This art icle was downloaded by: [ Kok Kwang Ng] On: 03 January 2013, At : 19: 55 Publisher: Taylor & Francis I nform a Lt d Regist ered in England and Wales Regist ered Num ber: 1072954 Regist ered office: Mort im er House, 37- 41 Mort im er St reet , London W1T 3JH, UK Separation Science and Technology Publicat ion det ails, including inst ruct ions f or aut hors and subscript ion inf ormat ion: ht t p: / / www. t andf online. com/ loi/ lsst 20 Effect of Ultrasound on Membrane Filtration and Cleaning Operations Kok-Kwang Ng Lin a a , Chien-Ju Wu , Pui-Kwan Andy Hong b a , Hsiu-Lan Yang , Chung-Hsin Wu c a , Sri Chandana Panchangam & Cheng-Fang Lin a , Yen-Ching a a Graduat e Inst it ut e of Environment al Engineering, Nat ional Taiwan Universit y, Taipei, Taiwan b Depart ment of Civil and Environment al Engineering, Universit y of Ut ah, Salt Lake Cit y, UT, USA c Depart ment of Chemical and Mat erials Engineering, Nat ional Kaohsiung Universit y of Applied Sciences, Kaohsiung, Taiwan Accept ed aut hor version post ed online: 29 May 2012. To cite this article: Kok-Kwang Ng , Chien-Ju Wu , Hsiu-Lan Yang , Sri Chandana Panchangam , Yen-Ching Lin , Pui-Kwan Andy Hong , Chung-Hsin Wu & Cheng-Fang Lin (2012): Ef f ect of Ult rasound on Membrane Filt rat ion and Cleaning Operat ions, Separat ion Science and Technology, 48: 2, 215-222 To link to this article: ht t p: / / dx. doi. org/ 10. 1080/ 01496395. 2012. 682289 PLEASE SCROLL DOWN FOR ARTI CLE Full t erm s and condit ions of use: ht t p: / / www.t andfonline.com / page/ t erm s- and- condit ions This art icle m ay be used for research, t eaching, and privat e st udy purposes. Any subst ant ial or syst em at ic reproduct ion, redist ribut ion, reselling, loan, sub- licensing, syst em at ic supply, or dist ribut ion in any form t o anyone is expressly forbidden. The publisher does not give any warrant y express or im plied or m ake any represent at ion t hat t he cont ent s will be com plet e or accurat e or up t o dat e. The accuracy of any inst ruct ions, form ulae, and drug doses should be independent ly verified wit h prim ary sources. The publisher shall not be liable for any loss, act ions, claim s, proceedings, dem and, or cost s or dam ages what soever or howsoever caused arising direct ly or indirect ly in connect ion wit h or arising out of t he use of t his m at erial. Separation Science and Technology, 48: 215–222, 2013 Copyright # Taylor & Francis Group, LLC ISSN: 0149-6395 print=1520-5754 online DOI: 10.1080/01496395.2012.682289 Effect of Ultrasound on Membrane Filtration and Cleaning Operations Kok-Kwang Ng,1 Chien-Ju Wu,1 Hsiu-Lan Yang,1 Sri Chandana Panchangam,1 Yen-Ching Lin,1 Pui-Kwan Andy Hong,2 Chung-Hsin Wu,3 and Cheng-Fang Lin1 1 Graduate Institute of Environmental Engineering, National Taiwan University, Taipei, Taiwan Department of Civil and Environmental Engineering, University of Utah, Salt Lake City, UT, USA 3 Department of Chemical and Materials Engineering, National Kaohsiung University of Applied Sciences, Kaohsiung, Taiwan Downloaded by [Kok Kwang Ng] at 19:55 03 January 2013 2 The impact of ultrasound (US) on membrane filtration and cleaning were studied and compared at various operating parameters of nominal pore sizes of 10 and 100 kDa membrane, trans-membrane pressure (TMP) of 100 and 140 kPa, and US frequencies of 20 kHz and 40 kHz. An average of 15%–20% increase of permeability was observed when US (20 kHz) was applied to assist membrane filtration on 10 kDa membrane and 100 kPa transmembrane pressure (TMP). However, an insignificant improvement was observed in the case of larger pore size membrane at higher TMP (140 kPa). US also augmented the membrane cleaning process effectively. Lower frequency 20 kHz US exhibited a higher flux recovery (>90%) than the high frequency 40 kHz (59%) using the 10 kDa pore size membrane with US-assisted membrane cleaning. Important factors influencing optimization of US effectiveness lie heavily on its configuration and operation. The experimental results as supported with SEM images demonstrate that US-assisted filtration and cleaning are most effective when membrane pore size, US frequency, and TMP are lower. Keywords flux; fouling; membrane cleaning; membrane process; ultrasound INTRODUCTION Because of the effectiveness of membrane filtration in removing suspended particles and health concerned contaminants, it has been widely used to treat drinking water, industrial wastewaters, and subsequently for reclamation purposes. In treatment of drinking water, membrane processes play a vital role which complements traditional water treatment processes in order to meet increasingly stringent drinking water standards and secure public confidence (1). In treatment of industrial water, membranes can also be configured in conjunction with other