Desalination 365 (2015) 70–78

Contents lists available at ScienceDirect

Desalination

j ourna l homepage: www.e lsev ie r .com/ locate /desa l

Fabrication and anti-biofouling properties of alumina and zeolitenanoparticle embedded ultrafiltration membranes

Lei-xi Dong a, Hong-wei Yang a, Shi-ting Liu a, Xiao-mao Wang a,⁎, Yuefeng F. Xie a,b

a State Key Joint Laboratory of Environmental Simulation and Pollution Control, School of Environment, Tsinghua University, Beijing 100084, Chinab Environmental Engineering Programs, The Pennsylvania State University, Middletown, PA 17057, USA

H I G H L I G H T S G R A P H I C A L A B S T R A C T

• Alumina and LTL zeolite nanoparticlessuited the in situ embedment method.

• The nanoparticle embedded mem-branes had higher hydrophilicity andfilterability.

• Embedded nanoparticles exhibited anti-adhesion ability but no bacteriocidalfunction.

• Anti-adhesion ability was responsiblefor the anti-biofouling performance.

⁎ Corresponding author.E-mail address: wangxiaomao@tsinghua.edu.cn (X. W

http://dx.doi.org/10.1016/j.desal.2015.02.0230011-9164/© 2015 Elsevier B.V. All rights reserved.

a b s t r a c t

a r t i c l e i n f o

Article history:Received 3 November 2014Received in revised form 11 February 2015Accepted 12 February 2015Available online 18 February 2015

Keywords:Ultrafiltration membraneSurface modificationNanomaterialsAnti-biofoulingAnti-adhesionHydrophilicity

Nanoparticle embedded polysulfone ultrafiltration (UF) membranes were prepared by using the in situ embed-ment method, and the anti-biofouling properties of the prepared membranes were evaluated by conductingbacteria adhesion test, bacterium inactivation test and biofilm formation test separately. Among the several alu-minum and/or silicon oxide nanoparticles tested, alumina (Al2O3) and Linda type L (LTL) zeolite nanoparticleswere successfully embedded which could be evenly dispersed on membrane surface with high coverage ratio(38% and 49%, respectively) and were resistant to hydraulic shear detachment. The water contact angles forthe nanoparticle embedded membranes (UF-Al2O3 and UF-LTL) and the control membrane (UF-C) were 57°,40° and 66°, respectively. Owing to the higher surface hydrophilicity, both UF-Al2O3 and UF-LTL demonstrateda higher filterability than UF-C. Biofouling was inhibited on both UF-Al2O3 and UF-LTL, indicated by the lowerPseudomonas aeruginosa biofilm formation rate. Further investigation showed that both UF-Al2O3 and UF-LTL ex-hibited a high anti-adhesion efficiency to both Escherichia coli and P. aeruginosa, but no bacteriocidal effect onE. coli. The anti-biofouling ability of UF-Al2O3 and UF-LTL mainly benefited from the anti-adhesion ability attrib-uted to the embedded nanoparticles. The improved anti-adhesion ability could not be simply explained by theenhanced hydrophilicity.

© 2015 Elsevier B.V. All rights reserved.

ang).

1. Introduction

Low-pressure membrane filtration has experienced acceleratedgrowth over the past decades in applications for drinkingwater produc-tion and wastewater treatment. Membrane fouling, however, restricts

71L. Dong et al. / Desalination 365 (2015) 70–78

its wider application [1]. Membrane fouling can be roughly divided intoorganic fouling, inorganic fouling and biological fouling (i.e., biofilmformation). Previous studies showed that membrane fouling rate isclosely related with the hydrophilicity of the membrane. Under other-wise identical conditions, a more hydrophilic membrane is usually lesssusceptible to fouling. As such, a lot of efforts from both the industrialand scientific societies have been devoted to increase the membranehydrophilicity. It was estimated that over 50% of the commercial low-pressure membranes were surface-modified [2], mainly for increasinghydrophilicity.

In general, membrane modification approaches can be grouped intofour categories, i.e., blending additives, physical coating, chemical coating,and heterogeneous reactions [2,3].Minor amount of hydrophilic additivessuch as polyvinylpyrrolidone (PVP) and polyethylene glycol (PEG) blend-ed into themembrane cast solution seems to be the simplestmodificationmethod. Physical coating refers to the coating process by which a coatinglayer is formedbut no chemical bonds exist between the coating layer andthe basemembrane. Due to its high hydrophilicity and good film-formingproperties, polyvinyl alcohol (PVA) is often used in physical coating toimprove membrane hydrophilicity forming a thin layer on the originalsurface [4]. Chemical coating, in contrast to physical coating, is a processbywhich a polymer layer is coated onmembrane surfacewhere chemicalbonds form through “grafting to” or “grafting from” approaches. For in-stance, low-fouling UF membranes could be achieved by simultaneousphoto-graft (UV irradiation) copolymerization of PEG methacrylate(PEGMA) onto membrane surface [5]. Heterogeneous reactions, in-cluding sulfonation, carboxylation, hydroxylation and physical acti-vation by high energy radiation, plasma and UV irradiation, takeplace on the membrane surface and alter membrane surface propertiesusually not generating a coating layer. A synergy of differentmodificationstrategies broadens the dimensions to improve the hydrophilicity of themodified membranes.

With thedevelopment of nanotechnology, nanomaterials emerge upas a novel “functional material” in membrane modification area. Thenanomaterials are either blended in the membrane matrix or coatedon the membrane surface usually with the aid of tethering chemicals.Previous studies showed that inorganic nanoparticles, including alumina(Al2O3) [6–10], zinc oxide (ZnO) [11–14], silicon dioxide (SiO2) [15–17],titanium dioxide (TiO2) [18–20] and zeolites [21,22], could endowmem-branes with more hydrophilic surface. In addition, nanomaterials mightbe able to impart membranes with excellent anti-biofouling ability [23].Silver nanoparticles are oneof themost extensively investigatednanopar-ticles that might show the bacteriocidal ability [24,25]. The bacteriocidalability of silver nanoparticle-modifiedmembraneswas primarily attribut-ed to the release of silver ions from the membranes. Copper, long knownfor its bacteriocidal activity and relatively low cost, was also used formembrane modification [26,27]. In contrast, the antibacterial potentialof TiO2 hybrid membranes is manifested when exposed to UV illumina-tion. Recently, carbon nanotubes [28] and graphene oxide nanosheets[29,30] are the hotspots in membrane modification fields. It is reportedthat both nanomaterials could show antimicrobial abilities.

