INNOVATIVE WATER REUSE PROCESS USING PHOTOCATALYSIS AND CATALYTIC CERAMIC MEMBRANE FILTRATION By Xiaoyu Wang A DISSERTATION Submitted to Michigan State University i n partial fulfillment of the requirements for the degree of Environmental Engineering - Doctor of Philosophy 201 8 ABSTRACT INNOVATIVE WATER REUSE PROCESS USING PHOTOCATALYSIS AND CATALYTIC CERAMIC MEMBRANE FILTRATION By Xiaoyu Wang Water scarcity has become one of the most challenging problems that humanity must address. Co st - effective water and wastewater treatment technologies are imperative to protect public health and the environment. The objectives of this work were to assess catalytic ceramic membrane filtration - ozonation processes for the mitigation of fouling and cal culate the energy cost and environmental impacts of the process; to efficiently use sunlight as energy source for photodegradation and the removal of persistent organics in wastewater; finally, to propose an innovative process that combines photodegradatio n and membrane filtration process for emerging contaminants removal and industrial water reuse. The first part of the study analyzed the effect of ozone dosage on membrane fouling using catalytic ceramic membranes, and used life - cycle - assessment to assess the energy consumption and environmental impacts of the catalytic ceramic membrane system and compared that to hollow fiber membrane filtration. Results showed that membrane fouling was effectively controlled at ozone dosages of 10 µg/s or greater using ma nganese oxide coated membranes. At least 15 µg ozone/s was necessary to control membrane fouling with uncoated titianium oxide membranes. Catalytic ceramic membrane filtration resulted in less energy consumption of pressurization and backwashing as compar ed with hollow fiber membrane filtration, and had a slightly lower environmental impact than hollow fiber membrane filtration. In photodegradation study, three photocatalysts were synthesized and characterized. Photodegradation abilities were compared usi ng with methylene blue as a model contaminant. Moreover, the deactivation mechanism and photo - deactivation of photocatalysts were also studied. The order of photocatalysis degradation efficacies of methylene blue during illumination was CdS > NiFe 2 O 4 ZnF e 2 O 4 . It was found that the methylene blue removal efficacy is affected by the absorption range of the photocatalysts, initial dye concentrations, amount of photocatalysts added, and photoreactor conditions. The proposed water reuse process combining photo catalytic reaction and ceramic membrane filtration to achieve safe and high efficient water reuse and reduce energy consumption and operating costs was evaluated using caffeine as a target contaminant. Results show that by using photodegradation and membra ne filtration, caffeine could be degraded and photocatalysts could be recovered. Copyright by XIAOYU WANG 2018 v ACKNOWLEDGEMENTS I support, I receive important guidance, and I obtained so much happiness from my friends. I really hope to express my appreciation to all the support from so many people. F irst of all, I would like to express my thanks to my advisor, Dr. Susan Masten. Th ank you for your excellent guidance, unconditional support, and all the patience. You are always willing to help and give me all kinds of supports. You make me improve skills throughout the process and teach me stay true. I learned so much from you! I hop e to thank my committee members, Dr. Annick Anctil, Dr. Rebecca Lahr, and Dr. Joan Rose for providing their valuable feedback on my research and suggestions. I would express my thanks to Dr. Anctil again, for offering help and guidance in my research, and expending my research interests. I would also like to thank Dr. Simon Davies, for his fully support and advice , and for his important guidance at the beginning of my research period. I would like to thank our department research administrator Lori Larner , and secretary Laura Post. To all the faculty and staff in department of Civil and Environmental Engineering at Michigan State University, thank you for support and useful courses . I also want to thank the students in my TA classes and the other TAs in th e classes, I learned a lot from you. vi I hope to thank my parents, for your love, courage and support. Although there are thousands of miles of distance between us right now, I always feel your love! Love you! Also, thank my husband, for your company, suppor t and encouragement. Last, I would like to thank the source of funding, the MSU Discretionary Funding Initiative, MSU College of Engineering and Department of Civil and Environmental Engineering. vii TABLE OF CONTENTS LIST OF TABLES ................................ ................................ ................................ .......... ix LIST OF FIGURES ................................ ................................ ................................ ........ x CHAPTER ONE ................................ ................................ ................................ ............. 1 Introduction and Dissertation Overview ................................ ................................ ......... 1 1.1 Background ................................ ................................ ................................ ......... 1 1.2 Water reuse process with photocatalytic reactors and hybrid membrane - ozonation filtration ................................ ................................ ............................... 3 1.3 Ozonation catalytic ceramic membrane filtration ................................ ................. 6 1.4 Photocatalytic degradation of organic contaminants in wastewater treatment application ................................ ................................ ................................ ........... 9 1.5 General purpose of the study ................................ ................................ ............ 13 REFERENCES ................................ ................................ ................................ ............ 15 CHAPTER TWO ................................ ................................ ................................ .......... 21 Analysis of energy costs for catalytic ozone membrane filtration ................................ . 21 Abstract ................................ ................................ ................................ ................... 22 2.1 Introduction ................................ ................................ ................................ ........ 23 2.2 Experimental methods ................................ ................................ ....................... 25 2.2.1 Feed w ater ................................ ................................ ................................ . 25 2.2.2 Membrane module ................................ ................................ ..................... 25 2.2.3 Ozone contactor ................................ ................................ ......................... 26 2.2.4 Hybrid ozonation filtration setup ................................ .............................. 26 2.2.5 Data Analysis ................................ ................................ ............................. 28 2.3 Result and discussion ................................ ................................ ........................ 32 2.3.1 Comparison of permeate flux and resistance fouling ................................ . 32 2.3.2 Energy and economic analysis of membrane systems for design applications ................................ ................................ ................................ 37 2.4 Conclusion ................................ ................................ ................................ ......... 40 APPENDIX ................................ ................................ ................................ ................... 41 REFERENCES ................................ ................................ ................................ ............ 45 CHAPTER THREE ................................ ................................ ................................ ...... 49 Energy consumption and environmental impact of ozonation catalytic membrane filtration system and comparison with hollow fiber membrane for water treatmen t ...... 49 Abstract ................................ ................................ ................................ ................... 50 3.1 Introduction ................................ ................................ ................................ ........ 51 3.2 Experimental ................................ ................................ ................................ ...... 54 3.2.1 Goal and scope definition ................................ ................................ ........... 54 3.2.2 Data Collection ................................ ................................ ........................... 58 3.2.3 Impac t Assessment ................................ ................................ .................... 59 viii 3.3 Results and discussion ................................ ................................ ...................... 60 3.3.1 Catalytic ceramic membrane with ozonation energy and environmental impacts a nalysis ................................ ................................ ......................... 60 3.3.2 Comparison between catalytic ceramic membrane and hollow fiber membrane in energy consumption and environmental impacts ................. 63 3.4 Conclusions ................................ ................................ ................................ ....... 67 APPENDIX ................................ ................................ ................................ ................... 69 REFERENCES ................................ ................................ ................................ ............ 98 CHAPTER FOUR ................................ ................................ ................................ ...... 102 Effects of methylene blue photocatalytic degradation under visible light irradiation by ZnFe 2 O 4 , NiFe 2 O 4 and CdS ................................ ................................ ....................... 102 A bstract ................................ ................................ ................................ ................. 103 4.1 Introduction ................................ ................................ ................................ ...... 104 4.2 Materials and Methods ................................ ................................ .................... 106 4 .2.1 Materials preparation ................................ ................................ ................ 106 4.2.2 Photocatalysts characterization ................................ ................................ 108 4.2.3 Evaluation of photocatalysts in degradation abili ty ................................ ... 108 4.3 Results and Discussion ................................ ................................ ................... 109 4.3.1 XRD analysis of ZnFe 2 O 4 ................................ ................................ ......... 109 4.3.2 SEM analysis ................................ ................................ ........................... 110 4.3.3 UV - vis analysis ................................ ................................ ......................... 112 4.3.4 Photocatalytic degradation of methylene blue test ................................ ... 115 4.3.5 Methylene blue degradation kinetics by CdS ................................ ........... 120 4.3.6 Methylene blue degradation mechanism by CdS ................................ ..... 122 4.3.7 Photo - deactivation of CdS ................................ ................................ ........ 124 4.3.8 SEM and EDS of used CdS ................................ ................................ ..... 125 4.4 Conclusions ................................ ................................ ................................ ..... 129 REFERENCES ................................ ................................ ................................ .......... 130 CHAPTER FIVE ................................ ................................ ................................ ......... 134 Innovative Water Reuse Process with Ph otocatalytic Reactors and Catalytic Ceramic Membrane Filtration ................................ ................................ ................................ ... 134 Abstract ................................ ................................ ................................ ................. 134 5.1 Introduction ................................ ................................ ................................ ...... 135 5.2 Materials and Methods ................................ ................................ .................... 138 5.2.1 Photocatalysts preparation and characterization ................................ ...... 138 5. 2.2 Ceramic membrane filtration process set up ................................ ............ 138 5.2.3 Test of the designed water reuse process ................................ ................ 138 5.3. Results ................................ ................................ ................................ ........... 139 5.4 Conclusions ................................ ................................ ................................ ..... 141 REFERENCES ................................ ................................ ................................ .......... 143 CHAPTER SIX ................................ ................................ ................................ ........... 147 Conclusions and Future Perspective ................................ ................................ ......... 147 ix LIST OF TABLES Table 2.1 Backwash process operation parameters ................................ ..................... 3 1 Table 2.2 Energy consumption for Mn oxide coated membra ne filtration with different ozone dosing rates and operation modes with one membrane module ........................ 39 Table 3.1 Main processes in the membrane filtration life cycle and materials/energy consumption ................................ ................................ ................................ .................. 5 9 Table 3.2 Characterized environmental impacts in catalytic me mbrane (CC) and hollow fiber (HF) membrane filtration processes ................................ ................................ ...... 6 7 Table 3.A.1 Catalytic ceramic membrane energy network ................................ .......... 9 1 Table 3.A.2 Process contribution in energy consumption ................................ ............ 9 1 Table 3.A.3 Catalytic ceramic membrane environmental impacts normali zation ....... 9 2 Table 3.A.4 Catalytic ceramic membrane environmental impacts normalization ....... 9 3 Table 3.A.5 Comparison of catalytic membrane and hollow fiber membrane energy network ................................ ................................ ................................ .......................... 9 3 Table 3.A.6 Comparison of catalytic membrane and hollow fiber membrane envi ronmental impacts characterization ................................ ................................ ..... 9 5 Table 3.A.7 Comparison of catalytic membrane and hollow fiber membrane environmental impacts normalization ................................ ................................ ..... 100 x LIST OF FIGURES Figure 1.1 Designed water reuse process flow diagram ................................ .................. 4 Figure 1.2 Schematic representation of photocatalysts [39] ................................ ............ 10 Figure 2.1 Schematic of the ozonation - membrane filtration system .............................. 28 Figure 2.2 Effect of ozone dosage on permeability using manganese oxide coated catalytic membrane filtration dead - end operation mode (Conditions: p ermeability in L/m²·h·bar, Feed TOC=12.1 mg/L, temperature=22±2 ) ................................ ............ 33 Figure 2.3 Effect of ozone dosage on permeate flux during manganese oxide coated catalytic membrane filtration crossflow operation mode (Conditions: permeability in L/m²·h·bar, Feed TOC= 12.1 mg/L, cross flow velocity = 0.5±0.1 m/s temperature=22±2 ) ................................ ................................ ................................ ................................ . 34 Figure 2.4 Effect of ozone dosage on permeability during titania oxide coated catalytic membrane filtration (a) dead - end op eration mode and (b) crossflow operation mode (Conditions: permeability in L/m²·h·bar, Feed TOC=12.1 mg/L, cross flow velocity = 0.5±0.1 m/s temperature=22±2 ) ................................ ................................ ................ 35 Figure 2.5 Effect of ozone dos age on fouling resistance in manganese oxide coated and titania oxide coated catalytic membrane filtration system ................................ .............. 37 Figure 2.6 Energy consumed during the treatment of in 1,000 liters of water (including backwash cycles) ................................ ................................ ................................ .......... 38 Figure 3.1 System boundaries of the study in (a) catalytic ceramic ozonation membrane filtration processes and (b) hollow fiber membrane processes with chlorine a s disinfectants ................................ ................................ ................................ .................. 55 Figure 3.2 Embodied energy consumption of total and different processes in catalytic ceramic membrane ................................ ................................ ................................ ........ 60 Figure 3.3 Environmental impacts normalization results for catalytic ceramic membrane filtration processes ................................ ................................ ................................ ........ 61 Figure 3.4 Comparison of energy consumption with high ozone injection rate in catalytic c eramic membrane filtration system. ................................ ................................ ............. 63 Figure 3.5 Embodied energy consumption in catalytic ceramic membrane (CC) and hollow fiber membrane (HF) filtration in water treatment application ............................. 64 Figure 3.6. Comparison of environmental impacts normalization results for catalytic xi ceramic membrane (CC) and hollow fiber membrane (HF) filtration processes with HF impacts processes percentage ................................ ................................ ...................... 66 Figure 3.A.1. System boundaries of hollow fiber membrane processes with chlorine as disinfectants ................................ ................................ ................................ .................. 71 Figure 3.A.2. System boundari es of the study in catalytic ceramic ozonation membrane filtration processes ................................ ................................ ................................ ........ 79 Figure 4.1 XRD patterns of the obtained ZnFe 2 O 4 ................................ ...................... 110 Figure 4.2 SEM images of (a) CdS, (b) ZnFe 2 O 4 , and (c) NiFe 2 O 4 ............................. 111 Figure 4.3 UV vis absorption spectra of different photocatalysts (a) CdS (b) ZnFe 2 O 4 and (c) NiFe 2 O 4 ................................ ................................ ................................ .................. 113 Figure 4.4 Degradation curves of methylene blue over different photocatalysts, initial methylene blue concentration of 8 µmol/L, with the amount of photocatalyst used controlled at 1.0 g/L. ................................ ................................ ................................ .... 115 Figure 4.5. Comparison of removal efficiency of different photocatalysts, initial concentration of MB at 8 µmol/L, at 25 ºC ................................ ................................ ... 116 Figure 4.6 Different concentrations of methylene blue removal efficacy with ZnFe 2 O 4 under visible light, at 25 ºC ................................ ................................ .......................... 117 Figure 4.7 Methylene blue removal efficacy with different (a) reactors and (b) amounts of nanop articles under visible light, initial MB concentration of 8µmol/L, at 25 ºC ........... 118 Figure 4.8 Degradation curve of methylene blue with CdS ................................ ......... 120 Figure 4.9 Kinetics plots for methylene blue degradation by CdS ............................... 121 Effect of radical scavengers on the degradation of MB with CdS ............. 123 Figure 4.11 Methylene blue degradation over time by continued treatment using CdS ................................ ................................ ................................ ................................ .... 124 Figure 4.12 First - order plots of 2 nd illumination treatment with CdS ............................ 124 Figure 4.13 SEM images of CdS before and after photo - treatment ............................ 126 Figure 4.14 EDS analysis of used CdS and NiFe 2 O 4 ................................ .................. 127 Figure 5.1 Schematic of the proposed water reuse process ................................ ....... 139 Figure 5.2 Caffeine removal efficiency by designed treatment process with no oxidation during membrane filtration ................................ ................................ ........................... 140 1 CHAPTER ONE Introduction and Dissertation Overview 1.1 Background only about 2.53% of the total water is fr esh. Moreover , the potable fresh water in rivers, lakes and ground water that is easily accessible is approximately 0.3% of the total fresh water. Water scarcity affects almost every continent and nearly 1.2 billion people live in areas of physical scarcit y . Another 500 million people live in areas that are approaching conditions resulting in water scarcity. 