Chemical Engineering Journal 221 (2013) 111–123

Contents lists available at SciVerse ScienceDirect

Chemical Engineering Journal

journal homepage: www.elsevier .com/locate /cej

PAA grafting onto new acrylate-alumoxane/PES mixed matrixnano-enhanced membrane: Preparation, characterization andperformance in dye removal

1385-8947/$ - see front matter � 2013 Elsevier B.V. All rights reserved.http://dx.doi.org/10.1016/j.cej.2013.01.079

⇑ Corresponding author. Tel.: +98 831 4274530; fax: +98 831 4274542.E-mail address: smadaeni@yahoo.com (S.S. Madaeni).

Parisa Daraei a, Sayed Siavash Madaeni a,⇑, Negin Ghaemi b, Mohammad Ali Khadivi a, Laleh Rajabi c,Ali Ashraf Derakhshan c, Fatemeh Seyedpour c

a Membrane Research Center, Department of Chemical Engineering, Razi University, Kermanshah, Iranb Department of Chemical Engineering, Kermanshah University of Technology, Kermanshah, Iranc Polymer Research Center, Department of Chemical Engineering, Razi University, Kermanshah, Iran

h i g h l i g h t s

" Nano-sized acrylate-alumoxane was synthesized using acrylic acid and nanoboehmite." Boehmite and acrylate-alumoxane were added to PES casting dope." Grafting of PAA on acrylate-alumoxane/PES membrane results in high flux NF membrane." Grafted nanocomposite membrane shows superior dye removal and antifouling property." Grafting on nanocomposite membrane is a new way to obtain preferable modification.

a r t i c l e i n f o

Article history:Received 2 November 2012Received in revised form 22 January 2013Accepted 24 January 2013Available online 6 February 2013

Keywords:NanofiltrationNanoboehmiteAcrylate-alumoxanePAA graftingAcid blue

a b s t r a c t

Nanoboehmite particles were modified by acrylic acid to produce new alumoxane nanoparticles. Scan-ning electron microscopy (SEM) and Fourier transform infrared spectroscopy (FTIR) were used to realizeshape, size and functional groups of synthesized acrylate-alumoxane nanoparticles. The analysesdeclared that spherical acrylate-alumoxane nanoparticles were formed with hydroxyl and acrylategroups on their surface. Boehmite and acrylate-alumoxane were successfully introduced into polyether-sulfone (PES) membrane matrix by the phase inversion method. Fabricated membranes were examinedfor water permeability, dye (acid blue 193) retention capability and fouling resistance against whey pro-teins. The field emission scanning electron microscopy (FE-SEM) images were used to estimate thechanges in skin-layer morphology and bulk porosity of the prepared membranes. As a result, the directlyarrayed finger-like macro-voids as well as bulk porosity were gained by adding acrylate-alumoxanenanoparticles compared to pristine PES membrane. However, different quantities of acrylate-alumoxanein the casting solution induced no noticeable alteration in the membranes bulk porosity. The membranescontaining 1 wt.% of acrylate-alumoxane and 1 wt.% of nanoboehmite were selected to be grafted bypolyacrylic acid (PAA). Comparison of grafting efficiency for pristine PES membrane and nanofillerblended membranes proved that acrylate-alumoxane offered more effective grafted membrane by pro-viding polymerization initiation sites on mixed matrix membrane surface. In addition to high water per-meability (around 19 kg/m2 h bar), the acrylate-alumoxane mixed/PAA grafted membrane showedsuperior dye removal and fouling resistance. Atomic force microscopy (AFM) as well as water contactangle test was applied for investigation of membranes surface properties.

� 2013 Elsevier B.V. All rights reserved.

1. Introduction

Nanofiltration (NF) membranes as a growing technology in sep-aration processes was mostly studied to be improved in antifouling

properties as well as separation of the salts and the pollutants withmolecular weights ranged between 100 and 1000 Da (e.g. dyes)[1–4]. This pursues using changes in the casting solution contentsand preparation procedure, as well as applying additives, nanopar-ticles and grafting hydrophilic polymers [2,5–9].

The dye removal from effluents, as one of the important stagesof water and waste water treatment, is mostly conducted by

Nomenclature

List of symbolsAA acrylic acidAFM atomic force microscopyATR-FTIR attenuated total reflectance Fourier transform infrared

spectroscopyB1 mixed matrix membrane with 1 wt.% nanoboehmiteBAA0.5 mixed matrix membrane with 0.5 wt.% acrylate-alu-

moxaneBAA1 mixed matrix membrane with 1 wt.% acrylate-alumox-

aneBAA2 mixed matrix membrane with 2 wt.% acrylate-alumox-

aneBP bulk porosityCf dye concentration in feed (mg/l)Cp dye concentration in permeate (mg/l)DMAc N,N-dimethylacetamideEG ethylene glycolFE-SEM field emission scanning electron microscopyFR flux recovery (%)FTIR Fourier transform infrared spectroscopyGB1 PAA grafted B1