techniques to achieve various levels of water quality requirements needed Received 14 December 2011; accepted 1 April 2012. Address correspondence to Cheng-Fang Lin, Graduate Institute of Environmental Engineering, National Taiwan University, 71 Chou-Shan Rd., Taipei 106, Taiwan. Tel.: þ8862-3366-7427; Fax: þ886-2-2392-8830. E-mail: cflin@ntu.edu.tw by different industries. Moreover, as clean and easily accessible water sources became scarcer, membrane processes have proven to be a very reliable technology for water reuse in many parts of the world. Because of their exhibited qualification membrane processes have become indispensable as they: 1. produce effluents of high quality, 2. have small footprints and use indoor space that is immune to harsh weather conditions, 3. are resilient and easy to operate and maintain, 4. allow expansion of existing facilities, and 5. reduce chemical uses and sludge production (2,3). While incorporation of membrane filtration in water treatment processes offers the above advantages, reducing membrane fouling and flux decline has been a vital challenge in membrane applications that can impact water production efficiency, membrane life, and operating costs (4). The mechanisms involved in membrane fouling of ultrafiltration (UF) and microfiltration (MF) are pore blocking (complete blocking), direct adsorption (standard blocking), long term adsorption (intermediate blocking), boundary layer resistance (cake filtration), and concentration polarization layer that leads to reversible or irreversible fouling conditions (5–7). Reversible fouling blockage may be cleaned relatively easily by physical means such as backwashing or air stripping. However, cleaning for irreversible fouling usually requires chemicals such as a strong acid, base, enzyme, or surfactant. The results of chemical cleaning are often less than satisfactory. In addition, the membrane surface can be gradually degraded by successive chemical cleanings thereby reducing their lifespan. Moreover chemical cleaning also generates chemical wastes with associated disposal problems (4). Incorporation of ultrasound (US) irradiation for membrane cleaning is an interesting option that can improve membrane operation (4,8–12). The principle in using US 215 Downloaded by [Kok Kwang Ng] at 19:55 03 January 2013 216 K. K. NG ET AL. during membrane filtration relies primarily upon with the cavitation effect of US. When the ultrasonic device emits a sufficiently large supersonic wave to overcome the cohesive forces within water, water undergoes constant pressure extrusion and creates a lot of micro-scale bubbles around the membrane surface. As these bubbles form and burst under different pressures, they release energy and cause vibration that loosens particles on the membrane surface that can reduce membrane fouling without compromising treatment efficiency (4). Sabri et al. (13) studied intermittent US at 45 kHz during filtration of pulp and paper mill wastewaters and they reported up to 400% flux enhancement. Kobayashi et al. (9) found stronger permeation enhancement by US at low frequency (28 kHz) than at high frequency (45 or 100 kHz). Kobayashi et al. (9) and Kokugan et al. (14) found that US enhanced permeation by increasing mass transfer coefficient of water across microfiltration ceramic membrane as well as ultrafiltration organic membrane. US irradiation studies identified the prevention of concentration polarization and membrane fouling due to cake layer stripping, providing a less compressible fouling cake, and decreasing solute concentration nears the membrane surface (4,9,15). Lamminen et al. (4) categorized US cleaning mechanisms into four distinct pathways: 1. 2. 3. 4. acoustic streaming, microstreaming, microstreamers, and micro-jets. Only acoustic streaming does not involve the collapse of cavitation bubbles. The ultrasonic waves cause bulk water movement to and from the membrane surface that generates forces to push and pull the attached particles away from the surface. Microstreaming occurs when bubbles oscillate in size and create significant shear forces with an effective range of bubble diameter. The microstreamers phenomenon is defined as the cavitation bubbles travelling in ribbon-like paths as these bubbles travel vigorously around the membrane surface. They tend to coalesce and scour away particles while translating to antinodes. Microjets occur when two asymmetrical cavitation bubbles collapse at the same time or just one bubble collapses near a solid interface and creates a strong jet of water that scours particles away from the membrane surface. Numerous researchers have demonstrated the effectiveness of US-associated membrane cleaning with the purpose of increasing permeate flux and extending the membrane’s lifespan. Li et al. (16) reported that US with