1.6 billion people are facing economic water shortage where countries lack the necessary infrastructure to safety treat and distribute water from river s and aquifers [1] . Some scientists estimated that, by 2025, two - thirds of the world's population could be living under water stressed conditions [2] . With the existing climate change scenario, almost half the world's population will be living in areas of high water stress by 2030, including between 75 million and 250 million people in Africa. In addition, water scarcity in some arid and semi - arid places will displace between 24 million and 700 million people [2] . Water scarcity has become one of the most c hallenging problems to humanity mainly because of the population over - growth. Safe, clean drinking water access is important for public health. In 2010, the UN recognized the human right to water and sanitation: everyone has the right to sufficient, cont inuous, safe, acceptable, physically accessible and affordable water for personal and domestic use [3] . The key issue for the treatment and reuse of wastewater is the effective 2 removal of toxic contaminants, particularly organic pollutants, of low concentr ations. The demands for high quality water keep driving research for more effective, economical and energy efficient processes for the treatment of water sources [4] . Feasible water reuse process es that produce safe, clean water are greatly need ed . Water treatment is the process to achieve the water quality that meets specified goals or standards set by the end user or a community through its regulatory agencies. Goals and standards can include the requirements of regulatory agencies, additional requireme nts set by a local community, and requirements associated with specific industrial processes [5] . Municipal water treatment system s ha ve a history of more than 200 years, and during the last 30 to 40 years there are many changes in both opinion s regarding and technologies used for water treatment. Conventional drinking water treatment technologies and processes include lime - soda, ion - exchange, coagulation/flocculation, sedimentation. Choosing the effective water treatment processes and designing water treat ment plant should be based on the source water characteristics. The source water includes groundwater, lakes and reservoirs, rivers, seawater, and wastewater impaired waters. The relative concentrations of source water, and other water quality parameters a ffect treatment depend heavily on local conditions of geology, climate, and human activity [6] . Wastewater sources can be classified into domestic wastewater, industrial wastewater, infiltration/inflow, and storm water [7] . Like with water treatment desig n, it is important to choose the appropriate treatment processes based on wastewater characteristics. The 3 main task of wastewater treatment is to remove hazard s from wastewater that would affect the environment . While b iological wastewater treatment is eff ective at reducing the biochemical oxygen demand to levels around 20 - 30 mg/L and is inexpensive compared to chemical tr eatment , but there are limitations due to the low concentration of organic contamina n ts in the influent, the amount of energy needed to p roduce oxygen to oxidize organics , or the amount of land required for lagoon systems . In the developed world, industries are typically required to pretreat their wastes before discharge to the sanitary sewers. These same industries often obtain their wa ter from municipal treatment systems. As a result, water is pumped from lakes or underground to the water treatment plant where it is processed and then pumped factories . T he wastewater is pretreated and then discharged to the sanitary sewer. After treatme nt, the water is discharged back to the environment. Energy and cost s can be saved if local water reuse process can be adopted, thereby reducing the energy requirements for wastewater/water pumping. These energy savings are significant, as drinking water a nd wastewater plants typically are the largest energy consumers, accounting for 30 to 40 % of the total energy consumption in the US [8] . Moreover, d rinking water and wastewater systems account for the release of more than 45 million tons of greenhouse gas es annually in the US [8] . 1.2 W ater reuse process with photocatalytic reactors and hybrid membrane - ozonation filtration Conventional biological treatment methods are not very effective for some resistant 4 organic pollutants such as organic dyes, emerging co ntaminants and organic acids. Advanced oxidation processes (AOPs) can be described as aqueous phase oxidation methods based on the use of highly reactive species for the destruction of the target pollutant. Hydroxyl radicals are the predominant reactive sp ecies that are used for the degradation of organic pollutants [ 9 ] . The water reuse process is designed to achieve safe and high efficient water reuse process and save considerable amounts of energy. The objective of this research is to design and investiga te an innovative process combining photodegradation with membrane filtration for the removal of emerging contaminants and the reuse of industrial water. The process has the potential to degrade pharmaceuticals and personal care chemicals, and therefore cou ld be used for the reuse of water in these industries. If successful this would achieve the goal of making wastewater into water source , while saving energy and reduce environmental impacts of water reuse. Figure 1.1 Designed water reuse process flow d iagram 5 This designed photodegradation process , which uses sunlight as an energy source to treat wastewater by photocatalysis , could be an effective technique for the removal of organic pollutants and some sulfate inorganics in wastewater. Photocatalysis c ould significantly decrease energy consumption in waste water treatment. In addition, the production of hydrogen would be expected by utilizing organics in the wastewater [10 - 13] . Given the expected increase in the cost of fossil fuels and concern about the environmental impact of emissions from the combustion of carbon - based fuels this may be become feasible in the future. The objectives of the research were t o efficiently use sunlight as energy source for the photodegradation of persistent organic chemic al s; t o reduce energy consumption and save costs for the water reuse process compared to water and wastewater treatment processes. The sub - objectives of the photodegradation are: 1. To evaluate the efficacy of several photocatalysts ( CdS, ZnFe 2 O 4 , NiFe2O 4). CdS is the most widely studied photocatalys t for hydrogen production and solar energy storage. ZnFe 2 O 4 and NiFe 2 O 4 can also act as photocatalysts, and they are comparably more environmentally friendly , and b oth have a narrow gap band. 2. To determine the mechanism of photodegradation of methylene blue, and find out the factors that affect photodegradation efficiency. 3 . To analyze the efficiency of contaminant efficiency and the reasons for photo - deactivation effects during photocatalytic processes. 6 Th e sub - objectives of the membrane - ozonation filtration are: 1. To compare membrane performance between manganese oxide catalytic membrane and virgin titania oxide catalytic membrane. 2. To test the difference in cake resistance and permeate flux between cr oss - flow and dead - end operation mode. 3. To determine the most economical operational mode in membrane system for decreasing fouling and enhancing permeate flux. To determine the ozone dosage for both cross - flow and dead - end of different membrane system un der applicable flux/pressure ratio, and calculate the cost for both modes in ozone generation, extra pump operation, and some other cost. 4. To analyze the energy consumption and environmental impacts of the catalytic ceramic membrane system by life - cycle - assessment, which were then compared to that for hollow fiber membrane filtration. 1.3 Ozonation catalytic ceramic membrane filtration Membrane filtration is an effective technology for water treatment. However, membrane fouling limits the application of micr ofiltration/ultrafiltration technology . Research has shown that NOM (natural organic matter) is the main component of the membrane surface fouling cake [ 14 ] . The use of ozone in combination with membrane filtration has been reported to reduce membrane foul ing [ 15 - 18 ] . Ozone is a powerful oxidant that preferentially oxidizes electron - rich moieties containing carbon carbon double bonds and aromatic alcohols, with high oxidation potential and high reactivity with NOM [ 14 ] . Ozonation can also be used as a disin fection process to reduce THMs in drinking water formed during 7 chlorination [ 19 ] . Minimizing ozonation dosages also results in the formation of less organic carbon and few halogenated acetic acids (upon chlorination for secondary disin f ection). However, o zonation application to membrane filtration process is limited by the low chemical resistance of most polymeric membrane materials to ozone [ 20 ] . Because ceramic membranes are chemically stable they can be used in conjunction with ozonation without degrada tion [ 21 ] . They also have longer lifetimes than polymeric membranes [ 22 , 23 ] . Moreover, the surface properties of ceramic membranes allow for the use of coating techniques to improve membrane performance. Metal oxides, such as iron oxide, manganese oxide, titania, alumina, and zirconia, act as catalysts for degradation of ozone and the hydroxyl radicals formation [ 16 ] . These catalytic reactions are assumed to occur when ozone gas is adsorbed on the metal oxide surface [ 14 ] . Karnik et al. [ 21 ] used salicyli c acid as a probe compound to demonstrate the catalytic reaction during ozone membrane filtration. Kim et al. [ 24 ] studied the effect of ozone dosage and hydrodynamic conditions on permeate flux in a hybrid ozonation ceramic ultrafiltration membrane system by treating natural water, and observed that fouling was mitigated at higher cross - flow velocities at the same ozone dosage and lower transmembrane pressures (TMP). Studies have been conducted about membrane coating and catalytic membranes. Karnik et al . [ 25 - 26 ] conducted research about the AFM and SEM characterization of alumina zirconia titania (AZT) ceramic membranes coated with iron oxide nanoparticles: they used atomic force microscopy (AFM), scanning electron microscopy (SEM) and energy dispersive X - ray spectroscopy (EDS) to characterize the effects of sintering and coating 8 layer thickness on the microstructure of the ceramic membranes, and found the optimum coating for iron oxide particles on the underlying AZT which meets the stringent EPA regulat ory requirements. Davies et al. [ 27 ] described the use of the layer - by - layer technique to coat ceramic membranes with iron and manganese oxide nanoparticles and figured out the coating layers are thin and relatively uniform and should be enhanced with th e performance of the hybrid ozonation - filtration process. With the coated membranes it is possible to significantly reduce fouling, improve the removal of DBP precursors and more effectively kill Escherichia coli bacteria. Corneal et al. [ 28 ] conducted ex periments to test the removal efficiency of the coated membranes used to treat water containing natural organic matter (NOM) in a hybrid ozonation membrane water filtration system and 41 % of total organic carbon (TOC) was reduced in the water samples. T he manganese oxide surface layer acted as a catalyst for the oxidation of suspended and dissolved organic carbon when the membrane was used in a combined ozonation - membrane filtration system treating natural water. This catalytic activity manifested itself as improved recovery of the permeate flux due to the oxidation of organic contaminants that deposited on the membrane surface. [ 29 ] Kim et al. [ 30 ] observed less fouling at higher cross - flow velocities at the same ozone dosage and lower transmembrane pressu res (TMP). Ozone transfer efficiency from the gas phase to the bulk liquid phase was found to be higher at higher ozone concentrations, 9 cross - flow velocities , and TMPs. The accessibility of foulants to ozone at the catalytic membrane surface is believed to be a key factor affecting fouling behavior. 1.4 Photocatalytic degradation of organic contaminants in wastewater treatment application Over the last few decades, the application of photocatalysis to utilize solar light energy and for the photodegradation of organic compounds present in water and wastewaters is getting more and more popular [ 31 ] . This is a desired and vital process for both water purification and renewable hydrogen energy production [ 32 - 35 ] . Electrons in valence band of photocatalysts are a the conduction band by the absorption of a photon with sufficient energy. During this process, electrons and holes are generated in the conduction band (CB) and valence band (VB) , respectively [ 36 ] . The b and gap of a photocatalyst is defined as energy gap extending from the top of the filled valence band to the bottom of empty conduction band [ 37 ] . Considering the overall potential losses, there should be energy provided to excite the electron to overcome bandgap energy of a semicondu ctor to achieve organic compounds degradation [ 3 8 ] . A schematic of semiconductor is shown in fig. 1 .2 . The process shows on the semiconductor particle and how electrons and holes are generated at valence band and conduction band. 10 Fig ure 1 . 2 Schematic representation of photocatalysts [ 39 ] Generally, semiconductor catalysts have low quantum efficiency because of the high recombination rate of light - induced electron hole pairs at or near the catalyst surface, which is considered one of the major limit ations to obtaining high photocatalytic efficiency [ 40 4 2 ] . Therefore, many efforts have been made to improve the characteristics of semiconductor - based photocatalysts to improve their activity. [ 4 3 ] arises due to the formation of an intermediate radical that can inject electrons into the CB . as ethanol and 1 - - position to OH group [ 29 ] . Other compounds can also ge nerate reducing radicals that react to form the radical anion CO 2 - . Besides oxidation of a scavenger by holes, the scavenging surface OH groups or adsorbed water molecules [ 3 2 ] . Using sunlight to degrade organic compounds by a semiconductor requires that the conduction band of the photocatalysts be more negative than the potential of the 11 co mpounds , such as methanol, are better electron donors than water . They will also produce H + , and when they are reduced, hydrogen would be produced. The rate of this hole recombination reaction can be retarded by the deposition of noble metals on the photoc atalyst surface, the role of which is to scavenge electrons from the semiconductor , thereby increasing the electron transfer rate to the adsorbed species and decreasing the possibility of their recombination with holes [ 33 ] . O ver 35 organic compounds or w aste products that could potentially be photodegraded [34] , including sugars, alcohols, aldehydes, ketones and organic acids are commonly found in wastes derived from biomass. The degradation of organic pollutants often obeys an overall first order kinetic s and its efficiency can be quantitatively characterized with the apparent reaction rate constant of the reaction [33] . The rate constant depends on the pollutant concentration and light source intensity, among other things. Moreover, the process is econom ically advantageous if the photocatalyst can treat the pollutant rapidly and efficiently at high pollutant concentrations. It is thus highly desired to develop Fenton - like photocatalysts that work efficiently at high pollutant concentrations under sunlight illumination. Many factors can affect the ability of a hole scavenger to donate an electron into the valence band of the semiconductor. In particular, the electron - hole recombination rate in the photocatalyst is very important in determining if the hole can migrate to the surface and react with the scavenger. They can enhance the ability of the photocatalytic process to harvest solar energy, by preventing electron - hole recombination. 12 The reaction efficiency will increase if the concentration of organic co mpound increases in water. However, if the concentration is too high, there will be negative effects on reaction. For example, in experiment s of oxalic acid as a hole scavenger with zinc ferrite as photocatalysts, excessive oxalic oxide will oxide photocat alysts and stop the reaction [ 4 4 - 4 5 ] . The pH is an important factor in this system, as it can affect the charge of both organic compounds (i.e., the scavenger s) , and the surface properties of the photocatalyst. The point of zero charge (zpc) describes the condition where the electrical charge density on a surface is zero [ 4 6 ] . When the solution pH is lower than the zpc, the immobilized surface of the photocatalyst becomes positive ly charge d and the attraction force increases the adsorption of molecules ont o the activated surface and hence increases the tendency of subsequent photocatalytic reactions [ 4 7 - 4 9 ] . Characteristics of organic pollutants in wastewater differ greatly in several parameters, particularly in their speciation behavior, solubility in wate r , and hydrophobicity. Wastewater pH varies significantly and can play an important role in the photocatalytic degradation of organic contaminants. Electrostatic interaction between a semiconductor surface, solvent molecules, substrate and charged radicals formed during photocatalytic oxidation strongly depends on the solution pH. In addition protonation and deprotonation of the organic pollutants can take place depending on the solution pH [ 50 ] . Zinc ferrite (ZnFe 2 O 4 ) contains earth abundant metals such a s ferric, nickel, can be considered as stable and environmental friendly. It is a promising catalyst for both the 13 production of hydrogen from sunlight and photodegradation [ 51 - 5 4 ] . It has a comparatively low 1.9 eV band gap energy [ 5 2 ] and its CB is locate d at a more negative potential than that for the reduction of water [ 4 5 , 5 3 ] . Existing studies have been conducted to utilize spinel photocatalysts or related mixed metal oxides as catalysts for photohydrogen production [ 51 - 55 ] . M etal - free semiconductors a s photocatalysts in air purification [ 5 6 ] and hydrogen generation [ 5 7 ] have also been investigated In order for photodegradation to be effective, the target chemical must be able to donate electrons. This can be assessed by considering the energy level of the highest occupied molecular orbital [ 56 ] . Another way is to determine the redox potential for the reaction in which an electron donor, RH, loses an electron to form a radical [ 4 4 ] . If the redox pot ential electron transfer is possible. In photocatalytic systems, organic compounds are oxidized and decomposed by the photogenerated holes. 1.5 General purpose of the st udy This dissertation investigates two types of advanced oxidation processes : photodegradation and membrane ozonation filtration to provide safe, low cost and high quality of water. C hapter two describes the investigation of catalytic ozone membrane filt ration . The effect of ozone dosage on membrane fouling was studied in both cross - flow and dead - end configurations. The performance of a manganese oxide coated membrane was compared 14 with that of uncoated titanium oxide membrane. C hapter three describes the comparison of the operation of hybrid ceramic membrane filtration with conventional polymeric membrane filtration in terms of energy consumption and environmental impact. Treatment specifications for hollow fiber (polymeric) membrane filtration were emplo yed. Life cycle assessment (LCA) was used to assess the environmental impact of the two processes along with the energy consumption. The investigation of the characteristics of CdS, ZnFe 2 O 4 , and NiFe 2 O 4 as photocatalysts and their ability to degrade methy lene blue is described in chapter 4 . T he mechanism, kinetics and deactivation effects of methylene blue degradation by photocatalysts was also studied . C hapter five presents the water reuse process that combine the two processes together for caffeine remo val and photocatalysts regeneration. The experiment al results of this designed process showed that both CdS and ZnFe 2 O 4 could remove caffeine in feed water. C eramic membrane could effectively filter nanoparticles from water for future water reuse and catal ysts recovery. 15 REFERENCES 16 REFERENCES [1] Scarcity, Decade, Water for Life, 2015, UN - Water, MDG, Water, Sanitation, Financing, Gender, IWRM, Human Right, Transboundary, Cities, Quality, Food Security." United Nations. Nov. 24 , 2014. [2] Coping with Water Scarcity: challenge of the twenty - first century. UN Water, 2007. United Nations, 22 Mar. 2007. [3] Drinking - water fact sheet. World Health Organization. Nov. 2016. [4] Shannon, M.A., Bohn, P.W., Elimelech, M., Georgiadis, J.G. , Marinas, B.J., Mayes, A.M., 2008. Science and technology for water purification in the coming decades. Nature 452 (20), 301 - 310. [5] Howe, Kerry J, John C. Crittenden, Kerry J. Howe, David W. Hand, George Tchobanoglous, and R R. Trussell. Principles of Water Treatment. Hoboken, New Jersey: John Wiley & Sons, Inc, 2012. pp. 1. [6] Howe, Kerry J, John C. Crittenden, Kerry J. Howe, David W. Hand, George Tchobanoglous, and R R. Trussell. Principles of Water Treatment. Hoboken, New Jersey: John Wiley & Sons, Inc, 2012. pp. 9. [7] Davis, Mackenzie Leo. Water and wastewater engineering. McGraw - Hill, 2010. Page 18 - 2. [8] https://www.epa.gov/sustainable - water - infrastructure/energy - efficiency - water - utilities [9] Comninellis, Christos, et al. Advanced oxidation pro cesses for water treatment: advances and trends for R&D. Journal of Chemical Technology and Biotechnology 83.6 (2008): 769 - 776. [10] Lyubina, T.P., D.V. Markovskaya, E.A. Kozlova, and V.N. Parmon. photocatalytic hydrogen evolution from aqueous solutions of glycerol under visible light irradiation. International Journal of Hydrogen Energy, 2013. [11] Kondarides, D.I., V.M. Daskalaki, A. Patsoura, and X.E. Verykios, Hydrogen production by photo - induced reforming of biomass components and derivatives at ambien t conditions. Catalysis Letters, 2008. 122: 26 - 32. [12] Ashokkumar, M., An overview on semiconductor particulate systems for photoproduction of hydrogen. International Journal of Hydrogen Energy, 1998. 23: 427 - 438. [13] Shimura , K and H. Yoshida, Hetero geneous photocatalytic hydrogen production from water and biomass derivatives. Energy and Environmental Science, 2011. 4: 2467 - 17 2481. [14] U. von Gunten. Ozonation of drinking water. Part I. Oxidation kinetics and product formation, Water Res. 37 (2003) 144 3 1467. [15] S.Van Geluwe, L. Braeken, B. Van der Bruggen. Ozone oxidation for the alleviation of membrane fouling by natural organic matter: a review. Water Res. 45(2011), 3551 - 3570. [16] J. Kim, W. Shan, S. Davies, M. Baumann, S. J. Masten, V. Tarabara. Interactions of aqueous NOM with nanoscale TiO2: implications for ceramic membrane filtration - ozonation hybrid process. Environ. Sci. Technol. 