Nanostructured Ceramic Photocatalytic Membrane Modified with a Polymer Template for Textile Wastewater Treatment

Photocatalytic ceramic membranes have attracted considerable attention for industrial wastewater treatment. However, morphological control of the membrane surface to improve its photocatalytic reactivity for the degradation of organic pollutants remains a challenge. Herein, we report a new nanostructured TiO2/Al2O3 composite ceramic membrane prepared from a poly(oxyethylene methacrylate) (POEM) template through a sol–gel method and its photocatalytic performance in the treatment of a model dye compound. The POEM polymeric template allowed the homogeneous distribution of catalytic sites, i.e., the TiO2 layer, on the Al2O3 membrane surface, resulting in improved organic dye degradation along with effective fouling mitigation. The immobilization of a TiO2 layer on the Al2O3 membrane support also significantly enhanced the membrane adsorption capacity toward dye organic compounds. An organic removal efficiency of over 96% was achieved with the TiO2/Al2O3 composite membrane under Ultraviolet (UV) irradiation. In addition, the self-cleaning efficiency of the TiO2/Al2O3 composite membrane was remarkably improved by the degradation of organic foulants on the membrane under UV illumination.


Introduction
The demand for advanced water treatment technologies is increasing for the treatment of high-strength wastewater, such as industrial wastewater including complex water pollutants.Membrane technology is a promising way to enhance wastewater treatment efficiency as it can produce great effluent (permeate) quality with a smaller footprint than that afforded by conventional wastewater treatment processes [1][2][3].However, the polymeric membranes widely adopted in wastewater treatment applications display relatively low thermal and chemical resistance.There has been an upsurge of interest in ceramic membranes for the treatment of high-strength wastewater, for which polymeric membranes are not suitable [4,5].Ceramic membranes consisting of metal oxides usually exhibit high membrane porosity, membrane permeability, and narrow distribution of the membrane pore size.These unique properties allow for superior separation characteristics and antifouling behavior than those of polymeric membranes.
Photocatalytic membrane reactors (PMRs) combine photocatalytic reactors with membrane filtration units [6,7].As the photocatalytic membrane is combined with oxidation promoters, such as UV irradiation, the membrane plays an important role as a support for the photocatalytic materials and as a selective barrier to degrade organic contaminants.The TiO 2 materials are often considered as coating materials on ceramic support for the photocatalytic membrane because of their high photocatalytic activity and thermal/chemical stability at low environmental toxicity [8][9][10].The TiO 2 coating on the ceramic support has been reported to provide high water permeability with relatively low fouling rate to polymeric membrane [11,12].The photo-degradation activity of nanostructured ceramic membranes functionalized in situ have been found to be very effective for wastewater treatment [9].Thus, the photocatalytic layer on the membrane is expected to reduce the extent of membrane fouling while producing an excellent permeate quality due to the selective photocatalytic degradation of organic compounds on the catalytic membrane surface [13].In addition, the intrinsic properties of the ceramic membrane against certain chemical reagents allow for the photocatalytic membrane to withstand the concomitant attack of the strong oxidants produced during photocatalytic processes, e.g., hydroxyl radicals [14].
The hierarchically porous nanostructure of the catalyst film on the ceramic membrane plays a critical role in determining the photocatalytic membrane performance.The morphology of the photocatalyst film on the membrane support has been found to be an important factor determining the membrane surface reactivity [15].The introduction of surface-directing agents on the ceramic support has allowed the control of its porous structure [6,8,15,16].For example, the porous structure of a photocatalyst film was enhanced using a Pluronic block copolymer for high organic dye removal efficiency [16].Choi et al. [17] found that the mesoporous and narrow pore size distribution of photocatalyst films improve the wastewater treatment efficiency significantly.A grafted copolymer-based composite membrane exhibited excellent properties in terms of organic-compound degradation efficiency [15].Homopolymers can also be used with great potential to improve the photocatalytic membrane reactivity.The photocatalytic activity of TiO 2 composite membranes has been significantly enhanced by introducing hydrophobic homopolymers consisting of polyvinyl chloride (PVC) as a polymer template in the membrane support [6].Nevertheless, developing photocatalytic membranes with uniform distribution of catalytic materials on ceramic support with fine-tuned nanostructure is still major challenge.
Herein, we report the development of a nanostructured TiO 2 /Al 2 O 3 composite ceramic membrane using a poly(oxyethylene methacrylate) (POEM) template through a sol-gel method to improve the photocatalytic membrane reactivity for textile wastewater treatment application.The prepared membrane was characterized by scanning electron microscopy (SEM), atomic force microscopy (AFM), X-ray diffraction (XRD), and transmission electron microscopy (TEM).The adsorption behavior and antifouling performance were also systematically evaluated by filtering a model dye solution.