GBAA1 PAA grafted BAA1

GPPES PAA grafted PPESJw water flux (kg/m2 h)

Jw,1 initial water flux (kg/m2 h)Jw,2 after fouling water flux (kg/m2 h)Jwh flux of whey solution (kg/m2 h)KPS potassium persulfateNF nanofiltrationPAA polyacrylic acidPPES pristine polyethersulfone membranePVP polyvinylpyrrolidoneR dye removalRir irreversible fouling ratio (%)RO reverse osmosisRr reversible fouling ratio (%)Rt total fouling ratio (%)Sa average roughness (nm)SEM scanning electron microscopySq root mean square of the Z data (nm)Sz mean difference between highest peaks and lowest val-

leys (nm)Vm membrane pieces volume (cm3)Wd dry membrane weight (g)Ww wet membrane weight (g)

112 P. Daraei et al. / Chemical Engineering Journal 221 (2013) 111–123

reverse osmosis (RO) and nanofiltration processes; however, work-ing at lower pressures rather than RO along with acceptable yieldof dye retention are the commonly mentioned advantages of NFcompared to RO processes [10–14]. Applying combined methodssuch as coagulation and NF [15] and ion exchange membraneswith an adsorptive mechanism were studied in this regard [16].Ceramic membranes prepared by ZnAl2O3–TiO2 coated on raw claywere indicated high efficiency in dye removal and high permeateflux as reported [17]. Graft polymerization methods were utilizedfor production of high flux NF membranes with high efficiency indye removal [4–6].

The other valuable attempt in membrane preparation isimproving the antifouling properties and increasing the life timeof the membranes. Addition of nanoparticles on the membranesurface (coating) or in their matrix (blending in casting solutionor nonsolvent) [18], graft polymerization methods [6,19] andsimultaneous use of these methods [20,21] are introduced as mul-tiple ways to achieve membranes with more effective antifoulingcapability. Additionally, in the case of using nanofillers as additivein NF membranes, boehmite, organically modified clays, TiO2, alu-mina, carbon nanotubes, etc., were reported to be effective in anti-fouling property, higher contaminant rejection and improvedmechanical strength of the modified membranes [22–26].

c-AlOOH (boehmite) nanoparticles mixed with casting solutionwere recently used to induce antifouling property to polyethersulf-one (PES) membranes [22]. It was revealed that the extra hydroxylgroups of nanoboehmite was responsible for decreasing membranefouling by increasing membrane hydrophilicity and changingsurface roughness [22]. Immersion of casted membrane innanoparticles-nonsolvent emulsion bath followed by thermaltreatment for growing boehmite and gibbsite crystals was also car-ried out for embedding aluminum oxide nanoparticles in mem-brane matrix. The modified membrane showed decreasedattachment of bacteria on the membrane surface (antibiofoulingproperty) [27]; however, in this modification method (mixing ananofiller), reduction of permeate flux might be observed with in-

crease of nanofiller quantity because of nanoparticles agglomera-tion [22] which can be considered as a shortcoming of nanofillersaddition into the membrane matrix [28].

Boehmite and modified boehmite nanoparticles are widely em-ployed in fabrication of nanocomposites, metal doped aluminumoxide, catalysts, ceramic membranes and active coating layer onthe membranes [29–33]. Boehmite nanoparticles can be modifiedusing carboxylic acids because carboxyl group is able to be coordi-nated with aluminum as unidentate and bidentate ligands[32,34,35]. Modification of boehmite nanoparticles is carried outfor changing the physical/chemical properties by targeted selectionof a special carboxylic acid [34]. Although, there are several typesof carboxylate modified boehmite nanoparticles (e.g. parahydroxybenzoate, salicylate, para aminobenzoate and fumarate)which contain the coordination of carboxyl group with the basicformula of [Al(O)x(OH)y(O2CR)z]n where 2x + y + z is equal to 3, x,y and z can include any integer or fraction values (for instance, xcan be 0–1.5, y can be 0–3, z can be 0–3, and n can be 1–6 or great-er) and R refers to C1 to C14 hydrocarbons [29,36]. By our knowl-edge, acrylate modified boehmite (acrylate-alumoxane) has notbeen synthesized in any research regarding membrane preparationor modification.

In this study, a new procedure is presented for modification ofPES membrane by mixing self-synthesized acrylate-alumoxanenanoparticles in the casting solution to achieve a support withlinking sites for an effective grafting of polyacrylic acid on themembrane surface. Indeed, the main approach is to investigatethe possibility of combination of graft polymerization and nanofill-er addition for superior modification of PES NF membranes. Thefabricated membranes are tested for water flux, acid blue 193(831.68 Da) dye removal capability and antifouling properties.The scanning electron microscopy (SEM), atomic force microscopy(AFM), Fourier transform infrared spectroscopy (FTIR) and attenu-ated total reflectance FTIR (ATR-FTIR) are utilized for characteriza-tion of modified boehmite nanoparticles and nanocomposite NFmembrane.

