in light attenuation with catalyst loading was confirmed in this work. The optimum catalyst loading in the different reactor configurations was 0.15 g/L (tubular solar), 0.2 g/L (annular solar), 0.4 g/L (annular UV lamp), and 0.4 g/L (hybrid light). This resulted in experimental reaction rates of 0.337 mgL-1min-1 (tubular solar), 0.584 mgL-1min-1 (annular UV lamp), and 0.93 mgL-1min-1 (hybrid light). An analysis of the local volumetric rate of energy absorption (LVREA) and reaction rate profiles along the radial coordinate showed a non-uniformity which worsened with an increase in catalyst loading. The reaction order with respect to the volumetric rate of energy absorption (VREA) indicated that solar illumination resulted in a higher electron-hole recombination as compared to UV illumination. This, combined with the higher intensity of the UV lamp, resulted in a higher reaction rate under UV light as compared to sunlight, demonstrating that the UV lamp could be used to supplement sunlight. For a typical sunny day, a lamp controller was designed that could adjust the UV lamp output as a function of the solar intensity to maintain the reaction rate at a reference level while ensuring less energy consumption than an ON/OFF lamp controller. This work demonstrated the feasibility of hybrid solar/UV lamp photocatalysis reactor which could maintain the advantages of solar photocatalysis while mitigating its drawbacks.

A presença de fármacos em águas superficiais, subterrâneas, solo e efluentes tem gerado preocupações no que diz respeito à ineficiência dos tratamentos convencionais e os riscos ambientais associados ao descarte inadequado destes micropoluentes emergentes. Diante disto, o objetivo deste trabalho foi avaliar a adsorção e a dessorção de ácido acetilsalicílico (AAS) em solução aquosa em carvão ativado de babaçu in natura (CAB) e funcionalizado quimicamente com HNO3 (CAA). Foram determinados aos adsorventes os teores de cinzas, umidade e de material volátil, granulometria, análise elementar, aspectos morfológicos dos adsorventes, por meio de microscopia eletrônica de varredura (MEV), pH no ponto de carga zero (pHPCZ), características da superfície dos adsorventes por meio de adsorção/dessorção de N2 e os grupos funcionais por espectroscopia na região no infravermelho (FTIR) e método de Boehm. Ensaios cinéticos, de equilíbrio e termodinâmicos foram conduzidos para a adsorção e dessorção em batelada. Ensaios em coluna de leito fixo foram conduzidos preliminarmente. Os resultados dos ensaios cinéticos de adsorção e dessorção foram ajustados aos modelos de pseudo-primeira e pseudo-segunda ordem e os resultados de equilíbrio de adsorção aos modelos matemáticos de Langmuir e Freundlich e de dessorção ao modelo de Langmuir-Freundlich. Aos resultados termodinâmicos de adsorção e dessorção foram determinados os valores de ΔG º, ΔH º e ΔS º. Os adsorventes apresentaram características microporosas, com valores de pHPCZ de 6,4 para CAB e 4,5 para CAA que interferem diretamente na adsorção. Os dados da cinética de adsorção foram melhor ajustados ao modelo de pseudo-segunda ordem e os do equilíbrio de adsorção ao modelo de Langmuir. Os resultados do ΔG º e ΔH º resultaram negativos e o do ΔSº positivo nas temperaturas de 25 ºC, 45 ºC e 55 ºC, indicando que o processo é exotérmico e espontâneo. O aumento da temperatura favoreceu a adsorção com o CAB, e reduziu a capacidade adsortiva do CAA. Em pH ácido a quantidade máxima adsorvida do AAS foi de 119,80 mg g -1 para CAB e 115,08 mg g -1 para CAA. Nos ensaios de dessorção os melhores resultados foram observados com o CAB em pHPCZ = 6,4. Os resultados cinéticos tiveram melhor ajuste ao modelo de pseudo-primeira ordem. As isotermas de dessorção ajustadas ao modelo de Langmuir-Freundlich. Adicionalmente, com os ciclos foram obtidos os percentuais médios de adsorção e dessorção de 92% e 75% no CAB e de 72% e 10% no CAA, respectivamente. O conjunto de dados de pH e temperatura avaliados nos ensaios em coluna de leito fixo determinaram que a reação de adsorção é melhor definida em pH 2,0 e temperatura de 40 ºC. / The presence of drugs in surface water, groundwater, soil and effluents has raised concerns regarding the inefficiency of conventional treatments and the environmental risks associated with the inappropriate disposal of these emerging micropollutants. The objective of this work was to evaluate the adsorption and desorption of acetylsalicylic acid (AAS) in aqueous solution of in natura activated babassu carbon (CAB) and chemically functionalized with HNO3 (CAA). The ash adsorbents, moisture and volatile matter, granulometry, elemental analysis, morphological aspects of the adsorbents were determined by scanning electron microscopy (SEM), pH at the zero load point (pHPCZ), surface characteristics of the adsorbents. adsorbents by adsorption/desorption of N2 and functional groups by infrared spectroscopy (FTIR) and Boehm method. Kinetic, equilibrium and thermodynamic tests were conducted for batch adsorption and desorption. Preliminary tests were conducted in a fixed bed column. The adsorption and desorption kinetics results were adjusted to the pseudofirst and pseudo-second order models and the adsorption equilibrium results to the Langmuir and Freundlich mathematical models. The values of ΔG º, ΔH º and ΔS º were determined for the thermodynamic adsorption and desorption results. The desorption equilibrium results were fitted to the Langmuir-Freundlich model. The adsorbents presented microporous characteristics, with values of pHPCZ of 6.4 for CAB and 4,5 for CAA that interfere directly in the adsorption. Adsorption was favored at acidic pH in both adsorbents. The adsorption kinetics data were better fitted to the pseudo second order model and those of the adsorption equilibrium to the Langmuir model, indicating the influence of the chemisorption on the adsorption. The results of ΔGº and ΔHº were negative and that of ΔSº positive at temperatures of 25 ºC, 45 ºC and 55 ºC, indicating the influence of the chemisorption in the process that is exothermic and spontaneous. The increase in temperature favored the adsorption with the CAB, and reduced the adsorptive capacity of the CAA. At acid pH the maximum adsorbed amount of ASA was 119.80 mg g -1 for CAB and 115.08 mg g -1 for CAA. In the desorption experiments the best results were observed with the CAB at pHPCZ = 6.4. The kinetic results had better fit to the pseudo first order model. Desorption isotherms fitted to the Langmuir-Freundlich model indicated affinity between the CAB, ASA and the chosen regenerant. Additionally, the average adsorption and desorption percentages of 92% and 75% in CAB and 72% and 10% in CAA, respectively, were obtained with the cycles. The set of pH and temperature data evaluated in the fixed bed column assays determined that the adsorption reaction is best defined at pH 2.0 and temperature of 40 ºC.