forward flushing recovered flux on a 0.2 mm nylon membrane fouled by a Kraft paper mill effluent. Matsumoto et al. (17) claimed that US-assisted backwashing cleaned fouled membrane more thoroughly than conventional backwashing. The use of US frequency from 28 kHz to 1 MHz for membrane cleaning was studied recently (4,9,18,19) and the results showed benefits of US-assisted cleaning (20). Several researchers (4,9,18,19) agreed that lower US frequencies exhibited higher flux recoveries than higher frequencies during US-assisted cleaning. Kobayashi et al. (19) observed virtually no cleaning effect at 100 kHz when compared to lower frequencies at 45 and 28 kHz. Using the MF membrane, Wakeman and Tarleton (21) concluded that US at 23 kHz offered not only better cleaning efficiency but also less energy consumption than at 38 kHz. In addition, Lozier and Sierka (22) combined US with ozone to reduce suspended solids on the membrane surface. US has been successfully combined with membrane backwashing in the laboratory. However, it has not been widely commercialized because of membrane damages in several occasions (13,23,24). Nevertheless, Lamminen et al. (4) and Muthukumaran et al. (15) stated that US-assisted cleaning did not affect membrane integrity. US irradiation can be applied wither in intermittent mode or continuous mode. Using intermittent US application is likely to be less energy intensive than continuous US irradiation and pulsed US has been discussed to be effective at removing Bovine serum albumin (BSA) fouling layers from a cross-flow MF (25). However, Simon et al. (26) indicated that the use of continuous application was more effective than intermittent application when applied to polymer solutions. Clearly, more studies are required to resolve the conflicting results. Owing to the advantages offered by US-assisted membrane filtration and cleaning in the present study, optimal US-assisted cross-flow membrane operation conditions were conducted. Both the filtration and cleaning processes were evaluated with the aid of intermittent (10 sec every 10 min of operation) ultrasonic irradiation. Hypotheses of membrane fouling at two different trans-membrane pressures (TMPs) 100 and 140 kPa and membrane pore sizes (10 and 100 kDa) were discussed. Regular backwashing and US-assisted backwashing were performed. The overall objective of this research was to reduce membrane fouling and enhance membrane cleaning by exploring 1. the ability of US to clean fouling associated with two specific membrane pore sizes (10 and 100 kDa) in a flat-sheet cell, 2. the ability of US to clean the membrane under two different trans-membrane pressures and 3. two different US frequencies in cleaning membrane fouling. MATERIALS AND METHODS Feed Water Humic acid (sodium salt) from Aldrich was used to prepare a stock feed solution of 9 mg L
1 by dissolving 1 g of 217 ULTRASOUND IMPACT ON MEMBRANE FILTRATION AND CLEANING Downloaded by [Kok Kwang Ng] at 19:55 03 January 2013 humic acid in 1 L deionized water, which then was filtered through a 0.45 mm membrane to remove particles from the solution. A calibration curve of 0 to 10 mg L
1 was prepared from a working solution of 100 mg L
1 potassium hydrogen phthalate (C8H5KO4). Humic acid (feed solution) was analyzed for its total organic carbon (TOC) concentration by using a LACHAT (Model IL-550 TOCTN) and a concentration of 9.02 mg L
1 was recorded with calibration curve. 1 N of sodium hydroxide (NaOH) and 1 N of hydrochloric acid (HCl) were used to adjust the pH of the solution to 7.2. Feed solutions were supplemented with sodium chlorite (NaCl) to maintain conductivity near 300 mS cm
1. The stock humic acid solution and feed solutions of 9 mg L
1 were stored at 4 C for later use. Ultrafiltration (UF) Membrane Negatively charged Pall-polyethersulfone (PES) flat sheet ultrafiltration membranes (effective filtration area ¼ 25 cm2; pore size ¼ 10 kDa and 100 kDa) were used in this study at constant trans-membrane pressures (TMPs) of 100 kPa and 140 kPa. Membranes of two different pore sizes were compared in US-assisted operations. Prior to experiment, the membrane was washed with methyl alcohol (CH3OH, anhydrous) purchased from Macron Chemicals for 5 min and rinsed with distilled water for 8 h. The PES flat sheet membrane was then operated with distilled water for 2 h until a near-constant flux was reached to eliminate pre-compaction at the same pressure. The membranes were operated for 3 h with and without US operation and were analyzed for fouling. All fouled membranes were cleaned by forward flushing, backwashing, and US-assisted backwashing. Ultrasound Module The US device consisted of a generator, frequency converter, amplifier, fixed stand and transducer manufactured by Bensonic (BW-2015P). During the US-assisted filtration experiment, the power intensity was measured at 105 W with 20 kHz. The US-assisted filtration setup is shown in Fig. 1. The US transducer was kept at 3.2 cm directly above membrane surface. The feed solution was pre-warmed to room temperature (25  1 C) before each membrane test. The temperature after the sonication was measured slightly increased to 26 C. Therefore, the influence of temperature fluctuation on the effect of US-assisted filtration and the feed water properties are considered to be negligible. After 30 min of filtration, US irradiation was intermittently switched on for 10 s at every 10 min operation. Filtration tests were performed in cross-flow mode with a cross-flow velocity of 0.13 ms
