1. Introduction
Photocatalytic oxidation, which is one of remarkable technologies of water pollution control, with the characteristics of high efficiency, low energy consumption and a wide range of application, can oxidize most organic compounds, especially non-biodegradable organic contaminants, by mineralizing them to small inorganic molecules. For this reason, photocatalytic oxidation technology has broad prospects for application. Among various semiconductor photocatalysts, there is a general consensus among researchers that TiO2 is more superior because of its high activity, large stability to light illumination, and low price[1-4]. In photocatalytic degradation, two modes of TiO2 application are adopted: (1) TiO2 immobilized on support materials, and (2) TiO2 suspended in aqueous medium[5,6]. Application of TiO2 in suspension instead of immobilizing the TiO2 on solid carriers has shown an improvement in organic degradation efficiencies due to the uniform distribution and large specific surface area. However, classical solid- liquid separation processes such as sedimentation, centrifugation and coagulation used for separation of the fine TiO2 particles (typically less than 1 μm), are not effective[7]. In addition to the low reutilization rate, there is also a chance of secondary pollution caused by fine TiO2 particles in the effluent. Therefore, the recovery of the photocatalysts is one of the main concerns that affect its engineering application on a large scale. A lot of investigations have been conducted aiming at solving this problem[8-12].
In recent 15 years, titanium microsphere has also been proposed as a means to recover TiO2 photocatalyst. Generally the core-shell structured TiO2 microspheres with a mesoporous surface made of nano-TiO2, have low density, high specific surface area and large size favorable for separation[13-16]. But the preparation method and operating conditions have great effect on particle morphology associated with photocatalysis, and many problems existing in the current preparation methods need to be solved by optimizing the process conditions or developing new ones[17].
Membrane separation process for separation and purification has been developed dramatically during the past few decades. It can simultaneously separate and concentrate all pollutants in water by the retention of its microspores without secondary pollution and phase change. In addition, with the advantage of low energy consumption, its equipment is compact, easy to operate and capable of continuous operation at room temperature[ 18]. However, membrane fouling due to the adsorption- precipitation of organic and inorganic compounds onto membranes leads to a decrease in the permeate flux, an increase in membrane cleaning costs and a reduction of the life of the membrane. Although considerable progress has been made in membrane fouling[19,20], techniques for controlling membrane fouling remain inadequate, which is the major obstacle in the successful implementation of membrane separation technology. The TiO2 photocatalysis-membrane separation coupling technology emerged recently can solve the two problems mentioned above effectively [21]. The coupling technology not only keeps the characteristics and capacity of the two technologies, but also produces some synergistic effects to overcome the drawbacks of the single technology. On the one hand, the pollutants are oxidized by the photocatalysis, and the selected membranes show the capability not only to retain the photocatalyst, but also to reject partially organic species by controlling the residence time in the reacting system. In other words, the membrane is also a selective barrier for the molecules to be degraded, thus the coupling technology could enhance the photocatalytic efficiency and achieve excellent effluent quality. On the other hand, the coupling of photocatalysis and membrane separation could solve or alleviate the problem of flux decline associated with membrane fouling[22].
In this review article, the researches for hybrid water/ wastewater treatment process of membrane and photocatalyst were summarized from various journal publications.
2. Membrane Photoreactor (MPR)
K. Azrague et al.[10] described the combination of a dialysis membrane and a photochemical reactor into a MPR to mineralize organic compounds. With this process, a pollutant model (2,4-dihydroxy benzoic acid) was mineralized from turbid waters in the photoreactor by using V-UV (vacuum ultra violet) irradiation (172 nm). This “advanced oxidation processes” produced high concentrations of hydroxyl radicals and was a good method to mineralize organic compounds. A model based on diffusion through the membrane and first order reaction in the reactor was in good agreement with the experimental data, in a wide range of operating conditions.