P. Daraei et al. / Chemical Engineering Journal 221 (2013) 111–123 113

2. Materials and methods

2.1. Materials

Acrylic acid (AA), acetone, methanol, n-hexane, polyvinylpyrrol-idone (PVP) and dimethylacetamide (DMAc) were obtained fromMerck and used without further purification. The applied polye-thersulfone for the preparation of polymeric membranes was pur-chased from Ultrason E6020P, MW = 58,000 g/mol, BASF.Potassium persulfate (KPS) and ethylene glycol from Merck wereutilized in surface grafting reaction as received. The dye used inseparation test was acid blue 193 which obtained from BoyakhSaz, Iran (MW = 831.68 Da).

2.2. Preparation of acrylate-alumoxane

The base boehmite nanoparticles were prepared at room tem-perature by the same procedure explained in detail elsewhere[22,37]. The acrylate groups were attached on boehmite surfaceduring 72 h reflux. For this purpose 5 g of nanoboehmite wasmixed with 10 g of acrylic acid. After ultrasonicating for 15 min,the mixture was agitated by a magnetic stirrer at 400 rpm andheated to reflux get started. After 72 h of reflux, the product waswashed with acetone and methanol for several times and filteredby a filter paper. Then it was dried at 60 �C in an oven for 24 h.The obtained solid was comminuted, washed with n-hexane forsix times and let to precipitate after mechanical agitation to easilyaspirate n-hexane in every washing step. Next, the nanoparticleswere dried again at 50 �C.

2.3. Fabrication of mixed matrix membranes

Phase inversion method was applied for preparing mixed ma-trix membranes including simple blending of nanofillers in thecasting solution [22,28]. Thus, this method was selected to preparethe acrylate-alumoxane/PES nanocomposite membranes. Table 1represents the casting solution composition and the code for eachfabricated membrane. 1 wt.% of PVP was added to each castingsolution as pore former agent. Ultrasound mixing (30 min) wasused to homogenize the nanofillers in DMAc solvent. The next40 min sonication was performed after dissolving the PES inDMAc-nanofillers mixture and stirring for 12 h which resulted indegassing of the solution and adequate dispersion of nanoparticlesin the polymer solution. Casting was carried out on glass plates bya home-made multi-thickness knife adjusted in 150 lm. Distilledwater was applied as coagulant for casted membranes. The tem-perature of casting solution and nonsolvent bath was kept at 20 �C.

2.4. PAA grafting onto nanocomposite membranes

A gentle procedure including low concentration of reactants(monomer and initiator) and reaction temperature of 70 �C wasconducted for graft polymerization of acrylic acid on bare andnanocomposite membranes. 200 ml of aqueous reactants solutionincluding 3 wt.% AA, 0.015 wt.% KPS and 0.5 wt.% EG as cross-linkerwas prepared. The oxygen gas is the inhibitor of thermal

Table 1Casting solution composition for prepared membranes.

Membrane code PES Boehmite Acrylate-alumoxane PVP DMAc

PPES 20 – – 1 79BAA0.5 20 – 0.5 1 78.5BAA1 20 – 1 1 78BAA2 20 – 2 1 77B1 20 1 – 1 78

polymerization and terminates the polymerization cycle by react-ing with radicals created by KPS. Thus, the solution was deoxyge-nized by nitrogen gas bubbling for 15 min, poured in reactiontank where the membranes were stack on its bottom. The solutionwas removed after 20 min under nitrogen gas blowing. The reac-tion was initiated by heating the tank and completed after 4 h at70–75 �C. Next, the membranes were rinsed with distilled waterand kept in an oven to dry at 50 �C. PPES, B1 and BAA1 were chosento be grafted. The grafted membranes are represented by GPPES,GB1 and GBAA1. The total procedure of preparation of mixed matrixmembranes and grafting the PAA on the membrane surface isbriefly shown in Fig. 1.

2.5. Characterization methods

2.5.1. Type of acrylate-alumoxane nanoparticlesABB Bomem MB-104 spectrophotometer (Canada) was em-

ployed to assess the production of acrylate-alumoxane and identifythe coordination type of carboxylic groups to aluminum atoms.

The particle size and shape of synthesized acrylate-alumoxanewas monitored by a scanning electron microscope (SEM, TESCANVEGA TS 5136MM, USA).

2.5.2. Membrane characterizationSEM technique was used to probe the morphology of prepared

membranes including skin-layer and support structure. Thus, thepieces of all prepared membranes were cut and precisely brokenin liquid nitrogen bath. The fractured section of samples werecoated with gold and scanned by field emission scanning electronmicroscopy (FE-SEM, Hitachi S-4160, Japan). The bulk porosity ofmembranes can be estimated via image analysis. The cross sec-tional images were successfully analyzed by Image J, version1.45s free software. Furthermore, the skin-layer thickness wasmeasured on the live images before capturing the snapshots duringscanning which FE-SEM apparatus equipped for.

Fig. 1. Summarized procedure of membrane preparation method.

Fig. 2. Acid blue 193 dye structure.