1 and fresh membrane was used for each analysis. The feed solution passed through the membrane installed on the membrane system and permeate was collected on the outside of the membrane system. Flux was determined by weighing permeate with an electronic FIG. 1. Schematic diagram of the cross-flow experimental setup for USassisted membrane operation controlled with a PC-PLC (Personal Computed-Programmable Logic Controller). (Color figure available online) balance at timed intervals. All permeation and cleaning experiments were monitored and controlled by a computer with a programmable logic controller (PLC) and operating software Gensis 32 automation suite. The six continuous operating steps include: 1. 2. 3. 4. 5. 6. humic acid solution filtration, DI water forward flushing, DI water backwashing, DI water filtration, chemical cleaning (soaking), and DI water flushing. This semi-automatic system was operated consistently to minimize experimental variations. The normalized flux was calculated based on initial flux permeation as follows: Normalized Flux ¼ Jf  100% JI ð1Þ where Jf ¼ flux measured at the point of operation JI ¼ initial permeation flux Ultrasound-Assisted Membrane Cleaning The cleaning processes include regular and US-assisted backwashing. The membrane was operated for 8 h without US operation before the cleaning experiments. After filtration, the fouled membrane was subjected to forward 218 K. K. NG ET AL. TABLE 1 Specifications of FEG-SEM Content SEM specification Downloaded by [Kok Kwang Ng] at 19:55 03 January 2013 Manufacturer Acceleration Voltage Magnification SEI resolution Functions Stand holder JEOL, JSM-6330 F 0.5 kv  30 kv X 10  X 500 k 1.5 nm (at 15 kv), 5.0 nm (at 1 kv) Auto focus, lightening, contrast X: 50 mm, Y: 70 mm, Z: 4,8,15,25,39 mm flushing with DI water for 30 min. The cleaning processes were regular backwashing and US-assisted backwashing. The US-assisted backwashing was operated with ultrasound at 20 or 40 kHz intermittently on for 10 s at every 10 min operation with a fixed power intensity of 105 W. During US-assisted cleaning, permeate flux was recorded every 3 min. The flux recovery rate was calculated with the following equation: Flux recovery ¼ Jc  100% Jo ð2Þ Where: Jc ¼ final flux after cleaning process Jo ¼ initial flux with DI water Scanning Electron Microscopy (SEM) Analysis The fouled membranes were sampled for morphological observation using a JEOL JSM-6330 F scanning electron microscopy (FEG-SEM). The scanning electron microscopy (SEM) device consisted of an acceleration voltage, stand holder with (X: 50 mm, Y: 70 mm, Z: 4, 8, 15, 25, 39 mm), and resolution 1.5 nm (at 15 kv), 5.0 nm (at 1 kv). The SEM was operated at 20 keV and with a magnification of 3,000X. The images of the membrane surfaces before and after US-assisted filtration and cleaning were taken for analysis. The methods were similar to those described previously by Lin et al. (27). The detailed SEM analysis is listed in Table 1. RESULTS AND DISCUSSION Impact of US-Assistance on Membrane Filtration at Different TMPs TMP is a primary factor that affects permeation flux and membrane fouling. The cross-flow UF (100 kDa) permeate fluxes at TMPs 100 and 140 kPa with and without intermittent US irradiation at 20 kHz after 30 min of filtration were evaluated. The results are presented in Fig. 2. Both TMP experiments lasted 180 min and the permeation fluxes were recorded every 3 min. At TMP of 100 and FIG. 2. Comparison of permeation fluxes over time at different TMP and MWCO of (a) 100 kPa with 100 kDa membrane pore size, (b) 140 kPa with 100 kDa membrane pore size and (c) 100 kPa with 10 kDa membrane pore size, influent DOC ¼ 9.0 mg L
1, US frequency ¼ 20 kHz, kHz, filtration time ¼ 180 min, cross flow velocity ¼ 0.13 ms
1, US irradiation for 10 s every 10 min starting at 30 min of filtration. 140 kPa, the initial permeate fluxes were monitored at 0.54 mL min
1 cm
2 and 0.56 mL min
1 cm
2, respectively. The final permeation of 100 kPa without US was recorded at 0.211 mL min
1 cm
2, while it was measured at 0.3 mL min
1 cm