43, (2009) 5488 - 5494. [17] Mozia, S., Tomaszewska, M., Morawski, A.W., 2006. Application of an ozonation - adsorpti on - ultrafiltration system for surface water treatment. Desalination 190, 308 - 314 [18] B. Schlichter, V. Mavorv, H. Chmiel. Study of a hybrid process combining ozonation and membrane filtration - filtration of model solutions. Desalination 156 (2003), 257 - 2 65. [19] Y.G. Park. Effect of Ozonation for Reducing Membrane - Fouling in the UF Membrane. Elsevier Science B.V., P.O. Box 211 Amsterdam 1000 AE Netherlands (2002). [20] Y. Mori, T. Oota, M. Hashino, M. Takamura, Y. Fujii. Ozone - microfiltration system, Desa lination 117 (1998) 211 218. [21] B.S. Karnik, S.H.R. Davis, M.J. Baumann, S.J. Masten, Use of salicylic acid as a model compound to investigate hydroxyl radical reaction in ozonation membrane filtration hybrid processes, Environ. Eng. Sci. 24 (6) (2007) 1 9. [22] M. Mulder. Basic Principles of Membrane Technology, 2nd ed. Kluwer Academic Publishers, Netherlands, 2003. [23] Durant, John L., et al. Mutagenicity of C 24 H 14 PAH in human cells expressing CYP1A1. Mutation Research/Genetic Toxicology and Envir onmental Mutagenesis 446.1 (1999): 1 - 14. [24] J. Kim, S.H. Davies, M.J. Baumann, V.V. Tarabara, S.J. Masten. Effect of ozone dosage and hydrodynamic conditions on the permeate flux in a hybrid ozonatione ceramic ultrafiltration system treating natural wate rs. Journal of Membrane Science (2008) 311, 165 - 172 [25] Karnik, B. S., et al. AFM and SEM Characterization of Iron Oxide Coated Ceramic Membranes. Journal of Materials Science 41.20 (2006): 6861 - 70 [26] Karnik, B. S., et al. TEM Characterization of Iron - O xide - Coated Ceramic Membranes. Journal of Materials Science 44.15 (2009): 4148 - 54. 18 [27] Davies, S. H., et al. Fabrication of Catalytic Ceramic Membranes for Water Filtration. Water Science & Technology: Water Supply 10.1 (2010): 81 - 6. [28] Corneal, Lindsay M., Susan J. Masten, Simon H.R. Davies, Volodymyr V. Tarabara, Seokjong Byun, Melissa J. Baumann, AFM, SEM and EDS characterization of manganese oxide coated ceramic water filtration membranes, Journal of Membrane Science, Volume 360, Issues 1 2, 15 Septe mber 2010, Pages 292 - 302, ISSN 0376 - 7388 [29] Corneal, Lindsay M., et al. Mn Oxide Coated Catalytic Membranes for Hybrid Ozonation - Membrane Filtration: Membrane Microstructural Characterization. Journal of Membrane Science 369.1 - 2 (2011): 182 - 7. [30] Kim, Hyun - Chul, and Brian A. Dempsey. Membrane fouling due to alginate, SMP, EfOM, humic acid, and NOM. Journal of membrane science 428 (2013): 190 - 197. DOI: 10.1016/j.memsci.2012.11.004. [31] Gao, R. Fundamental Reactions in TiO2 Nanocrystallite Aqueous Solu tions Studied by Pulse Radiolysis. Radiation Physics and Chemistry 65.6 (2002): 599 - 609. [32] Schneider, J. and D.W. Bahnemann, Undesired role of sacrificial reagents in photocatalysis. Journal of Physical Chemistry Letters, 2013. 4: 3479 - 3483. [33] Ahmed, S., et al., Influence of parameters on the heterogeneous photocatalytic degradation of pesticides and phenolic contaminants in wastewater: a short review. Journal of Environmental Management, 2011. 92(3): p. 311 - 330. [34] Chen, X.B., S.H. Shen, L.J. Guo, and S.S. Mao, Semiconductor - based photocatalytic hydrogen generation. Chemical Reviews, 2010. 110: 6503 - 6570. [35] Konstantinou, I.K. and T.A. Albanis, TiO2 - assisted photocatalytic degradation of azo dyes in aqueous solution: kinetic and mechanistic invest igations. Applied Catalysis B: Environmental, 2004. 49(1): p. 1 - 14. [36] Qi, Lifang, Jiaguo Yu, and Mietek Jaroniec. Preparation and enhanced visible - light photocatalytic H2 - production activity of CdS - sensitized Pt/TiO2 nanosheets with exposed (001) facets . Physical Chemistry Chemical Physics 13.19 (2011): 8915 - 8923. [37] Qi, Lifang, Jiaguo Yu, and Mietek Jaroniec. Preparation and enhanced visible - light photocatalytic H2 - production activity of CdS - sensitized Pt/TiO2 nanosheets with exposed (001) facets. Phy sical Chemistry Chemical Physics 13.19 (2011): 8915 - 8923. [38] Zhang, Nan, et al. Nanochemistry - derived Bi 2 WO 6 nanostructures: towards production of sustainable chemicals and fuels induced by visible light. Chemical Society Reviews 43.15 (2014): 5276 - 5287. [39] Mozia, S., Photocatalytic membrane reactors (PMRs) in water and wastewater treatment. A review. Separation and Purification Technology, 2010. 73(2): p. 71 - 91. 19 [40] Shimura, K and H. Yoshida, Heterogeneous photocatalytic hydrogen production from water and biomass derivatives. Energy and Environmental Science, 2011. 4: 2467 - 2481. [41] Grady, C.P.L.J., G.T. Daigger, and L. H.C., Biological Waste Treatment. 2nd ed. 1999, New York, NY: Marcel Dekker. [42] Bicakova, O. and P. Straka, Production of hydrogen from renewable resources and its effectiveness. International Journal of Hydrogen Energy, 2012. 37: 11563 - 11578. [43] Chong, Meng Nan, et al. Recent developments in photocatalytic water treatment technology: a review. Water research 44.10 (2010): 2997 - 3027 . [44] Zhang, X., Y. Wang, and G. Li, Effect of operating parameters on microwave assisted photocatalytic degradation of azo dye X - 3B with grain TiO2 catalyst. Journal of Molecular Catalysis A: Chemical, 2005. 237(1 - 2): p. 199 - 205. [45] Rabindranathan, S., S. Devipriya, and S. Yesodharan, Photocatalytic degradation of phosphamidon on semiconductor oxides. Journal of Hazardous Materials, 2003. 102(2 - 3): p. 217 - 229. [46] Russel, W.B., Saville, D.A., and Schowalter, W.R. (1989). Colloidal Dispersions. Cambridg e University Press. [47] Gogniat, G., et al., The bactericidal effect of TiO2 photocatalysis involves adsorption onto catalyst and the loss of membrane integrity. FEMS Microbiol Lett, 2006. 258(1): p. 18 - 24. [48] Ahmed, S., et al., Influence of parameters on the heterogeneous photocatalytic degradation of pesticides and phenolic contaminants in wastewater: a short review. Journal of Environmental Management, 2011. 92(3): p. 311 - 330. [49] Chong, M.N., et al., Recent developments in photocatalytic water treat ment technology: A review. Water Research, 2010. 44(10): p. 2997 - 3027. [50] Ahmed, Saber, et al. Influence of Parameters on the Heterogeneous Photocatalytic Degradation of Pesticides and Phenolic Contaminants in Wastewater: A Short Review. Journal of envir onmental management 92.3 (2011): 311 - 30. [51] Zhang, L., Y.M. He, P. Ye, Y. Wu, and T.H. Wu, Visible light photocatalytic activities of ZnFe2O4 loaded by Ag3VO4 heterojunction composites. Journal of Alloys and Compounds, 2013. 549: 105 - 113. [52] Boumaza, S ., A. Boudjemaa, A. Bouguelia, R. Bouarab, and M. Trari, Visible light induced hydrogen evolution on new hetero - system ZnFe2O4/SrTiO3. Applied Energy, 2010. 87: 2230 - 2236. [53] Dom, R., R. Subasri, N.Y. Hebalkar, A.S. Chary, and P.H. Borse, Synthesis of a 20 hydrogen producing nanocrystalline ZnFe2O4 visible light photocatalyst using a rapid microwave irradiation method. Rsc Advances, 2012. 2: 12782 - 12791. [54] Chae, S.Y., J.B. Yadav, and O.S. Joo, Electrospray - deposited nickel ferrite thin film electrode for hydrogen production in PV - assisted water electrolysis system. International Journal of Energy Research, 2012. 36: 1044 - 1050. [55] Xu, X.X., A.K. Azad, and J.T.S. Irvine, Photocatalytic H2 generation from spinels ZnFe2O4, ZnFeGaO4 and ZnGa2O4. Catalysis Tod ay, 2013. 199: 22 - 26. [56] Zhang, X.H., M. Yang, J.G. Zhao, and L.J. Guo, Photocatalytic hydrogen evolution with simultaneous degradation of organics over Cu0.2In0.2Zn1.6S2 solid solution. International Journal of Hydrogen Energy, 2013. 38: 15985 - 15991. [ 57] Kambe, S., Fujii, M., Kawai, T., Kawai, S., and Nakahara, F. Photocatalytic hydrogen - production with Cd(S, Se) solid - solution particles - determining factors for the highly efficient photocatalyst. Chemical Physics Letters, 1984. 109:105 - 109. 21 CHAPTER TWO Analysis of energy costs for catalytic ozone membrane filtration The work in Chapter TWO has been published: Wang, X., Davies, S. H., Masten, S. J.. Analysis of energy costs for catalytic ozone membrane filtration. Separation and Purif ication Technology 186 (2017): 182 - 187. 22 Abstract Membrane fouling can be reduced through shear stress generated by cross - flow at the membrane surface. Previous work has shown that presence of ozone can reduce membrane f ouling. In this work, the effect of ozonation and cross - flow on both membrane fouling on ceramic membranes and the energy cost for the process was studied. The effect of ozone dosage on membrane fouling was studied in both cross - flow and dead - end configura tions. The performance of a manganese oxide coated membrane was compared with that of uncoated titanium oxide membrane. Membrane fouling decreased with increasing ozone dosages in the manganese oxide coated catalytic membrane, although increasing the dosag e beyond 10 most energy - efficient mode of operation was found to be dead - end filtration using a 23 2.1 Introduction Ultrafiltration (UF) membrane fil tration can effectively remove suspended particles, colloids, bacteria, viruses and significant fraction of natural organic matter (NOM) from feed water [1] . There are a number of advantages of UF membrane treatment over conventional water treatment techno logies, including reduced sludge production [2] , smaller footprint [3] , and the physical exclusion of microorganisms from the product water. However, membrane fouling continues to be one of the main factors that limit the application and increase the opera tional costs of membrane technologies [4] . Fouling results in a decrease in the production rate of purified water and a concomitant increase in the requisite transmembrane pressure [5] . In water treatment, NOM is the predominant cause of UF membrane foul ing [6 - 9] . The use of ozone in combination with membrane filtration has been reported to reduce membrane fouling [10 - 1 4] . Ozone, a powerful oxidant, is reactive with NOM [15] . Pre - o zonation can also reduce formation of Trihalomethanes (THMs) during subsequ ent chlorination [16] . Ceramic membranes are chemically stable, have longer lifetimes than polymeric membranes, and they can be used in conjunction with ozonation without degradation of the membrane [17, 18] . Moreover, the surface properties of ceramic m embranes may be altered to improve membrane performance. Metal oxides, such as titania [10,11] , iron oxide [11] , and manganese oxide [11, 19, 20] , can act as catalysts for the degradation of ozone and the formation of hydroxyl radicals. These catalytic rea ctions are assumed to occur at the metal oxide surface. Kim et al. [12] studied the performance of nanoparticle - enhanced 24 membranes and concluded fouling was dependent upon the physicochemical aspects of nanoparticles, such as particle size, hydrophilicity, and surface charge. Byun [11] reported that when used with ozone the performance of the manganese oxide coated membrane was superior to that of Ti and Fe oxide membranes in terms of fouling mitigation and the reduction of TOC in the permeate. In addition, the coating of the membranes with manganese oxide nanoparticles significantly reduced the concentrations of THM and Haloacetic acids (HAA) precursors found in the permeate [19] . Szymanska et al [13] and Zouboulis et al. [21] showed that with hybrid ozonat ion ceramic membrane microfiltration, mitigation of membrane fouling was well controlled. Wei et al. [22, 23] demonstrated that pre - ozonation mitigated fouling on a UF membrane caused by soluble extracellular organic matters released from Microcystis aerug inosa. They f ound that ozone reacted with the cake fouling and gel layers and prevented membrane pore blocking. The objective of this study was to analyze the energy costs for membrane filtration, in the presence of ozone, using either a manganese oxide coated ceramic membrane or titania oxide coated membrane (virgin membrane). Studies were conducted to determine the optimum ozone dosage and hydrodynamic conditions for this hybrid membrane system. The energy costs for both ozone generation and pumping w ere evaluated to determine the operating conditions for which the energy costs were minimized. 25 2.2 Experimental methods 2.2.1 Feed water The source water was obtained from Lake Lansing (Haslett, Michigan), a borderline eutrophic lake. The water was s tored in the dark at 4 °C until use. Water samples were pre - filtered through a 0.5 - µm ceramic cartridge micro - filter (Doulton USA, Southfield, MI) to remove larger particles. After filtration, the average TOC concentration in the test water was 10 ± 1 mg C /L. Before commencing each experiment, the temperature of the feed water was adjusted to room temperature (20 ± 3°C). 2.2.2 Membrane module The nominal molecular weight cut - off of the virgin membrane (Inside CeRAM, TAMI North America, Saint - Laurent, Que bec, Canada) was 5 KDa. The seven - channel membrane had a total filtering surface area of 131.9 cm 2 , an active length of 25 cm, and an external diameter of 10 mm. The grain size within the support layer and the filtration layer of the virgin membrane varied between 132 nm and 296 nm and between 1.05 and 6.64 nm [11] , supported with titanium oxide filtration layer. The clean water permeability of the coated membrane was 80 ± 2 L/m 2 h. The virgin membrane was coated with manganese oxide according to the proce dure described by Corneal et al. [20] . In this study, the performance of a membrane coated twenty times manganese oxide nanoparticles was examined, as higher permeate fluxes were obtained with membrane coated twenty times coated than with those coated thir ty or forty times. Byun et al. [11] found the Mn oxide coating to be crystalline Mn 2 O 3 , and that 26 for the membrane coated 20 times the thickness of coating varied between 14 and 54 nm with 20 times coatings. With the manganese oxide coated membranes, the lo wer operating pressure was used because the permeability of the manganese oxide coated membrane was higher than that of the uncoated membrane, as sintering at 500 °C leads to coarsening of the grains within the filtration layer [19] . 2. 2.3 Ozone contacto r A schematic for the ozone contactor is shown in Figure 2. 1 . The water in the contactor was maintained at a constant level using a conductivity water level sensor connected to a programmable logic relay (SG2 PLR, B&B Electronics, IL), which opened or close d a solenoid valve (6013, Bürkert, Germany) to regulate water flow from the sample reservoir to the contactor. Ozone was generated from pure, dry oxygen (99.999%) using a corona discharge ozone generator (Absolute Ozone®, Absolute System Inc., Edmonton, A B, CA). An ozone monitor (Model 450H, Teledyne Technologies Inc., San Diego, CA) was used to measure ozone concentration. The flow rate of ozone gas was controlled at 10 mL/min by a rotameter (Cole - Parmer Inc., Vernon Hills, IL) installed between the ozone monitor and the membrane module. The ozone transfer efficiency in the contactor was determined before each experiment, by measuring the difference in the influent and effluent gaseous ozone concentrations and multiplying that difference by the gas flow r ate. 2. 2.4 Hybrid ozonation filtration setup The schematic of the ozone injection system is shown in Figure 2. 1. Nitrogen gas (99%, 27 Airgas) was used for pressurizing feed tanks. The membrane module housing (TAMI North America, St. Laurent, Québec, Cana da) was made of stainless steel. Both dead - end filtration and cross - flow filtration were performed at volumetric flow rate through the membrane of 18±1.5 mL/min. This corresponds to a flux of 81.2±2.8 L/m 2 h, which is typical of that used in full - scale wat er treatment. The flux was maintained using a peristaltic pump (Eldex®, Eldex Laboratories, Inc., Napa, CA). A recirculation pump (Materflex®, Cole Parmer Inc., Vernon Hills, IL) used when operating the system in cross - flow mode to obtain the constant cro ss - flow velocity of 0.5±0.1 m/s. The transmembrane pressure (TMP) was recorded by a multifunctional sensor (L Series, Alicat Scientific, Tuscon, AZ) every 60 s. Temperature and atmosphere pressure were also monitored every 60 s by the multifunctional senso r. The permeate flux was measured using an electronic balance (Adventure Pro Analytical Balance, Ohaus Corp., Pine Brook, NJ) every 60 s. The transmembrane pressure (TMP, bar) and permeate flux (J, L/m 2 h) were recorded by Flow Vision SC (Alicat Scientific, Tucson, AZ) data acquisition software. Experiments were conducted at ozone dosage injection rates of 5, 10, 15, and 20 µg/s. Control experiments were conducted to measure membrane flux without gas injection. The range of ozone dosages was selected based on prior work, which indicated that ozone dosages greater than 20 µg/s resulted in little enhancement in the permeate flux and reduction in membrane fouling [14, 21, 24] . Experiments were halted whe n the permeability of the membrane decreased to approximately 40% of the initial permeability. After each experiment, the fouled membranes were cleaned using deionized water and 28 20 µg/s gaseous ozone for 2 - 4 h until the initial clean water flux was restor ed. Before each filtration experiment, the initial flux was measured to ensure that the permeability of the membrane was within (98±2) % of its initial value [11] . After cleaning the initial TMP for the system varied between 1.2 and 1.5 bar. Fig ure 2. 1 Schematic of the ozonation - membrane filtration system 2. 2.5 Data Analysis 2. 2.5.1 Permeability and resistance fouling The permeability, L p , was calculated as shown in Equation 1: (1) where, L p is given in units of L/(m 2 ·h·bar). The greater the permeability, the higher the efficiency of the membrane. The total resistance across the membrane was calcu lated from the TMP and flux data 29 (2) (3) water at 22 2 (0.89 × 10 3 2 ). The total resistance, R t , is the sum of the intrinsic resistance of the membrane, R m , and the resistance due to the material deposited on the membrane surface, R c . R t was calculated from the final TMP and flux. 2. 2.5 .2 Calculation of energy consumption 2. 2.5.2.1 Energy consumed to pressurize f eed water The power necessary to pressurize the feed water was determined by Equation (4) [25] : (4) where, q (m 3 /s) is the average flow rate of the membrane module during the operational 2 /4 × v, v is the velocity of the water in filtration system; (kg/m 3 ) is the density of feed water at room temperature; g ( 9.81 m/s 2 ) is gravitational acceleration of the earth; h is the water head (m), TMP (bar) is the transmembrane pressure across membrane module, and is calculated using the equation TMP = is the efficiency of the pressurizing pump which was assumed t o be 75%. The power (P h ) is reported in units of kW. 2. 2.5.2.2 Energy consumed by r ecirculation pump The power required to recirculate water through the system was determined by Equation (5) [25] : 30 = (5) in which P r is in kW; the total water head ( h r = P/ g) of the recirculation syst em was is the efficiency of the pressurizat ion pump, which was determined to be 70% based on the pump curve. The pressure drop in the system was calculated using Darcy - Weisbach equation: (6) (7) (8) cross - flow velocity (0.5±0.1 m/s during experiment), f is the Darcy friction factor, d (m) is the hydraulic diameter of the pipe/membrane channels, (kg/m 3 ) is the density of feed water under room temperature, L (m) is the length of different tubing section s of tubing, valves, and membrane channel using an online pressure drop calculator ( Pressure Drop Online - Calculator. Software - Factory Norbert Schmitz. http://www.pressure - drop.com/Online - Calcul ator/ ) . 2. 2.5.2.3 Energy consumption during o zone generation in backwash and operation cycles Based on product data sheet ozone generator ( Primozone Production AB. http://www.water - proved.de/en/Downloads/Primozone - Ozone - Generator -- System - 31 controller.pdf ), Ozone Generators parameters (Lenntech BV. http://www.lenntech.com/otozone.htm ) and published liter ature [26] , approximately 10 kWh energy is consumed to produce 1 kg ozone from. The ozone transfer efficiency in the system was determined to be 95% based on the analysis of dissolved ozone concentrations and the ozone concentration in the gas phase befor e and after the contactor. The operational period was determined by the filter run time between backwashing. The filters were backwashed when transmembrane pressures reached 2.5 ±0.1 bar (>10 O 3 /s) or 3.8 ± 0.2 bar (<5 3 /s) . Gaseous ozone was used for membrane backwash cleaning. The operational parameters for the backwashing are shown in Table 2. 1. The energy needed for backwashing, including that required for pressurization during backwash process and that for generating ozone. The e nergy consum ed during operation was W = P×T, in which W is in kWh, P (kW) is the power required to pressurize the feed water or recirculate water, and T is the operation time. Table 2. 1 Backwash process operation parameters Criteria Parameters Duration 2 - 4 hours, t ake 3 hours as average Fouled membrane TMP 2.2 - 2.6 bar for high ozone injection rate (>10 g/s); 3.6 - 4.0 bar for no ozone injection or 5 g/s ozonation Cleaned membrane TMP 1.2 - 1.5 bar Average membrane TMP during backwash process Approximately 1.9 bar for high ozone injection rate (>10 g/s); 2.58 bar for no ozone injection or 5 g/s ozonation Feed water flow rate ~18 mL/min Ozone injection rate 20 g/s 32 2.3 Result and discussion 2.3.1 Comparison of permeate flux and resistance fouling Fig ure 2 .2 displays the eff ect of ozone dosage on permeate flux during catalytic membrane filtration when a manganese oxide coated membrane was employed using dead - end operation mode. As shown in this figure 2.2 , irrespective of ozone dosage, the permeability initially declined rapi dly during first 30 minutes. In this stage, the fouling cake begins to form at membrane surface. This stage was followed by one in which the permeability decreased at a much slower pace. At an ozone dosing rate of 20 µg/s the permeability was eventually nearly constant. When the permeability reached steady state condition, the rate of formation of the fouling cake due to deposition of natural organic matter on the surface of the membrane is presumed to be similar to the rate of reaction of the accumulated foulants with ozone and OH radicals formed as a result of the catalytic degradation of ozone by manganese oxide [19] . Increasing the ozone dosing rate improved the performance of the system as catalytic degradation of the filter cake became increasingly f avored over formation of the filter cake. 33 Fig ure 2. 2 Effect of ozone dosage on permeability using manganese oxide coated catalytic membrane filtration dead - end operation mode (Conditions: permeability in L/m²·h·bar, Feed TOC=12.1 mg/L, temperature=2 2±2 ) Fig ure 2. 