Except for silver and copper nanoparticles, the anti-biofouling prop-erties of surfacemodifiedmembranes byusing other nanoparticles haveseldom been investigated. The less expensive and more chemically sta-ble nanoparticles such as alumina and zeolite might also exhibit anti-biofouling capability owing to the high specific surface area and someother unique but unknown physico-chemical properties [31]. In addi-tion, our previous study [6] showed that alumina nanoparticles couldbe easily embedded onto polyvinylidene fluoride (PVDF) membranesurface simultaneously when the membrane was formed, by using thedistinct in situ embedment method. The method is easy to apply, andthe embedded particleswere all exposed on themembrane surface like-ly by physical anchoring (rather than buried in the membrane matrix).It was expected that alumina and other similar nanoparticles (e.g., zeo-lite) can also be embedded onto membrane surface of other materialssuch as polysulfone (PSf). Both PVDF and PSf are the common polymers

used for low-pressuremembrane fabrication. The anti-biofouling proper-ties were disintegrated into the anti-adhesion and bacteriocidal proper-ties and evaluated by a number of different techniques such as bacteriaadhesion test, bacterium inactivation test and biofilm formation test.The primary purpose of this study is to further elucidate the functions ofthe embedded nanoparticles on the membrane surface.

2. Materials and methods

2.1. Membrane fabrication

Each nanoparticle embedded ultrafiltrationmembranewas fabricatedvia the in situ embedment method [6], which is a slight modification ofthemost popularly adopted non-solvent induced phase separationmeth-od. The cast solutionwas prepared by dissolving 15 g of polysulfone (PSf)(averagemolecularweight (MW) ~ 22 kDa, Aldrich, USA) and 8 g of poly-vinylpyrrolidone (PVP) (average MW ~ 10 kDa, Sigma-Aldrich, USA) into77 g of anhydrous 1-methyl-2-pyrrolidinone (NMP) (Sigma-Aldrich,USA) under vigorous stirring at room temperature for 24 h. The air bub-bles in the cast solution, if any, could be dissipated after about 12 h atroom temperature. A thin film (200 μm) of the obtained cast solutionwas casted onto a flat glass plate (DP-GB-2B, Paul N. Gardner, USA) byusing a doctor blade (AP-G06/10, Paul N. Gardner, USA). The filmwas ex-posed in the air for about 10 s before it was immersed into thewater bathin which nanoparticles were uniformly dispersed (after sonication for2 h) at a weight concentration of 500mg/L. In this way, the nanoparticleswere expected to spontaneously embed onto the membrane top surface.The temperature of the water bath was controlled at 35 °C. The mem-brane was removed from the water bath, thoroughly rinsed with Milli-Q water and stored in regularly replaced Milli-Q water (4 °C) prior tofuture tests. Nanoparticles tested in this study included alumina(Al2O3, b50 nm according to the supplier, Aldrich, USA), threetypes of zeolitewhichwere Linde typeA zeolite (LTA, 100 nm,NanoScapeAG, Germany), Linde type L zeolite (LTL, 80 nm, NanoScape AG,Germany), and Faujasite type X zeolite (FTX, 150 nm, NanoScapeAG, Germany), silicon dioxide (SiO2, b50 nm, Aldrich, USA), and sin-gle wall carbon nanotube (SWNT, b50 nm, Aldrich, USA). The ultra-filtration membranes fabricated with nanoparticles embedded onmembrane surface were assigned as UF-X, where X stood for thename or acronym of the nanoparticles (Al2O3, LTA, LTL, FTX, SiO2

and SWNT). The possible agglomeration of the nanoparticles in thewater bath was evaluated by comparing the mean particle size inMilli-Q water and that provided by the suppliers. The particle sizedistribution (PSD) of the nanoparticles in water was determinedby using a PSD analyzer (Delsa Nano Analyzer, Beckman Coulter,USA).

Control ultrafiltration membrane (with no nanoparticle embedded)was fabricated according to the same procedures described above,except that the water bath contained only Milli-Q water but no nano-particles. The membrane was denoted as UF-C.

2.2. Membrane characterization

Membrane resistance (or filterability) was determined by conductingdead-end filtration of Milli-Q water using a stirred filtration cell (Amicon8050, Millipore, USA) operated at 0.1 MPa of applied pressure. Themembrane coupons were first pre-compacted at 0.15 MPa for 30 min.The filtration flux (Jw) was obtained from the recorded filtrate weight asa function of thefiltration time. Themembrane resistance (Rm)was calcu-lated according to the Darcy equation,

Rm ¼ ΔPμ Jw

ð1Þ

where ΔP is the applied pressure and μ is the water viscosity.

72 L. Dong et al. / Desalination 365 (2015) 70–78

The membrane surface hydrophilicity was indicated by water con-tact angle, which was measured by using a video-supported measuringinstrument (OCA20, Dataphysics, Germany) based on the sessile-dropmethod [12]. Membranes were dehydrated in a desiccator with silicagel used as the desiccant prior to the measurement. A volume of 1 μLMilli-Q water was dropped onto the membrane surface and the EllipseFitting method was employed to fit the shape profile of the waterdrop so as to calculate the value of contact angle. For each membrane,more than 10 measurements were carried out at different spots andthe average of all contact angle values was reported in this paper.