Irradiation was carried out in a DEMA (Mangels, Born-heim-Roisdorf, Germany) 13/12 Solidex glass annular photoreactor containing 350 cm3 of solution. K. Azrague et al.[10] used Xe-excimer light source, so-called high-pressure dielectric barrier discharge lamp. This quasi-monochromatic xenon excimer lamp is of cylindrical geometry, built of two concentric Suprasil quartz tubes with a total length of 25 cm and external diameter of 3 cm. It was positioned in the axis of the photoreactor and cooled water was circulated in the lamp jacked[10], as shown in Fig. 1.
A typical degradation curve is presented in Fig. 2 and corresponds to the degradation of the pollutant model in batch mode. The apparent constant of degradation (kapp) = 1.05 × 10-3 /s has been obtained from this curve (pseudo-first order) and this value will be used when modelling the degradation of 2,4-DHBA in the MPR[10].
Fig. 3 shows the membrane photoreactor system. The photoreactor was the same as for the batch irradiation experiments. A pump continuously re-circulated the turbid water from the feed tank tangentially to the membrane and an other pump continuously re-circulated the clear water through the photoreactor inside the hollow fibres. Recirculation rates are adjusted in such a way that no convection was observed during one experiment[ 10]. The membrane used in all experiments was made of polyacrilonitrile hollow-fibers (PAN 650 SF Asahi Medical) of 200 μm in diameter and 35 mm in length. The molecular weight cut-off was 11,800 g/mol and the module is 1.3 m2 in area[10].
To mimic turbid water, they chose a bentonite suspension, and obtained almost the same results as with clear water and the good modelling obtained showed that neither the transfer and kinetics parameters nor the effective membrane area A was noticeably changed in the presence of that level of clays. At this low concentration of bentonite the deposit between fibers is not very significant and then the MPR model works smoothly[10], as shown in Fig. 4.
3. Fouling Control of a Membrane Coupled Photocatalytic Process
Fouling in membrane coupled photocatalytic reactors was investigated in the case of greywater treatment by establishing the link between product type, dose, irradiation time and fouling rates in a cross flow membrane cell fitted with a 0.4 μm pore sized polyethylene membrane. Rapid fouling occurred only with shower gels and conditioners and was linked to changes in the organo-TiO2 aggregate size postulated to be caused by polymers within the products. Fouling was reduced to a negligible level when sufficient irradiation was applied demonstrating that the membrane component of the process is not the issue and that scale up and implementation of the process relates to effective design of the UV reactor[11].
Fouling rates of the high fouling systems were reduced to a level similar to those observed for the other systems after irradiation under UV light for 16 h, a time period previously observed to ensure complete irradiation. Similar results were observed for all three organo-TiO2 complexes that caused fouling[11], as shown in Fig. 5.
4. Photocatalytic Membrane Reactors for Degradation of Organic Pollutants
Different flowsheets (batch and continuous) of photocatalytic membrane reactors, to be used for degradation of organic pollutants present in water, together to some experimental results, are reported. 4-Nitrophenol (4NP) was used as a probe polluting agent and titanium dioxide in suspension was the catalyst. The photodegradation tests in the batch system were carried out without membrane changing the characteristic variables of the process (light intensity, TiO2 concentration, 4NP concentration, O2 concentration, pH) to find their optimum values. The batch system consisted of a water jacket thermostatted and stirred beaker irradiated from above with a UV-Vis lamp (light intensity on the surface of the liquid 4.4 mW/cm2). An empirical predictive equation was obtained describing the reaction rate as a function of the reported variables[12].
Photodegradation tests in the membrane reactors (total volume from 400 to 700 mL) were carried out coupling the batch to a re-circulation cell containing various types of flat sheet membranes which were able to retain the suspended catalyst and partially selective to the pollutant. The membranes were : NTR7410 and NTR7450 (Nitto Denko); N30F and NF-PES-010 (Hoechst); MPCB0000R98 (SEPAREM). The measured permeate flux was in the range 5-30 l/h m2 at 4 bar and all membranes showed both a rejection and a capacity to adsorb the pollutant with a transitory phase varying from 80 to 400 min at 4 bar. This behaviour could be a benefit for the process because oscillations in the pollutant concentration are not transmitted in the permeate. Three factors: rejection, photocatalytic degradation and adsorption were able to maintain the 4NP concentration in the permeate at very low values. For the continuous system, the lowest 4NP concentration in the permeate was 6-7% (w/w) of the initial 4NP concentration (40 mg/L) after a transient period of 300 min. Further improvements of this process are under investigation[12].