3 shows the effect of ozone dosage on permeate flux during cross flow operation. While the observed permeability trends for cross - flow and dead - end operational modes are similar, the permeability fluctuated more during cross - flow operation t han in dead - end operation. As was the case with dead - end filtration, increasing the ozone dosage resulted in less fouling. In the absence of ozone injection and with ozone injection at dosages of 5 - 10 µg/s, the permeability values after 240 min were greate r in cross - flow mode than that in dead - end mode. Under these conditions shear created by the recirculation of the retentate appears to disrupt fouling cake formation. However, at higher (15 and 20 µg/s) ozone dosages the permeabilities observed during ope ration were similar in cross - flow and dead - end filtration. This suggests that the reaction of ozone and foulants affects the foulants in such a way that the fouling cake less easily dislodged from the membrane surface. 0 10 20 30 40 50 60 70 80 0 50 100 150 200 250 Time (min) no ozone 5 µg/s 10 µg/s 15 µg/s 20 µg/s 34 Fig ure 2. 3 Effect of ozone dosage on permeate flux during manganese oxide coated catalytic membrane filtration crossflow operation mode (Conditions: permeability in L/m²·h·bar, Feed TOC=12.1 mg/L, cross flow velocity = 0.5±0.1 m/s temperature=22±2 ) The permeability of the virgin titanium oxide coated catalytic membrane during operation of membrane filtration in both dead end and cross flow filtration mode is shown in Figure 2. 4. The trends were similar to that observed with the manganese oxide co ated membrane, as the extent of fouling decreased with increasing ozone dosage. In the absence of ozone, the extent of fouling was greater on the virgin titania membrane than on the coated membrane. This is likely the result of the electrostatic characteri stic of the membrane surface. The point of zero charge of manganese oxide (pH zpc = 2.8 - 4.5 [27] ) is lower than that of titania oxide (pH zpc = 4.1 - 6.2 [10] ). At the pH of the treated water is ~8, the hydroxyl groups on the Mn oxide surface would be fully deprotonated. The surface hydroxyl groups on the TiO 2 surface are not completely deprotonated; thus the manganese oxide surface is likely to have higher charge density and the repulsive forces in the Mn oxide membrane system between the negatively charged components of the 0 10 20 30 40 50 60 70 80 0 50 100 150 200 250 Time (min) no ozone 5 µg/s 10 µg/s 15 µg/s 20 µg/s 35 NOM and the oxide are probably higher than that observed with the titania oxide membrane. With higher repulsive electrostatic forces, the deposition of organics would be reduced, and with that, the degree of membrane fouling would also be diminished. (a) Figure 2.4 Effect of ozone dosage on permeability during titania oxide coated catalytic membrane filtration (a) dead - end operation mode and (b) crossflow operation mode (Conditions: permeability in L/m²·h·bar, Feed TOC=12.1 mg/L, cros s flow velocity = 0.5±0.1 m/s temperature=22±2 ) 0 10 20 30 40 50 60 70 80 0 50 100 150 200 250 Time (min) no ozone 5 µg/s 10 µg/s 15 µg/s 20 µg/s 36 (b) The effect of ozone dosage and mode of operation on fouling resistance is shown in figure 2.5. The fouling resistance was determined at 80 min (steady state) for manganese oxide coated catalytic membrane filtr ation and at 120 min for titania oxide coated catalytic membrane filtration, since as mentioned early, it took longer for steady state conditions to be achieved with the titania oxide coated catalytic membrane filtration. As figure 2.5 shows, the fouling resistance decreased with increasing ozone dosage for both operational modes and membrane types. Recirculation was not as effective as ozonation at reducing fouling resistance. Under identical operational conditions, the resistance of the manganese oxide c oated membrane was less than that of the titania oxide coated membrane. 0 10 20 30 40 50 60 70 80 0 50 100 150 200 250 Time (min) no ozone g/s g/s g/s g/s 37 Fig ure 2. 5 Effect of ozone dosage on fouling resistance in manganese oxide coated and titania oxide coated catalytic membrane filtration system 2.3.2 Energy and economic analysis of membrane systems for design applications As overall resistance of the manganese oxide membrane, and the operational cost would be less with this membrane under same ozone dosage and operation mode compared to the virgin titania membrane. Using experime ntal results for the coated membrane the energy necessary to filter 1,000 liters of water was calculated (see Section 2.5 and Table 2 .2 ). The energy costs include energy consumption for pumping water across the membrane, ozone generation, and recirculating through the system. It also includes minor losses through the system. As shown in f igure 2. 6 for the manganese oxide coated membrane the optimal ozone dosing rate is 10 µg/s for dead end operation and for cross - flow mode it is 5 - 10 µg/s. At these dosing rates the energy consumed is similar for dead - end and crossflow systems. Dead - end operation is simpler and does not require a recirculation pump thus mode of operation is more desirable. Thus, under the conditions 0.00E+00 2.00E+12 4.00E+12 6.00E+12 8.00E+12 1.00E+13 1.20E+13 1.40E+13 1.60E+13 Dead-end Cross-flow Dead-end Cross-flow Manganese oxide coated membrane Titania oxide coated membrane Resistance (m - 1) Membrane type and operation mode g/s g/s g/s g/s 38 studied for Lake Lansing water, dead - en d catalytic membrane filtration using the man be the most efficient operational mode. Fig ure 2. 6 Energy consumed during the treatment of in 1,000 liters of water (including backwash cycles) Calculations for energy consu mption for the hybrid - ozonation filtration system are given in section 2.5 data analysis, and results are shown in Table 2 .2 . The energy needed to filter 1,000 liters of water was estimated assuming there was no deterioration in performance over an extend ed period of time. Based on these calculations the minimum amount of energy required to filter this volume of Lake Lansing water is 0.97kWh. 0.00 0.20 0.40 0.60 0.80 1.00 1.20 1.40 1.60 1.80 2.00 0 5 10 15 20 Energy consumed in treating 1000 liters of water (kWh) Ozone injection ( g/s) Dead-end Cross-flow 39 Table 2. 2 Energy consumption for Mn oxide coated membrane filtration with different ozone dosing rates and op eration modes with one membrane module Operation mode Ozone dosage Operation period during 1 cycle (min) Water filtered during 1 cycle (L) Energy consumed during operation Approximate energy needed for 1 cycle backwash cycle (kWh) Time required to treat 1,000 L of water (hours) Numbers of operation periods Total energy consumed (kWh) Energy consumed in pressurizati on (kWh) Energy consumed in ozonation (kWh) Energy consumed in recirculation (kWh) Total energy consumed during operation (kWh) Dead - end 0 93.0 1.563 1.453 × 10 - 4 0 0 1.453× 10 - 4 2.583 × 10 - 3 1023.7 640 1.75 5 98.0 1.839 1.728 × 10 - 4 3.158 × 10 - 4 4.886× 10 - 4 915.6 544 1.67 10 255.0 4.638 3.530 × 10 - 4 1.629 × 10 - 3 1.982× 10 - 3 2.502 × 10 - 3 927.2 216 0.97 15 233.0 4.659 4.05 4 × 10 - 4 2.236 × 10 - 3 2.641× 10 - 3 844.2 215 1.10 20 241.0 4.721 3.592 × 10 - 4 3.082 × 10 - 3 3.441× 10 - 3 861.3 212 1.26 Cross - flow 0 82.0 1.535 1.594 × 10 - 4 0 7.291 × 10 - 5 2.323× 10 - 4 2.583 × 10 - 3 923.0 652 1.83 5 175.0 3.541 3.585 × 10 - 4 5.526 × 10 - 4 1.556 × 10 - 4 1.067× 10 - 3 837.8 282 1.03 10 222.0 4.294 3.679 × 10 - 4 1.402 × 10 - 3 1.974 × 10 - 4 1.967× 10 - 3 2.502 × 10 - 3 873.3 233 1.04 15 250.0 4.238 2.952 × 10 - 4 2.368 × 10 - 3 2.223 × 10 - 4 2.886× 10 - 3 995.0 236 1.27 20 240.0 4.887 4.066 × 10 - 4 3.0 32 × 10 - 3 2.134 × 10 - 4 3.652× 10 - 3 828.7 205 1.26 40 2.4 Conclusion The energy costs for operation of catalytic membrane filtration were calculated. Membrane fouling was effectively controlled using catalytic membrane filtration. For the manganese oxide co ated membrane, membrane fouling was effectively controlled at ozone dosages of 10 µg/s or greater. With the virgin membrane, an ozone dosage of at least 15 µg/s was necessary to control membrane fouling. The overall resistance of the manganese oxide membr ane was less with manganese oxide coated membrane under same ozone dosage and operation mode compared to that observed for the virgin titania membrane. Using the flux data obtained, it was determined that for the water studied the most energy - efficient op - end operation, with the manganese oxide coated membrane. Future work will include a life - cycle assessment (LCA) to evaluate the environmental impacts of membrane filtration, and compare this technology to other water treatment processes, to determine whether this technology is more energy efficient than conventional processes for treating water. Acknowledgements This study was supported by the Department of Civil and Environmental Engineering, Michigan State Uni versity, and American Water Works Association, Michigan Section. The authors are grateful for the comments from the reviewers of this manuscript. 41 APPENDIX 42 APPENDIX Supplemental Section: Samples calculat ion of energy consumption Using 5 Average TMP during 175 min operation period was 2.723 bar, the water head was: h = = 27.78 m The power needed at the average TMP was: P h = = 0.1229 watt = 1.229 × 10 - 4 kW The work/energy provided during the operation period was: P h T = (1.229 × 10 - 4 kW) × (175 min) (1 h/60 min) = 3.585 × 10 - 4 kWh 0.00389 bar, obtained from pressur e drop online calculator [26] ; so the power required during this cycle was: = 0.02182 watt = 2.182 × 10 - 4 kW membrane length and cut - off area), so the power demand for this part was: 43 = 0.03153 watt = 3.153 × 10 - 4 kW The power demand for this pressure drop was: = 5.335 × 10 - 4 kW Energy demand during operation period for this pressure drop was: P r T = 5.335 × 10 - 4 kW × (175/60) = 1.556 × 10 - 3 kW Energy demand for ozone generation was: E O3 = (5 3 /s) (175 min) (60s/1 min) (10 kwh/1kg O 3 ) (1kg/10 - 9 10 - 4 kWh Energy cost in one operation period was obtained: Q= P h T+ P r T+ E O3 = 1.067 × 10 - 4 kwh During this process, 3.54 L feed water had been treated. Gaseous ozone was used for membrane backwash cleaning, and the operation parameters are shown in table 2. 1. Energy needed for 1 cycle backwash inclu ded energy for 20 µg/s ozonation and energy for pressurization during backwash process, there was difference in energy consumption for assorted ozone injection rate as the average TMP varied during backwash cycles: P backwash 1(>10 ug/s) = = 0.076 watt = 7.6 × 10 - 5 kW 44 P backwash 2(<5 ug/s) = = 0.1032 watt = 1.032 × 10 - 4 kW E backwash = = 2.274 × 10 - 3 kWh Energy cost during backwash: Q backwash 1(>10 ug/s) = P backwash 1(>1 0 ug/s) T backwash + E backwash = 2.583 × 10 - 3 kWh , or Q backwash 1(<5 ug/s) = P backwash 1(<5 ug/s) T backwash + E backwash = 2.502 × 10 - 3 kWh It would take 282 cycles to treat 1,000 L of water in one membrane module, so the total energy for treating 1000 liters o f water using 5 operation mode would be 1.03 kW . 45 REFERENCES 46 REFERENCES [1] Peter - Varbanets, Maryna, et al. Mechanisms of Membrane Fouling during Ultra - Low Pressure Ultrafiltration. Journal of Membrane Scienc e 377.1 - 2 (2011): 42 - 53. DOI: 10.1016/j.memsci.2011.03.029. [2] Cleveland, Christopher T. Big Advantages in Membrane Filtration. American Water Works Association. Journal 91.6 (1999): 10. [3] Reardon, R. D. Membrane Treatment of Secondary Effluent for Sub sequent use. Water Intelligence Online 6 (2007). DOI: 10.2166/9781843397595 [4] Kenari, Seyedeh Laleh Dashtban, and Benoit Barbeau. Understanding Ultrafiltration Fouling of Ceramic and Polymeric Membranes Caused by Oxidized Iron and Manganese in Water Tre atment. Journal of Membrane Science 516 (2016): 1 - 12. DOI: http://dx.doi.org.proxy2.cl.msu.edu/10.1016/j.memsci.2016.06.003 [5] Huang, Haiou. Microfiltration Membrane Fouling in Water Treatment: Impact of Chemical Attachments. Dissertation. 2007. [6] Van Geluwe, Steven, Leen Braeken, and der Bruggen Van. Ozone Oxidation for the Alleviation of Membrane Fouling by Natural Organic Matter: A Review. Water research 45.12 (2011): 3551 - 70. DOI: 10.1016/j.watres.2011.04.016 . [7] Kim, Hyun - Chul, and Brian A. Dempsey. Membrane fouling due to alginate, SMP, EfOM, humic acid, and NOM. Journal of membrane science 428 (2013): 190 - 197. DOI: 10.1016/j.memsci.2012.11.004. [8] Cho, Jaeweon, G ary, Amy, and Pellegrino, John. Membrane filtration of natural organic matter: comparison of flux decline, NOM rejection, and foulants during filtration with three UF membranes. Desalination 127.3 (2000): 283 - 298. DOI:10.1016/S0011 - 9164(00)00017 - 5 [9] How e, Kerry J., and Clark, Mark M.. Fouling of microfiltration and ultrafiltration membranes by natural waters. Environmental science & technology 36.16 (2002): 3571 - 3576. DOI: 10.1021/es025587r. [10] Kim, Jeonghwan, et al. Interactions of aqueous NOM with nanoscale TiO 2 : Implications for ceramic membrane filtration - ozonation hybrid process. Environmental science & technology 43.14 (2009): 5488 - 5494. DOI: 10.1021/es900342q. [11] Byun, S., et al. Mn oxide coated catalytic membranes for a hybrid ozonation mem brane filtration: comparison of Ti, Fe and Mn oxide coated membranes for water quality. Water research 45.1 (2011): 163 - 170. DOI: 10.1016/j.watres.2010.08.031. [12] Kim, Jeonghwan, and Van der Bruggen, Bart. The use of nanoparticles in polymeric 47 and ceram ic membrane structures: review of manufacturing procedures and performance improvement for water treatment. Environmental Pollution 158.7 (2010): 2335 - 2349. DOI: 10.1016/j.envpol.2010.03.024. [13] Szymanska, K., Zouboulis, A. I., and Zamboulis, D.. Hybri d ozonation microfiltration system for the treatment of surface water using ceramic membrane. Journal of Membrane Science 468 (2014): 163 - 171. DOI: 10.1016/j.memsci.2014.05.056. [14] Karnik, Bhavana S., et al. Effects of ozonation on the permeate flux of nanocrystalline ceramic membranes. Water Research 39.4 (2005): 728 - 734. DOI: 10.1016/j.watres.2004.11.017. [15] Von Gunten, Urs. Ozonation of drinking water: Part I. Oxidation kinetics and product formation. Water research 37.7 (2003): 1443 - 1467. DOI: 10. 1016/S0043 - 1354(02)00457 - 8. [16] Mori, Y., et al. Ozone - microfiltration system. Desalination 117.1 (1998): 211 - 218. DOI:10.1016/S0011 - 9164(98)00098 - 8. [17] Davies, S. H., et al. Fabrication of catalytic ceramic membranes for water filtration. Water Scien ce and Technology: Water Supply 10.1 (2010): 81 - 86. DOI: 10.2166/ws.2010.789. [18] Karnik, Bhavana S., et al. Fabrication of catalytic membranes for the treatment of drinking water using combined ozonation and ultrafiltration. Environmental science & tech nology 39.19 (2005): 7656 - 7661. DOI: 10.1021/es0503938. [19] Corneal, Lindsay M., et al. Mn oxide coated catalytic membranes for hybrid ozonation - membrane filtration: Membrane microstructural characterization. Journal of membrane science 369.1 (2011): 182 - 187. DOI: 10.1016/j.memsci.2010.11.071. [20] Corneal, Lindsay M., et al. AFM, SEM and EDS characterization of manganese oxide coated ceramic water filtration membranes. Journal of Membrane Science 360.1 (2010): 292 - 302. DOI: 10.1016/j.memsci.2010.05.026. [21] Zouboulis, A., Zamboulis, D., and Szymanska, K.. Hybrid membrane processes for the treatment of surface water and mitigation of membrane fouling. Separation and Purification Technology 137 (2014): 43 - 52. DOI: 10.1016/j.seppur.2014.09.023. [22] Wei , Dequan, et al. Effect of pre - ozonation on mitigation of ceramic UF membrane fouling caused by algal extracellular organic matters. Chemical Engineering Journal 294 (2016): 157 - 166. DOI: 10.1016/j.cej.2016.02.110 [23] Wei, Dequan, et al. Effect of in - sit u ozonation on ceramic UF membrane fouling mitigation in algal - rich water treatment. Journal of Membrane Science 498 (2016): 116 - 124. DOI: 10.1016/j.memsci.2015.09.063. . 48 [24] Kim, Jeonghwan, et al. Effect of ozone dosage and hydrodynamic conditions on the permeate flux in a hybrid ozonation ceramic ultrafiltration system treating natural waters. Journal of Membrane Science 311.1 (2008): 165 - 172. DOI: 10.1016/j.memsci.2007.12.010. [25] Chin, David A. Fluid Mechanics for Engineers. Hoboken, NJ: Pearson High er Education, 2015. 614 - 15. [26] Allam, Rodney J. Improved oxygen production technologies. Energy Procedia 1.1 (2009): 461 - 470. DOI: 10.1016/j.egypro.2009.01.062. [27] Tan, Wen - feng, et al. Determination of the point - of - zero charge of manganese oxides wit h different methods including an improved salt titration method. Soil Science 173.4 (2008): 277 - 286. DOI: 10.1097/SS.0b013e31816d1f12. 49 CHAPTER THREE Energy consumption and environmental impact of ozonation catalytic membrane filtration system and compa rison with hollow fiber membrane for water treatment The work in Chapter Three has been accepted: Wang, X., Anctil, A., Masten, S. J.. Energy consumption and environmental impact analysis of ozonation catalytic membrane filtration system for water treat ment. Environmental Engineering Science 50 Abstract The productivity of water filtration can be improved with ceramic membranes as compared to polymeric membranes because of their higher stability, longer lifetime, and higher pe rmeability. The use of ozonation in combination with manganese oxide coated ceramic membrane filtration has been shown to reduce membrane fouling. In this study, life - cycle - assessment was conducted to analyze the energy consumption and environmental impact s of the catalytic ceramic membrane system, which were then compared to that for hollow fiber membrane filtration. For both systems, the analysis was accomplished for a treatment plant with a capacity of 9 MGD. Energy consumption and environmental impacts were determined for membrane pressurization, backwashing, chlorine injection for hollow fiber membrane treatment disinfection or ozone injection for ceramic membrane fouling control and disinfection processes, and membrane modules manufacturing. The result s showed that ceramic membrane combined with ozonation could save energy costs for pressurization and backwashing, but ozone generation consumed more energy than chlorine disinfection. The total energy consumption for catalytic ceramic membrane filtration with ozonation was 7% greater than for hollow fiber membrane filtration. However, catalytic membrane filtration had a slightly lower normalized environmental impact than polymeric membrane filtration. Innovative technologies to reduce energy consumption f or production of ozone, or the use of less energy intensive oxidants for fouling mitigation are expected to make the ceramic membrane filtration processes more energy - efficient and environmental - friendly. 51 3. 1 Introduction Membrane technology has become more economical, and the use of membrane filtration has increased rapidly [1] . There are a number of advantages in membrane process over conventional drinking water treatment technologies, including less sludge production, smaller footprint, and the physic al exclusion of microorganisms from the product water [2] . However, during membrane filtration process, the deposition of materials, such as rejected colloids, chemicals, microorganism on the membrane surface r esults in a decrease in the production of puri fied water or a concomitant increase in the requisite pressure [3] . Compared to polymeric membranes, ceramic membranes are chemically stable, have longer lifetimes [4, 5] , low operation cost, greater recyclability, and higher permeability [6] . Moreover, the surfaces of ceramic membranes can be coated with metal oxides, which serve as catalysts and thereby improve membrane performance [2] . The chemical stability of ceramic membranes allows its application in conjunction with ozonation [5] . Metal oxides, su ch as iron oxide, manganese oxide, titania, alumina, and zirconia, act as catalysts for degradation of ozone and the formation of hydroxyl radical radicals [2] . These catalytic reactions have been shown to occur when ozone molecules are adsorbed on the met al oxide surface [7] . However, the capital costs of ceramic membranes are higher than polymeric ultrafiltration/microfiltration membranes, and the materials are brittle [6] . The use of ozone in combination with membrane filtration has been reported to re duce membrane fouling [7 - 9] . Byun [10] evaluated the performance of the coated and uncoated 52 membranes, showing that manganese oxide coated membrane was superior to that of the other membranes tested. Song et al. [11] studied the effect of ozonation on prot ein - based ceramic membrane fouling and confirmed that the bovine serum albumin membrane fouling mechanism changed from cake - intermediate to cake - standard with ozonation at high ozone dosage. Song et al. also found [12] that pre - ozonation and in - situ ozonat ion were both effective in controlling membrane fouling, and in - situ ozonation with ceramic membrane filtration increased the generation of hydroxyl radical, thus improving the oxidation of the accumulated foulants on the membrane surface and effectively r educing both reversible and irreversible membrane fouling. Yu et al. [13] evaluated the performance of coagulation followed by ozonation with a low dosage for pretreatment before ultrafiltration. They found that the addition of ozone into the UF membrane tank reduced membrane fouling, increased the suspend solids concentration in the membrane tank, and reduced the concentration of bacteria. Wang et al. [14] demonstrated that hybrid manganese oxide coated ceramic membrane with ozonation system, dead - end fi ltration with ozone injection was more energy - efficient than crossflow operation mode or either mode without ozone injection. Membrane fouling was also shown to decrease with increasing ozone dosages [14] . Ozonation has the advantage over chlorination, i n that it is a more powerful disinfectant, and its use with membrane filtration has been shown to reduce the formation of trihalomethanes (THMs) formation in drinking water [15, 16 - 19] and achieve a greater level of microorganism inactivation [20] . 53 In thi s study, catalytic ceramic membrane filtration was compared to conventional polymeric membrane for water treatment. In a conventional polymeric membrane filtration plant, water passes through pretreatment, polymeric membrane filtration, disinfection, usua lly with chlorine as disinfectant, and then stored in tanks and distributed for use. Filtration service runs usually last 30 - 90 minutes (depend of feed water properties and flow rate), after which the membranes are cleaned by backwashing with water, and in some cases with water and air [17] . With hybrid catalytic ceramic membrane filtration involving ozonation, the source water flows through pretreatment, and ozone is injected at the membrane inlet. Ozone serves as disinfectant, and at the same time, reacts with organic matter in feed water so that fouling on membrane surface is controlled. The objective of this study was to determine whether ceramic membrane combined ozonation has advantages in saving energy and reducing environmental impacts in water tre atment application compared with conventional polymeric membrane filtration. Treatment specifications for hollow fiber (polymeric) membrane filtration were employed. Additionally, the identification of the most energy consumptive and least environmentally friendly aspects of the two processes can be used to guide future research efforts. Life cycle assessment (LCA) was used to assess the environmental impact of the two processes along with the energy consumption. 