2 with the assistance of US which accounted to 13% of permeation flux improvement. On the other hand, 0.12 (without US) and 0.14 mL min
1 cm
2 (with US) were observed at the end of membrane filtration with 140 kPa. Figure 2a demonstrates that when US irradiation is applied during the membrane filtration (after 30 min of operation) on 100 kDa pore size membrane at constant Downloaded by [Kok Kwang Ng] at 19:55 03 January 2013 ULTRASOUND IMPACT ON MEMBRANE FILTRATION AND CLEANING 100 kPa TMP, a consistent 15%–20% higher normalized flux was recorded for the subsequent 150 min of operation. This improvement can be attributed to US irradiation of water that forms streams of micro-scale air bubbles around the membrane surface. These streams of air bubbles continually fluctuate around the membrane surface in a way of simulating the air-stripping phenomenon that scours particulates off the surface thus enhancing membrane permeation (4). Contrarily, the effect of US was barely noticeable (<1%) at TMP of 140 kPa (Fig. 2b). These experimental results support the hypothesis of US-assisted membrane filtration is more effective at lower TMP. This finding agrees well with results of Chen et al. (11). Higher filtration pressure would increase compressive force and decrease the quantity of cavitation bubbles formed. This can be understood in light of the fact that high filtration pressure increases liquid solution cohesiveness that hinders cavitation formation and therefore reduces the quantity of cavitation bubbles and their impact on permeability. Impact of US-Assisted Filtration on Membranes of Different Pore Sizes Molecular weight cut-off (MWCO), as the term suggests, is an indication of the lowest molecular weight of solute that is retained by the membrane by 90%. In many instances, MWCO is a good assessment of how well a particular compound will be rejected during filtration based on its molecular weight. The effects of MWCO on membrane are strongly related to the feed solution characteristics and the particle size distribution. However, in the comprehensive literature review, Le-Clech et al. (28) found many contradicting conclusions on MWCO and particle size distribution. Therefore, in the present work, membranes of pore sizes of both 10 kDa and 100 kDa were examined for US-assisted permeation tests at TMP of 100 kPa and cross-flow velocity of 0.13 m s
1. As indicated in Fig. 2a and Fig. 2c, the initial permeation flux (100% normalized flux) for 100 kDa was recorded at 0.54 mL min
1 cm
2 and for 10 kDa at 0.33 mL min
1 cm
2. The permeate flux declined faster with the 10 kDa membrane during filtration and particularly severely in the first 60 min. He et al. (29) tested an anaerobic MBR at constant TMP and found that the smallest MWCO (20 kDa) showed the largest permeability loss within the first 15 min of filtration when compared to membranes of 30, 50, and 70 kDa. Membranes with smaller pores reject a wider range of materials and are expected to show higher flux resistance as compared to membranes with large pores. After 180 min of filtration without US, permeation of 10 kDa membrane dropped to approximately 22% of its initial flux, while the permeation of 100 kDa membrane dropped to 38.9% of its initial value as shown in Figs. 2c and 2a, respectively. With the assistance of US, the 10 kDa membrane exhibited stronger improvement in performance than that of 219 100 kDa membrane. As indicated in Figs. 2a and 2c, ultrasonic irradiation was started after 30 min of filtration and the flux was immediately improved for the 10 kDa membrane but not as noticeably for the 100 kDa membrane. A consistent 20% flux enhancement was recorded throughout filtration with the 10 kDa membrane. At the end of the permeation experiment, ultrasonic irradiation was able to restore the permeation flux from 22% to 50% of the initial flux for the 10 kDa membrane but negligible permeation improvement (<5%) for the 100 kDa membrane. Matsumoto et al. (17) investigated the effect of pore size on ceramic membranes in US-assisted filtration. Their results consistently showed flux enhancement in the case of small pore size (0.2 mm) rather than in larger pore sizes (0.8 mm and 1.5 mm). This phenomenon led to a hypothesis that the primary fouling mechanism for the 10 kDa membrane was the gel layer formation and the major mechanism for the membrane of 100 kDa membrane was inner pore blocking. Lin et al. (30) observed that the fraction of humic materials with MWCO <1 kDa prepared from Aldrich stock is less than 25% of the total dissolved organic carbon in the feed solution. They hypothesized that the primary fouling mechanism on the 10 kDa membrane was possibly contributed by gel layer formation with very little inner pore adsorption or clogging (Fig. 3a). On the other hand, a much larger portion of the humic materials was more likely to enter the