Membrane Fabrication
To prevent the penetration of the polymer/TiO 2 precursor mixture solution into the membrane pores, a PVP ethanol solution (10 wt %) was spin-coated on the bare Al 2 O 3 membrane at 500 rpm for 20 s and subsequently dried at 50 • C in an oven for the complete removal of ethanol.To prepare the TiO 2 precursor solution, HCl was slowly added to TTIP dropwise under vigorous stirring.Deionized water was added to the solution and stirred for at least 30 min.POEM (0.03 g) was dissolved in 1.5 mL of THF (2 wt %) and mixed with 0.2 mL of the as-prepared TiO 2 precursor solution over 3 h for the sol-gel process.The mixed polymer/TiO 2 precursor mixture solution was spin-coated on the PVP-coated Al 2 O 3 membrane, followed by calcination at 450 • C for 30 min.After calcination, the organic materials such as POEM, PVP, and the solvents were removed and a crystalline TiO 2 film was obtained on the Al 2 O 3 membrane.

Experimental Set-Up
The laboratory-scale, dead-end PMR is shown in Figure 1.The experimental set-up consisted of a membrane module integrated with a UV lamp, feed reservoir, and data acquisition system.A Congo red dye solution (100 mg•L −1 ) was used to simulate the dye wastewater.The feed solution was prepared by diluting a stock solution of 500 mg•L −1 .The test solution was placed in a 5 L feed reservoir under continuous stirring at a constant pressure of 1.5 bar applied by a pressurized oxygen tank to push the feed solution toward the membrane module equipped with a disk-type ceramic membrane (effective surface area of 4.5 cm 2 ).The aim of oxygen gas was to enhance the photocatalytic activity as it can act as an electron acceptor and thus can reduce the probability of electron hole recombination.A UV-C lamp (λ max = 254 nm, Philips TUV 4W SLM, Seoul, Korea) was installed inside the membrane module to irradiate UV light on membrane surface directly.The thickness of water channel between UV lamp and membrane surface was approximately 2 cm.The permeate water produced by the membrane module was collected in a permeate vessel and weighed on an electronic balance (Ohaus Corporation, Pine Brook, NJ, USA).The data acquisition system was used to monitor the permeate weight on the electronic balance with the filtration time.The permeate flux (LMH) (L m −2 h −1 ) was calculated by the permeate volume (L) divided by the effective membrane surface area (m 2 ) and filtration time (h).The concentration of Congo red in the membrane permeate was analyzed by measuring its absorbance (λ max ) at 496 nm using a UV-Visible spectrophotometer (SCINCO, S-3100, Seoul, Korea).Pure deionized water from an ultrapure water production system (STS-8l, Human Science, Seoul, Korea) was used in all the experimental work performed in this study.
for 20 s and subsequently dried at 50 °C in an oven for the complete removal of ethanol.To prepare the TiO2 precursor solution, HCl was slowly added to TTIP dropwise under vigorous stirring.Deionized water was added to the solution and stirred for at least 30 min.POEM (0.03 g) was dissolved in 1.5 mL of THF (2 wt.%) and mixed with 0.2 mL of the as-prepared TiO2 precursor solution over 3 h for the sol-gel process.The mixed polymer/TiO2 precursor mixture solution was spin-coated on the PVP-coated Al2O3 membrane, followed by calcination at 450 °C for 30 min.After calcination, the organic materials such as POEM, PVP, and the solvents were removed and a crystalline TiO2 film was obtained on the Al2O3 membrane.

Experimental Set-Up
The laboratory-scale, dead-end PMR is shown in Figure 1.The experimental set-up consisted of a membrane module integrated with a UV lamp, feed reservoir, and data acquisition system.A Congo red dye solution (100 mg L −1 ) was used to simulate the dye wastewater.The feed solution was prepared by diluting a stock solution of 500 mg L −1 .The test solution was placed in a 5 L feed reservoir under continuous stirring at a constant pressure of 1.5 bar applied by a pressurized oxygen tank to push the feed solution toward the membrane module equipped with a disk-type ceramic membrane (effective surface area of 4.5 cm 2 ).The aim of oxygen gas was to enhance the photocatalytic activity as it can act as an electron acceptor and thus can reduce the probability of electron hole recombination.A UV-C lamp (λmax = 254 nm, Philips TUV 4W SLM, Seoul, Korea) was installed inside the membrane module to irradiate UV light on membrane surface directly.The thickness of water channel between UV lamp and membrane surface was approximately 2 cm.The permeate water produced by the membrane module was collected in a permeate vessel and weighed on an electronic balance (Ohaus Corporation, Pine Brook, NJ, USA).The data acquisition system was used to monitor the permeate weight on the electronic balance with the filtration time.The permeate flux (LMH) (L m −2 h −1 ) was calculated by the permeate volume (L) divided by the effective membrane surface area (m 2 ) and filtration time (h).The concentration of Congo red in the membrane permeate was analyzed by measuring its absorbance (λmax) at 496 nm using a UV-Visible spectrophotometer (SCINCO, S-3100, Seoul, Korea).Pure deionized water from an ultrapure water production system (STS-8l, Human Science, Seoul, Korea) was used in all the experimental work performed in this study.