54 3. 2 Experimental 3. 2.1 Goal and scope de finition 3. 2.1.1 Goal The goals of this study were to determine the cumulative energy demand, as well as environmental impact, of the hybrid manganese oxide coated catalytic ceramic membrane system (CC) , and compare these results with that of a hollow fibe r membrane (HF) filtration system. Calculations were performed assuming that both systems would be used for drinking water treatment plant over their entire life cycle. The results are used to make recommendations regarding future improvements for ceramic membrane, and to interpret whether there is a benefit of using catalytic ceramic membrane combined with ozonation vs hollow fiber membrane filtration. 3 . 2.1.2 Assumptions The following assumptions were made for all scenarios: The source of the feed wat er is surface water. The quality of the source water is constant over the course of the system life. For the scenarios, treated water meets all regulatory requirements [21] . The lifetimes of the treatment plants are 20 years. Hollow fiber membranes are re placed every 4 years, and ceramic membranes are typically reported to have a 20 - year lifetime. For catalytic membrane, the calculations are based on laboratory data [14] and scaled to industrial - scale. Both membrane treatment processes would be installed a t the same location. 55 3. 2.1.3 System definition and boundaries As figure 3. 1 shows, the focused stages were the water treatment process involving membrane filtration as well as membrane module. The life cycle assessment was evaluated via four main categorie s: Feed water pressurization, disinfection/ozone injection, filter backwashing, and membrane manufacturing. Neither pre - nor post - treatment was considered. (a) Figure 3.1 System boundaries of the study in (a) catalytic ceramic ozonation membrane filtrat ion processes and (b) hollow fiber membrane processes with chlorine as disinfectants 56 (b) Table 3. 1 lists the main processes in the membrane filtration life cycle and materials/energy consumption. Maintenance and labor were not con sidered because of the uncertainty in costs and needs. As it was assumed that both treatment processes were installed at the same location in Michigan, US, transportation of materials was not considered. Feed pumps, feed piping, and ancillary equipment wer e not considered in the comparison. Neither were pretreatment, storage and distribution processes considered. The membrane filtration systems were assumed to more than 95% of recovery and minimal sludge production. As such, solids disposal was not conside red in the system evaluation [22] . 57 Table 3. 1 Main processes in the membrane filtration life cycle and materials/energy consumption Processes CC HF Pressurization Energy (electricity) Energy (electricity) Ozone injection (CC) / Disinfection (HF) Oxygen , energy (electricity) Chlorine Backwash Treated water, cleaning chemicals (sodium bisulfate, citric acid), energy (electricity) Treated water, cleaning chemicals (sodium bisulfate, citric acid), energy (electricity) Membrane Manufacture Energy (electric ity), ceramic, manganese chloride, ozone, potassium nitrate, Poly(diallyldimethylammoniu m chloride) (PDDA), sodium hydroxide Energy (electricity), polypropylene, polyvinylidene fluoride 3. 2.1.4 Functional Unit In this study, a daily clean water producti on rate of 9,000,000 gallons per day (9 MGD) was used as the functional unit. This functional unit (FU) corresponded to a reference flow to which all flows of the system are related. The membrane manufacturing energy and materials consumption were distribu ted over the life time on a daily basis. 58 3. 2.2 Data Collection For catalytic ceramic membrane filtration, feed water was assumed to be pressurized throughout the treatment processes, and energy was provided by electricity. Energy consumption for pressuriza tion was calculated from water flux, membrane filtration area, filtration rate, and transmembrane pressure, which were based on laboratory results and scaled up industrial data [14, 23] . Ozone injection rate was selected from former lab results [24] and sc aled up with membrane surface area ratio to a full scale 9 MGD system (4.55 x 10 - 5 lbs/gal), and ozone was assumed to be generated from pure oxygen using the corona - discharge method. The injection amount was scaled up by membrane area ratio [14] . The chemi cals used for backwashing were assumed to be the same for both ceramic and hollow fiber membrane; however the backwashing pressures were different and based on the literature [14, 22] . With ceramic membrane manufacturing, data obtained in our laboratory we re scaled up to industrial scale ceramic membrane filtration [23, 25] . For hollow fiber membrane filtration, energy consumption for pressurization was also calculated from water flux and transmembrane pressure, which were based on treatment plant capacity and the literature [22] . The chlorine injection rate and dosage were also based on published data [22] . Backwash chemicals and amounts were based on an existing hollow fiber water treatment plant [26] , and backwash pressure was obtained from literature [2 2] . Polypropylene (PP), polyvinylidene fluoride (PVDF), acrylonitrile butadiene styrene (ABS), polyurethane (PU) and energy are consumed during hollow fiber membrane manufacturing [27 - 30] and therefore, included in the analysis. 59 The life cycle assessment was analyzed by SimaPro (8.4.0.0, Pre Consultants, the Netherlands. 2018). SimaPro could help effectively apply life cycle assessment calculation, and measure the energy cost and environmental impacts across the entire processes. The electricity consumptio n for pumping, membranes manufacturing, ozone generation, backwash and other were determined using Michigan eGrid electricity (RFCM, 2010) and the USLCI inventory data base. Data for chemicals such as chlorine in disinfection process, oxygen for ozone gene ration, backwash chemicals, and materials for membrane manufacturing were obtained from Ecoinvent version 3.0. 3. 2.3 Impact Assessment Cumulative Energy Demand (CED) 1.09 was used to calculate the quantitative energy consumption, with the indicator of GJ and weighting factor of 1, by Ecoinvent version 3.0 in SimaPro 7 database. For assessing environmental impacts and ecotoxicity, TRACI 2.1 (US 2008) was used. TRACI is the midpoint oriented life cycle impact analysis ( LCIA ) methodology developed by the U.S . Environmental Protection Agency specifically for the US using input parameters consistent with US locations [31] . Environmental impacts, including ozone depletion, climate change, smog, acidification, eutrophication, carcinogenic, non - carcinogenic, respi ratory effects, ecotoxicity, fossil fuel depletion, were normalized and evaluated for both ceramic membrane and hollow fiber membrane filtration processes. Each characterized environmental impact was analyzed by its units, and the normalized impact was cal culated by characterized value times a given weight. Both normalized and characterized environmental impacts were compared between catalytic membrane and hollow fiber membrane filtration systems. To simplify and clarify 60 the study, the materials that repres ent less than 0.1% of the functional unit are not shown in the results. 3. 3 Results and discussion 3. 3.1 Catalytic ceramic membrane with ozonation energy and environmental impacts analysis For the catalytic ceramic membrane filtration system, pressuriz ation consumed the most of the energy, as Figure 3. 2 shows. On the other hand, ozone generation resulted in almost one - third of the energy consumed. This is not surprising as ozone generation by corona discharge, the common method of producing ozone, consu mes ~7 kW·hkWh electricity per kg of ozone generated [24, 32] . Figure 3. 2 Embodied energy consumption of total and different processes in catalytic ceramic membrane 61 Figure 3. 3 illustrates the environmental impacts of catalytic ceramic membrane filtrati on. The data are normalized to daily water production. Normalization results came from transforming by the selected reference values or information for each impact. Higher normalization results have greater impacts than lower ones [33] . They present that acidification should be paid attention in the desired processes as it was the highest normalized environmental impact. Carcinogenicity and ecotoxicity of pollutants generated, along with that due to smog, and global warming were also significant. Pressuriz ation contributed 49% of all the environmental impacts, and ozone injection resulted in 43% of the impacts. The environmental impacts of the ozone injection process were due to carcinogenicity and ecotoxicity of pollutants generated. Figure 3. 3 Environ mental impacts normalization results for catalytic ceramic membrane filtration processes 62 As discussed in the introduction, ozone can react with natural organic matter in water and that collected at the membrane surface, thus, effectively controlling membr ane fouling and reducing the costs of membrane pressurization [16, 34] . Wang et al. demonstrated that with hybrid ozone - membrane filtration, increasing the ozone injection rate resulted in a decrease in the transmembrane pressure, thereby reducing the ener gy required for pressurization and also extending the operation period and decreasing backwash times [14] . Although higher ozone injection rate has proved to effectively control membrane fouling, the total energy cost increased by 17.5% (10.85 GJ) when the ozone injection rate was doubled. As Figure 3. 4 shows, when ozone injection rate was doubled, the energy for pressurization decreased by 19.8% (6.4 GJ), which is consistent with previous work showing that fouling can be controlled using ozone injection [7 - 9,10] . However, as ozone generation consumes a significant amount of energy, the energy saved during pressurization and backwash was not sufficient to offset the increased energy consumption from ozone production. This underscores the importance of opera ting the hybrid system at the lowest possible ozone injection rate and optimizing ozone mass transfer so that ozone is not wasted in the system. 63 Figure 3. 4 Comparison of energy consumption with high ozone injection rate in catalytic ceramic membrane fi ltration system. 3. 3.2 Comparison between catalytic ceramic membrane and hollow fiber membrane in energy consumption and environmental impacts At the conditions studied, energy consumption for pressurization is 25% lower for catalytic ceramic membrane fi ltration process as compared to that for hollow fiber membrane filtration. As is apparent in Figure 3. 5, the largest fraction of energy consumed during operation is for membrane pressurization. This underscores the importance of operating a system at the lowest possible pressure required to achieve the desired treatment. When comparing the two processes, energy consumption for backwashing and membrane manufacturing is 15% and 30% lower, respectively, for hybrid membrane filter vs hollow 64 fiber membrane filt ration. The energy consumed for manufacturing of membranes is lower with CC than with HF. As a result of the large amount of energy consumed in ozone generation, the total energy consumption for catalytic membrane filtration is 7.6% higher than that for ho llow fiber membrane filtration. Improvements in energy - and mass transfer efficiency of ozone generation is critical to the reduction in energy consumption for CC applications. Electricity contributed to 95% of embodied energy usage for both treatment pro cesses. Figure 3. 5 Embodied energy consumption in catalytic ceramic membrane (CC) and hollow fiber membrane (HF) filtration in water treatment application In terms of environmental impacts, CC had a slightly lower total impacts. It had lower 65 environmen tal impacts on ozone depletion and carcinogenicity, non - carcinogenics, and ecotoxicity of pollutants, as shown in Figure 3. 6 and Table 3. 2 Based on published literature [19, 35 - 36] , CC can be used to minimize the formation of carcinogenic by - products uch a s THMs and bromate. Ozonation ceramic membrane filtration was also shown to effectively inactivate E. coli, eliminating the need for disinfection of backwash water, thereby further reducing its environmental impacts compared to polymeric membrane filtratio n. Although energy consumption for CC was greater than for HF, CC had a lower environmental impact because of more environmental - friendly membrane manufacturing processes, and the lower energy required for pressurization. 66 Figure 3. 6. Comparison of en vironmental impacts normalization results for catalytic ceramic membrane (CC) and hollow fiber membrane (HF) filtration processes with HF impacts processes percentage 67 Table 3. 2 Characterized environmental impacts in catalytic membrane (CC) and hollo w fiber (HF) membrane filtration processes Impact category Unit CC HF Ozone depletion kg CFC - 11 eq 1.40×10 - 5 9.18×10 - 5 Global warming kg CO 2 eq 4550 3960 Smog kg O 3 eq 296 259 Acidification kg SO 2 eq 38.5 33.6 Eutrophication kg N eq 1.44 1.18 Carcin ogenics CTUh 1.48×10 - 5 2.20×10 - 5 Non carcinogenics CTUh 1.45×10 - 4 1.55×10 - 4 Respiratory effects kg PM2.5 eq 2.32 1.90 Ecotoxicity CTUe 2200 2410 Fossil fuel depletion MJ surplus 1940 1770 3. 4 Conclusions In this study, LCA was conducted to assess th e energy consumption and environmental impacts of catalytic ceramic membrane with ozonation filtration at the industrial scale, and the results were compared to that for hollow fiber membrane filtration for water treatment. The results suggest that: For ca talytic ceramic membrane filtration, pressurization resulted in the majority of energy consumed (62%). Energy consumption for pressurization plus ozone 68 generation consumed 93% of total energy input. Acidification, carcinogenicity and ecotoxicity of polluta nts, smog, and climate change in catalytic ceramic membrane filtration processes had higher environmental impacts in normalization. Ozone injection resulted in 43% of environmental impacts while pressurization with 49%. Doubling the ozone injection rate re sulted in a 17.5% increase in the energy consumption for catalytic ceramic membrane filtration. Although the energy for pressurization and backwash were reduced, ozone generation also consumed significant amount of energy contributed in higher energy consu mption. Catalytic ceramic membrane filtration resulted in less energy consumption of pressurization and backwashing as compared with hollow fiber membrane filtration. Catalytic ceramic membrane filtration had a slightly lower environmental impacts than h ollow fiber membrane filtration. Energy consumption for ozone generation underscores the need for research to develop more efficient methods to generate ozone and to optimize mass transfer. Alternatively, alternative and less energy consumptive methods to reduce membrane fouling would less both energy consumption and environmental impacts of catalytic ceramic membrane filtration, making it clearly favorable over polymeric membrane filtration. 69 APPENDIX 70 APPENDIX S upporti ng Information Energy consumption and environmental impact analysis of ozonation catalytic membrane filtration system and comparison with hollow fiber membrane in water treatment application Hollow fiber membrane inventory analysis Process Diagram 71 Figure 3.A.1. System boundaries of hollow fiber membrane processes with chlorine as disinfectants 72 Overall process Output: Flow Amount Range Unit Description Hollow fiber membrane water treatment clean water production 9,000,000 8,000,000 10,000,00 0 gal Assum ed daily water production for a town with about 100,000 people. Each person uses about 80 - 100 gallons of water per day [1] . Input: (Daily input) Flow Quantity Range Unit Description Pressurization 9,000,000 gal Pressurization throughout the processes Membrane cartridges 1.315 items Total 1920 membrane cartridges [2] in hollow fiber membrane treatment plant. Treatment plant with 20 years lifetime. Cartridge would be replaced every 4 years. Disinfection 9,000,000 gal All water are going to be disinfec ted Backwash 351,000 gal 3.9% of total water flow consumed during backwashing [3] . 9000000*3.9% =35100 73 (A) Pressurization Output: Flow Amount Unit Description Pressurization 1 gal Input: Flow Amount Unit Description Electricity at eGrid, RFCM. 4.2069×10^ - 4 kwh Energy cost in pressurization could be calculated from water flow rate and transmembrane pressure (a) (b) (c) . [3 - 6] (a) Flux check Based on 9 MPG daily water production and hollow fiber membrane water flux capacity, 1720 cartridges, ea ch contained 7400 fibers can achieve the daily water production in 24 hours (one day). Average membrane filtration rate: 9 MPG = 34068.7 m 3 /day = 1419.529 m 3 /h 74 Total cartridge number = 1920 Each cartridge filtration rate q= (1419.529 m 3 /h) ÷ 1920 = 0.739 3 m 3 /h Effective areas = 7400 × 0.00102m 2 = 7.523 m 2 [4] [5] Permeate flux = (0.7393 m 3 /h) ÷ (7.523 m 2 ) = 0.09828 m/h Within the range of q = 0.03 - 0.17 m/h [3] , which is acceptable (b) Transmembrane pressure for 5 kda (same grain size as ceramic memb rane) TMP = 3.0 bar = 300 kPa = 3×10 5 N/m 2 [5, 6] (c) Energy calculation Ph 5 )/0.75 = 82.1625 Watt = 0.0821625 kW In which, q - flow in m 3 /h, TMP - transmembrane pressure, pa, and Ph - power pumping water, kW. So the average rate power transfer of each cartridge is 0.0821625 kW. Filtration time for 1 gal of water: t = (3.785×10 - 3 m 3 ) / (0.7393 m 3 /h) = 5.1202 × 10 - 3 h for every gal water, energy con sumption is 0.0821625 kW × 5.1265 × 10 - 3 h = 4.2069 × 10 - 4 kW·h 75 (B) Membrane cartridge Output: Flow Amount Unit Description Membrane cartridges 1 item Input: Flow Amount Unit Description Electricity at eGrid, RFCM 50 Wh Estimated electricity cost in manufacturing the cartridge. Estimated power of manufacturing machines 500 w with 1 min producing 1 cartridge unit. (similar estimation in ceramic membrane supporting) Membrane fibers 7400 items 1920 membrane cartridges, 7400 membrane fibers. Cartridges w ere replaced every 4 years, and membrane fibers were also replaced every 4 years [2] . 100% polypropylene media 6.923 kg Estimation of materials input in one membrane cartridge: each cartridge is 764 mm long with material density of 946 kg/m³, ID (based on 7400 fibers and 0.7 - 0.8 mm fiber diameter) = 70 mm, OD = 130 mm [7] . 76 (C) Membrane fibers Output: Flow Amount Unit Description Membrane fiber 1 item Input: Flow Amount Unit Description Electricity, at eGrid, RFCM 1 Wh Estimated electricity cost i n manufacturing the cartridge. Estimated power of manufacturing machines 500 w with 5 seconds producing 1 cartridge unit. PVDF 0.41546 g For each hollow fiber, OD = 700 - 800 um, ID = 400 - 500 um, length of membrane ~760 mm, so the volume of each hollow fibe r is 2.334×10 - 7 m 3 , and density of PVDF is 1.78g/cm 3 , so weight is 0.41546 g [8] [9] . 77 (D) Disinfection Output: Flow Amount Unit Description Disinfection 1 gal Input: Flow Amount Unit Description Chlorine 13.154 mg Average typical application do se for chlorine disinfection is 29 lb free cholrine/ MG water [10] [11] 29 lb × (1 mg/2.2046×10 - 6 lb) ÷ 1000000 gallons water = 13.154 mg chlorine/gal water. (E) Backwash Output: Flow Amount Unit Description Backwash 1 gal 78 Input: Flow Amount Unit De scription Hollow fiber membrane water treatment clean water production 1 gal Use treated water for membrane backwash. Sodium bisulfate 2.125 ug [10] Citric Acid 2.62 ug Sulfuric acid 0.125 ug Electricity, at eGrid, RFCM 5.4736×10 - 4 kWh For each memb rane module, assumed average backwash TMP wash = 1.3TMP operation , [3] and backwash Q= 1.4786 m 3 /h [3] , pump energy energy consumption is 5.4736 × 10 - 4 kW·h 79 Catalytic membrane with ozonation inventory analysis Process Diagram Figure 3.A.2. System boundaries of the stud y in catalytic ceramic ozonation membrane filtration processes 80 Overall process Output: Flow Amount Range Unit Description Catalytic ceramic membrane water treatment clean water production 9,000,000 8,000,000 10,000,00 0 gal Same assumption as hollow fi ber membrane: daily water production for a town with about 100,000 people. Each person uses about 80 - 100 gallons of water per day [1] . Input: Process Amount Unit Description Pressurization 9000000 gal Assumed daily water production amount Ozone injecti on 185.9 kg With 5 ug/s ozone injection, scaled up with membrane surface area scaled up ratio (131.9 cm 2 vs. 5 m 2 ). [13] [14] Backwash 315900 gal Catalytic ceramic membrane can filtrate water for 100 min between backwashes instead of 90 min in hollow fibe r membrane, So 3.9%*(9/10) = 3.51%. 9000000*3.51% =315900. [3] [13] Membrane Manufacture 0.311297 item Membrane water filtration rate = 1.25 m 3 /h, 9 MGD, 365 days, last 10 years for ceramic membrane, Replace every 10 years. Lab results scaled up to indust ial scale. [13] [14] 81 (A) Pressurization Output: Flow Amount Unit Description Pressurization 1 gal Input: Flow Amount Unit Description Electricity at eGrid, RFCM 3.50 × 10 - 4 kW·h Water flux = 0.25 m 2 /h, filtration area = 5 m 2 , filtration rate Q = 1. 25 m 3 /h, average TMP during treatment = 2.4621 bar, pump efficacy = 75% (same as hollow fiber membrane), Ph = QTMP/efficiency. When use the same pore size membrane, transmembrane pressure at the same flux will not change much after scaled up. [13] [14] ( B) Ceramic Membrane array Output: Flow Amount Unit Description Membrane cartridges 1 item 82 Input: Flow Amount Unit Description Electricity, at eGrid, RFCM 50 Wh Estimated electricity to manufacture the cartridge. (Similar as hollow fiber membrane) Ca talytic manganese oxide coated ceramic membrane 1 items 1 membrane array contain 1 membrane unit, but membrane need to be changed every 10 years 100% polypropylene media 13.58 kg Use the same membrane supporting material as hollow fiber membrane. Length: 864 mm. Out Diameter: 203 mm, inner diameter: 143 mm, density: 946 kg/m³. [7] (C) Ceramic membrane unit Output: Flow Amoun t Unit Description Catalytic ceramic membrane unit 1 item 83 Input: Flow Amount Unit Description Electricity, at eGrid, RFCM 15 40 Wh For every 20 membranes: 20 W·h (centrifuged) + 29 W·h (sonic) +1000 W·h (500 Celsius 45 min) + 50 W·h ceramic membrane manufacture (1100 W·h every 20 membrane). Scale up ratio (490.87 cm 3 vs. 1.37×10 7 ml): 28 [13] [14] . Energy cost: 55*28 =1540 W·h. [15] [13] Ceramic 5.582 kg Each ceramic membrane weighs 0.2 kg, times volume ratio Manganese chloride 9.55 g Solution to synthesis catalytic membrane, lab value times surface ratio (131.9 cm 2 vs. 5 m 2 ) [13] [14] . Ozone 538 g KNO3 1.9159 g Poly(dia llyldimethyla mmonium chloride) (PDDA) 379 g NaOH 151.6 g (D) Ozone injection Output: Flow Amount Unit Description Ozone injection 1 kg 84 Input: Flow Amount Unit Description Oxygen 1 kg Corona - discharge method. 