Concerning membrane photoreactors various flowsheet configurations were tested; the most interesting one was the continuous system presented in Fig. 6, where the catalyst is suspended in the aqueous solution and the irradiation is done on the recirculation reservoir[ 12].
The optimum choice of the ratio between the irradiated volume and the total volume (Vi/Vt) was important. When total suspension volume was increased from 400 to 700 mL (Fig. 7), the ratio Vi/Vt increased owing to a constant recycle volume and, consequently, the 4NP abatement was higher due to the increased percentage of irradiated with respect to recycled suspension[12].
5. Integration of Photocatalysis with Membrane Processes for Purification of Water
The aim of the presented work was the investigation on the possibility of application of the hybrid photocatalysis/ membrane processes system for removal of azo dyes (Acid Red 18, Direct Green 99 and Acid Yellow 36) from water. The photocatalytic reactions were conducted in the flow reactor with immobilized photocatalyst bed and in the suspended system integrated with ultrafiltration (UF). A commercially available titanium dioxide (Aeroxide P25, Degussa, Germany) was used as a photocatalyst. The solution after the photocatalytic reaction was applied as feed in nanofiltration (NF) or membrane distillation (MD). The changes of various parameters, including concentration of dyes, pH and conductivity of the solution, and TOC and TDS contents were analyzed during the process. It was found that the solutions containing the model azo dyes could be successfully decolorized during the photocatalytic processes applied in the studies. The application of UF process results in separation of photocatalyst from the treated solutions whereas during the NF and MD high retention of degradation products was obtained[23].
The experiments of the photocatalytic decomposition of dyes with application of the immobilized catalyst bed were conducted in the apparatus presented in Fig. 8. The main component of the system was the flow reactor with immobilized catalyst bed. The reactor was built of quartz tubes partly covered with TiO2 particles.
The photocatalytic reaction in the suspended system was conducted in the photocatalytic membrane reactor utilizing UF where the catalyst particles were retained in the feed solution by means of the membrane. The apparatus applied in the photocatalysis/UF experiments is presented in Fig. 9.
As can be seen from Fig. 10, the AY36 discoloration required the longest time of irradiation, which it took 30 h to obtain the complete color removal. For AR18 and DG99 the solutions were colorless after 21 and 25 h, respectively. Although the light fastness is the lowest for DG99, the highest rate of discoloration was observed for AR18. Acid Red 18 is a monoazodye so it is more easily degradable than DG99 which has five azo bonds in the structure. From the other hand, AY36 which is a monoazodye, likewise AR18, showed the lowest rate of discoloration. However, AY36 has the highest light fastness equal to 7 (in the 8-grade scale) and this factor can explain the longest time required for discoloration of AY36 solution[ 23].
6. Hybrid Photocatalysis and Ceramic Membrane Filtration Process for Humic acid Degradation
A hybrid photocatalysis with ceramic membrane filtration system is demonstrated for degradation of humic acids (HAs), which are typical refractory components of natural organic matter (NOM) present in water and wastewater. More specifically, the combination of chemical oxidation photocatalysis process with physical separation via a ceramic membrane filtration was explored. The effects of operating parameters such as transmembrane pressure and the membrane pore size on the permeate flux and organic removal was investigated. The interaction between the two solutes in the system, humic acids and TiO2 photocatalyst, played an important role in the observed flux decline during ceramic ultrafiltration (UF) and microfiltration (MF). Results showed that the MF membrane showed flux rates that were about 30% lower than the ones achieved with UF membranes. The dissolved organic carbon (DOC) removal was found to be higher in UF membrane (> 70%) compared to MF membrane (50%). Finally from the liquid chromatography (LC) analysis showed that after photocatalytic treatment, there is a change in the molecular weight distribution of the organic compounds and preferential adsorption of those compounds by TiO2 results in different fouling mechanisms in UF and MF membranes. It can be concluded that the use of ceramic membrane not only acts as a barrier in recovering the TiO2 photocatalyst but also assists in DOC reduction[24].