Photocatalytic Performance of the Membranes
The adsorption capacity of the bare Al 2 O 3 and TiO 2 /Al 2 O 3 composite membranes was examined by filtering a feed solution under dark conditions (i.e., without UV illumination).To evaluate the antifouling properties and photocatalytic activity of the membranes, filtrations were performed under UV illumination in dead-end filtration mode.All experimental works were carried out at least three times to confirm result reproducibility.The membranes were cleaned chemically after each experiment by placing them in a 1 mole/L NaOH solution (98% Beads, SAMCHUN Pure Chemicals Co., Ltd.Seoul, Korea) for 5 min and then in deionized water for 12 h.In order to know effect of temperature on membrane permeability, the permeability measured experimentally was compared with the one calculated by Darcy's equation below at elevated temperature in permeate line under UV radiation at the end of filtration.
where J represents membrane permeate flux (L m −2 h −1 ), ∆P is the trans-membrane pressure (bar), R m is the membrane resistance (m −1 ), and µ represents viscosity (Pa•s) of pure water at specific temperature.

Analytical Methods for Membrane Characterization
The surface and cross-section morphologies of the TiO 2 film on the Al 2 O 3 membrane and the bare Al 2 O 3 membrane were investigated by field emission SEM (FE-SEM, SUPRA 55VP, Carl Zeiss, Germany/FE-SEM, JSM-7001F, JEOL Ltd., Tokyo, Japan).The crystalline phase of the composite membranes was characterized by XRD (generator: 40 kV, 40 mA, D8 ADVANCE with DAVINCI, BRUKER, Bremen, Germany, wavelength (λ): Cu kα1 (Å)-1.5418Å, two theta range: 10-80 • ).The morphology and crystallinity of the TiO 2 structure detached from the fluorine-doped tin oxide (FTO) glass by scratching the slide glass were analyzed by high resolution-TEM (HR-TEM, JEM-3010, JEOL, Tokyo, Japan).The 3D surface morphology of the TiO 2 film on the Al 2 O 3 membrane was characterized by AFM (XE-Bio, Park Systems, Suwon, Korea).