114 P. Daraei et al. / Chemical Engineering Journal 221 (2013) 111–123

The surface roughness as a major characteristic of fabricatedmembranes, affecting the antifouling behavior, was determinedby AFM (Nanosurf� Mobile S scanning probe-optical microscope,Switzerland). The small pieces (1 � 1 cm2) of prepared membraneswere staked on a glassy substrate and scanned in noncontactmode. The roughness parameters including Sa, Sq and Sz was mea-sured using Nanosurf� Mobile S software (version 1.8). Moreover,SPM software (version 1.4.0.6) was employed to determine themean pore size of the prepared membranes [23].

PAA grafted membranes (e.g. GPPES, GB1 and GBAA1) wereprobed by ATR-FTIR (Bruker Vertex 80, Germany) spectroscopy toshow the success of surface grafting.

A test was carried out for seeking the membranes affinity to getdrenched and also estimating the approximate bulk porosity asverification for image analysis results. The applied procedure toestimate the trend of bulk porosity changes was same as what per-formed for swelling test. The test was conducted for having a bet-ter comparison of the image analyses result for the bulk voidspaces. The certain dimensions (2 � 2 cm2) of membranes werecut and immersed in distilled water for 24 h. Then, they were driedonly on top and bottom surface by sandwiching between filter pa-pers and rapidly weighed. The membrane pieces were completelydried at 40–50 �C and the dry weights were recorded. Considering1 g cm�3 for water density, the related equation for calculating thebulk porosity (BP) via weight gain degree can be represented by[28]:

BPð%Þ ¼Ww �Wd

Vm� 100 ð1Þ

where Ww and Wd are the wet and dry weight (g) of membranepieces, and Vm is the dry membrane volume (cm3).

The modification procedure for the PES membrane was per-formed for increasing the membranes hydrophilicity. Hence, thewater contact angle test helps to investigate the membrane surfacehydrophilicity. The small droplets of distilled water were droppedon membrane surface in various locations. A digital microscope(USB Digital Microscope 1.3 Mega Pixel Color Video Camera400X, China) was applied to record the droplets status immedi-ately after dropping. The contact angle was simply measured usingImage J image analyzer freeware.

A simple gravimetric method was used to estimate the graftingyield for various fabricated membranes. The difference betweengrafted membrane weight and naked support can indicate thegrafting efficiency.

2.5.3. Membrane performance and water flux testThe permeability of the membranes was determined by testing

pure water flux. The circular pieces of each membrane were cutwith around 5 cm diameter for providing an operating surface areaof 19.5 cm2. The membrane pieces were placed in a cylindricalbatch test setup i.e., dead-end cell (100 ml volume) equipped witha magnetic stirrer and a 3 l reservoir tank. The reservoir was filledwith distilled water and pressurized at 5 bars using a nitrogen gasinlet on the top of the tank. By this way, the feed solution was in-jected into the dead-end cell and the membrane was compacted.After 30 min of compaction, the driving pressure was decreasedand adjusted at 4 bars to start collecting the permeated waterand measuring the water flux (kg/m2 h).

The membrane performance was tested using acid blue 193 dye(see Fig. 2 for dye structure) solution. The feed was prepared bydissolving 300 mg of dye in 1 l of distilled water. UV–Vis spectro-photometer (UNICO UV-2100 spectrophotometer) was used tomeasure the concentration of feed and permeate samples at564 nm wave length. For preparation of standard calibration curve,the absorbance of samples containing dye at several known con-centrations was measured at particular wavelength 564 nm.

Permeated water was collected and examined for dye concentra-tion during certain time intervals. The dye removal efficiencywas obtained using the following equation:

Rð%Þ ¼ 1� Cp

Cf

� �� 100 ð2Þ

where R, Cp and Cf are dye retention, dye concentration in permeateand feed solutions, respectively. The variation of feed concentrationand increasing the dye content in feed cannot be neglected; how-ever, for achieving more precise results for dye removal, the test cellwas continuously replenished by fresh dye solution using a reser-voir. Furthermore, the final dye concentration in feed was measuredand compared with initial concentration. The final dye removal wascalculated considering the final dye concentration in the remainingfeed.

Antifouling capability of fabricated membranes was investi-gated using self-made whey solution with high concentration ofproteins which causes a rapid and severe fouling in a short time.The fouled membrane by whey proteins was washed with water(without any cleaning agent) to investigate the water flux recoveryafter getting fouled during the filtration process. Therefore, the testsequence was measuring pure water flux (Jw,1) during 60 min, cal-culating the final permeated whey flux (Jwh) after 120 min (thetime that the permeate flux reaches to a constant value) under agi-tation at 400 rpm, immersing membrane in distilled water for30 min, rinsing membrane and testing after-fouling water flux forthe membranes (Jw,2) [22]. Then, the flux recovery was obtainedas follows:

