C. S. Ong
*a,
W. J. Lau
*b,
P. S. Goh
b,
B. C. Ng
b,
A. F. Ismail
b and
C. M. Choo
a
aDiscipline of Chemical Engineering, Faculty of Engineering and the Built Environment, SEGi University, 47810 Petaling Jaya, Selangor, Malaysia. E-mail: ongchisiang@segi.edu.my
bAdvanced Membrane Technology Research Centre (AMTEC), Universiti Teknologi Malaysia, 81310 Skudai, Johor, Malaysia. E-mail: lwoeijye@utm.my
First published on 22nd October 2015
The rapid expansion and development of membrane based wastewater treatment in recent decades have led to the emerging technology of submerged membrane photocatalytic reactors (SMPR), which exhibit not only a lower degree of fouling but are also capable of separating and degrading organic pollutants simultaneously during the treatment process. This review intends to provide an update on the influence of several key operational parameters, i.e. photocatalyst loading (both suspended and immobilized), feed pH and concentration, wavelength and intensity of UV light, membrane module packing density and air bubble flow rate on the efficiencies of SMPR in treating degradable organic pollutants. The structure and properties of the photocatalytic membrane as well as membrane performance stability under UV irradiation are also discussed. Understanding the effect of each operational parameter is of paramount importance towards achieving optimum SMPR performance and addressing the challenges encountered in the development of SMPR. Strategies and approaches are also recommended in this review to overcome the persistent problems and facilitate the research and development of SMPR.
Fig. 1 Schematic diagram of a laboratory-scale PMR with zinc oxide (ZnO) photocatalysts in suspension.2 |
There are numerous configurations of PMRs being pursued by researchers. These include pressurized and depressurized (submerged) systems that work in either batch or continuous mode. The pressurized PMR system can achieve a higher flux compared to the submerged system due to the higher driving force applied to the system. However, the major drawbacks of the pressurized PMR system are the decline of permeation rate as a function of filtration time coupled with a higher degree of surface fouling caused by remaining pollutants and/or suspended catalysts in the feed solution. The impact of suspended catalysts on membrane flux has been previously investigated by several research groups and their results have shown that suspended nanoparticles in the PMR system have a negative effect on membrane permeability.3–5 In this regards, the depressurized PMR system, also known as the submerged membrane photocatalytic reactor (SMPR), has been generally agreed as an alternative approach to overcome the problems encountered by the pressurized system. Compared to pressurized conventional PMRs, the most notable advantages of SMPR are its (1) possibility of operating at high flux with relatively low energy consumption and fouling tendency and (2) enhanced mass transfer between UV light and targeted pollutants (for greater photodegradation efficiency) due to the generation of air bubbles in the system.1
In PMRs, the catalyst can be either immobilized on/in a membrane or suspended in the reaction mixture. Two main configurations for PMR are generally pursued, namely, (i) a reactor with catalysts suspended in the feed solution and (ii) a reactor with catalysts immobilized in/on the membrane. The advantages and disadvantages of suspended and immobilized titanium dioxide (TiO2) catalysts are summarized in Table 1. Although a suspended system offers a more intensive treatment, it requires an additional process to separate the catalysts from the suspended reactor. Many methods have been proposed to improve the recovery of the suspended catalysts. Some of them are (1) improving the catalysts aggregation through pH adjustment and (2) enhancing the separation of magnetic catalysts in a magnetic field. Nevertheless, these methods were only able to enhance the sedimentation and facilitate the recovery of catalyst in the batch system.6–9
TiO2 suspended reactor | TiO2 immobilized reactor |
---|---|
Advantages | Advantages |
1 Higher photocatalytic area to reactor volume ratio | 1 No separation and recycle of the catalyst |
2 Higher mass transfer and degradation efficiency | 2 Less membrane fouling due to enhanced hydrophilicity of membrane surface upon nanoparticle incorporation |
3 Fairly uniform catalyst distribution | 3 Pollutants could be degraded either in feed or in permeate |
4 Adjustable amount of nanoparticle suspension in the reactor to deal with different compositions of treated solution | |
Disadvantages | Disadvantages |
1 Higher operating cost and requires additional treatment after degradation process | 1 Degradation efficiency is lower than that of suspension mode |
2 Impossible to adjust the catalyst loading to deal with different compositions of treated solution | |
3 Replacement of membrane is required when catalyst loses its activity |
The membrane photocatalytic reactor has attracted considerable attention among membrane scientists mainly due to its powerful and efficient capability for degrading and mineralizing recalcitrant compounds under vacuum pressure conditions. This great achievement was initiated by Molinari in 2002,10 and presently the SMPR has been extensively applied for the purification and disinfection of contaminated groundwater, surface water, and wastewater that contain recalcitrant, inhibitory, and toxic compounds with low biodegradability.