Membrane Morphology
The nanostructure morphology of a membrane has a great effect on its photocatalytic performance for wastewater treatment.Polymer templates are considered effective structure-directing agents for the control of the nanostructure morphology.They not only effectively induce the organized structure of a membrane, but also have no negative impact on the membrane performance due to its complete degradation upon calcination.Since the hydrophilic TiO 2 precursor TTIP has high affinity toward hydrophilic materials, it interacts strongly with the hydrophilic POEM chains.A POEM/TTIP hybrid was thus successfully formed during the sol-gel process.Since the low-viscosity POEM/TTIP solution can easily penetrate the macroporous Al 2 O 3 membrane, its presence will influence the nanostructure morphology and, subsequently, the membrane performance.Our previous study showed that polymeric template/TTIP solution could be penetrated into the pore structure of macroporous Al 2 O 3 support and it caused non-uniform photocatalytic film on the ceramic support [15].With the help of a PVP coating on the Al 2 O 3 membrane [6,15], the penetration of the POEM/TTIP solution deep inside the membrane was prevented and a TiO 2 thin film was thereby successfully obtained.During calcination at 450 • C, TTIP was transformed into crystalline TiO 2 , while the organic POEM, PVP, and solvents were all completely decomposed.The synthesized TiO 2 film on the Al 2 O 3 membrane was found to be highly crystalline and well organized [15,18].
The surface of the TiO 2 structure on the FTO glass and Al 2 O 3 membrane was characterized by SEM analysis (Figure 2).A uniform TiO 2 film prepared with the aid of POEM was successfully synthesized on a flat FTO glass substrate.Owing to the nonporous substrate and strong interactions between TTIP and POEM, the TiO 2 film exhibits a well-organized morphology, as shown in Figure 2a.The bare membrane with a pore size of approximately 100 nm is clearly observed in Figure 2b.Large Al 2 O 3 crystals are also observed in the magnified surface image in Figure 2c.The morphology of the POEM-directed TiO 2 structure on the Al 2 O 3 membrane in Figure 2d is similar to that of the bare membrane.This is attributable to the formation of a very thin TiO 2 film resulting from the low-viscosity POEM/TTIP solution.In the enlarged image shown in Figure 2e, the Al 2 O 3 crystals appear well covered with uniform and well-connected TiO 2 nanoparticles consisting of the POEM/TTIP hybrid.
Appl.Sci.2017, 7, 1284 5 of 12 between TTIP and POEM, the TiO2 film exhibits a well-organized morphology, as shown in Figure 2a.The bare membrane with a pore size of approximately 100 nm is clearly observed in Figure 2b.Large Al2O3 crystals are also observed in the magnified surface image in Figure 2c.The morphology of the POEM-directed TiO2 structure on the Al2O3 membrane in Figure 2d is similar to that of the bare membrane.This is attributable to the formation of a very thin TiO2 film resulting from the low-viscosity POEM/TTIP solution.In the enlarged image shown in Figure 2e, the Al2O3 crystals appear well covered with uniform and well-connected TiO2 nanoparticles consisting of the POEM/TTIP hybrid.The distribution of TiO2 on Al2O3 membrane was investigated with EDS analysis (Figure 3).The Ti elemental mapping image of TiO2 structure prepared by the POEM template revealed that TiO2 was homogeneously distributed on the entire surface of Al2O3 membrane.The aggregation of photocatalysts on membrane should be prevented for effective photocatalysis and high permeate flux.It is widely known that the uniform distribution of photocatalyst without aggregation increases the surface area, thereby leading to the large number of active sites for photocatalysis.In addition, it is favorable for high permeate flux.Thus, the uniform distribution of TiO2 on membrane without severe aggregation might contribute to the enhanced performance of photocatalytic membrane.The cross-sectional SEM images in Figure 4a,b show the macroporous structure of the bare Al2O3 membrane.Figure 4c,d reveals that the cross-section of the TiO2-deposited Al2O3 membrane is not significantly different from that of the bare membrane as a result of the deposition of a very thin TiO2 film, as seen in the magnified image in Figure 4d.The distribution of TiO 2 on Al 2 O 3 membrane was investigated with EDS analysis (Figure 3).The Ti elemental mapping image of TiO 2 structure prepared by the POEM template revealed that TiO 2 was homogeneously distributed on the entire surface of Al 2 O 3 membrane.The aggregation of photocatalysts on membrane should be prevented for effective photocatalysis and high permeate flux.It is widely known that the uniform distribution of photocatalyst without aggregation increases the surface area, thereby leading to the large number of active sites for photocatalysis.In addition, it is favorable for high permeate flux.Thus, the uniform distribution of TiO 2 on membrane without severe aggregation might contribute to the enhanced performance of photocatalytic membrane.The cross-sectional SEM images in Figure 4a,b show the macroporous structure of the bare Al 2 O 3 membrane.Figure 4c,d reveals that the cross-section of the TiO 2 -deposited Al 2 O 3 membrane is not significantly different from that of the bare membrane as a result of the deposition of a very thin TiO 2 film, as seen in the magnified image in Figure 4d.TEM and HR-TEM analyses were performed to characterize the crystalline properties and morphology of the TiO2 film.The TEM image in Figure 5a confirmed that the POEM-templated TiO2 structure is composed of uniform and well-interconnected nanoparticles.The HR-TEM and selected area electron diffraction (SAED) pattern in Figure 5b reveal that the TiO2 structure is highly crystalline with a d-spacing value of approximately 0.35 nm, consistent with the (101) plane of the anatase TiO2 phase.In general, these results indicate that the POEM-assisted TiO2 structure is highly crystalline and well organized.TEM and HR-TEM analyses were performed to characterize the crystalline properties and morphology of the TiO2 film.The TEM image in Figure 5a confirmed that the POEM-templated TiO2 structure is composed of uniform and well-interconnected nanoparticles.The HR-TEM and selected area electron diffraction (SAED) pattern in Figure 5b reveal that the TiO2 structure is highly crystalline with a d-spacing value of approximately 0.35 nm, consistent with the (101) plane of the anatase TiO2 phase.In general, these results indicate that the POEM-assisted TiO2 structure is highly crystalline and well organized.TEM and HR-TEM analyses were performed to characterize the crystalline properties and morphology of the TiO 2 film.The TEM image in Figure 5a confirmed that the POEM-templated TiO 2 structure is composed of uniform and well-interconnected nanoparticles.The HR-TEM and selected area electron diffraction (SAED) pattern in Figure 5b reveal that the TiO 2 structure is highly crystalline with a d-spacing value of approximately 0.35 nm, consistent with the (101) plane of the anatase TiO 2 phase.In general, these results indicate that the POEM-assisted TiO 2 structure is highly crystalline and well organized.For detailed characterization of the crystalline properties, XRD analysis was conducted for the POEM-directed TiO2 film deposited on the Al2O3 membrane, as shown in Figure 6.The sharp diffraction peaks are attributed to the TiO2 and Al2O3 phases.The highly crystalline Al2O3 patterns marked in Figure 6 match the α-phase of alumina (Al2O3) (JCPDS Card No. 10-0173).The TiO2 peaks at 2θ = 25.5, 37.7, and 61.2° correspond to the (101), (004), and (115) planes of anatase TiO2 (JCPDS Card No. 21-1272).The anatase phase is widely known to be more effective for photocatalytic reactions than the rutile and brookite phases [9,19].From the XRD results, it was concluded that the organic species were completely removed and that a highly crystalline anatase TiO2 structure was obtained after calcination at 450 °C.The three-dimensional morphology and structure of the POEM-directed TiO2 film on the Al2O3 membrane and the bare Al2O3 membrane were investigated by AFM analysis.The topographical images and root-mean-square roughness (Rrms) of the membranes were obtained, as shown in Figure 7.The Rrms value for the bare Al2O3 membrane was 24.83 nm, resulting from its macroporous morphology.With the introduction of a TiO2 film templated by POEM, the value for the TiO2/Al2O3 ).The anatase phase is widely known to be more effective for photocatalytic reactions than the rutile and brookite phases [9,19].From the XRD results, it was concluded that the organic species were completely removed and that a highly crystalline anatase TiO 2 structure was obtained after calcination at 450 • C. For detailed characterization of the crystalline properties, XRD analysis was conducted for the POEM-directed TiO2 film deposited on the Al2O3 membrane, as shown in Figure 6.The sharp diffraction peaks are attributed to the TiO2 and Al2O3 phases.The highly crystalline Al2O3 patterns marked in Figure 6 match the α-phase of alumina (Al2O3) (JCPDS Card No. 10-0173).The TiO2 peaks at 2θ = 25.5, 37.7, and 61.2° correspond to the (101), (004), and (115) planes of anatase TiO2 (JCPDS Card No. 21-1272).The anatase phase is widely known to be more effective for photocatalytic reactions than the rutile and brookite phases [9,19].From the XRD results, it was concluded that the organic species were completely removed and that a highly crystalline anatase TiO2 structure was obtained after calcination at 450 °C.The three-dimensional morphology and structure of the POEM-directed TiO2 film on the Al2O3 membrane and the bare Al2O3 membrane were investigated by AFM analysis.The topographical images and root-mean-square roughness (Rrms) of the membranes were obtained, as shown in Figure 7.The Rrms value for the bare Al2O3 membrane was 24.83 nm, resulting from its macroporous morphology.With the introduction of a TiO2 film templated by POEM, the value for the TiO2/Al2O3