pore channels of the 100 kDa membrane, which would subsequently aggregate and stick together resulting in a pore blocking mechanism (Fig. 3b). Katsoufidou et al. (31) supported this hypothesis and stated that irreversible fouling was caused by internal pore adsorption that led to progressive deterioration of membrane performance. US irradiation in water produces acoustic streaming, microstreaming, microstreamers, and micro-jets that loosen only the accumulated particulates on surface of membranes of small pores. It does little to membranes of large pores in which constriction within open channels occur (32). Therefore, US offered no permeation enhancement for the membrane of 100 kDa but great enhancement FIG. 3. Schematic of Fouling mechanisms of UF membrane of different pore sizes. 220 K. K. NG ET AL. Downloaded by [Kok Kwang Ng] at 19:55 03 January 2013 for membranes with pore size smaller than the fouling particles in feed solution. During the course of this experiment, no evident damages by the intermittent US irradiation was observed since the permeate flux of membranes did not increase during filtration, even though Masselin et al. (23) indicated that the PES membranes could be damaged after about 5 min exposure at 47 kHz US irradiation. On the other hand, Cai et al. (33) reported no damage was found on the membrane surface after US irradiation from SEM images taken and recommended the use of US irradiation intermittently because it is more convenient and economical to remain as less energy dependent as possible. Ultrasound-Assisted Enhanced Membrane Cleaning Process Volumetric flux of PES membranes with pore sizes 10 kDa and 100 kDa were recorded for 480 min at feed solution concentration of 9 mg L
1. The fouled membranes were subjected to 30 min of forward flushing and backwashing. Conventional cleaning was compared with USassisted cleaning at 20 kHz and 40 kHz. As seen in Fig. 4, initial permeation fluxes of DI water were recorded at 0.68 and 2.14 mL min
1 cm
2 for membranes with pore sizes of 10 and 100 kDa, respectively. Once filtration FIG. 4. US-assisted cleaning of membranes of (a) 10 kDa and (b) 100 kDa for flux recovery, influent humic acid concentration ¼ 9.0 mg L
1, membrane MWCO ¼ 100 kDa, TMP ¼ 100 kPa, cross flow velocity ¼ 0.13 ms
1. began, the fluxes dropped sharply to 0.11 and 0.24 mL min
1 cm
2, respectively. During conventional cleaning, the forward flushing and backwashing achieved 28% and 41% of flux recovery for the membrane of 10 kDa, and 19% and 24% of flux recovery for the membrane of 100 kDa. These flux recoveries were due to the removal of foulants on the membrane surface and in the membrane pores by water pressure (34,35). A more significant improvement in flux recovery was observed with the assistance of US during membrane cleaning. Dramatic improvements were observed, namely, more than 90% (0.67 mL min
1 cm
2) of flux recovery with US at 20 kHz and 59% (0.42 mL min
1 cm
2) of flux recovery with US at 40 kHz when forward flushing and backwashing were performed on the membrane of 10 kDa with US. For the membrane of 100 kDa, flux recovery was restored to 81% of the initial flux (1.75 mL min
1 cm
2) with US at 20 kHz and to 69% of the initial flux (1.49 mL min
1 cm
2) with US at 40 kHz. The results demonstrated the effectiveness of US in enhancing membrane cleaning. As US irradiation forms micro-scales of cavitation in water, energy released from collapses of cavitation bubbles can produce approximately 110 ms
1 of micro jet stream that assists in cleavages of large particles and detachment or erosion of smaller particles adsorbed on the membrane surface (36), thus enhancing the effectiveness of membrane cleaning. Lamminen et al. (4) agreed that the detachment of particles from membrane pores was primarily due to the cavitation effects, while acoustic streaming provided a means to transport the loosened particles away from membrane surface. In terms of different US frequencies, it was agreed that lower US frequency offers better cleaning efficiency than does high US frequency (4,9,19). The same phenomena were observed in the current studies as well. This may be attributed to the fact that higher US frequencies create smaller cavitation bubbles, which generate weaker water extrusion to loosen the particles from the membrane surface (18). Mason and Lorimer (37) stated that the production and intensity of cavitation in liquids increased with decreasing US frequency. This is because the rarefaction cycles are too short to allow a bubble to grow to a large size at high frequency and the effect of US-assisted cleaning was more effective with lower US frequency. SEM on Fouled Membrane Surfaces after US-Assisted Cleaning To validate the effectiveness of US-assisted cleaning, SEM pictures were taken as shown in Fig 5. Surface conditions