FRð%Þ ¼Jw;2

Jw;1� 100 ð3Þ

Flux recovery provides an estimation for membrane reuse aftergetting fouled by contaminated water with whey proteins as wellas the tendency of proteins to permanently adhere on the mem-brane surface. Moreover, total fouling ratio (Rt), reversible foulingratio (Rr) and irreversible fouling ratio (Rir) as the most investigatedfouling parameters were calculated using the known equations asfollows [22,38,39]:

Rtð%Þ ¼Jw;1 � Jwh

Jw;1� 100 ð4Þ

Rrð%Þ ¼Jw;2 � Jwh

Jw;1� 100 ð5Þ

Rirð%Þ ¼Jw;1 � Jw;2

Jw;1� 100 ð6Þ

P. Daraei et al. / Chemical Engineering Journal 221 (2013) 111–123 115

These parameters help to better understand the membrane foulingbehavior. Obviously, higher reversible fouling and lower total foul-ing ratio result in lower fouling and flux reduction in subsequentutilization of membrane after fouling [22].

3. Results and discussion

3.1. Characterization of AA modified nanoboehmite

The FTIR technique proves the formation of acrylate-alumoxaneas demonstrated in Fig. 3. According to the spectra, due to the reac-tion of boehmite and AA, position of basic bonds of carboxylicgroup in AA is changed. In AA spectrum, the peak related to acidiccarbonyl shifted from the original value (1700 cm�1) to 1690 cm�1

because of conjugation with carbon–carbon double bond. The reac-tion of AA with boehmite may result in formation of both uniden-tate and bidentate structures as depicted in Fig. 4. The peaksappeared in 1454 and 1556 cm�1 refer to symmetric and asymmet-ric stretching vibrations of bidentate acrylate groups, respectively.

Fig. 3. FTIR spectra of acrylic acid, bo

The strong bond in 1725 cm�1 represents stretching vibration ofunidentate ones [35]. The higher intensity of unidentate mode re-veals that the AA molecules are mostly bonded to aluminum atomof boehmite in unidentate mode. This may be due to relative stabil-ity of the conjugated system of AA. In the AA spectrum, the broadbond in 2400–3400 cm�1 is attributed to the stretching vibrationof acidic hydroxyl group, which is faded out in 2400 cm�1 for acry-late-alumoxane. This confirms formation of carboxylate and elim-ination of acidic OH groups. The bonds in 3072 and 3288 cm�1

represent the stretching vibration of boehmite hydroxyl groups,which still exist in acrylate-alumoxane spectrum. It means thatthe surface of nanoboehmite is not completely reacted with AAand there are some boehmite hydroxyl groups in the producedacrylate-alumoxane. Furthermore, the peaks appeared at 1176and 1066 cm�1 in acrylate-alumoxane and boehmite spectra corre-spond to hydrogen bonds. Additionally, nitrate impurities show astrong bond in 1382 cm�1.

SEM images of acrylate-alumoxane nanoparticles proved thatthe particles were almost in spherical shape (Fig. 5). The particles

ehmite and acrylate-alumoxane.

Fig. 4. Demonstration of unidentate and bidentate coordination of acrylate withaluminum atom in acrylate-alumoxane nanoparticles.

116 P. Daraei et al. / Chemical Engineering Journal 221 (2013) 111–123

size was obtained between 20 and 50 nm using several images.According to the SEM and FTIR results, the proposed structure fornew acrylate modified nanoboehmite is schematically depicted inFig. 5. According to FTIR spectra, the nanoparticles surface is cov-ered by hydroxyl and acrylate groups compared with nanoboeh-mite which possessed only hydroxyl groups.

3.2. Membranes characteristics

The ATR-FTIR spectra for all PAA grafted membranes proved thesuccess of grafting process (Fig. 6). The spectra of BAA1 and GBAA1

in the range of 600–2000 cm�1 are presented in Fig. 6b for detailedinvestigation. The spectra of acrylate alumoxane and boehmiteblended membrane (BAA1 and B1) did not show any clear peakrepresenting alumoxane or boehmite due to very low concentra-tion of nanofiller in the membrane matrix. However, in GBAA1

spectra (Fig. 6b), the peaks in 1728 and 1245 cm�1, attributed tothe vibrations of carbonyl groups, undoubtedly verify the existenceof PAA on the membrane surface. Moreover, the bonds in 1143, 995and 815 cm�1, which correspond to C–O stretching coupled withO–H bending, CH2 stretching and COOH stretching vibrations, areother proofs for PAA formation. The other grafted membranesshowed similar spectra (Fig. 6a).