Pure Water Permeability and Antifouling Properties
Table 1 shows the permeability of the bare Al2O3 and TiO2/Al2O3 composite membranes for DI water after 2.5 h of filtration.A significant decline in the membrane permeability was observed after depositing a TiO2 layer on the Al2O3 membrane support.In the absence of UV illumination, the permeability of the Al2O3 membrane was 65.2 L m −2 h −1 bar −1 , which was reduced to 9.9 L m −2 h −1 bar −1 for the POEM-directed TiO2/Al2O3 composite membrane.Such a reduction of the membrane permeability is attributed to the increased membrane resistance against the water flow upon significant reduction of the membrane pore size [20,21].However, exposure to UV illumination enhanced the membrane permeability of both membranes.The permeability of the bare Al2O3 membrane and TiO2/Al2O3 composite membrane under UV illumination increased to 71.5 and 14.2 L m −2 h −1 bar −1 , respectively.Such a membrane permeability enhancement can be resulted from improvement of the membrane wettability or higher water temperature induced by the UV irradiation [15,22].The TiO2 coating layer can improve surface hydrophilicity under UV radiation due to the production of the hydroxyl groups on membrane [23].In addition, temperature increase can reduce water viscosity, thereby increasing membrane permeability [22].To distinguish this effect, membrane permeability measured experimentally was compared with the calculated one by Darcy's equation (Equation ( 1)) at elevated temperature under UV irradiation at the end of filtration.For the bare membrane under UV irradiation, there is no difference in permeability between measured and calculated value (71.4 vs. 71.5 L m −2 h −1 bar −1 ).This indicates that temperature played a dominant role in determining membrane permeability.For the TiO2/Al2O3 composite membrane, membrane permeability measured was 14.2 L m −2 h −1 bar −1 , but the one calculated at the temperature increased by UV irradiation was 10.5 L m −2 h −1 bar −1 .Considering membrane permeability measured with the composite membrane (9.9L m −2 h −1 bar −1 ), contribution of temperature increase to membrane permeability was relatively low compared to the bare membrane.Measurements of water temperature in the permeation line revealed a 5 ± 1 °C increase with bare Al2O3 membrane (about 31 °C).Increased water temperatures reduce the viscosity, thereby increasing the membrane permeability [22].For the TiO2/Al2O3 composite membrane, the water temperature increase was relatively small as about 3 °C under UV illumination (about 29 °C).A possible explanation is that the TiO2 film on the Al2O3 membrane can support absorbing the UV light and reduce heat dissipation [15,22].