of the membranes of 10 kDa and 100 kDa prior to the filtration are depicted in Fig. 5a and Fig. 5b. These images reveal that after 480 min of filtration, both membranes were covered with clusters of particles as shown in Fig. 5c and Fig. 5d. Backwashing was performed on ULTRASOUND IMPACT ON MEMBRANE FILTRATION AND CLEANING 221 Downloaded by [Kok Kwang Ng] at 19:55 03 January 2013 US frequency (20 kHz) in Fig. 6a exhibited better foulant removal on the membrane surface than higher US frequency (40 kHz) in Fig. 6c. US irradiation has been demonstrated to enhance membrane backwashing although important factors on optimizing US effectiveness rely heavily on its configuration and operation. FIG. 5. Surfaces of membranes before filtration: (a) 10 kDa and (b) 100 kDa; and surfaces after 480 min of filtration: (c) 10 kDa and (d) 100 kDa. Magnification: 3,000X. membrane of 10 kDa and 100 kDa as assisted by US at 20 kHz and 40 kHz. All SEM images were taken after 30 minutes of US-assisted backwashing. Figure 6a and Fig. 6b showed the 10 kDa and 100 kDa membrane surface conditions, respectively, after US-assisted cleaning at 20 kHz. US-assisted cleaning was more effective on membrane of 10 kDa than on membrane of 100 kDa. These results correlate our permeation records that show higher flux recovery for membrane of 10 kDa. Figure 6a and Fig. 6c are images taken to compare the impact of US frequencies on membrane cleaning efficiency. Evidently, lower CONCLUSION Ultrasound-assisted membrane filtration=cleaning do not alter the membrane’s intrinsic permeability. Instead it enhances the permeate flux through decreasing solute concentration near the membrane. The technique is highly influenced by several factors of feed solution properties, operating pressure (TMP), US frequency, and membrane pore size. This current research has demonstrated that different intermittent US configurations could significantly impact the effectiveness of the technique in improving membrane filtration and cleaning. Under the experimental conditions, an increment of 15%–20% permeate flux was observed when 20 kHz of US was applied to assist membrane permeation on 10 kDa pore size membrane with 100 kPa TMP. On the other hand, insignificant improvement was observed in larger pore size membrane (100 kDa) at higher TMP (140 kPa). Lower frequency of US (20 kHz) was found to clean the fouled membrane efficiently and exhibited more than 80% flux recovery than higher US frequency (40 kHz). Clearly, US-assisted filtration and cleaning are most effective when TMP, US frequency, and membrane pore sizes are smaller. Our experimental results suggest the potential of incorporating ultrasound in membrane applications and incorporation of intermittent US exhibits advantages by extending membrane filtration operating time without damaging the membrane surface and thereby exhibiting excellent flux recovery while improving membrane cleaning. ACKNOWLEDGEMENTS The authors would like to thank the National Science Council of the Republic of China (Contract No. NSC 96-2221-E-002-056-MY2) for financially supporting this research. REFERENCES FIG. 6. Comparison of US-assisted cleaning on different pore size membranes and frequencies: (a)10 kDa membrane cleaned with US at 20 kHz, (b) 100 kDa membrane cleaned with US at 20 kHz, (c) 10 kDa membrane cleaned with US at 40 kHz. Magnification: 3,000X. 1. Ng, K.K.; Lin, C.F.; Lateef, S.K.; Panchangam, S.C.; Hong, P.K.A.; Yang, P.Y. (2010) The effect of soluble microbial products in a fixed carrier biological system. Sep. Purif. Technol., 72: 98–104. 2. Cho, J.W.; Amy, G.; Pellegrino, J. (1999) Membrane filtration of natural organic matter: initial comparison of rejection and flux decline characteristics with ultrafiltration and nanofiltration membrane. Water Res., 33: 2517–2526. 3. Nakatsuka, S.; Nakate, I.; Miyano, T. (1996) Drinking water treatment by using ultrafiltration hollow fiber membranes. Desalination, 106: 55–61. 4. Lamminen, M.O.; Walker, H.W.; Weaver, L.K. (2007) Mechanisms and factors influencing the ultrasonic cleaning of particle-fouled ceramic membranes. J. Membr. Sci., 237: 213–223. Downloaded by [Kok Kwang Ng] at 19:55 03 January 2013 222 K. K. NG ET AL. 5. Bowen, W.R.; Calvo, J.I.; Hernandez, A. (1995) Steps of membrane blocking in flux decline during protein microfiltration. J. Membr. Sci., 101: 153–165. 6. Xu-Jiang, J.; Dodds, J.; Leclerc, D.; Lenoel, M. (1995) A technique for the study of the fouling of microfiltration membranes using two membranes in series. J. Membr. Sci., 105: 23–30. 7. Lindau, J.; Jonsson, A.S. (1999) Adsorptive fouling of modified and unmodified commercial polymeric ultrafiltration membranes. J. Membr. Sci., 160: 65–76. 8. Chai, X.; Kobayashi, T.; Fujii, N. (1998) Ultrasound effect on cross-flow filtration of polyacrylonitrile ultrafiltration membranes. J. Membr. Sci., 148: 129–135. 