Morphologies of the top and bottom surfaces as well as the crosssection of eachmembranewere observed under a scanning electronmi-croscope (SEM) (Quanta 200F, FEI, USA) in high vacuum mode aftersputter-coatingwith ~20 Å of Au/Pd (Precision Etching and Coating Sys-tem 682, Gatan, USA). The coverage ratio of nanoparticles embedded onmembrane surface was determined by analyzing the SEM images usingthe ImageJ software.

Atomic force microscope (AFM, Veeco NanoScope 5, USA) was usedto measure the membrane surface roughness. Related parameters (in-cluding Ra and Rq) were determined through the AFM analysis software(NanoScope Software 7.20, UA). Ra is the arithmetic average of the abso-lute values of the surface height deviations (Z)measured from themeanplane, while Rq is the standard deviations of the Z values.

The stability of nanoparticles on membrane surface was evaluatedby comparing the nanoparticle coverages before and after strong hy-draulic shear applied to the membrane surface. The membrane wasfixed in a stirred filtration cell (Amicon 8200, Millipore, USA) in which150mLMilli-Qwaterwas vigorously stirred for 2 d. Themembrane cou-pons were observed under SEM (Quanta 200F, FEI, USA).

2.3. Anti-biofouling property evaluation

2.3.1. Model bacteria suspensionTwo model bacteria species were used for the evaluation of anti-

biofouling property of the different membranes. Escherichia coli(E. coli, ATCC 25922) was selected in that E. coli is among the mostcommonly found bacteria in natural water and wastewater, andPseudomonas aeruginosa (P. aeruginosa, ATCC 27853) was chosenbecause it has been often reported to promote or accelerate mem-brane biofouling via biofilm formation [32,33].

Luria–Bertani (LB) liquid medium was used for the cultivation ofboth bacteria species. E. coli was cultivated in fresh LB medium at37 °C for 3–4 h to the mid-log-growth phase (with an optical densityvalue at 600 nm (OD600) of 0.5–0.6). The cultivated E. coliwere separat-ed from the LB medium by 5 min centrifugation (5000 rpm, 4 °C). Thepellet was then re-suspended in PBS (phosphate buffered saline) solu-tion (KH2PO4 0.24 g/L, Na2HPO4 1.44 g/L, NaCl 8.0 g/L, KCl 0.2 g/L).This centrifugation and re-suspension procedure was repeated for threetimes to remove any residual LB medium. The obtained medium-freeE. coli suspension, with concentration approximately at 0.5 × 109 CFU/mL,is hereafter denoted as E. coli suspension. P. aeruginosa was cultivatedwith shaking (120 rpm) in LB medium at 37 °C for 12 h to reach themid-log-growth phase (with an OD600 value of 2.0–3.0). The cultivatedbacteriawere also harvested by the centrifugation and re-suspension pro-cedure that was repeated for three times. The medium-free P. aeruginosasuspension, with concentration approximately at 1.5 × 109 CFU/mL, ishereafter denoted as P. aeruginosa suspension. Both bacteria suspensionswere prepared no more than 1 h prior to use.

2.3.2. Bacteria adhesion testThe adhesion test of model bacteria E. coli and P. aeruginosa onto

the membrane surface was conducted by immersing the membranespecimen into the respective bacteria suspension. A piece of eachmem-brane (approximately 15 × 40 mm2), which had been sterilized by UVradiation (40W, 253.7 nm) for 30 min, was placed into a test tube con-taining 10 mL bacteria suspension and incubated at 37 °C for 12 h. The

membrane specimen was taken out of the suspension and gentlywashed with PBS to remove any loosely attached bacteria on the mem-brane. The adhered bacteria on the membrane were observed underSEM after the sample pretreatment procedure. In detail, the bacteriaonmembrane surfacewerefixedwith 2.5% (v/v) glutaraldehyde PBS so-lution at 4 °C for 5 h. After the fixation, the membrane specimen wasrinsed with PBS to remove any remaining glutaraldehyde. The mem-brane was then step-dehydrated with 25%, 50%, 75% and 100% ethanol,respectively. It was ready for SEM observation after an overnight air-drying. Bacteria coverage on the membrane surface, expressed as celldensity, was determined by counting cell numbers per unit area. Thebacteria coverages on the control membrane and on the nanoparticleembedded membranes were compared to evaluate the anti-adhesionproperty of the nanoparticle embedded membranes [32].

2.3.3. Bacterium inactivation testThe bacterium inactivation test was conducted by using E. coli. The

E. coli suspension was diluted with PBS to the concentration approxi-mately at 50 CFU/mL. A volume of the diluted suspension (10 mL) wasfiltered onto the membrane surface, which was then transferredinto a petri dish in which 15 mL chromogenic solid medium forE. coli (CHROMagar™ ECC) had been added. After incubation at 37 °Cfor 24 h, the blue colonies grown on membrane surface (about 1 mm indiameter) were counted. The colony number on nanoparticle embeddedmembranes was compared with that on the control membrane to indi-cate the possible bacteriocidal ability of the embedded nanoparticles.Four replications were performed.

2.3.4. Biofilm formation testThe rate of biofilm formation on membrane surface was evaluated

using P. aeruginosa [34]. Each membrane specimen, approximately10 × 10 mm2 in size, was affixed to a glass slide with the top surfacefacing up, which was immersed into the P. aeruginosa suspension(approximately at 1.5 × 109 CFU/mL) that was then incubated at 37 °Cfor 24 h or 72 h. During the 72 h incubation, P. aeruginosa suspensionwas replaced every 24 h to maintain a high concentration of bacteria.Following the gentle rinse with PBS, themembrane sample was stainedwith SYTO® 9 green fluorescent nucleic acid dye (LIVE/DEAD® BiofilmViability Kit, Invitrogen, Ltd., UK) for 15 min and then washed by PBSfor three times to remove any remaining dye. The formed biofilm, ifany, was observed under a confocal laser scanning microscope (CLSM,LSM710, Zeiss, Germany). Imageswere taken fromat leastfive random-ly selected areas of each membrane. The average thickness of theP. aeruginosa biofilm on the nanoparticle embedded membrane wascompared with that on the control membrane to indicate the anti-biofouling ability of the nanoparticle embedded membrane.