Pure Water Permeability and Antifouling Properties
Table 1 shows the permeability of the bare Al 2 O 3 and TiO 2 /Al 2 O 3 composite membranes for DI water after 2.5 h of filtration.A significant decline in the membrane permeability was observed after depositing a TiO 2 layer on the Al 2 O 3 membrane support.In the absence of UV illumination, the permeability of the Al 2 O 3 membrane was 65.2 L m −2 h −1 bar −1 , which was reduced to 9.9 L m −2 h −1 bar −1 for the POEM-directed TiO 2 /Al 2 O 3 composite membrane.Such a reduction of the membrane permeability is attributed to the increased membrane resistance against the water flow upon significant reduction of the membrane pore size [20,21].However, exposure to UV illumination enhanced the membrane permeability of both membranes.The permeability of the bare Al 2 O 3 membrane and TiO 2 /Al 2 O 3 composite membrane under UV illumination increased to 71.5 and 14.2 L m −2 h −1 bar −1 , respectively.Such a membrane permeability enhancement can be resulted from improvement of the membrane wettability or higher water temperature induced by the UV irradiation [15,22].The TiO 2 coating layer can improve surface hydrophilicity under UV radiation due to the production of the hydroxyl groups on membrane [23].In addition, temperature increase can reduce water viscosity, thereby increasing membrane permeability [22].To distinguish this effect, membrane permeability measured experimentally was compared with the calculated one by Darcy's equation (Equation ( 1)) at elevated temperature under UV irradiation at the end of filtration.For the bare membrane under UV irradiation, there is no difference in permeability between measured and calculated value (71.4 vs. 71.5 L m −2 h −1 bar −1 ).This indicates that temperature played a dominant role in determining membrane permeability.For the TiO 2 /Al 2 O 3 composite membrane, membrane permeability measured was 14.2 L m −2 h −1 bar −1 , but the one calculated at the temperature increased by UV irradiation was 10.5 L m −2 h −1 bar −1 .Considering membrane permeability measured with the composite membrane (9.9L m −2 h −1 bar −1 ), contribution of temperature increase to membrane permeability was relatively low compared to the bare membrane.Measurements of water temperature in the permeation line revealed a 5 ± 1 • C increase with bare Al 2 O 3 membrane (about 31 • C).Increased water temperatures reduce the viscosity, thereby increasing the membrane permeability [22].For the TiO 2 /Al 2 O 3 composite membrane, the water temperature increase was relatively small as about 3 • C under UV illumination (about 29 • C).A possible explanation is that the TiO 2 film on the Al 2 O 3 membrane can support absorbing the UV light and reduce heat dissipation [15,22].The higher fouling rate in the bare Al 2 O 3 membrane results from the higher probability of pore fouling in the larger pores of the bare membrane than in the smaller pores of the TiO 2 /Al 2 O 3 composite membrane [23].The previous results also confirmed that the membrane permeability decreased significantly after immobilizing a TiO 2 layer on the Al 2 O 3 membrane support, as shown in Table 1.The benefits of applying the POEM template become obvious upon exposing the membranes to UV illumination.In fact, no flux decline was observed for the POEM-directed TiO 2 /Al 2 O 3 composite membrane under UV illumination.In contrast, severe membrane fouling was observed in the bare Al 2 O 3 membrane under UV illumination, with a ~51.2% decay in the permeate flux relative to that in the absence of UV illumination (2.57 vs. 1.25 LMH h −1 ).The higher fouling rate on the bare Al 2 O 3 membrane under UV illumination is attributed to the formation of a cake layer on the membrane surface [15].Although photolysis phenomena are likely involved during organic degradation on the bare Al 2 O 3 membrane [24,25], photolysis itself may not be sufficient to remove the cake layer on the membrane.Figure 8 compares the effect of the permeate flux on the membrane fouling for both the bare Al2O3 and TiO2/Al2O3 composite membranes in the absence and presence of UV illumination after filtration for 1 day.In the absence of UV illumination, the extent of the flux decline was lower for the TiO2/Al2O3 composite membrane than for the bare Al2O3 membrane.The flux decline was ~89.6% lower on the TiO2/Al2O3 composite membrane than on the bare Al2O3 membrane (1.25 vs. 0.13 LMH h −1 ).The higher fouling rate in the bare Al2O3 membrane results from the higher probability of pore fouling in the larger pores of the bare membrane than in the smaller pores of the TiO2/Al2O3 composite membrane [23].The previous results also confirmed that the membrane permeability decreased significantly after immobilizing a TiO2 layer on the Al2O3 membrane support, as shown in Table 1.The benefits of applying the POEM template become obvious upon exposing the membranes to UV illumination.In fact, no flux decline was observed for the POEM-directed TiO2/Al2O3 composite membrane under UV illumination.In contrast, severe membrane fouling was observed in the bare Al2O3 membrane under UV illumination, with a ~51.2% decay in the permeate flux relative to that in the absence of UV illumination (2.57 vs. 1.25 LMH h −1 ).The higher fouling rate on the bare Al2O3 membrane under UV illumination is attributed to the formation of a cake layer on the membrane surface [15].Although photolysis phenomena are likely involved during organic degradation on the bare Al2O3 membrane [24,25], photolysis itself may not be sufficient to remove the cake layer on the membrane.