The skin-layer thicknesses for the fabricated membranes arelisted in Table 2. The presented values are average of several mea-surements by FE-SEM technique and image analysis software withignorable errors. Additionally, the top-layer structure for nano-composite membranes including the grafted ones and theproposed skin-layers are clearly demonstrated in Fig. 7. The top-layer thickness sequence for the membranes is BAA2 > GBAA1 >BAA1 > GB1 > GPPES > PPES > B1 > BAA0.5. Accordingly, the presenceof nanoparticles with hydrophilic hydroxyl groups in the castingsolution leads to a reduction in the skin-layer thickness. Fromthe values tabulated in Table 2, addition of nanoboehmite (whichpossesses extra hydroxyl groups rather than acrylate-alumoxane)leads to increase of casting solution hydrophilicity and decreaseof top-layer thickness in B1 relative to PPES and other preparedmixed matrix membranes [26,40]. Addition of acrylate-alumoxanenanoparticles with concentrations more than 0.5 wt.% results ingrowth of top-layer thickness due to agglomeration of nanoparti-cles (see Fig. 7) in the surface of the membrane during precipita-tion in distilled water (nonsolvent). Additionally, enhancement ofnanofiller content in the casting solution can rise the demixingtime and allow more migration of nanoparticles to the membranetop surface and leads to membranes with thicker skin-layer[22,41]. The results for BAA2, containing the most quantity of acry-late-alumoxane nanoparticles, confirm this explanation (Table 2and Fig. 7). On the other hand, the lowest amount of nanoparticlesleads to the highest decrease of top-layer thickness in BAA0.5. FromTable 2, the thickness of top-layer in all grafted membranes wasmore than thickness of the support layer indicating the creationof grafted PAA layer on the membranes surface.

According to FE-SEM images of cross section of fabricated mem-branes (Fig. 8), addition of boehmite and acrylate-alumoxanenanoparticles significantly increased the bulk porosity relative topristine PES membrane; however, increase of nanoparticles quan-tity could not enhance the porosity. Similar results have been re-ported in another work [22]. The top-layer was influenced bynanoparticles presence in a different way compared to the mem-brane bulk. The macro-voids shape is the most visible characteris-tic changed in mixed matrix membranes rather than bare PESmembrane [22]. In BAA0.5, the directly aligned finger-like macro-voids is clearly observed. Increasing the amount of nanoparticlesslightly decreases the size of macro-voids in BAA1 and BAA2. Thesame effect was detected in cross-section image of B1 (Fig. 9). Thismay be attributed to delayed demixing because of higher viscosityof casting solution in nanoparticles concentration more than0.5 wt.% [22,23,26].

Similar to the recently published article [22], the variousamounts of nanoboehmite or modified nanoboehmite induced nochanges in membrane bulk porosity. This was accurately investi-gated for prepared membranes by image analysis for cross-sectionmicrophotographs (Table 2). The results demonstrated that rela-tively unchanged bulk porosity was obtained for various amountsof acrylate-alumoxane nanoparticles. The change in mean pore sizeof the membrane surface is a parameter affecting separation andpermeability. The mean pore size values obtained from AFMimages are tabulated in Table 2. Addition of acrylate-alumoxaneis not able to critically change the mean pore diameter. However,a slight increase of pore size may be attributed to nanofilleragglomeration, which is enhanced by nanofiller increment. More-over, all grafted membranes possessed smaller mean pore size

Fig. 5. SEM microphotograph, scheme of spherical prepared nanoparticles and their surface arrangement of functional groups.

P. Daraei et al. / Chemical Engineering Journal 221 (2013) 111–123 117

compared to original membranes. The highest reduction in poresize was observed for boehmite mixed membrane. This indicatesthe positive effect of blending hydrophilic nanofiller in membranematrix to achieve smaller pore size [22,23]. Permeability and dyeremoval capability of prepared membranes present more informa-tion on the membranes morphological variations.

The hydrophilicity of the prepared membranes was changed byadding the nanofillers and grafting the membranes (Table 2). Theexistence of the modified nanoparticles reduces the water contactangle meaning increment of the membrane hydrophilicity. How-ever, addition of 1 wt.% of modified boehmite (54.3�) did not de-crease the water contact angle as much as the same quantity ofboehmite (45.1�). This can be explained by greater number ofhydroxyl groups on boehmite rather than acrylate-alumoxane. Byincreasing the alumoxane content in the membrane matrix nonoticeable decrement in contact angle is obtained. Higher mem-brane hydrophilicity is achieved by PAA grafting onto the mem-branes. All the grafted membranes showed smaller contact anglerather than their naked supports. From the data in Table 2, GBAA1

is the most hydrophilic membrane (contact angle around 40.6�)due to more efficient grafting of PAA. Therefore, grafting of themembrane mixed with the properly modified nanofiller might re-sult in higher hydrophilicity compared to naked mixed matrixmembrane. Comparing the contact angle values for GBAA1 andGB1 verifies that mixing acrylate-alumoxane with PES along withgrafting PAA results in superior hydrophilicity rather than graftednanoboehmite-mixed membrane.