In addition to CLSM, the crystal violet staining method [35] couldalso be used for the quantification of the P. aeruginosa biofilm formationon the various membranes. According to the method, the formedbiofilm was stained by crystal violet, which was then extracted byethanol/acetone mixture followed by absorbance determination at595nm. Themethodwaswidely applied to the evaluation of the antimi-crobial activity ofmany new (nano)materials [35]. In our case, however,the membrane material (primarily PSf) could also be strongly stained bycrystal violet, which greatly impaired the quantification accuracy espe-cially when the biofilm thickness was low. It was therefore not possibleto supplement the CLSM method by using the crystal violet stainingmethod.

3. Results and discussion

3.1. In situ embedment of nanoparticles

The in situ embedment approach was employed for fabrication ofnanoparticle embedded ultrafiltration membranes. Among the severaltypes of nanoparticles tested in this study, only Al2O3 and LTL zeolite

73L. Dong et al. / Desalination 365 (2015) 70–78

suited the approach most. A first requisite when applying the approachis that the nanoparticles must be well dispersed in the water bath (i.e.,coagulation bath for phase separation). Probably due to the highhydrophobicity, SWNT could not be dispersed well in water. ThoughSiO2 nanoparticles would not settle from the water in a couple hours,particle size distribution (PSD) measurement result showed that theyagglomerated to a high extent; the measured mean particle sizewas 1191 nm, much larger than that for the primary nanoparticles(b50 nm). The remaining four types of nanoparticles were very stablydispersed in thewater (some ofwhichwould not settle in a fewmonths).Nevertheless, PSD measurements indicated that these nanoparticlesmight slightly aggregate in the water. The measured mean particle sizesfor Al2O3, and LTA, LTL and FTX zeolite were 302, 390, 211 and 340 nm,respectively, 2–6 times larger than that for the primary nanoparticlesspecified by the suppliers. It appears that the spontaneous aggregationof nanoparticles in water is fairly common, which might be driven by

Fig. 1. SEM images of ultrafiltration membranes in situ embedded with (a) Al2O3, (b) LTA, (c)subjected to 2 d of hydraulic shear.

the long-range surface–surface interactions (e.g., van der Waals at-traction, electrostatic interaction and acid–base interaction), hydro-gen bonding and other forces. The rate of aggregation howeverdiffered substantially among the nanoparticles. In this study, sonica-tion was applied to disperse the nanoparticles in the water bath justbefore cast film immersion.

The second requisite when applying the approach is that the nano-particles in the water bath could transport to the cast solution film sur-face and be tightly attached when the polymer (PSf in this study) wassolidified. SEM observation showed that all four types of nanoparticleswhich dispersedwell inwater could be embedded to the PSfmembranesurface. However, the nanoparticle coverage ratios differed substantial-ly among the different nanoparticles, although the nanoparticle concen-trations in the water bath were all 500 mg/L. In comparison, Al2O3

(Fig. 1a) and LTL zeolite (Fig. 1c) nanoparticles showed excellent disper-sion on membrane surface with fairly high coverage ratio at 38% and

LTL and (d) FTX nanoparticles, and (e) UF-Al2O3 and (f) UF-LTL membranes after being

74 L. Dong et al. / Desalination 365 (2015) 70–78

49%, respectively, while LTA zeolite (Fig. 1b) and FTX zeolite (Fig. 1d)nanoparticles poorly distributed and the coverage ratios were only4.7% and 6.4%, respectively. In our previous study [6], γ-alumina wassuccessfully embedded to PVDF membrane surface. It was argued thatthe free energy change for the nanoparticle attachment determinedwhether or not the nanoparticles could be embedded. A negative freeenergy changewas favorable. However, the surface tension and its com-ponents could not be accurately determined for any type of the zeolitenanoparticles in this study. It was possibly due to the porous structureof zeolite. Therefore, the exact underlying reason for that LTL zeolitenanoparticles differed from the other two zeolite nanoparticles couldnot be proposed. Although theoretical analysis could not be made, thisstudy experimentally showed that alumina and LTL zeolite nanoparti-cles were suitable for PSf membrane surface modification by using thein situ embedment approach.

The stability of the embedded nanoparticles on the membrane sur-face was assessed by exposing the membrane to hydraulic shear forcegenerated by magnetic stirring of the above water in a filtration cell.Results (Fig. 1e and f) showed that no discernable nanoparticle detach-mentwas observed after a duration of 2 d. Itwas argued thatmost of thenanoparticles were physically anchored into the membrane matrix andas such they were resistant to hydraulic shear force.

3.2. Membrane characterization

Ultrafiltration membranes were expected to form under the con-ditions adopted for membrane fabrication (cast solution recipe, filmthickness, water bath temperature, etc.). Based on the SEM images,the membrane pore size and thickness for UF-C were determined tobe 30 nm and 180 μm, respectively. The membrane resistance to filtra-tion was determined to be 1.3 × 1012 m−1 (Fig. 2). The water dropletcontact angle was measured to be 66°, indicating that the membranewas fairly hydrophilic. It was due to the blending of PVP as hydrophilicadditive in the cast solution. Three-dimensional AFM images of themembrane top surface (Fig. 3) showed that the membrane had a fairlysmooth surface, with Ra and Rq values being 15.4 ± 3.0 and 19.0 ±3.6 nm, respectively.