A comprehensive overview and detailed discussion on the development of PMR have been previously published in several review articles.11–13 A broad range of subject matters, such as fundamental mechanism, reactor configurations, operational parameters, kinetics and modelling, and water quality analysis as well as the related life cycle assessment, have also been covered, but are mainly focused on the development between 1990 and 2010. Moreover, a brief description on the types of photocatalytic membranes, PMR configuration and potential applications could be found in several recently published books.14–17 However, the rapid expansion and increasing demand of employing SMPR in water and wastewater treatment over the past several years have motivated us to provide an update on the recent progresses of SMPR applications. The most significant contribution of this review is to highlight and emphasize the importance of operating conditions on the performance of SMPR followed by a discussion on the existing challenges and strategies that can be implemented to heighten the current performance of SMPR.
Type of photocatalyst/max loading | Membrane configuration/polymera | Operating conditions | System/reactor volume | Targeted pollutant(s)b | Initial concentration/range of concentration | Type of UV/power intensity | Time required to degrade at least 50% of pollutants (h) | Ref. |
---|---|---|---|---|---|---|---|---|
a PTFE – polytetrafluoroethylene, PVDF – polyvinylidene fluoride, PEI – polyethylenimine and PP – polypropylene.b RB5 – reactive black 5, HA – humic acid, CBZ – carbamazepine, AO7 – acid orange 7, DFC – diclofenac, TrOC – trace organic compound, AR1 – acid red 1 azo dye. | ||||||||
Degussa P25 TiO2/0.5 g L−1 | Flat sheet/PTFE | Submerged membrane with TiO2 catalyst suspension ABFR: 1.5 L min−1, pH 6.4–6.9 | 3 L | RB5 dye | 125 ppm | UVC (15 W, intensity: N/A) | ∼120 min (with 0.5 g L−1 TiO2 at 100 ppm) | Damodar et al.25 |
Degussa P25 TiO2/1 g L−1 | Hollow fiber/N/A | Submerged membrane with TiO2 catalyst suspension. ABFR: null; Pohang seawater: pH 8.1; Masan seawater: pH 8.2 Mooncheon lake water: pH 7.98 | 800 mL | Two seawater sources (from city of Pohang and city of Masan) and Mooncheon lake water | Pohang seawater: 0.198 ppm (TOC), Masan seawater: 2.03 ppm (TOC) Mooncheon lake water: 4.91 ppm (TOC) | UVA (8 W, intensity: N/A) | Mooncheon lake water: ∼90 min (with 1 g L−1 TiO2 at 4.91 ppm); no significant TOC reduction for Pohang and Masan seawater | Kim et al.26 |
Degussa P25 TiO2/0.6 g L−1 | Hollow fiber/polypropylene (PP) or PVDF | Submerged membrane with TiO2 catalyst suspension. ABFR: 5 L min−1, pH: N/A | 4 L | HA | TOC: 10 ppm | UVA (8 W, intensity: N/A) | ∼15 min (with 0.6 g L−1 TiO2 at 10 ppm) | Halim et al.27 |
Degussa P25 TiO2/0.5 g L−1 | Flat sheet/PVDF | Flat submerged membrane with TiO2 suspension. ABFR: 4 L min−1, pH 7 | 8 L | HA | DOC: 10 ppm | UVC (16 W, 1.17 mW cm−2) | ∼30 min (with 0.5 g L−1 TiO2 at 10 ppm) | Yong et al.28 |