9. Kobayashi, T.; Chai, X.; Fujii, N. (1999) Ultrasound enhanced cross-flow membrane filtration. Sep. Purif. Technol., 17: 31–40. 10. Pirkonen, P. (2001) Ultrasound in Filtration and Sludge Dewatering. In: Advances in Sonochemistry, Mason, T.; Tiehm, A., eds.; Elsevier Science, Vol. 6, 221–246. 11. Chen, D.; Weavers, L.K.; Walker, H.W. (2006) Ultrasonic control of ceramic membrane fouling by particles: Effect of ultrasonic factors. Ultrasonics Sonochemistry, 13: 379–387. 12. Lamminen, M.O.; Walker, H.W.; Weavers, L.K. (2006) Cleaning of particle-fouled membranes during cross-flow filtration using an embedded ultrasonic transducer system. J. Membr. Sci., 283: 225–232. 13. Sabri, N.; Pirkonen, P.; Sekki, H. (1997) 2nd Applications of Power ultrasound in physical and chemical processing. Toulense, 99–104. 14. Kokugan, T.; Kaseno, S.; Fufiware, S.; Shimizu, M. (1999) Ultrasonic effect on ultrafiltration properties of ceramic membrane. Membrane, 20: 213–223. 15. Muthukumaran, S.; Yang, K.; Seuren, A.; Kentish, S.; Ashokkumar, M.; Steven, G.W.; Grieser, F. (2004) The use of ultrasonic cleaning for ultrafiltration membranes in the dairy industry. Sep. Purif. Technol., 39: 99–107. 16. Li, J.; Sanderson, R.D.; Jacobs, E.P. (2005) Ultrasonic cleaning of nylon microfiltration membrane fouled by Kraft paper mill effluent. J. Membr. Sci., 205: 247–257. 17. Matsumoto, Y.; Miwa, T.; Nakao, S.I.; Kimura, S. (1996) Improvement of membrane permeation performance by ultrasonic microfiltration. J. Chem. Eng. Jpn., 29: 561–567. 18. Kyllönen, H.M.; Pirkonen, P.; Nystrom, M. (2005) Membrane filtration enhanced by ultrasound: A review. Desalination, 181: 319–335. 19. Kobayashi, T.; Kobayashi, T.; Hosaka, Y.; Fujii, N. (2003) Ultrasound-enhanced membrane-cleaning processes applied water treatments: Influence of sonic frequency on filtration treatments. Ultrasounics, 41: 185–190. 20. Chai, X.; Kobayashi, T.; Fujii, N. (1999) Ultrasound-associated cleaning of polymeric membranes for water treatment. Sep. Purif. Technol., 15: 139–146. 21. Wakeman, R.J.; Tarleton, E.S. (1991) An experimental study of electroacoustic cross-flow microfiltration. Chem. Eng. Res. Design, 69: 387–397. 22. Lozier, J.; Sierka, R. (1985) Using ozone and ultrasound to reduce RO membrane fouling. J. AWWA, 77: 60–65. 23. Masselin, I.; Chasseray, X.; Durand-Bourlier, L.; Laine, J.M.; Syzaref, P.Y.; Laemordant, D. (2001) Effect of sonicationon polymeric membranes. J. Membr. Sci., 213–200. 24. Juang, R.S.; Lin, K.H. (2004) Flux recovery in the ultrafiltration of suspended solutions with ultrasound. J. Membr. Sci., 115–124. 25. Yuk, Y.J.; Youm, K.H. (2003) Enhancement of ultrafiltration performance using ultrasound. Memburein, 13: 283–290. 26. Simon, A.; Gondrexon, N.; Taha, S.; Cabon, J.; Dorange, G. (2000) Low frequency ultrasound to improve dead-end ultrafiltration performance. Sep. Sci. Technol., 35: 2619–2637. 27. Lin, T.T.; Lin, C.F.; Wei, W.C.J.; Lo, S.J. (1993) Mechanisms of metal stabilization in cementitous metals: Interaction of tricalcium aluminate and copper oxide=hydroxide. Environ. Sci. Technol., 27: 1312–1316. 28. Le-Clech, P.; Chen, V.; Fane, T.A.G. (2006) Fouling in membrane bioreactors used in wastewater treatment. J. Membr. Sci., 284: 17–53. 29. He, Y.; Xu, P.; Li, C.; Zhang, B. (2005) High-concentration food wastewater treatment by an anaerobic membrane bioreactor. Water Res., 39: 4110–4118. 30. Lin, C.F.; Lin, A.Y.C.; Chandana, P.S.; Tsai, C.Y. (2009) Effect of mass retention of dissolved organic matter and membrane pore size on membrane fouling and flux decline. Water Res., 43: 389–394. 31. Katsoufidou, K.; Yiantsios, S.G.; Karabelas, A.J. (2005) A study of ultrafiltration membrane fouling by humic acids and flux recovery by backwashing: Experiments and modeling. J. Membr. Sci., 266: 40–50. 32. Wang, F.; Tarabara, V.V. (2008) Pore blocking mechanisms during early stages of membrane fouling by colloids. J. Coll. Inter. Sci., 328: 464–469. 33. Cai, M., Wang, S., Zheng, Y., Liang, H. (2009) Effect of ultrasound of Radix astraglus extract and cleaning of fouled membrane. Sep. Purif. Technol., 68: 351–356. 34. Wu, J.L.; Le-Clech, P., Stuetz, R.M.; Fane, A.G.; Chen, V. (2001) Effects of relaxation and backwashing conditions on fouling in membrane bioreactor. J. Membr. Sci., 324: 26–32. 35. Bourgeous, K.N.; Darby, J.L.; Tchobanoglous, G. (2001) Ultrafiltration of wastewater: Effect of particles, mode of operation and backwash effectiveness. Water Res., 35: 77–90. 36. Mason, T.J.; Lorimer, J.P. (2006) Applied Sonochemistry: The Uses of Power Ultrasound in Chemical and Processing Ellis; Wiley-VCH VerlagGmbh: Weinheim. 37. Mason, T.J.; Lorimer, J.P. (1988) Sonochemistry: Theory, Applications and Uses of Ultrasound in Chemistry; Ellis Horwood: Chichester.