Embedment of nanoparticles could substantially increase thehydrophilicity of the membranes. The water droplet contact angles forUF-Al2O3 and UF-LTL were measured to be 57° and 40°, respectively(Fig. 2). The higher hydrophilicity for UF-LTL was primarily due to thehigher coverage ratio of LTL zeolite nanoparticles on themembrane sur-face (see above). Embedment of nanoparticles did not change themem-brane thickness andmorphology significantly (SEM images not shown).As SEM observation revealed, it might block the membrane pores to

Fig. 2. Comparison of water droplet contact angle and membrane resistance among UF-C,UF-Al2O3 and UF-LTL. The insets were the images for contact angle measurement.

some extent. However, contrary to as expected, the membrane resis-tances for UF-Al2O3 and UF-LTL were determined to be 9.1 × 1011 and9.4 × 1011 m−1, respectively (Fig. 2), both of which were less thanthat forUF-C. It indicates that the embedded particlesmight not actuallyblock the membrane pores for water passage. In addition, membraneresistance is determined not only by membrane pore size and mem-brane structure (i.e., microvoids and macrovoids) but also by mem-brane hydrophilicity [21]. The increased permeability of nanoparticleembedded membranes could be attributed to the enhanced membranehydrophilicity.

In addition to increasing membrane permeability, hydrophilicity isone of the most important properties of membranes that determinethe antifouling ability [16,36,37]. A number of previous studies haveattempted to increase membrane hydrophilicity by doping nanoparti-cles mostly through the blending additive method. For example, Liaoet al. [38] blended with NaY zeolite particles and Ag+ exchanged NaYzeolite particles at a concentration of 0.8 wt.%, which resulted in thecontact angle value dropping from 91° to 82°. Leo et al. [21] illustratedthat by increasing the concentration of SAPO-44 zeolites from 5 wt.%to 15 wt.%, the contact angle value decreased from 70.6° to 64.5°. Italso enhanced the antifouling ability by reducing the adsorption ofhumic acid in membrane pores and thus alleviating fouling initiatedby pore blocking. Compared to the blending method adopted in previ-ous studies [9,39,40], the in situ embedment approach adopted in thisstudy demonstrated more potential in the improvement of membranehydrophilicity (especially when LTL zeolite nanoparticles are used).

The surface roughness of the nanoparticle embedded membraneswas also measured from the AFM images (Fig. 3). The Ra and Rq valuesfor UF-Al2O3 were 85.3 ± 11.7 and 111.4 ± 18.6 nm, respectively;while those for UF-LTL were 29.8± 6.6 and 37.4± 7.6 nm, respectively.It revealed that the embedment of nanoparticles would lead to a morerough membrane surface. It was as expected because the nanoparticleswere embedded onto rather than buried into the membrane polymermatrix. Noted is that the particle size for the primary Al2O3 and LTLzeolite nanoparticleswas b50 and 80 nm, respectively. The higher surfaceroughness of UF-Al2O3 was due to the agglomeration of the Al2O3 nano-particles before they were embedded to the membrane surface (see theabove measured particle size in water). It appears that agglomeration ofthe LTL zeolite particles was minimal. Noted is that in addition to surfaceroughness, surface rigidity and elasticity could be determined by usingAFM as well [41]. However, only surface roughness was measured be-cause it is more related tomembrane fouling. Though themembrane sur-face rigidity and elasticity could be changed due to the embedment ofnanoparticles, the effect was however not considered in this study.

3.3. Anti-adhesion property of the membranes

The anti-adhesion propertywas assessed by comparing the coverageratios of model bacteria on the nanoparticle embedded membraneswith that on the control membrane (UF-C). The attachment of modelbacteria E. coli and P. aeruginosa on membrane surface were observedunder SEM.

It was found that both E. coli and P. aeruginosa could easily adhereonto the control membrane (UF-C) (Fig. 4a and d). After a 12 h contactwith each bacterium suspension, the membrane surface was almostfully covered by the bacteria cells (ca. 2.4 × 1011 cells/m2 of E. coli and5.4 × 1011 cells/m2 of P. aeruginosa) (Fig. 5). Embedment of nanoparti-cles could greatly reduce the adhesion of both types of bacteria ontothe membrane surface (Fig. 4b, c, e and f), demonstrated by the resultsthat the coverage ratios of either E. coli or P. aeruginosa cells on UF-Al2O3

and UF-LTL were substantially less than that on UF-C. If we define ananti-adhesion efficiency as the reduced percentage of the bacteria cov-erage rate compared with the control membrane, the UF-Al2O3 mem-brane exhibited an anti-adhesion efficiency of 78% to E. coli and 94% toP. aeruginosa, while UF-LTL membrane manifested an anti-adhesion ef-ficiency of 61% to E. coli and 97% to P. aeruginosa (Fig. 5). Though it may

Fig. 3. Three-dimensional AFM surface images of the UF-C, UF-Al2O3 and UF-LTL membranes.

75L. Dong et al. / Desalination 365 (2015) 70–78

not be appropriate to simply compare the anti-adhesion efficiencies, theabove results showed that Al2O3 and LTL zeolite nanoparticles had sim-ilar anti-adhesion ability to both E. coli and P. aeruginosa. The improvedanti-adhesion ability could be attributed to the increased membranesurface hydrophilicity due to nanoparticle embedment. A number ofprevious studies [33,42–44] have reported a clear correlation betweenthe increase in hydrophilicity and corresponding decrease of bacteriaattachment. Noted is that the control membrane was hydrophilic in na-ture. It was thus argued that the attachment/adhesion of bacteria on themembrane surface was a gradual process. The membrane surface prop-erty could bemodified by the bacteria and their secretedmatter. It couldbe expected that the nanoparticle embeddedmembraneswould be fullycovered by the bacteria if a longer contact time was allowed. It wasdemonstrated by using P. aeruginosa (see below). However, hydrophilic-ity onlymight not explain the enhanced anti-adhesion ability of the nano-particle embedded membranes, considering that the anti-adhesionefficiencies were similar but the water droplet contact angles differedsubstantially between UF-Al2O3 and UF-LTL.