Degussa P25 TiO2/1.5 g L−1 | Hollow fiber/PVDF | Submerged membrane with TiO2 catalyst suspension. ABFR: 20 L min−1, pH: N/A | 9 L | Polysaccharides | TOC: 2.5 ppm | Three UV-A (30 W, 8.3 mW cm−2) | <360 min (with 0.5 g L−1 TiO2 at 2.09 ppm) | Sarasidis et al.29 |
Degussa P25 TiO2/1 g L−1 | Hollow fiber/PVDF | Submerged hollow fiber membrane with P25 TiO2 suspension, pH: N/A | 1 L | Carbamazepine (CBZ) | 5 ppm | 240 units of vis-LED (<0.5 W m−2) | ∼120 min (with 1 g L−1 TiO2 at 5 ppm) | Wang et al.30 |
Degussa P25 TiO2/18 wt% | Hollow fiber/polyethylenimine (PEI) | Submerged membrane embedded with TiO2 catalyst; ABFR and pH: N/A | 25 mL | Acid orange 7 (AO7) | 20 ppm | Four UVA (8 W, intensity: N/A) | <60 min (with 18 wt% TiO2 at 20 ppm) | Zhang et al.31 |
Degussa P25 TiO2/0.75 g L−1 | Hollow fiber/PVDF | Submerged membrane with TiO2 catalyst suspension ABFR: 5 L min−1, pH 6.8 | 3 L | Diclofenac (DFC) | 2.5 ppm | Four UVA (24 W, 14.4 mW cm−2) | <300 min (with 0.5 g L−1 TiO2 at 2.5 ppm) | Sarasidis et al.23 |
Degussa P25 TiO2/0.5 g L−1 | Hollow fiber/PVDF | Submerged membrane with TiO2 catalyst suspension ABFR: 1.2 L min−1, pH 3, 8 | 4 L | Trace organic compound (TrOC) | 0.5 ppm | Seven UVA (8 W, intensity: N/A) | <240 min (with 0.5 g L−1 TiO2 at 0.5 ppm) | Fernández et al.32 |
Anatase TiO2/2.0 g L−1 | Hollow fiber/polypropylene (PP) | Submerged membrane with TiO2 catalyst suspension ABFR: 3 L min−1, pH 3, 7 and 10 | 5 L | Acid red 1 azo dye (AR1) | 15–75 ppm | UVC (8 W, 62.91 mW cm−2) | ∼80 min (with 0.5 g L−1 TiO2 at 15 ppm) | Kertèsz et al.33 |
Degussa P25 TiO2/4 wt% | Hollow fiber/PVDF | Submerged membrane embedded with TiO2 catalyst ABFR: 1–5 L min−1, pH 7 | 14 L | Oil molecules | 250–10000 ppm | UVA (8 W/0.333 mW cm−2) | ∼60 min (with 2 wt% TiO2 at 1000 ppm) | Ong et al.34 |
TiO2–ZrO2/0.15 g L−1 | Hollow fiber/PVDF | Submerged membrane with TiO2 catalyst suspension. ABFR: 4 L min−1, pH 4 | 2 L | HA | 50 ppm | UVC (4 W, intensity: N/A) | <240 min (with 0.15 g L−1 TiO2–ZrO2 at 50 ppm) | Khan et al.35 |
Furthermore, the performance of SMPR at an optimum catalyst loading can also be influenced by the dimension of the photoreactor. This is because different reactor designs tend to have different water flow hydrodynamics and photon absorption rates.11 A large reactor volume usually has a lower saturated catalyst loading and lower efficiency compared to a small reactor. A considerable amount of studies have reported the effect of TiO2 loading on the process efficiency, as summarized in Table 3.18–21 The observed discrepancies in photodegradation and membrane performance can be attributed to the differences in the reactor configurations, light sources and contaminant properties as well as the interaction between operating conditions employed. To avoid an excessive use of photocatalysts and to ensure the highest photodegradation efficiency, it is very important to determine the optimum catalyst loading based on the characteristics of wastewater.