7. Effect of TiO2 Nanoparticles on Fouling Mitigation of Ultrafiltration Membranes for Activated Sludge Filtration
Membrane bioreactors (MBRs) have been widely used as advanced wastewater treatment process in recent years. However, MBR system has a membrane fouling problem, which makes the system less competitive. Thus there have been great efforts for fouling mitigation. In this study, two types of TiO2 immobilized ultrafiltration membranes (TiO2 entrapped and deposited membranes) were prepared and applied to activated sludge filtration in order to evaluate their fouling mitigation effect. Membrane performances were changed by addition of TiO2 nanoparticles to the casting solution. TiO2 entrapped membrane showed lower flux decline compared to that of neat polymeric membrane. Fouling mitigation effect increased with nanoparticle content, but it reached limit content above which fouling mitigation did not increase. Regardless of polymeric materials, membrane fouling was mitigated by TiO2 immobilization. TiO2 deposited membrane showed greater fouling mitigation effect compared to that of TiO2 entrapped membrane, since larger amount of nanoparticle was located on membrane surface. It can be concluded that TiO2 immobilized membranes are simple and powerful alternative for fouling mitigation in MBR application[25].
Flux decline behavior was measured using membrane cell and surface area was 18.1 cm2. Schematic diagram of the filtration system used in this study was shown in Fig. 11. Filtration tests were performed under the condition of 100 kPa at 20 ± 2°C. Crossflow velocity was controlled at 1.2 m/s and the flow rate was 2.5 L/min. Each set of experiment was performed simultaneously at the same condition. Feed tank (reactor) was consistently aerated during the filtration test for oxygen supply to biomass. Flux decline was measured by weighing permeate on top-loading balance at timed intervals[ 25].
The results of activated sludge filtration tests were shown in Fig. 12. In this study, filtration test was performed at the condition of high flux and low crossflow velocity. Furthermore, the MLSS concentration of mixed liquor was high (7,000 mg/L). Thus flux decline progressed rapidly at the beginning of filtration for all membranes. The result clearly showed that fouling of TiO2 entrapped membrane was significantly reduced compared to that of neat PSf membrane. It is well known that membrane fouling can be influenced by hydrodynamic condition, such as permeation drag and back transport, and chemical interaction between foulants and membranes. Since all the membranes were tested at the same hydrodynamic condition, the different fouling behavior means that surface property of membrane was changed by nanoparticle entrapment. Surface of TiO2 entrapped membrane can be more hydrophilic than that of neat polymeric membrane due to the higher affinity of metal oxides to water. Therefore, hydrophobic adsorption between sludge particle and TiO2 entrapped membrane was reduced, and deposited particles were readily removed by crossflow.
As shown in Fig. 11(b), with increase of TiO2 content in casting solution, fouling mitigation effect was gradually improved until TiO2/PSf ratio reached 0.3. However, beyond the ratio 0.3, flux decline behavior was not changed, although TiO2 content increased, as shown in Fig. 11(a). It means that surface property could not be improved beyond certain value. Furthermore, membrane performance and stability of casting solution can be poorer, since nanoparticles plug membrane pores and hinder the interaction between polymer and solvent molecules. Therefore, in case of TiO2 entrapped membrane, amount of TiO2 should be adjusted appropriately.