3.4. Bacteriocidal property of the membranes

It was proposed that nanoparticle may have bacteriocidal ability byphysically disrupting the bacterial cell wall due to its extremely smallparticle size and high surface tension energy [45]. The bacteriocidalproperty of the embedded nanoparticles was evaluated by comparingthe colony numbers formed on different membranes. It was hypo-thesized that if a bacterial cell was inactivated by its contacted nano-particle(s), the bacterial cell could not develop into a colony in the

Fig. 4. SEM images of membranes (a) UF-C, (b) UF-Al2O3, (c) UF-LTL attached w

subsequent cultivation step. E. coliwas used for the assessment. Results(Figs. 6 and 7) showed that E. coli could grow on either UF-Al2O3 or UF-LTL as well as on the control membrane (UF-C). A total of four replicateswere conducted, and the normalized survival rate was calculated by di-viding the colony numbers on the nanoparticle embedded membraneby that on the control membrane which was set as a control for everyfiltration process. The normalized survival rates were both equal to100%, indicating that the embedded nanoparticles had no bacteriocidalability.

3.5. Anti-biofouling property of the membranes

Biofouling is initiated by the adhesion of bacteria on the membranesurface followed by the development into a biofilm. The anti-biofoulingproperty of the membranes was assessed by comparing P. aeruginosabiofilm formation rates on the different membranes. The membranespecimens were immersed in P. aeruginosa suspension at 37 °C for upto 72 h to allow biofilm formation. The P. aeruginosa biofilmwas stainedand then observed under CLSM. As shown in Figs. 8 and 9, P. aeruginosacould easily develop into a biofilm on the control membrane; the bio-film thickness was measured to be 34.0 μm after incubation for 24 h.In contrast, the biofilm formation on either UF-Al2O3 or UF-LTLwas sub-stantially inhibited; the biofilm thicknesseswere 3.7 and 8.5 μm, respec-tively. When the incubation was extended to 72 h (during which theP. aeruginosa suspensionwas replaced every 24 h), the biofilm thicknesson the control membrane was further increased to 55.8 μm, while thebiofilm thickness on UF-Al2O3 and UF-LTL was kept almost unchangedat 6.3 and 5.8 μm, respectively. In accordance with the fact that either

ith E. coli and (d) UF-C, (e) UF-Al2O3, (f) UF-LTL attached with P. aeruginosa.

Fig. 5. Cell densities of E. coli and P. aeruginosa attached on membrane surface afterincubation at 37 °C for 12 h.

Fig. 7. The normalized survival rate of E. coli filtered ontomembrane surface after cultivationat 37 °C for 24 h.

76 L. Dong et al. / Desalination 365 (2015) 70–78

Al2O3 or LTL zeolite nanoparticles had no bacteriocidal ability, the em-bedment of these nanoparticles could not prevent the formation ofP. aeruginosa biofilm (i.e., biofouling). The inhibited biofilm formationon the nanoparticle embedded membranes was mostly due to the en-hanced anti-adhesion ability caused by the embedment of nanoparti-cles. In actuality, improvement of anti-adhesion ability which couldslow down the initial attachment of bacteria on membrane surface

Fig. 6. The growth of E. coli on the various membrane surfaces after cultivation at 37 °C for24 h.

might be a more effective approach to reducing biofouling than usingbacteriocidal agents which aim at killing/inactivating bacteria alreadyattached on membrane surface. The dead or inactivated bacteria mayserve as anchors for the next layer of bacteria, promoting further devel-opment of biofouling. Although the biofilm thickness on UF-LTL wasslightly higher than that on UF-Al2O3, it may not be appropriate to con-clude that Al2O3 nanoparticles had a better anti-biofouling performancethan LTL zeolite nanoparticles. The results described earlier showed thatthe two types of nanoparticles had similar anti-adhesion ability.

4. Conclusions

The in situ embedment approach was employed for nanoparticleembedded ultrafiltration (UF) membrane fabrication, and the anti-biofouling ability of the fabricated membranes was then assessed. Re-sults showed that Al2O3 and LTL zeolite nanoparticles could be success-fully embeddedwith high surface coverage (38% and 49%, respectively),which were fairly resistant to hydraulic shear detachment. Water con-tact angles for UF-Al2O3 and UF-LTL were measured as 57° and 40°, re-spectively, significantly lower than that for the control membrane(UF-C) (66°). UF-Al2O3 and UF-LTL had lower intrinsic membrane resis-tance than UF-C, which was attributed to the higher membrane surface

Fig. 8. The thickness of P. aeruginosa biofilm formed on various membrane surfaces afterincubation at 37 °C for 24 h and 72 h, respectively.

Fig. 9. CLSM images of biofilm formation of P. aeruginosa on various membrane surfaces after incubation at 37 °C for 72 h.

77L. Dong et al. / Desalination 365 (2015) 70–78

hydrophilicity. However, the membrane surface roughness was higherdue to the embedded nanoparticles. UF-Al2O3 and UF-LTL exhibitedsimilarly high anti-adhesion efficiency to either E. coli or P. aeruginosa.The adhesion of E. coli could be reduced by 78% and 61%, respectively,while the adhesion of P. aeruginosa could be reduced by 94% and 97%.In contrast, the embedded Al2O3 and LTL nanoparticles showed no bac-teriocidal effect on E. coli. Nevertheless, embedment of nanoparticlescould inhibit biofouling development, demonstrated by the resultsthat the biofilm thickness of P. aeruginosa grown on UF-Al2O3 and UF-LTL membranes was much less than that on UF-C. It is concluded thatthe enhanced anti-biofouling ability of nanoparticle embedded mem-branes mainly benefited from the anti-adhesion ability attributed tothe embedded nanoparticles. The improved ability might not be simplydue to the enhanced hydrophilicity, given that the water contact anglesfor UF-Al2O3 and UF-LTL were far from identical.

Acknowledgments

We acknowledge the funding for this research provided by theNational Natural Science Foundation of China (No. 51290284 and No.51278268).

References

[1] W. Gao, H. Liang, J. Ma, M. Han, Z.-L. Chen, Z.-S. Han, G.-B. Li, Membrane fouling con-trol in ultrafiltration technology for drinking water production: a review, J. Membr.Sci. 272 (2011) 1–8.