3O 2 3 . [16] Electricity 7 kW·h Energy consumption is less than 7 kW per kilo ozone produced. [16] [17] (E) Backwash Output: Flow Amount Unit Description Backwash 1 gal 85 Input: Flow Amount Unit Description Ceramic membrane water treatment cle an water production 1 gal Sodium bisulfate 2.125 ug Same as hollow fiber membrane filtration. Citric Acid 2.62 ug Sulfuric acid 0.125 ug Electricity, at eGrid, RFCM 3.7345 × 10 - 4 kWh Ph = Q backwash · TMP wash 7 watt, for every gal water, calculated energy consumption is 3.7345 × 10 - 4 kW·h [13] [14] Higher ozonation - catalytic membrane with ozonation inventory analysis Overall process Output: Flow Amount Range Unit Description Catalytic ceramic membrane wate r treatment clean water production 9,000,000 8,000,000 10,000,00 0 gal Same assumption as hollow fiber membrane: daily water production for a town with about 100,000 people. Each person uses about 80 - 100 gallons of water per day [1] . 86 Input: Process Amoun t Unit Description Pressurization 9000000 gal Assumed daily water production amount Ozone injection 371.87 kg With 10 ug/s ozone injection, scaled up with membrane surface area scaled up ratio (131.9 cm 2 vs. 5 m 2 ). [13] [14] Backwash 315900 gal Catalyti c ceramic membrane can filtrate water for 100 min between backwashes instead of 90 min in hollow fiber membrane, So 3.9%*(9/10) = 3.51%. 9000000*3.51% =315900. [3] [13] Membrane Manufacture 0.311297 item Membrane water filtration rate = 1.25 m 3 /h, 9 MGD, 365 days, last 10 years for ceramic membrane, Replace every 10 years. Lab results scaled up to industial scale. [13] [14] (A) Pressurization Output: Flow Amount Unit Description Pressurization 1 gal 87 Input: Flow Amount Unit Description Electricity at eGrid, RFCM 2.877 × 10 - 4 kW·h Water flux = 0.25 m 2 /h, filtration area = 5 m 2 , filtration rate Q = 1.25 m 3 /h, average TMP during treatment = 2.0515 bar, pump efficacy = 75% (same as hollow fiber membrane), Ph = QTMP/efficiency. When use the same pore siz e membrane, transmembrane pressure at the same flux will not change much after scaled up. [13] [14] (B) Ceramic Membrane array Output: Flow Amount Unit Description Membrane cartridges 1 item 88 Input: Flow Amount Unit Description Electricity, at eGrid, RFCM 50 Wh Estimated electricity to manufacture the cartridge. (Similar as hollow fiber membrane) Catalytic manganese oxide coated ceramic membrane 1 items 1 membrane array contain 1 membrane unit, but membrane need to be changed every 10 years 10 0% polypropylene media 13.58 kg Use the same membrane supporting material as hollow fiber membrane. Length: 864 mm. Out Diameter: 203 mm, inner diameter: 143 mm, density: 946 kg/m³. [7] (C) Ceramic membrane unit Output: Flow Amount Unit Description Cat alytic ceramic membrane unit 1 item 89 Input: Flow Amount Unit Description Electricity, at eGrid, RFCM 1540 Wh For every 20 membranes: 20 W·h (centrifuged) + 29 W·h (sonic) +1000 W·h (500 Celsius 45 min) + 50 W·h ceramic membrane manufacture (1100 W·h every 20 membrane). Scale up ratio (490.87 cm 3 vs. 1.37×10 7 ml): 28 [13] [14] . Energy cost: 55*28 =1540 W·h. [15] [13] Ceramic 5.582 kg Each ceramic membrane weighs 0.2 kg, times volume ratio Manganese chloride 9.55 g Solution to synthesis catalytic mem brane, lab value times surface ratio (131.9 cm 2 vs. 5 m 2 ) [13] [14] . Ozone 538 g KNO 3 1.9159 g Poly(diallyldimethyl ammonium chloride) (PDDA) 379 g NaOH 151.6 g (D) Ozone injection Output: Flow Amount Unit Description Ozone injection 1 kg 90 Inpu t: Flow Amount Unit Description Oxygen 1 kg Corona - discharge method. 3O 2 3 . [16] Electricity 7 kW·h Energy consumption is less than 7 kW per kilo ozone produced. [16] (E) Backwash Output: Flow Amount Unit Description Backwash 1 gal Input: Flow Amount Unit Description Ceramic membrane water treatment clean water pro duction 1 gal Ph = Q backwash · TMP wash water, calculated energy consumption is 3.109 × 10 - 4 kW·h [13] [14] Sodium bisulfate 2.125 ug Citric Acid 2.62 ug Sulfuric acid 0.125 ug Electricity, at eGrid, RF CM 3.109 × 10 - 4 kWh 91 Results: Catalytic ceramic membrane with ozonation energy network Table 3.A.1 Catalytic ceramic membrane energy network Inflows (4) Quantity Unit Total 61.9 GJ Pressurization 38.7 GJ Ozone Injection 19 GJ Backwash 3.91 GJ Membra ne Manufacturing 0.38 GJ Table 3.A.2 Process contribution in energy consumption Process Unit Total C atalytic pressurization C atalytic membrane array O zone injection O zone backwash Remaining processes MJ 79.30 30.47 32.07 12.59 4.17 Bituminous coal MJ 4 9468.08 32758.82 5.14 13535.43 3168.68 Crude oil MJ 1528.35 1012.11 0.16 418.19 97.90 Natural gas MJ 7363.65 4876.37 0.77 2014.84 471.68 Oxygen MJ 2368.43 0 0 2210.92 157.51 Polypropylene MJ 313.41 0 302.63 0 10.78 92 Results: Catalytic ceramic membrane with ozonation environmental impacts score Table 3.A.3 Catalytic ceramic membrane environmental impacts - normalization Impact category Total Impact category Pressurization Ozone injection Backwash Membrane Manufacturing Ozone depletion 8.70E - 05 Ozone dep letion 3.03E - 07 7.85E - 05 5.71E - 06 2.57E - 06 Global warming 1.88E - 01 Global warming 1.18E - 01 5.71E - 02 1.20E - 02 4.83E - 04 Smog 2.13E - 01 Smog 1.35E - 01 6.33E - 02 1.36E - 02 3.99E - 04 Acidification 4.24E - 01 Acidification 2.73E - 01 1.24E - 01 2.72E - 02 5.02E - 04 Eutrop hication 6.67E - 02 Eutrophication 1.63E - 02 4.53E - 02 4.35E - 03 7.72E - 04 Carcinogenics 2.81E - 01 Carcinogenics 5.98E - 02 1.92E - 01 1.81E - 02 1.10E - 02 Non carcinogenics 1.38E - 01 Non carcinogenics 6.78E - 02 5.98E - 02 8.86E - 03 1.23E - 03 Respiratory effects 9.56E - 02 R espiratory effects 5.06E - 02 3.81E - 02 6.14E - 03 8.31E - 04 Ecotoxicity 1.99E - 01 Ecotoxicity 5.38E - 02 1.28E - 01 1.29E - 02 4.17E - 03 Fossil fuel depletion 1.03E - 01 Fossil fuel depletion 6.16E - 02 3.25E - 02 6.55E - 03 2.45E - 03 93 Results: Comparison of catalytic membr ane with higher ozonation energy inputs Table 3.A.4 Catalytic ceramic membrane environmental impacts - normalization Processes Unit Original ozone injection rate Double ozone injection rate Pressurization GJ 38.7 32.3 Ozone Injection GJ 19 38.5 Backwash GJ 3.91 1.65 Membrane Manufacture GJ 0.378 0.383 Results: Comparison of catalytic membrane with ozonation and hollow fiber membrane energy inputs Table 3.A.5 Comparison of catalytic membrane and hollow fiber membrane energy network Processes Unit Cera mic membrane Hollow Fiber membrane Pressurization GJ 38.7 48.4 Ozone Injection /Disinfection GJ 19 2.34 Backwash GJ 3.91 4.48 Membrane Manufacture GJ 0.378 1.24 94 Results: Comparison of catalytic membrane and hollow fiber membrane environmental impact s Table 3.A.6 Comparison of catalytic membrane and hollow fiber membrane environmental impacts - characterization Impact category Unit Catalytic Ceramic membrane Hollow Fiber membrane Ozone depletion kg CFC - 11 eq 1.40E - 05 9.18E - 05 Global warming kg CO2 e q 4550.55 3964.06 Smog kg O3 eq 296.17 258.65 Acidification kg SO2 eq 38.50 33.60 Eutrophication kg N eq 1.44 1.18 Carcinogenics CTUh 1.48E - 05 2.20E - 05 Non carcinogenics CTUh 1.45E - 04 1.55E - 04 Respiratory effects kg PM2.5 eq 2.32 1.90 Ecotoxicity CT Ue 2198.94 2413.77 Fossil fuel depletion MJ surplus 1940.28 1767.86 95 Table 3.A.7 Comparison of catalytic membrane and hollow fiber membrane environmental impacts - normalization Impact category Catalytic Ceramic membrane Hollow Fiber Membrane Ozone depletion 8.70E - 05 5.69E - 04 Global warming 1.88E - 01 1.64E - 01 Smog 2.13E - 01 1.86E - 01 Acidification 4.24E - 01 3.70E - 01 Eutrophication 6.67E - 02 5.45E - 02 Carcinogenics 2.81E - 01 4.18E - 01 Non carcinogenics 1.38E - 01 1.48E - 01 Respiratory effects 9.56E - 02 7.8 3E - 02 Ecotoxicity 1.99E - 01 2.18E - 01 Fossil fuel depletion 1.03E - 01 9.39E - 02 Literature cited: [1] USGS. Water Questions & Answers. https://water.usgs.gov/edu/qa - home - percapita.html . 02 Dec. 2016. [2] S. Easterh. Bay Area Water Treatment Plant 'puts out some awesome water,' official says. MLive Media Group. September 15, 2015. http:/ /www.mlive.com/news/bay - city/index.ssf/2015/09/bay_area_water_treatment_plant_3.html 96 [3] K.J. Howe, C.C. John, D.W. Hand, G. Tchobanoglous, R.R. Trussell, Principles of Water Treatment. Hoboken, John Wiley & Sons Inc., New Jersey, 2012. pp. 285 - 286. [4] R Ramli, N Bolong, Effects of Pressure and Temperature on Ultrafiltration Hollow Fiber Membrane in Mobile Water Treatment System, J. Eng. Sci. Tech. 11 (2016) 1031 - 1040. [5] S. Nakatsuka, I. Nakate, T. Miyano, Drinking water treatment by using ultrafiltrat ion hollow fiber membranes, Desalination 106.1 - 3 (1996): 55 - 61. [6] A.I. Schäfer, G.F. Anthony, T.D. Waite, Cost factors and chemical pretreatment effects in the membrane filtration of waters containing natural organic matter, Water Res. 35 (2001): 1509 - 15 17. [7] Brother Filtration. PP pleated cartridge. http://www.brotherfiltration.com/pp - pleated - filter - cartridge/ , 2017 (accessed 6 Apr. 2017). [8] C. Conidi, F. Destani, A. Cassan o, Performance of hollow fiber ultrafiltration membranes in the clarification of blood orange juice, Beverages 1.4 (2015): 341 - 353. [9] D. Wang, K. Li, W.K. Teo, Preparation and characterization of polyvinylidene fluoride (PVDF) hollow fiber membranes, J. Membr. Sci. 163 (1999) 211 - 220. [10] Bay county department of water and sewer. Bay water treatment plant memo. May 18, 2013. [11] K. J. Howe, D.W. Hand, J.C. Crittenden, R.R. Trussell, G. Tchobanoglous, Principles of Water Treatment. Hoboken, John Wiley & Sons, N.J, 2012. pp. 286. [12] Ultrafiltration. http://msdssearch.dow.com/PublishedLiteratureDOWCOM/dh_0914/0901b80380914f2a.pdf?filepath=liquidseps/pdfs/no reg/795 - 00022.pdf&fromPage=GetDoc, 2011 (Accessed 1 April, 2017) [13] X. Wang, S.H. Davies, S.J. Ma sten, Analysis of energy costs for catalytic ozone membrane filtration, Sep. Purif. Technol. 186 (2017) 182 - 187. [14] Veolia Water Solutions & Technologies, CeraMem Industrial Ceramic Membrane Technology. https://www.veoliawatertechnologies.com/sites/g/files/dvc471/f/assets/documents/2015/01/34577VWS_CeraMem_4p_1 97 1 - 6.pdf , 2015 (accessed 6 Apr. 2017). [15] L.M. Corneal, M.J. Bau mann, S.J. Masten, S.H. Davies, V.V. Tarabara, S. Byun, Mn oxide coated catalytic membranes for hybrid ozonation - membrane filtration: Membrane microstructural characterization, J. Membr. Sci. 369 (2011) 182 - 187. [16] Lenntch B.V., Ozone generators. https://www.lenntech.com/otozone.htm , 2015 (accessed 6 Apr. 2017). [17] L.I. Smith, F.L. Greenwood, O. Hudrlik, Ozone (a laboratory ozonizer), Organic Syn. 26 (1946) 63. 98 REFERENCES 99 REFER ENCES [1] W. Gao, H. Liang, J. Ma, M. Han, Z. Chen, Z. Han, G. Li, Membrane fouling control in ultrafiltration technology for drinking water production: a review, Desalination 272 (2011) 1 - 8. [2] M. Peter - Varbanets, J. Margot, J. Traber, W. Pronk, Mechan isms of membrane fouling during ultra - low pressure ultrafiltration, J. Membr. Sci.377 (2011) 42 - 53. [3] A. Braghetta, F.A. DiGiano, W.P. Ball, NOM accumulation at NF membrane surface: impact of chemistry and shear, J. Environ. Eng.124 (1998) 1087 - 1098. [ 4] S.H. Davies, M.J. Baumann, S. Byun, L.M. Corneal, V.V. Tarabara, S.J. Masten, Fabrication of catalytic ceramic membranes for water filtration. Water Sci. Technol. Water Supply 10 (2010) 81 - 86. [5] B.S. Karnik, S.H. Davies, M.J. Baumann, S.J. Masten, Fa brication of catalytic membranes for the treatment of drinking water using combined ozonation and ultrafiltration, Environ. Sci. Technol. 39 (2005) 7656 - 7661. [6] A. Hög, J. Ludwig, M. Beery, The use of integrated flotation and ceramic membrane filtratio n for surface water treatment with high loads of suspended and dissolved organic matter, J. Water Process Eng. 6 (2015) 129 - 135. [7] J. Kim, W. Shan, S.H. Davies, M.J. Baumann, S.J. Masten, V.V. Tarabara, Interactions of aqueous NOM with nanoscale TiO2: Im plications for ceramic membrane filtration - ozonation hybrid process, Environ. Sci. Technol. 43 (2009) 5488 - 5494. [8] S. Van Geluwe, L. Braeken, B. Van der Bruggen, Ozone oxidation for the alleviation of membrane fouling by natural organic matter: A review , Water Res. 45 (2011) 3551 - 3570. [9] B. Schlichter, V. Mavrov, H. Chmiel, Study of a hybrid process combining ozonation and membrane filtration - filtration of model solutions, Desalination 156 (2003) 257 - 265. [10] S. Byun, S.H. Davies, A.L. Alpatova, L. M. Corneal, M.J. Baumann, V.V. Tarabara, S.J. Masten, Mn oxide coated catalytic membranes for a hybrid ozonation membrane filtration: comparison of Ti, Fe and Mn oxide coated membranes for water quality, Water Res. 45 (2011) 163 - 170. [11] J. Song, Z. Zhan g, S. Tang, Y. Tan, X. Zhang, Does pre - ozonation or in - situ ozonation really mitigate the protein - based ceramic membrane fouling in the integrated process of ozonation coupled with ceramic membrane filtration? J. Membr. Sci. 548 (2018) 254 - 262. [12] J. So ng, Z. Zhang, X. Zhang, A comparative study of pre - ozonation and in - situ 100 ozonation on mitigation of ceramic UF membrane fouling caused by alginate, J. Membr. Sci. 538 (2017) 50 - 57. [13] W. Yu, N.J. Graham, G.D. Fowler, Coagulation and oxidation for contro lling ultrafiltration membrane fouling in drinking water treatment: application of ozone at low dose in submerged membrane tank, Water Res. 95 (2016) 1 - 10. [14] X. Wang, S.H. Davies, S.J. Masten, Analysis of energy costs for catalytic ozone membrane filtra tion, Sep. Purif. Technol. 186 (2017) 182 - 187. [15] S. Mozia, M. Tomaszewska, A.W. Morawski, Application of an ozonation adsorption ultrafiltration system for surface water treatment, Desalination 190 (2006) 308 - 314. [16] Y.G. Park, Effect of ozonation for reducing membrane - fouling in the UF membrane, Desalination 147 (2002) 43 - 48. [17] M. Davis, Water and Wastewater Engineering: Design Principles and Practice. McGraw - Hill, New York 2010. pp. 12 - 18. [18] B.S. Karnik, S.H. Davies, K.C. Chen, D.R. Jaglowski, M.J. Baumann, S.J. Masten, Effects of ozonation on the permeate flux of nanocrystalline ceramic membranes, Water Res. 39 (2005) 728 - 734. [19] B.S. Karnik, S.H. Davies, M.J. Bauma nn, S.J. Masten, The effects of combined ozonation and filtration on disinfection by - product formation, Water Res. 39 (2005) 2839 - 2850. [20] M.W. LeChevallier, K.K. Au, Water treatment and pathogen control, IWA Publishing, London, 2004. pp. 56. [21] B. Sc hlichter, V. Mavrov, H. Chmiel, Study of a hybrid process combining ozonation and microfiltration/ultrafiltration for drinking water production from surface water, Desalination 168 (2004) 307 - 317. [22] K. J. Howe, D.W. Hand, J.C. Crittenden, R.R. Trussell, G. Tchobanoglous, Principles of Water Treatment. Hoboken, John Wiley & Sons, N.J, 2012. pp. 286. [23] Veolia Water Solutions & Technologies, CeraMem Industrial Ceramic Membrane Technology. https://www.veoliawatertechnologies.com/sites/g/files/dvc471/f/assets/documents/2015/ 01/34577VWS_CeraMem_4p_11 - 6.pdf , 2015 (accessed 6 Apr. 2017). [24] Lenntch B.V., Ozone generators. https://www.lenntech.com/otozone.htm , 2015 (accessed 6 Apr. 2017). [25] L.M. Corneal, M.J. Baumann, S.J. Masten, S.H. Davies, V.V. Tarabara, S. Byun, Mn oxide coated catalytic membranes for hybrid ozonation - m embrane filtration: Membrane 101 microstructural characterization, J. Membr. Sci. 369 (2011) 182 - 187. [26] Bay county department of water and sewer. Bay water treatment plant memo. May 18, 2013. [27] S . Easter, Bay Area Water Treatment Plant 'puts out some aw esome water,' official says. MLive Media Group. http://www.mlive.com/news/bay - city/index.ssf/2015/09/bay_area_water_treatment_plant_3.html, 2015 (accessed 6 Apr. 2017). [28] Brother Filtration. PP pleated cartridge. http://www.brotherfiltration.com/pp - pleated - filter - cartridge/ , 2017 (accessed 6 Apr. 2017). [29] C. Conidi, F. Destani, A. Cassano, Performance of hollow fiber ultrafiltration membranes in the clarification of blood o range juice, Beverages 1 (2015) 341 - 353. [30] D. Wang, K. Li, W.K. Teo, Preparation and characterization of polyvinylidene fluoride (PVDF) hollow fiber membranes, J. Membr. Sci. 163 (1999) 211 - 220. [31] M. Puettmann, R. Bergman, E. Oneil, Cradle - To - Gate L ife Cycle Assessment of North American Hardboard and Engineered Wood Siding and Trim Production. CORRIM Final Report. University of Washington. Seattle, WA. Jul. 2016, (2016) 1 - 77. [32] H. Clarke, J.B. Conant, O. Kamm, R. Adams, Organic syntheses: An annua l publication of satisfactory methods for the preparation of organic chemicals. John Wiley & Sons, 2000. [33] R. Heijungs, J. Guinée, R. Kleijn, V. Rovers, Bias in normalization: causes, consequences, detection and remedies, Int. J. Life Cycle Assess. 12 (2007) 211. [34] U. Von Gunten, Ozonation of drinking water: Part I. Oxidation kinetics and product formation, Water Res. 37 (2003) 1443 - 1467. [35] M. Moslemi, S.H. Davies, S.J. Masten, Bromate formation in a hybrid ozonation - ceramic membrane filtration sy stem, Water Res. 45 (2011) 5529 - 5534. [36] F. Fiessinger, Y. Richard, A. Montiel, P. Musquere, Advantages and disadvantages of chemical oxidation and disinfection by ozone and chlorine dioxide, Sci. Total Environ. 18 (1981) 245 - 261. 102 CHAPTER FOUR Effects of methylene blue photocatalytic degradation under visible light irradiation by ZnFe 2 O 4 , NiFe 2 O 4 and CdS 103 Abstract C onventional biological and physical treatment technologies cannot effectively degrade many recalcitra nt contaminants in water. Research has been conducted to develop highly effective photocatalysts for the degradation of such recalcitrant contaminants. This study synthesized and evaluated at the characteristics of three photocatalysts: CdS, ZnFe 2 O 4 , and N iFe 2 O 4 and compared their photodegradation ability using methylene blue as a model contaminant. The results showed that methylene blue could be successfully degraded by all three catalysts under visible light at room temperature, and t he order of photocata ly tic efficacies was CdS > NiFe 2 O 4 ZnFe 2 O 4 . CdS had the widest absorption range, compared to NiFe 2 O 4 and ZnFe 2 O 4 . The absorption range of photocatalysts, initial dye concentrations, amount of photocatalysts addition, and photoreactor conditions affected methylene blue removal efficacy. 104 4.1 Introduction Up to 20% of the total mass of dyes produced worldwide is released in the textile effluents [1] . Organic dyes such as methylene blue contain aromatic rings, so conventional biological treatment methods are not very effective for dye degradation. Physical treatment technologies such as adsorption, coagulation, and membrane filtration transfer organic compounds from water but cannot degrade the molecules [2] . Additionally, the regeneration of adsorptive capac ity and the post - treatment of concentrated wastes are expensive [3] . Ozonation and chlorination are being used for the removal of dyes but limited due to their high operating costs [4] . Therefore, there is a need to develop more cost - effective methods for the degradation of dye effluents. Conventional wastewater treatment is energy - intensive [5] . Developing ways to effectively treat wastewater yet minimize energy consumption is of increasing interest [6] . The application of photocatalysis that utilize sol ar light energy for the photodegradation of organic compounds present in water and wastewaters is of particular importance. More than 50 types of organic compounds or waste products that potentially could be photodegraded, include organic dyes, sugars, alc ohols, aldehydes, ketones and organic acids [7] . Photocatalytic degradation is an oxidation - reduction reaction. When the catalyst is irradiated and absorbs photons of energy greater than or equal to its bandgap, electrons move from the valence band to the conduction band, resulting in the formation of electron (e ) and holes (h + ). The electrons (e ) are generated at conduction band and holes (h + ) are collected at valence band. The electrons can react with electron acceptors on the surface of catalysts. The generated holes can oxidize the molecules on the photocatalysts 105 surface, or oxidize OH - and H 2 O into ·OH radicals, a strong oxidizing agent that is capable of oxidizing many dyes [8] . Numerous studies have investigated the photocatalytic degradation of a variety of dyes. For example, Pirhashemi et al [9] prepared ZnO/Ag/Ag 2 WO 4 nano - particles by an ultrasonic - assisted method and investigated the photocatalytic degradation of RhB, and the results demonstrated that the oxidative ability of ·O2 - was greater than that of h+ and ·OH species for ZnO as photocatalysts. On the other hand, Daneshvar et al [10] studied the photocatalytic degradation of azo dye acid red 14 with ZnO and found that that hydroxyl radicals were the main reactive species and that positive holes were involved in the reactions. Moreover, the results of Daneshvar et al [10] also indicated that the degradation efficiency of acid red 14 was affected by illumination time, pH, and photocatalyst amount. The rate of photodegradation increased with the addition of hydrogen peroxide, but was inhibited with addition of ethanol. Houas et al [11] investigated the photocatalytic degradation of methylene blue (MB) under UV light with titanium oxide in aqueous heterogeneous suspensions. This study showed th at the cationic functional group of MB molecule was likely adsorbed to the TiO 2 surface, and MB was mineralized resulting in the formation of numerous inorganic byproducts (i.e., CO 2 , SO 4 , NH 4 + and NO 3 ) during the photodegradation process. Based on current research, the efficiency of photocatalysts in organic pollutants degradation are affected by many factors including, the type of photocatalyst, pH, illumination time, particle size of photocatalyst [12] , band gap, oxidizing agents, calcination 106 temperatures, and catalyst loadings [10,13] . In this study, the efficacy and kinetics of three photocatalysts (CdS, ZnFe 2 O 4 and NiFe 2 O 4 ) to degrade methylene blue were evaluated. Cadmium sulfid e is a photocatalyst that has been extensively studied as photocatalyst due to its activity under visible - light and potential to degrade organic contaminants resulting in hydrogen production [14, 15] . Spinel semiconductors such as ZnFe 2 O 4 and NiFe 2 O 4 are a group of minerals. Research have shown that certain spinels have excellent photocatalytic properties [16 - 20] . Spinels usually have more narrow band gap energies than does CdS and contain earth abundant metals. As spinels are not highly toxic, the ferrites are likely to be environmentally benign. The abilities of CdS, ZnFe 2 O 4 , and NiFe 2 O 4 in degrading methylene blue were compared in this study. Band gaps, SEM and absorption spectrum of the photocatalysts were determined. The impact of the concentration of m ethylene blue on removal efficacies was examined. 