Targeted pollutant | Photocatalyst | Range of catalyst loadings | Optimum catalyst loading | Light source, power and intensity | Ref. |
---|---|---|---|---|---|
Fulvic acid | Suspended | 0–0.6 g L−1 | 0.5 g L−1 | UVC 11 W, 0.75 mW cm−2 | 18 |
Bisphenol-A | Suspended | 0.2–2 g L−1 | 0.5 g L−1 | UVA 8 W, intensity: N/A | 19 |
Biologically treated sewage effluent | Suspended | 0.5–1.0 g L−1 | 1 g L−1 | UVA 10 W, 46.15–276.96 mW cm−2 | 21 |
Polysaccharide | Suspended | 0.25–1.5 g L−1 | 0.5 g L−1 | Three UV-A 30 W, 8.3 mW cm−2 | 23 |
Acid red 1 | Suspended | 0.01–2 g L−1 | 0.5 g L−1 | UVC 8 W, 62.91 mW cm−2 | 33 |
Oily wastewater | Immobilized | 0–4 wt% | 2 wt% | UVA 8 W, 0.333 mW cm−2 | 34 |
Carbamazepine | Suspended | 0.3–1 g L−1 | 1 g L−1 | 240 units of vis-LED with intensity <0.5 W m−2 | 30 |
(1) |
(2) |
Table 4 summarizes the findings from previous studies on feed concentration, in which optimum PMR performance can be achieved. According to Kertèsz et al.,33 as the dye concentration increases, the color of the irradiated solution becomes more intensive and concentrated, which in turn affects the penetration depth of UV light. The dye molecules also tend to absorb part of the light photons (UV-screening effect of the dye itself), which leads to insufficient photon energy for hydroxyl radical generation.33,36 Furthermore, the thick fouling layer formed at a high feed concentration would adversely affect photocatalytic degradation due to less active surface sites for UV irradiation. This is further supported by a recent study, where the degradation of oil under UV irradiation was highly efficient at low concentrations (see Fig. 2).24
Targeted pollutants | Range of feed concentration (ppm) | Feed concentration at which optimum performance was achieved (ppm) | Light source, power and intensity | Ref. |
---|---|---|---|---|
a The concentration of the organic components was determined based on total organic carbon (TOC). | ||||
Bisphenol-A | 10–50 | 10 | Four UVA 8 W, intensity: N/A | 37 |
Humic acid | 1–10 | 1 | UVA 8 W, intensity: N/A | 38 |
Biologically treated sewage effluent | 0–100 | 50 | UVA 10 W, 46.15–276.96 mW cm−2 | 21 |
Acid red 1 | 15–75 | 15 | UVC 8 W, 62.91 mW cm−2 | 33 |
Synthetic oily wastewater | 250–10000 | 250 | UVA 8 W, 0.333 mW cm−2 | 34 |
Carbamazepine with humic acid | 1.5–14.5a | 5 | 240 units of vis-LED 15–60 W, <0.5 W m−2 | 30 |
Humic acid | 5–50a | 10 | UVA 8 W, intensity: N/A | 27 |
Fig. 2 Effect of feed concentration on TOC degradation of PVDF–TiO2 composite membrane in the SMPR system (operating conditions: temperature = 25 °C, membrane type: PVDF with 2 wt% TiO2, module packing density = 35.3%, vacuum pump flow rate = 15 mL min−1 and pH = 7).24 |
On the contrary, Halim et al.27 reported that photocatalytic efficiency increased with increasing initial total organic carbon (TOC) concentration of HA from 5 to 50 ppm, due to the larger amount of reactants anticipated in the reaction mixture, which led to more frequent collisions between the organic molecules and the catalyst particles and higher adsorption rate. As implied in the abovementioned studies, PMR membranes achieve optimum performance depending on the characteristics of the targeted pollutants.
One recent study has reported that permeate flux increased with increasing the module packing density from 17.6% to 35.3%, due to the enhanced mass transfer between the water molecules and membrane surface (see Fig. 3). However, when the module packing density was further increased to 52.9%, the fibers were likely to attach to each other and this resulted in limited spaces available between adjacent fibers, thus reducing the active surface sites for UV irradiation. This as a consequence, deteriorated both photodegradation and membrane flux. This is in agreement with the experimental study conducted by Kiat et al.,46 where severe fouling tended to occur when the packing density of the module exceeded a critical value, i.e. 30.8%. Under an optimized packing density, Yeo et al.41 reported that promising membrane permeability could be achieved. However, it must be pointed out that a high density of fiber packing could cause foulant accumulation inside the fiber bundle and adversely decrease permeability. Contradictory results were reported by Gunther et al.,44 where the water flux increased by 25% when the fiber packing density was decreased from 80% to 40%. This enhancement was attributed to the suppression of cake layer formation when a low packing density of hollow fiber membranes was used.