[2] K.C. Khulbe, C. Feng, T. Matsuura, The art of surface modification of synthetic poly-meric membranes, J. Appl. Polym. Sci. 115 (2010) 855–895.

[3] M. Ulbricht, Advanced functional polymer membranes, Polymer 47 (2006)2217–2262.

[4] J.R. Du, S. Peldszus, P.M. Huck, X. Feng, Modification of poly(vinylidene fluoride)ultrafiltration membranes with poly(vinyl alcohol) for fouling control in drinkingwater treatment, Water Res. 43 (2009) 4559–4568.

[5] H. Susanto, M. Balakrishnan, M. Ulbricht, Via surface functionalization by photograftcopolymerization to low-fouling polyethersulfone-based ultrafiltration membranes,J. Membr. Sci. 288 (2007) 157–167.

[6] X.-M. Wang, X.-Y. Li, K. Shih, In situ embedment and growth of anhydrous andhydrated aluminum oxide particles on polyvinylidene fluoride (PVDF) membranes,J. Membr. Sci. 368 (2011) 134–143.

[7] P. Daraei, S.S. Madaeni, N. Ghaemi, M.A. Khadivi, L. Rajabi, PAA grafting onto newacrylate-alumoxane/PES mixed matrix nano-enhanced membrane: preparation,characterization and performance in dye removal, Chem. Eng. J. 221 (2013)111–123.

[8] J. Garcia-Ivars, M.-I. Alcaina-Miranda, M.-I. Iborra-Clar, J.-A. Mendoza-Roca, L.Pastor-Alcañiz, Enhancement in hydrophilicity of different polymer phase-inversion ultrafiltration membranes by introducing PEG/Al2O3 nanoparticles, Sep.Purif. Technol. 128 (2014) 45–57.

[9] V. Vatanpour, S.S. Madaeni, L. Rajabi, S. Zinadini, A.A. Derakhshan, Boehmite nano-particles as a new nanofiller for preparation of antifouling mixed matrix mem-branes, J. Membr. Sci. 401–402 (2012) 132–143.

[10] T.A. Saleh, V.K. Gupta, Synthesis and characterization of alumina nano-particlespolyamide membrane with enhanced flux rejection performance, Sep. Purif.Technol. 89 (2012) 245–251.

[11] S. Balta, A. Sotto, P. Luis, L. Benea, B. Van der Bruggen, J. Kim, A new outlook onmembrane enhancement with nanoparticles: the alternative of ZnO, J. Membr. Sci.389 (2012) 155–161.

[12] S. Liang, K. Xiao, Y. Mo, X. Huang, A novel ZnO nanoparticle blended polyvinylidenefluoride membrane for anti-irreversible fouling, J. Membr. Sci. 394–395 (2012)184–192.

[13] C.P. Leo, W.P. Cathie Lee, A.L. Ahmad, A.W. Mohammad, Polysulfone membranesblended with ZnO nanoparticles for reducing fouling by oleic acid, Sep. Purif.Technol. 89 (2012) 51–56.

[14] J. Hong, Y. He, Polyvinylidene fluoride ultrafiltration membrane blended with nano-ZnO particle for photo-catalysis self-cleaning, Desalination 332 (2014) 67–75.

[15] S. Liang, G. Qi, K. Xiao, J. Sun, E.P. Giannelis, X. Huang, M. Elimelech, Organic foulingbehavior of superhydrophilic polyvinylidene fluoride (PVDF) ultrafiltration mem-branes functionalized with surface-tailored nanoparticles: implications for organicfouling in membrane bioreactors, J. Membr. Sci. 463 (2014) 94–101.

[16] S. Liang, Y. Kang, A. Tiraferri, E.P. Giannelis, X. Huang, M. Elimelech, Highlyhydrophilic polyvinylidene fluoride (PVDF) ultrafiltration membranes viapostfabrication grafting of surface-tailored silica nanoparticles, ACS Appl.Mater. Interfaces 5 (2013) 6694–6703.

[17] H. Wu, J. Mansouri, V. Chen, Silica nanoparticles as carriers of antifouling ligands forPVDF ultrafiltration membranes, J. Membr. Sci. 433 (2013) 135–151.

[18] S.-Y. Kwak, S.H. Kim, S.S. Kim, Hybrid organic/inorganic reverse osmosis (RO) mem-brane for bactericidal antifouling. 1. Preparation and characterization of TiO2 nanopar-ticle self-assembled aromatic polyamide thin-film-composite (TFC) membrane,Environ. Sci. Technol. 35 (2001) 2388–2394.

[19] S.H. Kim, S.-Y. Kwak, B.-H. Sohn, T.H. Park, Design of TiO2 nanoparticle self-assembledaromatic polyamide thin-film-composite (TFC) membrane as an approach to solvebiofouling problem, J. Membr. Sci. 211 (2003) 157–165.

[20] H.S. Lee, S.J. Im, J.H. Kim, H.J. Kim, J.P. Kim, B.R. Min, Polyamide thin-film nanofiltrationmembranes containing TiO2 nanoparticles, Desalination 219 (2008) 48–56.

[21] C.P. Leo, N.H. Ahmad Kamil, M.U.M. Junaidi, S.N.M. Kamal, A.L. Ahmad, The potentialof SAPO-44 zeolite filler in fouling mitigation of polysulfone ultrafiltration mem-brane, Sep. Purif. Technol. 103 (2013) 84–91.

[22] M.U. Junaidi, C.P. Leo, S.N. Kamal, A.L. Ahmad, Foulingmitigation in humic acid ultra-filtration using polysulfone/SAPO-34 mixed matrix membrane, Water Sci. Technol.67 (2013) 2102–2109.

[23] J. Kim, B. Van der Bruggen, The use of nanoparticles in polymeric and ceramic mem-brane structures: review of manufacturing procedures and performance improve-ment for water treatment, Environ. Pollut. 158 (2010) 2335–2349.