3.3. Membranes performance

3.3.1. Water flux and dye removalFig. 9 indicates the steady results for permeate flux and acid

blue 193 dye retention after 2 h of experiment for all prepared

membranes. As it is expected, all grafted membranes show higherpermeated flux rather than the naked supports. This is due to pres-ence of hydrophilic PAA on the membrane surface and surface con-nected pores wall. The highest flux with a noticeable gap isattributed to GBAA1 offering a more efficient grafting of PAA onmembrane surface. The results for the grafting percentage alsoconfirmed this (Fig. 10). Whereas the acrylate-alumoxane nano-particles possess a linking site for polymer chains, the highestgrafting efficiency is presented by BAA1 according to the graftingyield results. Also, the pore size reduction is more critical forgrafted BAA1 (Table 2). In fact, the presence of superficial car-bon–carbon double bonds on the blended nanoparticles (especiallyon which located on the membrane surface and surface connectedpores wall) act as a proper initiation point for growing polyacrylicacid chains. The water permeability for novel grafted membrane(19 kg/m2 h bar) with a mixed matrix membrane as support iscompletely comparable with high flux NF membranes reported inliteratures [2,5–9].

The next higher flux value is attributed to GB1. It can be con-cluded that even boehmite nanoparticles can enhance grafting effi-ciency with their high hydrophilicity [22] which allows highervolume of aqueous reactant solution to be absorbed during graft-ing process.

The affinity of boehmite blended membrane to be drenchedwith water is clearly verified by swelling test. The data given inTable 2 for porosity measurements using swelling test revealedthat boehmite nanoparticles possessed great affinity to bedrenched. Hence, the porosity percent via swelling test (61.8%) isconsiderably more than what calculated by means of image analy-sis (45.2%). This makes the membrane more suitable for absorbingreaction mixture and represents higher development of top-layerthickness. The comparison of skin-layer thickness indicated thatboehmite mixed membrane (B1) had the highest increase in

600

500 700 900 1100 1300 1500 1700 1900

1100 1600 2100 2600 3100 3600

Rel

ativ

e ab

sorb

ance

Wavenumber (cm-1)

Wavenumber (cm-1)

PPES(a)

B1(b)

BAA1(c)

GPPES(d)

GB1(e)

GBAA1(f)

(a)

(b)

(c)

(d)

(e)

(f)A

bso

rban

ce

a

b

Fig. 6. ATR-FTIR spectra for all grafted membranes and the naked supports (a) and ATR-FTIR spectra of BAA1 and GBAA1 in the wave number range of 600–2000 cm�1 (b).

Table 2The data for fabricated membrane characterization.

Membranecode

Porosity(area%)

Porosity fromwater swelling (%)

Skin-layerthickness (nm)

Roughness parameter(Sa, nm)

Contactangle (�)

Mean poreradius(nm)

PPES 20.0 25.6 1110 ± 50 12.3 ± 0.5 69.1 ± 1.5 2.58 ± 0.23GPPES – – 1170 ± 30 11.6 ± 0.4 55.5 ± 2.3 2.31 ± 0.18BAA0.5 53.1 53.2 966 ± 30 19.8 ± 0.5 56.8 ± 1.7 2.55 ± 0.25BAA1 52.4 57.2 1800 ± 50 24.2 ± 0.5 54.3 ± 2.1 2.51 ± 0.17GBAA1 – – 2000 ± 20 20.5 ± 0.6 40.6 ± 1.6 2.10 ± 0.15BAA2 53.3 59.4 2100 ± 30 31.2 ± 0.4 53.2 ± 1.6 2.74 ± 0.22B1 45.2 61.8 1030 ± 20 2.5 ± 0.1 45.1 ± 1.9 2.13 ± 0.15GB1 – – 1700 ± 50 16.8 ± 0.4 43.4 ± 2.0 2.06 ± 0.19

118 P. Daraei et al. / Chemical Engineering Journal 221 (2013) 111–123

Fig. 7. FE-SEM microphotographs of skin-layer of fabricated membranes (the estimated skin-layer is specified by arrows).

P. Daraei et al. / Chemical Engineering Journal 221 (2013) 111–123 119

top-layer thickness (around 670 nm) when it was grafted. For thesame reason, the top-layer of bare PES was slightly thickened(around 80 nm). The skin-layer was gained around 200 nm inBAA1 which possessed lower hydroxyl groups and consequentlylower hydrophilicity rather than B1.

Some unpredictable results may be seen in flux sequence of pre-pared membranes that should be reasoned. B1 shows lower waterflux than that of PPES same as what observed by recent work [22].This effect was explained based on reduction of pore size (poreblockage) as a result of nanoparticles accumulation in the mem-brane surface pores [28]. BAA1 membrane also revealed this short-coming. Lower amount of nanoparticles (0.5 wt.%) leads to fasterdemixing during membrane coagulation which originate fromhydrophilicity of acrylate-alumoxane nanoparticles and also lowerviscosity of casting solution [22,31,33]. Higher water flux for BAA2

may be due to lower continuity and more rupture of membranetop-layer as clearly recognizable in Fig. 7.