8. Hybrid Photocatalysis/Submerged Microfiltration Membrane System for Drinking Water Treatment
In this study, the potential of UV/TiO2 photocatalytic oxidation method to control of membrane fouling caused by natural organic matter (NOM) was investigated under various conditions in submerged MF membrane system. Effect of TiO2 concentration, UV irradiation in the absence of TiO2, TiO2 in the absence of UV irradiation and combination of UV/TiO2 photocatalytic oxidation were investigated. Additionally, intermittent and continuous UV application and initial NOM concentration on the pressure increase and rejections were also studied. The results of synthetic and raw water experiments were compared. It was found that TiO2 concentration is very important parameter by means of permeate pressure increase and removal efficiencies. UV irradiation in the absence of TiO2 or TiO2 in the absence of UV irradiation was not effective and combination of UV/TiO2 photocatalytic oxidation gave better results. Also, intermittent UV application was not as effective as UV/TiO2. The increase in NOM concentration also increased the pressure increase. Synthetic and raw water experiments were compared and raw water experiments gave higher pressure increase and lower removal efficiencies[26].
9. Purification of Bilge Water by Hybrid Ultrafiltration and Photocatalytic Processes
Investigations have been performed on the purification of bilge water by a combination of ultrafiltration and photocatalytic processes. The separation of oil from bilge water obtained from Szczecin-Swinoujscie harbor was performed on a laboratory-scale ultrafiltration pilot plant with tubular membranes made from poly(vinyl chloride) (PVC), polyacrylonitrile (PAN) and polyvinylidene fluoride (PVDF). The examined membranes with MWCO of 70 kD for PVC and PAN and 100 kD for PVDF produced a permeate with an oil content less than 15 ppm. Rejection of chemical oxygen demand (COD) was in the range 92-96% for the ultrafiltration treatment. The permeate is generally of acceptable quality for direct discharge into the Baltic Sea district in accordance with present legislation. Further treatment of UF permeate has been performed by heterogeneous photocatalytic oxidation in order to remove the residual oil. The photocatalytic process was carried out using titanium dioxide based catalyst. The complete decomposition of oil was achieved after 2 h of UV illumination using a K-TiO2 photocatalyst with content amounting to 0.8 g/dm3 and after 3 h of UV illumination using 0.8 g/dm3 of KOH/TiO2 photocatalyst. The photocatalytic process was found to be very effective for the permeate reclamation[ 27].
Three different tubular membranes made from PVC, PAN and PVDF were used for the determination of the water permeability and rejection characteristics for dextrans with different molecular weights. All ultrafiltration experiments were performed in a laboratory pilot plant, which is outlined in Fig. 13[27].
The materials obtained by the calcination method appeared to be active in the reaction of photocatalytic decomposition of oil in water. The oil decomposition degree depends on both the calcination temperature and the kind of metal used. The best results were obtained on KOH/TiO2 photocatalyst calcined at 550°C, see Fig. 14. The experiment shows the potential of the photocatalytic process as a means of purification of UF permeate to completely remove the oil[27].
10. Hybrid Water Treatment Process of Membrane and Photocatalyst-Coated Polypropylene bead
The effect of titanium dioxide (TiO2) photocatalystcoated polypropylene (PP) beads concentration on membrane fouling and treatment efficiency was observed in a hybrid process of tubular alumina microfiltration (MF) and the PP beads with periodic air back-flushing for advanced water treatment. And the results were compared with the previous study of the hybrid process of the same MF and the photocatalyst- loaded polyethersulfone (PES) beads. The optimum concentration of PP beads could be 40 g/L in the experimental range; however, that of PES beads was 20-30 g/L in the previous work to prevent membrane fouling efficiently. The treatment efficiency of turbidity was almost constant from 99.2% to 99.5%, independent of the PP beads concentration; however, that of dissolved organic matters (DOM) was the maximal 87.6% at 40 g/L of the PP beads. Treatment portions of MF, photocatalyst adsorption, and photo-oxidation were investigated by comparing the treatment efficiencies of (MF), (MF+TiO2) and (MF+TiO2+UV) processes. The role of photo-oxidation by the PP beads and UV was the more dominant than that of adsorption by the PP beads in the DOM treatment, which was the same trend in the previous work for the hybrid process of the same MF and the PES beads[28].