Organic Dye Removal Efficiency
Figure 9 shows the dye adsorption curves for both the bare Al2O3 membrane and TiO2/Al2O3 composite membrane.In the absence of UV illumination, the TiO2/Al2O3 composite membrane shows a higher adsorption capacity than the bare Al2O3 membrane, which was almost saturated after 9 h of filtration.The TiO2/Al2O3 composite membrane presents a higher dye organic removal

Organic Dye Removal Efficiency
Figure 9 shows the dye adsorption curves for both the bare Al 2 O 3 membrane and TiO 2 /Al 2 O 3 composite membrane.In the absence of UV illumination, the TiO 2 /Al 2 O 3 composite membrane shows a higher adsorption capacity than the bare Al 2 O 3 membrane, which was almost saturated after 9 h of filtration.The TiO 2 /Al 2 O 3 composite membrane presents a higher dye organic removal efficiency (~20%) than the bare Al 2 O 3 membrane after the same filtration time.The higher adsorption capacity of the TiO 2 /Al 2 O 3 composite membrane arises mainly from complementary cavities and rebinding sites on the membrane assisted by the POEM template [26].The advantages of the presence of such a well-organized TiO 2 film were further observed upon carrying out the membrane filtration under UV illumination (Figure 9b).About 96% of dye removal efficiency was observed with the TiO 2 /Al 2 O 3 composite membrane after 24 h of filtration.The catalytic active sites offered by the POEM template favor the membrane photocatalytic activity.In contrast, only ~40% of the organic dye was removed by the bare Al 2 O 3 membrane under UV illumination [15].However, ~89% of organic removal efficiency was achieved after 6.5 h membrane filtration, probably due to photolysis phenomena occurring on the membrane surface [25].Moreover, the formed cake layer may act as a secondary membrane, thus improving the organic dye removal by membrane filtration [15].
Appl.Sci.2017, 7, 1284 10 of 12 adsorption capacity of the TiO2/Al2O3 composite membrane arises mainly from complementary cavities and rebinding sites on the membrane assisted by the POEM template [26].The advantages of the presence of such a well-organized TiO2 film were further observed upon carrying out the membrane filtration under UV illumination (Figure 9b).About 96% of dye removal efficiency was observed with the TiO2/Al2O3 composite membrane after 24 h of filtration.The catalytic active sites offered by the POEM template favor the membrane photocatalytic activity.In contrast, only ~40% of the organic dye was removed by the bare Al2O3 membrane under UV illumination [15].However, ~89% of organic removal efficiency was achieved after 6.5 h membrane filtration, probably due to photolysis phenomena occurring on the membrane surface [25].Moreover, the formed cake layer may act as a secondary membrane, thus improving the organic dye removal by membrane filtration [15].

Conclusions
A nanostructured photocatalytic TiO2/Al2O3 composite membrane was successfully fabricated using a POEM polymer template through a sol-gel process.The hydrophilic POEM chains interact strongly with the TiO2 precursor (TTIP) to form uniform POEM/TTIP hybrids.Upon calcination at 450 °C, TTIP is transformed into highly crystalline anatase TiO2 while the organic POEM is completely decomposed.The well-organized structure and morphology of the TiO2 film were confirmed by HR-TEM, SEM, XRD, and AFM analyses.A homogenous distribution of the catalytic sites on the membrane surface was achieved, exhibiting a higher adsorption capacity toward Congo red (a model organic dye) than that of the bare Al2O3 membrane.Under UV illumination, the TiO2/Al2O3 composite membrane showed 96% organic removal efficiency along with effective fouling mitigation.