[24] D.Y. Koseoglu-Imer, B. Kose, M. Altinbas, I. Koyuncu, The production of polysulfone(PS) membrane with silver nanoparticles (AgNP): physical properties, filtrationperformances, and biofouling resistances of membranes, J. Membr. Sci. 428 (2013)620–628.

[25] K. Zodrow, L. Brunet, S. Mahendra, D. Li, A. Zhang, Q. Li, P.J. Alvarez, Polysulfoneultrafiltration membranes impregnated with silver nanoparticles show improvedbiofouling resistance and virus removal, Water Res. 43 (2009) 715–723.

[26] M. Ben-Sasson, K.R. Zodrow, Q. Genggeng, Y. Kang, E.P. Giannelis, M. Elimelech, Sur-face functionalization of thin-film composite membranes with copper nanoparticlesfor antimicrobial surface properties, Environ. Sci. Technol. 48 (2014) 384–393.

[27] A.M. Isloor, B.M. Ganesh, S.M. Isloor, A.F. Ismail, H.S. Nagaraj, M. Pattabi, Studies oncopper coated polysulfone/modified poly isobutylene alt-maleic anhydride blendmembrane and its antibiofouling property, Desalination 308 (2013) 82–88.

[28] E. Celik, H. Park, H. Choi, H. Choi, Carbon nanotube blended polyethersulfone mem-branes for fouling control in water treatment, Water Res. 45 (2011) 274–282.

[29] S. Zinadini, A.A. Zinatizadeh, M. Rahimi, V. Vatanpour, Preparation of a novel anti-fouling mixed matrix PES membrane by embedding graphene oxide nanoplates, J.Membr. Sci. 453 (2014) 292–301.

[30] F. Perreault, M.E. Tousley, M. Elimelech, Thin-film composite polyamide membranesfunctionalized with biocidal graphene oxide nanosheets, Environ. Sci. Technol. Lett.1 (2014) 71–76.

[31] M.A. Ansari, H.M. Khan, A.A. Khan, R. Pal, S.S. Cameotra, Antibacterial potential ofAl2O3 nanoparticles against multidrug resistance strains of Staphylococcus aureusisolated from skin exudates, J. Nanopart. Res. 15 (2013) 1–12.

[32] X. Zhu, R. Bai, K.-H. Wee, C. Liu, S.-L. Tang, Membrane surfaces immobilized withionic or reduced silver and their anti-biofouling performances, J. Membr. Sci. 363(2010) 278–286.

[33] A. Alpatova, E.-S. Kim, X. Sun, G. Hwang, Y. Liu, M. Gamal El-Din, Fabrication of po-rous polymeric nanocomposite membranes with enhanced anti-fouling properties:effect of casting composition, J. Membr. Sci. 444 (2013) 449–460.

[34] E.-S. Kim, G. Hwang, M. Gamal El-Din, Y. Liu, Development of nanosilver and multi-walled carbon nanotubes thin-film nanocomposite membrane for enhanced watertreatment, J. Membr. Sci. 394-395 (2012) 37–48.

[35] S.K. Das, M.M.R. Khan, T. Parandhaman, F. Laffir, A.K. Guha, G. Sekaran, A.B. Mandal,Nano-silica fabricated with silver nanoparticles: antifouling adsorbent for efficient

78 L. Dong et al. / Desalination 365 (2015) 70–78

dye removal, effective water disinfection and biofouling control, Nanoscale 5 (2013)5549–5560.

[36] F. Liu, M.R.M. Abed, K. Li, Preparation and characterization of poly(vinylidene fluoride)(PVDF) based ultrafiltrationmembranes using nano γ-Al2O3, J. Membr. Sci. 366 (2011)97–103.

[37] Z. Yi, L.-P. Zhu, Y.-Y. Xu, Y.-F. Zhao, X.-T. Ma, B.-K. Zhu, Polysulfone-based amphiphilicpolymer for hydrophilicity and fouling-resistant modification of polyethersulfonemembranes, J. Membr. Sci. 365 (2010) 25–33.

[38] C. Liao, P. Yu, J. Zhao, L. Wang, Y. Luo, Preparation and characterization of NaY/PVDFhybrid ultrafiltration membranes containing silver ions as antibacterial materials,Desalination 272 (2011) 59–65.

[39] L. Yan, Y. Li, C. Xiang, S. Xianda, Effect of nano-sized Al2O3-particle addition on PVDFultrafiltration membrane performance, J. Membr. Sci. 276 (2006) 162–167.

[40] F. Liu, B.-R. Ma, D. Zhou, Y.-H. Xiang, L.-X. Xue, Breaking through tradeoff ofPolysulfone ultrafiltration membranes by zeolite 4A, Microporous MesoporousMater. 186 (2014) 113–120.

[41] S.K. Das, A.R. Das, A.K. Guha, Adsorption behavior of mercury on functionalizedAspergillus versicolor mycelia: atomic force microscopic study, Langmuir 25 (2009)360–366.

[42] J. Huang, H. Wang, K. Zhang, Modification of PES membrane with Ag–SiO2: reduc-tion of biofouling and improvement of filtration performance, Desalination 336(2014) 8–17.

[43] H. Karkhanechi, R. Takagi, Y. Ohmukai, H. Matsuyama, Enhancing the antibiofoulingperformance of RO membranes using Cu(OH)2 as an antibacterial agent, Desalination325 (2013) 40–47.

[44] C.X. Liu, D.R. Zhang, Y. He, X.S. Zhao, R. Bai, Modification of membrane surface foranti-biofouling performance: effect of anti-adhesion and anti-bacteria approaches,J. Membr. Sci. 346 (2010) 121–130.

[45] Q. Li, S. Mahendra, D.Y. Lyon, L. Brunet, M.V. Liga, D. Li, P.J. Alvarez, Antimicrobialnanomaterials for water disinfection and microbial control: potential applicationsand implications, Water Res. 42 (2008) 4591–4602.