Fig. 3 Effect of module packing density on (a) permeate flux and oil rejection and (b) TOC degradation of PVDF–TiO2 composite membrane in the SMPR system (operating conditions: temperature = 25 °C, membrane type: PVDF with 2 wt% TiO2, vacuum pump flow rate = 15 mL min−1 and pH = 7).24 |
Wu and Chen47 investigated the effect of flow distribution on shell-side mass transfer performance in randomly packed hollow fiber modules. They observed that the mass transfer coefficient was rapidly decreased with increasing packing density until 50% of the total volume fraction, and a further increase in packing density tended to increase the mass transfer coefficient. They attributed the improved mass transfer coefficient to the better orientation of the hollow fibers, which facilitated the water flowing through the adjacent fibers. As can be seen, the optimum packing density varies depending on the feed solution properties, module dimension and membrane material. Extensive investigations on this subject are certainly needed to provide a better understanding on how the module packing density governs the efficiency of photodegradation.
With air flow (O2 = ∼78%) under 365 nm irradiation:48
(3) |
Ozone (O3) → O (1D) + H2O → 2OH˙ (formation of hydroxyl radicals) | (4) |
(5) |
Oil + 3OH˙ → xCO2 + yH2O (mineralization) | (6) |
Fig. 4 Schematic diagram of photocatalytic reaction in the presence of photocatalyst TiO2 and air bubbles.49 |
The ABFR plays an important role in the photocatalytic reaction because the supplied oxygen could provide sufficient electron scavengers to trap excited conduction band electrons from recombination.11 The flux improvement at a higher ABFR is likely due to the generation of circulation flow in the SMPR, which limits the adsorption of pollutants onto the membrane surface and further reduces the membrane fouling tendency. The large airflow might enhance the mass transfer inside the treated wastewater and more OH− radicals would be produced in the feed solution, which in turn results in a higher photodegradation efficiency and membrane water flux. However, when excessive air bubbles are present in the system, the adsorption reaction of targeted pollutants onto the suspended catalyst is greatly reduced, which may result in a lower degradation rate because photocatalytic oxidation is a surface-oriented reaction. Chin et al.19 investigated the degradation of a TiO2 suspended submerged membrane reactor by varying the ABFR from 0.2 to 4 L min−1. The photocatalytic reaction rate increased with an increase in bubbling rate and reached a maximum value at a bubbling rate of 0.5 L min−1. Bubbling can not only increase the liquid film mass transfer coefficient around the aggregates, but may also provide a bubble cloud that can attenuate UV light transmission in the photoreactor. The balance of the competing effects between mass transfer and light attenuation might lead to an optimal bubbling rate, which can achieve the highest photodegradation efficiency.30
Fig. 5 Proposed mechanism of CBZ degradation with HA in a TiO2 suspension under vis-LED irradiation.30 |
It is worth mentioning that the concentration of OH˙ might increase with the increasing concentration of H+ in the acidic condition.18,51,52 This is in agreement with the study conducted by Chin et al.,37 where BPA degradation at a low pH was remarkably higher than that at a high pH. As pH was increased, the TiO2 surface became progressively more negative and this led to the development of greater repulsive forces between the TiO2 surface and BPA compounds, thus retarding the total degradation efficiency. Similarly, Khan et al.53 found that HA degradation was two times higher at a low pH compared to that obtained at a high pH (see Fig. 6). This is most likely due to the stronger electrostatic attractions between the HA and photocatalyst TiO2–ZrO2 in an acidic environment. Table 5 summarizes the effects of pH on the photodegradation of various pollutants using the SMPR treatment process. It is revealed that organic molecules tend to have different photocatalytic reactivities in different pH environments, depending on the nature of the respective pollutants.