4.2 Materials and Methods 4.2.1 Materials preparation All chemical reagents used in this experiment were from Alfa Aesar (Tewksbury, MA, US), of analytical grade. 4.2.1.1 Synthesis of CdS CdS was prepared accordin g to the hydrothermal method [15] . A Na 2 S solution (300 mL, 0.14 M) was slowly added to a cadmium acetate (Cd(OAc) 2 ) solution (300 mL, 0.14 M) under vigorous stirring. A yellow precipitate formed. The mixture was stirred for 24 h, settled for 1 day, and fi ltered through a 20 µm size filter. The yellow solid was then 107 suspended in pure water (120 mL) and transferred to a Teflon - lined stainless - steel solid was filtered, washe d with water and ethanol subsequently, filtered again, and kept in a desiccator for 24 h. 4.2.1.2 Synthesis of ZnFe 2 O 4 ZnFe 2 O 4 was prepared by the sol gel method [19] : 0.016 mol of Fe(NO 3 ) 3 ·9H 2 O and 0.008 mol of Zn(NO 3 ) 2 ·6H 2 O were dissolved in 20 mL deio nized water. Citric acid (0.036 mol) was then dissolved in the mixture. While magnetically stirred at 60 , 12 mL ammonia (25%) was added dropwise into the mixture to adjust the pH of the aqueous mixture to 7.0, and the mixture became a sol during this pro cess. After stirring for about 6 h, the sol became a black gel. The gel was then dried at 120 for 12 h, during which its volume expanded about 10 - fold. The dried gel was ground and calcined at 700 for 4 h until a brown product formed. 4.2.1.3 Synthesi s of NiFe 2 O 4 Synthesis of NiFe 2 O 4 was carried out by the hydrothermal method in a stainless - steel autoclave. Ni(NO 3 ) 2 ·6H 2 O (0.001 mol) and 0.002 mol of FeSO 4 ·6H 2 O was dissolved separately in 100 mL of deionized water. Then 0.036 mol of citric acid powder w as dissolved in the mixture. The mixed solution was precipitated with addition of 1 M NaOH solution until pH was 9.7, followed by the addition of 0.1 g of isopropyl alcohol under constant and vigorous magnetic stirring for 5 min. The suspension was poured into the stainless - steel autoclavable container for the hydrothermal treatment. The temperature 108 was controlled at 100°C for 3 h, and cooled to room temperature. Finally, the product was washed several times with absolute ethanol, and then dried in a desicc ator for 24 h to obtain Ni - ferrite nanoparticles [21] . 4.2.2 Photocatalysts characterization The photocatalysts samples were mounted on aluminum stubs using high vacuum carbon tubes (SPI Supplies, West Chester, PA). Samples were coated with iridium to an approx imate thickness of 2.7 nm in a Quorum Technologies/Electron Microscopy Sciences Q150T turbo pumped sputter coater (Quorum Technologies, Laughton, East Sussex, England BN8 6BN) purged with argon gas. The size and morphology of the photocatalyst samples were examined in a LEOL 7500F (field emission emitter) scanning electron microscope (JEOL Ltd., Tokyo, Japan) operated at accelerating voltage of 5 kV. For elemental analysis, energy dispersive X - ray spectroscopy was employed using an Oxford Instruments Aztec system (Oxford Instruments, High Wycomb, Bucks, England), software version 3.1 using a 150 mm 2 Silicon Drift Detector (JEOL 7500F SEM) and an ultra - thin window. The bandgap of the photocatalysts was investigated by a UV vis spectrophotometer (Shimadzu, UV - 2600, Japan) in the wavelength range from 200 nm to 800 nm. 4.2.3 Evaluation of photocatalysts in degradation ability Photoreactions were performed in a Pyrex glass reactor containing methylene blue solution . The initial concentration was varied between 2 and 20 µmol/L. A selected amount of photocatalyst (CdS, ZnFe 2 O 4 , NiFe 2 O 4 ) was added to the solution. The 109 SunLite TM , Milford, Connecticut, US), ozone - free Xe Arc Lamp of 100 W, with continuous stirring for 4 hours (batch system). Samples (10 mL) were collected at 0 min, 5 min, 10 min, 30 min, 1 h, 1.5h, 2 h, 3 h and 4 h. At the end of experiment, methylene blue concentrations were determined by visible spectroscopy (Shimad zu, UV - 2600, Japan) at its absorption peak (665 nm) after centrifugation at 3200 rpm for 10 min. (Clay ADAMS TM , COMPACT II CENTRIFUGE). 4.3 Results and Discussion 4 .3.1 XRD analysis of ZnFe 2 O 4 The structure of the ZnFe 2 O 4 composites was investigated by t he X - ray diffraction. Figure 4.1 shows the XRD patterns of ZnFe 2 O 4 . The pure ZnFe2O4 is in the spinel phase, and the distinctive peaks at 29.9, 35.2, 42.8, 53.1, 55.6, and 62.16 matched well with the (220), (311), (400), (422), (511) and (440) crystal plan es of ZnFe 2 O 4 . 110 Figure 4.1 XRD patterns of the obtained ZnFe 2 O 4 4.3. 2 SEM analysis The SEM images of CdS, ZnFe 2 O 4 , and NiFe 2 O 4 are shown in figure 4. 2 . The particle size for CdS ranged from 50 - 400 nm (figure 4. 2 (a)). ZnFe 2 O 4 was mainly composed of sphe rical grains with sizes of 50 - 400 nm (figure 4. 2 (b)), although the majority of particles were in range of 75 - 190 nm. NiFe 2 O 4 particles ranged in size from 50 - 700 nm (figure 1(c)). The surface of the NiFe 2 O 4 appeared to be more porous, and small particles a dhered to large irregular particles, probably due to magnetic characteristics of NiFe 2 O 4 . The size of NiFe 2 O 4 particles were a wide range; larger particles were likely the result of the different synthesis method. 111 (a) CdS (b) ZnFe 2 O 4 Figure 4.2 SEM i mages of (a) CdS, (b) ZnFe 2 O 4 , and (c) NiFe 2 O 4 112 (c) NiFe 2 O 4 4.3. 3 UV - vis analysis UV vis spectroscopy was used to investigate the absorptive properties of the catalysts. The results are shown in figure 4. 3 . All photocatalysts had vi sible light absorption. The absorption peaks of synthesized CdS, ZnFe 2 O 4 , and NiFe 2 O 4 , were at 506 nm, 385 nm and 280 nm, respectively . The absorption edge of CdS was at ~770 nm, and those of ZnFe 2 O 4 , and NiFe 2 O 4 were at 410 nm and 400 nm, respectively. Cd S nanoparticles had the widest absorption range. ZnFe 2 O 4 and NiFe 2 O 4 had a relatively narrow visible light absorption range and relative steep edge. According to reported literature, CdS had a band gap of 2.28 2.4 eV under different ther mal annealing tem perature [22] , while ZnFe 2 O 4 , and NiFe 2 O 4 as spinel n - type semiconductor, had a comparatively low band gap energy as approximately 1.9 eV [23] and 1.7 eV [24] , separately. 113 The spectra also showed that the formulated CdS had excellent visible - light absorpt ion, from 210 nm to 800 nm. ZnFe 2 O 4 had a relatively narrow visible light absorption range and relative steep edge. ZnFe 2 O 4 nanoparticles were characterized as spinels, with tetrahedral and octahedral sites occupied by Zn 2+ and Fe 3+ cations [25] . According to Lv et al [26] , the transition of photoexcited electrons from the 2p level of O into 3d level of Fe can explain the absorption of visible light with ZnFe 2 O 4 , with the assumption that the O 2p orbital serves as the valence band and the Fe 3d orbital acts as the conduction band. For NiFe 2 O 4 , Ni 2+ and Fe 3+ also occupy the tetrahedral and octahedral sites [27] . (a) CdS Figure 4.3 UV vis absorption spectra of different photocatalysts (a) CdS (b) ZnFe 2 O 4 and (c) NiFe 2 O 4 114 (b) ZnFe 2 O 4 (c) NiFe 2 O 4 115 4.3. 4 Photocatalytic degradation of methylene blue test 4.3.3.1 Methylene blue removal efficiencies by different photocatalysts The efficacy of all three photocatalysts were evaluated using methylene blue as the target chemical. The absorption pea k of methylene blue is at 665 nm. The initial concentration of the methylene blue was controlled at 8 µmol/L. The amount of photocatalyst used was 1.0 g/L. Figure 4. 4 Degradation curves of methylene blue over different photocatalysts, initial methylene blue concentration of 8 µmol/L, with the amount of photocatalyst used controlled at 1.0 g/L. The kinetics of disappearance of MB by various photocatalysts are shown in Figure 4. 4. During illumination, methylene blue concentrations decreased rapidly initia lly; the rate decreased after 100 min. According to Xu [11] , MB degradation is initiated by cleavage of C S+=C bonds in MB. The decrease in reaction rate is consistent with the hypothesis -0.2 0 0.2 0.4 0.6 0.8 1 0 40 80 120 160 200 240 % removal Irradiation Time (min) CdS ZnFe O NiFe O 116 that as products are formed, they also absorb light, resulting in a reduction in the rate of photolysis of MB. Figure 4. 5 . Comparison of removal efficiency of different photocatalysts , initial concentration of MB at 8 µmol/L, at 25 ºC As Figure 4. 5 shows, the removal efficacies of methylene blue were: CdS > NiFe 2 O 4 ZnFe 2 O 4 . The reason for the greater photolysis rate with CdS is likely due to the wide absorption range of CdS. As such, during irradiation, CdS nanoparticles can utilize a wider range of photon energy than the other photocatalysts. Although ZnFe 2 O 4 has a slightly wider absorption range than NiFe 2 O 4 , the band gap of ZnFe 2 O 4 is greater than that of NiFe 2 O 4 , thus the electrons in the valence band would be harder to excite during irradiation. 4.3.3.2 Methylene blue removal efficiencies by different initial co ncentrations The effect of initial concentration was determined by varying the initial concentration of 0 10 20 30 40 50 60 70 80 CdS ZnFe O NiFe O 4 hour removal efficiency % 117 methylene blue between 2 and 20 µmol/L. The mass of ZnFe 2 O 4 photocatalyst remained constant at 1 g/L. Figure 4. 6 Different concentrations of methylene blue removal efficacy with ZnFe 2 O 4 under visible light , at 25 ºC As shown in figure 4. 6 , the efficacy of photodegradation decreased as methylene blue concentration increased. As the methylene blue concentrations were increased, the concentration of phot ocatalysts available to provide the electron loop necessary for oxidation became rate limiting. The lowest concentration (2 µmol/L MB) had the highest removal efficacy (31.51%), and it was twice that observed when the initial methylene blue concentration w as 20 µmol/L. 4.3.3.3 Other factors that affected methylene blue removal efficiencies The effect of reactor volume and illumination area was studied using two Pyrex glass 0 5 10 15 20 25 30 35 40 2 8 15 20 removal efficacy by ZnFe O (%) Methylene blue start concentrations (µmol/L) 118 reactors with different diameters (9.8 cm outside diameter (OD) and 8.6 cm OD). Th e initial methylene blue and CdS concentrations were 8 µmol/L and 1 g/L CdS, respectively. All other conditions were the same. The photoreactors were illuminated with continuous stirring for 4 hours. A s figure 4. 7 (a) shows , the removal efficiency was sign ificantly greater in the reactor with a smaller diameter . Moreover, different amounts of CdS (1 g/L and 0.5 g/L) and equal mixed of two kinds of nanoparticles (1 g/L of equal mixed of CdS, ZnFe 2 O 4 , NiFe 2 O 4 ), separately, were in the same reactor under visib le light illumination with continuous stirring for 4 hours (batch system), shown in figure 4. 7 (b). ( a) Figure 4.7 Methylene blue removal efficacy with different (a) reactors and (b) amounts of nanoparticles under visible light, initial MB concentration of 8µmol/L, at 25 ºC 0 20 40 60 80 100 Reactor 1 (OD = 9.8 cm) Reactor 2 (OD = 8.6 cm) Methylene blue Removal efficacy by CdS (%) 119 ( b) A s figure 4. 7 (a) shows, reactor with a smaller diameter had a higher removal efficiency (~23% more). The reason for this could be that light would be easier to transmit through the reactor and there would be higher light intensity at the back of the reactor. Figure 4. 7 (b) showed that for CdS, 0.5 g/L nanoparticles had higher removal efficiency of methylene blue under visible light compared to 1 g/L of same nanoparticles. The photodegradation activiti es were affected availability of active sites on the catalyst surface and the penetration of light transmitted [28] . At 1 g/L CdS suspension, although the availability of active sites increased, there is also an increase in the turbidity and a decrease in photon penetration as a result of increased scattering effect, so the photodegradation activity decreased [29] . The initial objective of mixing two kinds of nanoparticles together and ran the tests was check if it would increase the methylene blue removal efficiency, as all nanoparticles had different absorption range, and the photo excitation of one kind of 0 10 20 30 40 50 60 70 80 90 100 Removal Efficiency (%) 120 nanoparticles might produce current and trigger the nanoparticles near them. However, as figure 4. 7 (b) shows, the trend with mixed nanoparticles was no t clear in this study. 4.3. 5 Methylene blue degradation kinetics by CdS Figure 4. 8 Degrada tion curve of methylene blue with CdS 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 50 100 150 200 250 1 - Ct/C0 Time (min) 121 (a) (b) F igure 4. 9 Kinetics plots for methylene blue degradation by CdS : (a) Zero - order plo ts for first 10 min methylene blue degradation ; (b) Pseudo - first - order plots for 10 min 4 h methylene blue degradation 0 0.2 0.4 0.6 0.8 1 0 2 4 6 8 10 Ct/C0 Time (min) zero order (0 - 10min) -3 -2.5 -2 -1.5 -1 -0.5 0 0 40 80 120 160 200 240 ln(Ct/Co) Time (min) 1st order (10 - 240 min) 122 4.3. 6 Methylene blue de gradation mechanism by CdS 123 Effect of radical scavengers on the degradation of MB with CdS 0 0.2 0.4 0.6 0.8 1 0 40 80 120 160 200 240 Ct/Co Time (min) MB MB with t-butanol t-butanol 124 4.3. 7 Photo - deactivation of CdS F igure 4.1 1 Methylene blue degradation over time by continued treatment using CdS F igure 4.1 2 First - order plots of 2 nd illumination treatment with CdS 0 0.2 0.4 0.6 0.8 1 1.2 0 100 200 300 400 500 Ct/Co Time (min) Dark control and experiment 1st and 2nd illumination treatment -0.9 -0.8 -0.7 -0.6 -0.5 -0.4 -0.3 -0.2 -0.1 0 0 40 80 120 160 200 240 ln (Ct/Co) Time (min) 125 4.3. 8 SEM and EDS of used CdS The SEM images of CdS before and after 4 hours illumination treatment are shown in figure 4.1 3 . As there is vacuumed dry after treatment, the photocatalysts became flat sheeted and formed cracks on the surface , and the particle sizes got larger (diameters range 0.45 - . That might be the reason for reduced removal efficiency. For both before and after, the particles were intended to accumulate together and form larger particles. 126 (a) CdS before photo - treatm ent (b) C dS after photo - treatment Figure 4.1 3 SEM images of CdS before and after photo - treatment 127 (a) EDS images of used CdS and NiFe 2 O 4 (b) E lement analysis of spectrum 1 F igure 4.14 EDS analysis of used CdS and NiFe 2 O 4 128 F igure 4.14 (c) E lement analys is of spectrum 2 (d ) E lement analysis of spectrum 5 129 F igure 4.14 showed the EDS images of used CdS and NiFe 2 O 4 catalysts. The image showed that there were molecules attached to the surface of photocatalysts, based on element analysis, it should be me thylene blue degradation products as there was carbon element detected in spectrum. This also proved that one mechanism way of photodegradation was that molecules attached to surface of catalysts. 4.4 Conclusions In this study, CdS, ZnFe 2 O 4 , and NiFe 2 O 4 m aterials were synthesized. Methylene blue could be successfully decolorized and degraded by all three photocatalysts under visible light at room temperature. The absorption, nanoparticle surface and degradation characteristics of the nanoparticles were stu died. All the synthesis nanoparticles had visible light absorption, while CdS had a great absorption range within invisible light spectra. T he order of photocatalysis degradation efficacies of methylene blue during illumination was CdS > NiFe 2 O 4 2 O 4 , the reason could be that the absorption spectra results showed that CdS had the widest absorption range, compared to NiFe 2 O 4 and ZnFe 2 O 4 . It was indicated that the methylene blue removal efficacy could be affected by absorption range of photocatalysts, in itial dye concentrations, amount of photocatalysts addition, and photoreactor conditions. The kinetics of degradation of MB by CdS prese nted a two - stage reaction. Removal efficiency of methylene blue decreased as methylene blue concentration went up, so wh en designing photodegradation application, it would not only to consider which photocatalyst to be chosen, but also be an important thing to figure out the feed water contaminants concentrations, shapes and sizes of photoreactor, and amounts of nanoparticl es added. 130 REFERENCES 131 REFERENCES [1] Weber, E. J., and Stickney, V. C.. Hydrolysis kinetics of reactive blue 19 - vinyl sulfone. Water Research 27.1 (1993): 63 - 67. [2] Janus, M., et al. New method of improving photocatalytic activity of commercial Degussa P25 for azo dyes decomposition. Applied Catalysis B: Environmental 75.1 - 2 (2007): 118 - 123. [3] Galindo, C., et al. Photooxidation of the phenylazonaphthol AO20 on TiO2: kinetic and mechanistic investigations. Chemosphere 45.6 - 7 (2001) : 997 - 1005. [4] Slokar, Y. M., et al. Methods of decoloration of textile wastewaters. Dyes and pigments 37.4 (1998): 335 - 356. [5] Liu, H., et al. Production of electricity during waste - water treatment using a single chamber microbial fuel cell. Environ Sci Technol, 38 (2004), pp. 2281 - 2285. [6] Sun, L., et al. Enhanced photodegradation of methyl orange synergistically by microcrystal facet cutting and flexible electrically - conducting channels. The Journal of Physical Chemistry C 118.48 (2014): 28063 - 28068. [7] Chen, X., et al. Semiconductor - based photocatalytic hydrogen generation. Chemical reviews 110.11 (2010): 6503 - 6570. [8] Konstantinou, I. K., et al. TiO2 - assisted photocatalytic degradation of azo dyes in aqueous solution: kinetic and mechanistic invest igations: a review. Applied Catalysis B: Environmental 49.1 (2004): 1 - 14. [9] Pirhashemi, M., et al. Ultrasonic - assisted preparation of plasmonic ZnO/Ag/Ag2WO4 nanocomposites with high visible - light photocatalytic performance for degradation of organic pol lutants. Journal of colloid and interface science 491 (2017): 216 - 229. [10] Daneshvar, N., et al. Photocatalytic degradation of azo dye acid red 14 in water on ZnO as an alternative catalyst to TiO 2 . Journal of photochemistry and photobiology A: chemistry 162.2 - 3 (2004): 317 - 322. [11] Houas, A., et al. Photocatalytic degradation pathway of methylene blue in water. Applied Catalysis B: Environmental 31.2 (2001): 145 - 157. [12] Xu, N., et al. Effects of particle size of TiO 2 on photocatalytic degradation of m ethylene blue in aqueous suspensions. Industrial & Engineering Chemistry Research 38.2 (1999): 373 - 379. [13] Akpan, U. G., et al. Parameters affecting the photocatalytic degradation of dyes using TiO 2 - based photocatalysts: a review. Journal of hazardous ma terials 170.2 - 3 132 (2009): 520 - 529. [14] Soltani, N., et al. Visible light - induced degradation of methylene blue in the presence of photocatalytic ZnS and CdS nanoparticles. International journal of molecular sciences 13.10 (2012): 12242 - 12258. [15] Yan, H., et al. Visible - light - driven hydrogen production with extremely high quantum efficiency on Pt PdS/CdS photocatalyst. Journal of Catalysis 266.2 (2009): 165 - 168. [16] Casbeer, E., V.K. Sharma, and X.Z. Li, Synthesis and photocatalytic activity of ferrites un der visible light: A review. Separation and Purification Technology, 2012. 87: 1 - 14. [17] Yuan, Z.H. and L.D. Zhang, Synthesis, characterization and photocatalytic activity of ZnFe 2 O 4 /TiO 2 nanocomposite. Journal of Materials Chemistry, 2001. 11: 1265 - 1268 . [18] Sharma, R., et al. Structural, magnetic and electrical properties of zinc doped nickel ferrite and their application in photo catalytic degradation of methylene blue. Physica B: Condensed Matter 414 (2013): 83 - 90. [19] Zhang, L., Y.M. He, P. Ye, Y. Wu, and T.H. Wu, Visible light photocatalytic activities of ZnFe 2 O 4 loaded by Ag 3 VO 4 heterojunction composites. Journal of Alloys and Compounds. 549 (2013): 105 - 113. [20] Jourvand, M., et al. Removal of methylene blue from aqueous solutions using modifie d clay. Journal of Basic Research in Medical Sciences 2.1 (2015): 32 - 41. [21] Zhou, Jun, et al. Low Temperature Synthesis of NiFe 2 O 4 by a Hydrothermal Method. Journal of the American Ceramic Society. 88.12 (2005): 3535 - 3537. [22] Zelaya Angel, O., et al. Band gap shift in CdS semiconductor by photoacoustic spectroscopy: Evidence of a cubic to hexagonal lattice transition. Applied Physics Letters 64.3 (1994): 291 - 293. [23] Borse, P.H., J.S. Jang, S.J. Hong, J.S. Lee, J.H. Jung, T.E. Hong, C.W. Ahn, E.D. Jeong, K.S. Hong, J.H. Yoon, and H.G. Kim, Photocatalytic hydrogen generation from water - methanol mixtu res using nanocrystalline ZnFe 2 O 4 under visible light irradiation. Journal of the Korean Physical Society, 2009. 55: 1472 - 1477. [24] Hong, D., Y. Yamada, M. Sheehan, S. Shikano, C. - H. Kuo, T. Ming, C. - K. Tsung and S. Fukuzumi,. Mesoporous nickel ferrites w ith spinel structure prepared by an aerosol - spray - pryrolysis method for photocatalytic hydrogen evolution. ACS Sustainable Chemistry and Engineering, 2014. [25] Botta, P. M., E. F. Aglietti, and JM Porto López. Kinetic study of ZnFe 2 O 4 formation from mecha nochemically activated Zn Fe 2 O 3 mixtures. Materials research bulletin 41.4 (2006): 714 - 723. 133 [26] Lv, Hongjin, et al. Synthesis of floriated ZnFe 2 O 4 with porous nanorod structures and its photocatalytic hydrogen production under visible light. Journal of Ma terials Chemistry 20.18 (2010): 3665 - 3672. [27] Ceylan, Abdullah, et al. Solid state reaction synthesis of NiFe 2 O 4 nanoparticles. Journal of Magnetism and Magnetic Materials 320.6 (2008): 857 - 863. [ 28] Goncalves, Maria ST, et al. Photochemical treatment of solutions of azo dyes containing TiO 2 . Chemosphere 39.5 (1999): 781 - 786. [ by TiO 2 powders. Chemical Engineering & Technology: Industrial Chemistry Plant Equipment Process Engineering Biotechnology 24.6 (2001): 645 - 649. [ 30] Xu, Lejin, and Jianl ong Wang. A heterogeneous Fenton - like system with nanoparticulate zero - valent iron for removal of 4 - chloro - 3 - methyl phenol. Journal of hazardous materials 186.1 (2011): 256 - 264. [ 31] Houas, Ammar, Hinda Lachheb, Mohamed Ksibi, Elimame Elaloui, Chantal Guil lard, and Jean - Marie Herrmann. "Photocatalytic degradation pathway of methylene blue in water." Applied Catalysis B: Environmental 31, no. 2 (2001): 145 - 157. [ 32] Zhang, Jun, Yupeng Zhang, Yinkai Lei, and Chunxu Pan. "Photocatalytic and degradation mechan isms of anatase TiO 2: a HRTEM study." Catalysis Science & Technology 1, no. 2 (2011): 273 - 278. 