The advanced water treatment system utilizing a hybrid module of the tubular ceramic MF and TiO2 photocatalyst- coated PP beads was demonstrated in Fig. 15. The cross-flow filtration was performed inside the tubular ceramic membrane with periodic air back-flushing utilizing nitrogen gas for protecting oxygen effect on water quality. The hybrid module was supplied with the PP beads fluidizing between the gap of ceramic membrane and the acryl module case[28].
The resistances of membrane fouling (Rf) decreased dramatically as increasing the photocatalyst-coated PP beads concentration from 5 g/L to 40 g/L, as compared in Fig. 16. However, in the results using the hybrid process of the same MF membrane and the photocatalyst- loaded PES beads with periodic air back-flushing, Rf showed the minimum point between 20 g/L and 30 g/L of the PP beads. It means that the optimum concentration of PP beads could be 40 g/L in the experimental range; however, that of PES beads was 20-30 g/L to prevent membrane fouling efficiently on the surface and inside the ceramic membrane. The treatment efficiency of turbidity was very high and almost constant beyond 99.2%, independent of the PP beads concentration, as shown in Table 5. It means that the PP beads could not affect the treatment of suspended particles like kaolin in this hybrid process. However, in the hybrid process of the same MF and PES beads, the treatment efficiency of turbidity increased a little from 95.7% to 98.7%, which were a little lower efficiency than those of the PP beads, as increasing the PES beads concentration from 5 g/L to 40 g/L. It means that the PP beads were the more efficient to remove the turbid materials than the PES beads in the hybrid water treatment process[28].
For evaluating the role of membrane filtration, adsorption, and photo-oxidation in the hybrid water treatment process of the tubular ceramic MF and the photocatalyst- coated PP beads with the periodic air back-flushing, the process with the PP beads without UV light (MF+TiO2), and only MF without any PP beads and UV (MF) were performed at HA 6 mg/L, respectively. And then, they were compared with the hybrid process of MF and the PP beads with UV light (MF+TiO2+UV). As shown in Fig. 17, the resistance of membrane fouling (Rf) of MF, (MF+TiO2) and (MF+TiO2+UV) processes at HA 6 mg/L were compared during 180 min’s operation. However, the MF process could be operated until 90 min, because Rf increased rapidly and the permeate flux could not be measured after 90 min. The Rf could maintain low at the (MF+TiO2+UV) process during 180 min, and increased dramatically as simplifying the process to MF. It means that the photocatalyst adsorption and photo- oxidation by the PP beads and UV irradiation could control the membrane fouling effectively in this hybrid water treatment process.
In the treatment efficiency of DOM (UV254 absorbance), the treatment portion of membrane filtration was still very high 67.9%; however, that of adsorption was a little 6.1% and that of photo-oxidation was high 12.3% at HA 6 mg/L. It means that the role of photo- oxidation by the photocatalyst-coated PP beads and UV light was the more dominant than that of adsorption by the PP beads for the DOM treatment in this hybrid process. The photo-oxidation by the PP beads and UV could reduce the membrane fouling effectively, because the photo-oxidation had the major role of DOM reduction in this hybrid water treatment process with the PP beads. In addition, in the previous work[29] for the hybrid process of the same MF and the PES beads, the role of photo-oxidation was the more important than that of adsorption for DOM treatment, which was the exact same trend with this study.
11. Summary
This article included (1) membrane photoreactor (MPR), (2) fouling control of a membrane coupled photocatalytic process, (3) photocatalytic membrane reactors for degradation of organic pollutants, (4) integration of photocatalysis with membrane processes for purification of water, (5) hybrid photocatalysis and ceramic membrane filtration process for humic acid degradation, (6) effect of TiO2 nanoparticles on fouling mitigation of ultrafiltration membranes for activated sludge filtration, (7) hybrid photocatalysis/submerged microfiltration membrane system for drinking water treatment, (8) purification of bilge water by hybrid ultrafiltration and photocatalytic processes, and (9) Hybrid water treatment process of membrane and photocatalyst- coated polypropylene bead. Finally, hybrid water/wastewater treatment processes of membrane and photocatalyst were researched actively as a hot issues in the worlds.