Figure 1 .
Figure 1.Experimental setup of the photocatalytic membrane reactor (PMR) with direct ultraviolet (UV) irradiation on membrane surface.

Figure 1 .
Figure 1.Experimental setup of the photocatalytic membrane reactor (PMR) with direct ultraviolet (UV) irradiation on membrane surface.

Figure 2 .
Figure 2. Surface scanning electron microscopy (SEM) image of a (a) Poly(oxyethylene methacrylate) (POEM)-directed TiO 2 film on FTO glass; (b,c) the bare Al 2 O 3 membrane; and (d,e) the POEM-directed TiO 2 film on the Al 2 O 3 membrane.

Figure 3 .
Figure 3. (a) Surface SEM image and (b) the corresponding EDS elemental (Ti) mapping image of POEM-directed TiO2 films on Al2O3 membrane.

Figure 4 .
Figure 4. Cross sectional SEM images of the (a,b) bare Al2O3 membrane and (c,d) POEM-directed TiO2 film on the Al2O3 membrane.

Figure 3 . 12 Figure 3 .
Figure 3. (a) Surface SEM image and (b) the corresponding EDS elemental (Ti) mapping image of POEM-directed TiO 2 films on Al 2 O 3 membrane.

Figure 4 .
Figure 4. Cross sectional SEM images of the (a,b) bare Al2O3 membrane and (c,d) POEM-directed TiO2 film on the Al2O3 membrane.

Figure 4 .
Figure 4. Cross sectional SEM images of the (a,b) bare Al 2 O 3 membrane and (c,d) POEM-directed TiO 2 film on the Al 2 O 3 membrane.

Figure 6 .
Figure 6.X-ray diffraction (XRD) pattern of the POEM-directed TiO2 film on the Al2O3 membrane.

Figure 6 .
Figure 6.X-ray diffraction (XRD) pattern of the POEM-directed TiO2 film on the Al2O3 membrane.

Figure 8
Figure 8 compares the effect of the permeate flux on the membrane fouling for both the bare Al 2 O 3 and TiO 2 /Al 2 O 3 composite membranes in the absence and presence of UV illumination after filtration for 1 day.In the absence of UV illumination, the extent of the flux decline was lower for the TiO 2 /Al 2 O 3 composite membrane than for the bare Al 2 O 3 membrane.The flux decline was ~89.6% lower on the TiO 2 /Al 2 O 3 composite membrane than on the bare Al 2 O 3 membrane (1.25 vs. 0.13 LMH h −1 ).The higher fouling rate in the bare Al 2 O 3 membrane results from the higher probability of pore fouling in the larger pores of the bare membrane than in the smaller pores of the TiO 2 /Al 2 O 3 composite membrane[23].The previous results also confirmed that the membrane permeability decreased significantly after immobilizing a TiO 2 layer on the Al 2 O 3 membrane support, as shown in Table1.The benefits of applying the POEM template become obvious upon exposing the membranes to UV illumination.In fact, no flux decline was observed for the POEM-directed TiO 2 /Al 2 O 3 composite membrane under UV illumination.In contrast, severe membrane fouling was observed in the bare Al 2 O 3 membrane under UV illumination, with a ~51.2% decay in the permeate flux relative to that in the absence of UV illumination (2.57 vs. 1.25 LMH h −1 ).The higher fouling rate on the bare Al 2 O 3 membrane under UV illumination is attributed to the formation of a cake layer on the membrane surface[15].Although photolysis phenomena are likely involved during organic degradation on the bare Al 2 O 3 membrane [24,25], photolysis itself may not be sufficient to remove the cake layer on the membrane.

Figure 8 .
Figure 8.Comparison of the membrane fouling rate for the bare Al2O3 membrane and Poly(oxyethylene methacrylate) (POEM)-directed TiO2/Al2O3 composite membrane during filtration in the presence (With UV) and absence (Without UV) of ultraviolet (UV) illumination (filtration time: 24 h).

Figure 8 .
Figure 8.Comparison of the membrane fouling rate for the bare Al 2 O 3 membrane and Poly(oxyethylene methacrylate) (POEM)-directed TiO 2 /Al 2 O 3 composite membrane during filtration in the presence (With UV) and absence (Without UV) of ultraviolet (UV) illumination (filtration time: 24 h).

Figure 9 .
Figure 9. Organic dye removal efficiency of bare Al2O3 membrane and POEM-directed TiO2/Al2O3 composite membrane in the (a) absence and (b) presence of UV illumination.

Table 1 .
Comparison of pure water permeability for bare Al 2 O 3 membrane and Poly(oxyethylene methacrylate) (POEM)-directed TiO 2 /Al 2 O 3 composite membrane during filtration (with and without ultraviolet (UV) illumination).