Fig. 6 Effects of pH solution on UV254 removal efficiency [HA] = 50 mg L−1, temperature = 28 °C using TiO2–ZrO2 particles.53 |
Targeted pollutants | Tested pH value | Optimum pH | Light source and its intensity | Ref. |
---|---|---|---|---|
FA | 3.4, 6.5, 8.2 and 10.3 | 3.4 | UVC 11 W, 0.75 mW cm−2 | 18 |
BPA | 4, 7 and 10 | 4 | UVA 8 W, intensity: N/A | 37 |
AR1 | 3, 7 and 11 | 11 | UVC 8 W, 62.91 mW cm−2 | 33 |
HA | 4, 7 and 10 | 4 | UVC 4 W, intensity: N/A | 53 |
CBZ | 3, 6, 9 and 12 | 12 | 240 units of visible LED with intensity <0.5 W m−2 | 30 |
In particular, light intensity is one of the few parameters that affect the degree of photocatalytic reaction on organic substrates. A relatively abundant light intensity is required to adequately provide the catalyst surface active sites with sufficient photons energy. The photoactivity of catalysts in the presence of UV wavelength (<400 nm) in many studies obeys the linear proportional correlation to the incident radiant flux and becomes steady at excessive radiant flux in the photoreactor. This phenomenon was observed by Ho and his co-workers,21 where the organic matters from biologically treated sewage effluent (BTSE) was treated using an SMPR system. As shown in Fig. 7, a similar tendency of photocatalytic degradation was observed irrespective of light intensity, where the degradation rate dropped significantly for the first 30 min of operation followed by an almost constant rate when all the particles absorbed photons and produced electron–hole pairs. Considering the energy consumption, the minimum intensity of 46.61 mW cm−2 was determined as the optimum intensity.
Fig. 7 Effect of UV light intensity on the photooxidation process (operating conditions: initial TOC = 12.47 mg L−1; TiO2 concentration = 1.0 g L−1).21 |
Recently, Wang et al.30 also reported the enhancement of CBZ photocatalytic degradation efficiency with an increase in visible-light intensity. Compared to 68% removal of CBZ under high intensity UV irradiation, only 28% of CBZ was degraded at low intensity UV irradiation. Furthermore, Kertèsz et al.33 compared the decolorization of an aqueous acid red 1 (AR1) solution under two different UV light wavelengths (254 nm and 366 nm). As can be seen from Fig. 8, no decolorization of the aqueous AR1 solution occurred in the absence of a catalyst using a lower UV wavelength. Complete decolorization of AR1 could only be achieved with the simultaneous presence of a catalyst and UV irradiation. Faster initial degradation was observed at 254 nm compared to 366 nm, but at the end of the irradiation experiments (at 90 min), the two decolorizations were almost equal. Fujishima et al.,55 however, indicated that the initiation of photocatalysis reaction rates is not highly dependent on light intensity because very few photons of energy can sufficiently induce the surface reaction. Based on these findings, it can be concluded that the impact of different operating parameters on PMR performance is very complicated and an optimum condition for a specified application should be selected on the basis of several preliminary studies with similar operational parameters.
Fig. 8 Influence of UV irradiation wavelength on the decolorization process (operating conditions: C0 = 15 mg L−1, CTiO2 = 0.5 g L−1, pH = 7, T = 25 °C, n = 400 rpm, IUV = 62.9 mW cm−2).33 |
Fig. 9 Redox potentials of the valence and conduction bands as well as band-gap energies for various metal oxides at pH 7. The redox potential positions of H+/H2 and OH˙/OH− at pH 7 are also illustrated (*NHE: Normal Hydrogen Electrode).57 |
Furthermore, Luisa et al.59 compared the photodegradation efficiency of diphenhydramine (DP) and methylene orange (MO) by incorporating three different photocatalysts, i.e. self-synthesized TiO2, graphene oxide combined with TiO2 (GO–TiO2) and commercial P25–TiO2 into a flat sheet membrane. The results indicated that the membranes prepared with the GO–TiO2 composite exhibited the highest photocatalytic activity, due to the lowest band gap energy coupled with highest surface area, as compared to the TiO2 catalyst. The current research priority of catalyst development is focused on reducing the band gap energy of particles, in addition to increasing the surface area, with the aim to achieve a faster degradation rate.