134 CHAPTER FIVE Innovative Water Reuse Process with Photocatalytic Reactors and Catalytic Ceramic Membrane Filtration Abstract Technically f easible and econ omical process es that produce safe, clean water is highly in demand. This chapter describes an innovative water reuse process that combines photocatalysis and ceramic membrane filtration . This proposed process replaces biological treatment with sunlight - ba sed photocatalysis to treat emerging contaminants in wastewater . This would reduce energy consumption and sludge production , and has the potential to generate hydrogen. A catalytic membrane filtration system was used in the water treatment process to remo ve turbidity, large molecules , and recover the catalysts. Experiments were conducted to determine the removal efficiency of caffeine using CdS and ZnFe 2 O 4 as photocatalysts in a batch system followed by manganese oxide coated ceramic membrane filtration. R esults showed that there was 20.7% caffeine removal by ZnFe 2 O 4 membrane system, and 23.3% for CdS membrane system. Turbidity of the water reach ed 0.16 NTU after membrane filtration for both photolytic systems . For future improvements, photodegradation could be combine d with an advanced oxidation process such as pre - ozonation or addition of hydrogen peroxide to improve the removal of caffeine and other emerging contaminants . 135 5 .1 Introduction E merging contaminants, including pharmaceuticals, personal car e products and endocrine disrupters, are persistent and recalcitrant [1] . These contaminants are largely extens ds ively and many are toxic . Conventional wastewater treatment is ineffective at removing many of the emerging contaminants , resulting in the rele ase of many of these compounds to the environment . Moreover, the risk of human exposure to most emerging contaminants is difficult to ascertain their ecotoxicological data are not available [2] . Therefore, appropriate and effective water reuse technology f or emerging contaminants is in high demand. After treatment, the reused water must meet water standards and demands. C affeine is widely used , commonly found in tea, coffee, drugs, etc . [3] It is observed in effluents from sewage treatment plants at concen [4] . C affeine residues have been even detected at remote locations far from human settlements [5] . Therefore, effective treatment processes for caffeine are necessary . Conventional biological treatment methods are in effective for rec alcitrant organic pollutants , such as organic dyes, emerging contaminants , and organic acids. Among the many existing wastewater treatment technologies and processes, oxidative degradation toolbox. Advanced oxidation processes (AOPs) have been used effectively for the degradation of many organic pollutants (e.g., see [ 6 ]) . Advanced oxidation processes can be described as aqueous phase oxidation methods based on the intermediacy of highly rea ctive species 136 in the mechanisms leading to the destruction of the target pollutant. One of the mainly used reactive species is hydroxyl radicals, used for the degradation of organic pollutants and yielding CO 2 and inorganic ions with high degradation effic iencies and sludge - free operation [6] . Photodegradation can also be used for contaminants oxidation in wastewater reuse process. Moreover, the need for energy efficient and sustainable processes have encouraged extensive research on solar energy utilizati on [ 7 , 8 ] . S emiconductor catalysts irradiated with light of an appropriate wavelength can be used to generate highly reactive transitory oxidative species (i.e. , OH, O 2 , and HO 2 ) for the mineralization of organic impurities and pollutants [ 7 , 8 ] . There are a range of strategies have been studied for the photocatalytic degradation of organic dyes using semiconductor photocatalysts [ 9 11 ] . This Fenton - like processes have been applied to and demonstrated with the degradation of a wide range of model pollutants, including RhB, methylene blue (MB), methyl orange (MO), nitrobenzene, 4 - chlorophenol, acid orange II, and nalidixic acid [ 12 - 15 ] . O rganic compounds commonly found in fermented biomass, such as organic acids are expected to significantly improve photocatalytic hydrogen production as compared to that in the absence of scavengers. C admium sulfide (CdS) has been extensively studied as a photocatalyst for hydrogen production [17 - 19] and photodegradation [20 - 22] . It has a direct band gap of about 2. 4 eV [ 16] and most of the absorption spectra is in the visible region. CdS has been used for the photocatalytic degradation of methylene blue [ 16, 2 3 ] with a relatively high removal efficacy ( see chapter four). 137 Studies showed that certain spinels semiconductors have excellent photocatalytic properties. They have band gap energies less than 2.2 eV and have good absorption in the visible light ra nge [24 - 26] . ZnFe 2 O 4 is one of several photoactive spinel semiconductors. It is advantageous over CdS as it contains earth abundant , non - toxic metals. Experimental work ( see chapter four) demonstrated that it degrades methylene blue , is stable, is environm entally friendly. Over the last decade , as membrane technology has improved and become more economical, the use of membrane filtration has increased rapidly [2 7 ] . The objective of membrane filtration is to remove microorganisms and other particles from wa ter. There are a number of advantages over conventional water treatment technologies in membrane process, including less sludge production, smaller footprint [2 8 ] , and the physical exclusion of microorganisms from the product water. However, membrane fouli ng continues to be one of the main factors limiting the application and increasing operational costs of membrane technology [2 9 ] . If membrane fouling can be well controlled, membrane filtration can be applied more generally . A dditionally, if the cost of me mbrane technology can be reduced, it is possible to choose it as one of the water reuse selections. Moreover, conventional wastewater treatment plant consumed significant amounts of energy . T his process could use sunlight as energy resource to oxidize or ganics and possibly produce hydrogen. Also, the photocatalysts could be recovered from surface of ceramic membrane for future reuse. In this study, the process was simulated in a laboratory scale to figure out if this process is effective for caffeine remo val. 138 5 .2 Materials and Methods 5 .2.1 Photocatalysts preparation and characterization S ynthesis procedures of CdS and ZnFe 2 O 4 were the same that described in chapter 4, follow ing hydrothermal method for CdS [30] and sol - gel method for ZnFe 2 O 4 [31] . Both Cd S and ZnFe 2 O 4 were investigated their UV - visible absorption abilities and scanning electron microscope images to determine their surface properties and particle size. 5 .2.2 Ceramic membrane filtration process set up The membrane system has been described in chapter 2. The virgin membrane used in this study was coated 20 times with manganese oxide. The grain size within the support layer and the filtration layer of the virgin membrane varied between 132 nm and 296 nm and between 1.05 and 6.64 nm . The membr ane module had a total filtering surface area of 131.9 cm 2 , an active length of 25 cm, and an external diameter of 10 mm. The grain size within the support layer and the filtration layer of the virgin membrane varied between 132 nm and 296 nm and between 1 .05 and 6.64 nm [32] . 5 .2. 3 Test of the designed water reuse process The water reuse treatment process included two processes, a batch system for photodegradation by either cadmium sulfide (CdS) or zinc ferr i te (ZnFe 2 O 4 ) for organic chemical oxidation an d ceramic membrane process . Photodegradation process was designed as an alternative to conventional secondary treatment to degrade caffein e . The c eramic membrane system achieved filtration and the recovery of the photocatalyst for reuse . 139 Figure 5.1 Sch ematic of the proposed water reuse process A Pyrex glass photoreactor containing caffeine about 10 mg/L was used as the photoreactor . A selected amount (0. 2 5 g) of photocatalyts (ZnFe 2 O 4 or CdS) were added in the container. The photoreactor was exposed to visible light by using 250 W Xe - lamp with continuous stirring for 4 hours. Water samples were collected at the beginning and end of illumination . After illumination, the water was pumped through a ceramic filter and collected outgoing water samples. At the end of experiment , the water samples were tested by UV - vis to determine the caffeine concentrations . 5. 3 . Results The removal efficiency of ZnFe 2 O 4 and Cds of the entire designed process is shown in figure 5.1. 140 (a) Caffeine treatment process using ZnF e 2 O 4 as photocatalysts (b) Caffeine treatment process using ZnFe 2 O 4 as photocatalysts F igure 5. 2 Caffeine removal efficiency by designed treatment process with no oxidation during membrane filtration 60 65 70 75 80 85 90 95 100 original 2 hour 3 hour 4 hour membrane filtration Caffeine % remaining ZnFe 2 O 4 60 65 70 75 80 85 90 95 100 0 1 2 3 4 after filtration Caffeine % remaining CdS 141 The results indicated that for ZnFe 2 O 4 , about 10% of ca ffeine was removed at the end of photodegradation, and total process remove d 20.73% caffeine. For CdS, 5% of caffeine was removed during photodegradation, and 23.3% at the end of process . For both systems, the turbidity reach ed about 0.16 NTU after membran e filtration. As a comparison, US EPA regulations require that the turbidity of drinking water be less than or equal to 0.3 NTUs in at least 95 percent of the samples in any month [33] . The photodegradation of caffeine was not very effective. The results s how that filtration removes more of the caffeine then does photocatalys is , however, the reported removal efficiencies might have been affected by the presence of nanoparticles in the solution, as caffeine concentration was detected by UV - vis spectra. Ozona tion has been shown to effectively degrade caffeine in water after 200 s with pH of 10 (90% removal) [34] . Pre - ozonation with light illumination is expected to achieve removal efficiency greater than this . Future research on this combination is recommended . 5.4 Conclusions This study focused on designing an innovative water reuse process using both photodegradation for caffeine oxidation and ceramic membrane filtration . In this designed water reuse process, wastewater from industries containing emerging contaminants could be treated onsite. During photodegradation, caffeine in wastewater was oxidized. Pre - oxidation was recommended to improve caffeine remove efficiency. The e ffluent from photoreactor was filtered through a 5 kDa ceramic membrane. The exper iment result s of this designed process showed that both CdS and ZnFe 2 O 4 could remove caffeine in feed water. C eramic membrane could effectively filter nanoparticles from water sample for 142 future water reuse and catalyst recovery. However, both removal proce ss were not very effective, with only about 20 % removal efficacy for caffeine . For future application, improvements in the photocatalysts are important to improve the organics removal efficiency and achieve a safe and highly efficient water reuse process and save considerable amounts of energy and money . 143 REFERENCES 144 REFERENCES [1] Miranda - García, N., et al. Photocatalytic degradation of emerging contaminants in municipal wastewater treatment pl a nt effluents using immobil ized TiO2 in a solar pilot plant. Applied Catalysis B: Environmental 103.3 - 4 (2011): 294 - 301. à Barcel ó . Analysis and removal of emerging contaminants in wastewater and drinking water. TrAC Trends in Analytical Chemistry 22.10 (2003): 685 - 696. [3] Seiler, Ralph L., Steven D. Zaugg, James M. Thomas, and D arcy L. Howcroft. "Caffeine and pharmaceuticals as indicators of waste water contamination in wells." Groundwater 37, no. 3 (1999): 405 - 410. [4] Deblonde, Tiphanie, Carole Cossu - Leguille, and Philippe Hartemann. Emerging pollutants in wastewater: a review of the literature. International journal of hygiene and environmental health 214, no. 6 (2011): 442 - 448. [5] Weigel, Stefan, Urs Berger, Einar Jensen, Roland Kallenborn, Hilde Thoresen, and Heinrich Hühnerfuss. Determination of selected pharmaceuticals and caffeine in sewage and seawater from Tromsø/Norway with emphasis on ibuprofen and its metabolites. Chemosphere 56, no. 6 (2004): 583 - 592. [6] Malato, Sixto, et al. Decontamination and disinfection of water by solar photocatalysis: recent overview and tren ds. Catalysis Today 147.1 (2009): 1 - 59. [7] Yu, Jiaguo, et al. Enhanced photocatalytic activity of bimodal mesoporous titania powders by C 60 modification. Dalton Transactions 40.25 (2011): 6635 - 6644. [8] Zhang, Nan, et al. Nanochemistry - derived Bi 2 WO 6 nanostructures: towards production of sustainable chemicals and fuels induced by visible light. Chemical Society Reviews 43.15 (2014): 5276 - 5287. [9] Ahmad, Akil, et al. Recent advances in new generation dye removal technologies: novel search for approache s to reprocess wastewater. RSC Advances 5.39 (2015): 30801 - 30818. [10] Zhang, Nan, Siqi Liu, and Yi - Jun Xu. Recent progress on metal core at semiconductor shell nanocomposites as a promising type of photocatalyst. Nanoscale 4.7 (2012): 2227 - 2238. [11] Qi , Lifang, Jiaguo Yu, and Mietek Jaroniec. Preparation and enhanced visible - light photocatalytic H 2 - production activity of CdS - sensitized Pt/TiO 2 nanosheets with exposed (001) facets. Physical Chemistry Chemical Physics 13.19 (2011): 8915 - 8923. [12] An, Jun jian, et al. Efficient visible light photo - Fenton - like degradation of organic 145 pollutants using in situ surface - modified BiFeO 3 as a catalyst. Journal of Environmental Sciences 25.6 (2013): 1213 - 1225. [13] Wang, Qian, Senlin Tian, and Ping Ning. Ferrocene - Catalyzed heterogeneous fenton - like degradation of methylene blue: influence of initial solution pH. Industrial & Engineering Chemistry Research 53.15 (2014): 6334 - 6340. [14] Xu, Zhihui, Jianru Liang, and Lixiang Zhou. Photo - Fenton - like degradation of azo dye methyl orange using synthetic ammonium and hydronium jarosite. Journal of Alloys and Compounds 546 (2013): 112 - 118. [15] ElShafei, Gamal MS, et al. Ultrasonic assisted - Fenton - like degradation of nitrobenzene at neutral pH using nanosized oxides of Fe and Cu. Ultrasonics sonochemistry 21.4 (2014): 1358 - 1365. [16] Soltani, Nayereh, Elias Saion, W. Mahmood Mat Yunus, Manizheh Navasery, Ghazaleh Bahmanrokh, Maryam Erfani, Mohammad Reza Zare, and Elham Gharibshahi. Photocatalytic degradation of methylene blue under visible light using PVP - capped ZnS and CdS nanoparticles. Solar Energy 97 (2013): 147 - 154. [17] Li, Qin, Beidou Guo, Jiaguo Yu, Jingrun Ran, Baohong Zhang, Huijuan Yan, and Jian Ru Gong. Highly efficient visible - light - driven photocatalytic hydro gen production of CdS - cluster - decorated graphene nanosheets. Journal of the American Chemical Society 133, no. 28 (2011): 10878 - 10884. [18] Matsumura, Michio, Yukinari Saho, and Hiroshi Tsubomura. Photocatalytic hydrogen production from solutions of sulfit e using platinized cadmium sulfide powder. The Journal of Physical Chemistry 87, no. 20 (1983): 3807 - 3808. [19] Kakuta, N., K. H. Park, M. F. Finlayson, A. Ueno, A. J. Bard, A. Campion, M. A. Fox, S. E. Webber, and JM D. White. Photoassisted hydrogen produ ction using visible light and coprecipitated zinc sulfide. cntdot. cadmium sulfide without a noble metal. The Journal of Physical Chemistry 89, no. 5 (1985): 732 - 734. [20] Li, Qin, Xin Li, S. Wageh, Ahmed Al Ghamdi, and Jiaguo Yu. CdS/graphene nanocomposit e photocatalysts. Advanced Energy Materials 5, no. 14 (2015). [21] Gao, Peng, Jincheng Liu, Tong Zhang, Darren Delai Sun, and Wunjern Ng. - antibacterial capability under visible l ight irradiation. Journal of hazardous materials 229 (2012): 209 - 216. [22] Chong, Meng Nan, Bo Jin, Christopher WK Chow, and Chris Saint. Recent developments in photocatalytic water treatment technology: a review. Water research 44, no. 10 (2010): 2997 - 302 7. [23] Soltani, Nayereh, Elias Saion, Mohd Zobir Hussein, Maryam Erfani, Alam Abedini, 146 Ghazaleh Bahmanrokh, Manizheh Navasery, and Parisa Vaziri. Visible light - induced degradation of methylene blue in the presence of photocatalytic ZnS and CdS nanoparticl es. International journal of molecular sciences 13, no. 10 (2012): 12242 - 12258. [24] Yuan, Z.H. and L.D. Zhang, Synthesis, characterization and photocatalytic activity of ZnFe2O4/TiO2 nanocomposite. Journal of Materials Chemistry, 2001. 11: 1265 - 1268. [25 ] Zhang, L., Y.M. He, P. Ye, Y. Wu, and T.H. Wu, Visible light photocatalytic activities of ZnFe2O4 loaded by Ag3VO4 heterojunction composites. Journal of Alloys and Compounds, 2013. 549: 105 - 113. [26] Dom, R., R. Subasri, N.Y. Hebalkar, A.S. Chary, and P .H. Borse, Synthesis of a hydrogen producing nanocrystalline ZnFe 2 O 4 visible light photocatalyst using a rapid microwave irradiation method. Rsc Advances, 2012. 2: 12782 - 12791. [27] T. Leiknes. The effect of coupling coagulation and flocculation with memb rane filtration in water treatment: A review. Journal of Environmental Sciences 21 (2009), pp. 8 - 12. [28] R.D. Reardon. Membrane Treatment of Secondary Effluent for Subsequent use. Water Intelligence Online 6 (2007). [29] A. Braghetta, F.A. DiGiano, W. P. Ball. NOM accumulation at NF membrane surface: impact of chemistry and shear. J. Environ. Environ. Eng. (ACSE) 124(1998), pp. 1087 1098. [30] Yan, H., et al. Visible - light - driven hydrogen production with extremely high quantum efficiency on Pt PdS/CdS ph otocatalyst. Journal of Catalysis 266.2 (2009): 165 - 168. [31] Zhang, L., Y.M. He, P. Ye, Y. Wu, and T.H. Wu, Visible light photocatalytic activities of ZnFe 2 O 4 loaded by Ag 3 VO 4 heterojunction composites. Journal of Alloys and Compounds. 549 (2013): 105 - 113 . [32] Davies, S. H., et al. Fabrication of Catalytic Ceramic Membranes for Water Filtration. Water Science & Technology: Water Supply 10.1 (2010): 81 - 6. [33] National Primary Drinking Water Regulations. US EPA. March 22, 2018. Accessed June 06, 2018. htt ps://www.epa.gov/ground - water - and - drinking - water/national - primary - drinking - water - regulations#one [ 34] Rosal, Roberto, et al. Degradation of caffeine and identification of the transformation products generated by ozonation. Chemosphere 74.6 (2009): 825 - 831. 147 CHAPTER SIX Conclusions and Future Perspective This research focused on photocatalysis and catalytic ceramic membrane filtration in water and wastewater treatment application . The objectives were to use photodegradation to efficiently oxidize the org anics in the wastewater using sunlight as energy source, and to reduce membrane fouling and analyze the energy cost savings in manganese oxide coated catalytic ceramic membrane filtration combined with ozone injection to replace regular membrane filtration . In ceramic membrane filtration ozonation system, membrane fouling was effectively controlled at ozone dosages of 10 µg/s or greater for the manganese oxide coated membrane . With the virgin membrane, an ozone dosage of at least 15 µg/s was necessary to control membrane fouling. The overall resistance of the manganese oxide membrane was less with manganese oxide coated membrane under same ozone dosage and operation mode compared to that observed for the virgin titania membrane. Using the flux data obtain ed, it was determined that for the water studied the most energy - - end operation, with the manganese oxide coated membrane. Based on life - cycle - assessment, pressurization resulted in the majority of energy consumed (62%) in ca talytic ceramic membrane filtration with ozonation . Energy consumption for pressurization plus ozone generation consumed 93% of total energy input. Acidification, carcinogenicity and ecotoxicity of pollutants, smog, and climate 148 change in catalytic ceramic membrane filtration processes had higher environmental impacts in normalization. Ozone injection resulted in 43% of environmental impacts while pressurization with 49%. Catalytic ceramic membrane filtration resulted in less energy consumption of pressuriza tion and backwashing as compared with hollow fiber membrane filtration. Catalytic ceramic membrane filtration had a slightly lower environmental impacts than hollow fiber membrane filtration. Energy consumption for ozone generation underscores the need for research to develop more efficient methods to generate ozone and to optimize mass transfer. Alternatively, alternative and less energy consumptive methods to reduce membrane fouling would less both energy consumption and environmental impacts of cataly tic ceramic membrane filtration, making it clearly favorable over polymeric membrane filtration. T he order of photocatalysis degradation efficacies of methylene blue during illumination was CdS > NiFe 2 O 4 2 O 4 . CdS had a great absorption range within invisible light spectra , and it was one of the important factor that it had a good photodegradation abilities for methylene blue . It was indicated that the methylene blue removal efficacy could be affected by absorption range of photocatalysts, initial dye concentrations, amount of photocatalysts addition, and photoreactor conditions. Removal efficiency of methylene blue decreased as methylene blue concentration went up, so when designing photodegradation application, it would not only to consider which pho tocatalyst to be chosen, but also be an important thing to figure out the feed water contaminants concentrations, shapes and sizes of photoreactor, and amounts of nanoparticles added. 149 When it came to the whole water reuse process, b oth CdS and ZnFe 2 O 4 coul d remove caffeine in feed water. And ceramic membrane could effectively filtrate nanoparticles from water sample for future water reuse and catalysts recovery. However, both removal process is not very effective, at about 20 % removal efficacy. For future application, improvements and revisions for photocatalysts would be important to improve the organics removal efficiency. It is supposed to reach a higher caffeine remov al efficiency to inject ozone or hydrogen peroxide during photodegradation. Research have been conducted on hydrogen production by photocatalysts, future studies could focus on how to use sunlight degradation contaminants while at the same time produce hyd rogen as energy source. Due to equipment limitations, collection of hydrogen was not set up. But it is to achieve safe and high efficient water reuse process and save considerable amounts of energy and money.