From energy point of view, utilization of renewable solar energy is attractive in water industry. Although several visible light induced photocatalysts have been developed,64–67 most of these photocatalytic membranes need to be initiated under UV irradiation. The effectiveness of catalysts within the membrane matrix is jeopardized by lower photocatalytic degradation and wide band gap energy. It is necessary to develop a novel catalyst that possesses a small band gap energy, high surface area, and excellent resistance against thermal shock. On the other hand, the transformation of organic pollutants might cause a variety of organic intermediates that can be toxic and more persistent than the original pollutants themselves.68,69 Therefore, attention also needs to be paid to understand the adverse effects caused by the harmful by-products generated from the partial degradation of organic pollutants. Besides the targeted pollutants, inorganic impurities that cannot be degraded by photocatalytic membranes might accumulate on the membrane surface and unfavourably reduce the photocatalytic activity.
With respect to UV light, the key factor limiting the feasibility of the process on the real scale is the short life of UV sources, which must be periodically replaced. In addition, the operation of UV lamps consumes a lot of energy, which is estimated to account for approximately 80% of operation costs.70 Although researchers have proposed the use of ultraviolet light emitting diodes (UV-LED), which have a longer life span and do not contain hazardous mercury, their purchase cost is still higher than UV lamps at the same electrical energy conversion. Solar photocatalytic oxidation is a very promising process; however, it is limited by the large working area required for solar light irradiation. The studies on the long term stability of novel photocatalysts with wavelengths in the visible or solar light range are still insufficient.
In terms of safety, SMPR should be properly covered to prevent direct exposure of UV light to the human body. A proper cooling device is also required as high temperatures may occur and cause overheating of the UV lamp in the long run. In addition, the destructive effect of UV light or hydroxyl radicals on polymeric membranes is another key issue when photocatalysis is involved in the separation. This is because the immobilized photocatalysts might absorb UV light energy, which causes membrane ageing and further alters its surface morphology and separation performance. It is in urgent need to find appropriate polymeric materials that are highly resistant towards UV irradiation and can be readily dissolved in a wide range of solvents.
Despite their assuring applications and significant achievement in recent decades, there are still some persistent problems that are encountered in operating SMPR systems. Some of these are reduced effectiveness of the catalysts embedded within the membrane matrix, polymer degradation under UV exposure and partial transformation of organic pollutants into hazardous by-products. Severe membrane fouling in the long term operation and its consequences on plant maintenance and operating cost have limited the viability of SMPR in the wastewater treatment industry. Gratefully, dedicated scientific investigations in recent years have offered several innovative approaches; for instance, the fabrication of dual-layer hollow fiber membranes with the catalyst immobilized in the outer surface layer. Dual-layer hollow fiber membranes have the advantages of maximizing membrane performance using an extremely high-performance or functional membrane material as the selective layer, while employing a low-cost material as the supporting layer. This approach significantly reduces the overall membrane material cost without compromising filtration and photocatalytic performance.
More research in this area is still required to resolve the aforementioned issues as well as to enhance the performance efficiency, reliability and stability of SMPR for industrial implementation. With the rapid progress made in materials science and engineering, it is expected that newly developed catalysts could outperform existing ones under visible light or solar illumination. In terms of membrane materials, polymers that can withstand UV irradiation and can be easily dissolved in common solvents, such as N-methyl-2-pyrrolidone and dimethylacetamide, are highly desired. The use of ceramic membranes that possess excellent thermal and chemical resistance and resistance to abrasion by suspended photocatalysts may be another promising option. With respect to the system configuration, the photocatalytic membrane reactor design should be further improved based on the total irradiated surface area of catalyst per unit volume, UV light power and its intensity as well as the light distribution within the reactor.
Last but not least, it is recommended that solar heterogeneous photocatalytic oxidation should be combined with photovoltaics to reduce the energy consumption of the process. To reduce fouling problems in SMPR systems, another photocatalyst removal pretreatment stage, such as sedimentation or precoat filtration, prior to membrane filtration might be advantageous, especially when solar photocatalytic oxidation is operated in batch mode. Although it might take years to resolve the remaining challenges in this field, it appears certain that SMPR will become universal and effective in dealing with a large variety of industrial wastewater applications in the future.
This journal is © The Royal Society of Chemistry 2015 |