As it seems, all the membranes including pristine PES (PPES)can efficiently remove dye from water because of small mean poresize and very tight and rather thick skin-layer as discussed before.This also can prove a defect free structure for prepared mem-branes. In B1 and BAA1 there is no significant improvement indye retention relative to PPES; however, a decrease in permeateflux is occurred. In BAA0.5 and BAA2, retention is decreased becauseof higher permeate flux of these membranes and probably somecracks on the membrane top-layer caused by accumulated nano-particles. The unchanged or even higher dye removal efficiencyof grafted membranes revealed that the grafted layer acted as extrabarrier against dye molecules in spite of offering higher hydrophi-licity of polyacrylic acid which led to higher water permeation.Furthermore, mean pore sizes of all grafted membranes are de-creased compared to naked supports. Hence, the dye retentiondoes not decrease by increasing the permeated flux especially forGBAA1. The other possible reason might be attributed to

Fig. 8. FE-SEM microphotographs of cross-section of prepared membranes.

96

98

96

9898

99

92

89

0

10

20

30

40

50

60

70

80

82

84

86

88

90

92

94

96

98

100

PPES GPES B1 GB1 BAA1 GBAA1 BAA0.5 BAA2

Per

mea

te f

lux

(kg

/m2 .

h)

Dey

rem

ova

l (%

)

Dye rejection (%) Permeate flux

Fig. 9. The average of three measurements for dye retention and permeated flux of all prepared membranes after 2 h of test at 4 bars trans-membrane pressure (the errorvalues were ignorable).

120 P. Daraei et al. / Chemical Engineering Journal 221 (2013) 111–123

24

35

28

0

5

10

15

20

25

30

35

40

PPES BAA1 B1

Gra

ftin

g y

ield

(%

)

Fig. 10. Grafting yield for various grafted membranes in this study.

P. Daraei et al. / Chemical Engineering Journal 221 (2013) 111–123 121

cross-linking between acrylate groups on the acrylate-alumoxanenanoparticles surface. It means that the polymerization of acrylategroups on the nanoparticles which lied near together may lead tofilling of possible cracks between nanoparticles. Moreover, basedon the small mean pore size of all prepared membranes, it canbe concluded that size exclusion is the most probable mechanismgoverning the dye removal. However, existence of hydroxyl andcarboxyl groups in modified (nano-enhanced or grafted) mem-branes may increase the electrostatic effects cooperating with sizeexclusion in removing negatively charged dye molecules. This maybe the reason for unchanged or high dye removal efficiency in spiteof significant high permeated fluxes of modified membranes.

Therefore, mixing the acrylate-alumoxane or boehmite nano-particles and then PAA grafting on the mixed matrix membranecan surprisingly enhance the permeated water flux without reduc-tion in dye removal performance of the membranes. The intro-duced combined method seems to be more efficient compared toindividual nanofiller blending or polymer grafting. However, per-formance improvement of blended membranes requires moreinvestigation considering that mixing nanoboehmite or acrylate-alumoxane may decrease the permeate flux. For this purpose, theantifouling behavior of all prepared membranes is discussed inthe next section.

0 10 20 30 40 50 60

PPES

BAA0.5

BAA1

BAA2

B1

GB1

GBAA1

GPPES

Fig. 11. Pure water flux and fouling param

3.3.2. Antifouling capabilityFig. 11 demonstrates the information on the fouling resistance

of all fabricated membranes. Flux recovery for fouled GBAA1 mem-brane is the highest and B1 ranks the second. The membrane sur-face roughness and hydrophilicity are among the mainparameters influencing the antifouling property [42]. Hence, AFMimages (Fig. 12) were analyzed and roughness parameters werecompared for all fabricated membranes. Only Sa values are pre-sented in Table 2 due to similar trends for Sq and Sz. B1 with thelowest Sa value possesses smoother surface compared to the othermembranes. This leads to easier elimination of surface trappedwhey proteins during simple washing with distilled water. In thecase of GBAA1, surface roughness is even higher compared to PPES(Sa = 12.3 nm) and reaches to 20.5 nm. Addition of the nanoparti-cles and their presence on the membrane surface may be consid-ered as a reason for roughness increment which leads to highertotal fouling ratio (Rt, see Fig. 11). However, the superior hydrophi-licity of PAA grafted membrane (with regard to very high waterflux and small contact angle) controls the fouling behavior of thismembrane. More hydrophilic surface offers lower irreversibleattachment of whey proteins [22,43]. The GPPES membrane showshigher fouling resistance compared to PPES in spite of unchangedsurface roughness. It can be concluded that grafting of PAA has po-sitive effect on softening of membranes except for the case of usingnanoboehmite (B1) which provides a significant increase in the sur-face roughness after being grafted. This may be due to un-uniformgrafting of B1 surface resulted from difference between graftingyield of nanoboehmite containing parts of the membrane surfaceand the parts having more PES. The nanoboehmite particles havegreater tendency to absorb the grafting reaction solution (an aque-ous solution). In other words grafting reaction solution is dissolvedinto the bulk of nanoboehmite particles. This is due to high hydro-philicity caused by hydroxyl groups. The surface of BAA1, contain-ing acrylate-alumoxane nanoparticles, which have less hydroxylgroups than boehmite, becomes smoother after graft polymeriza-tion. It is assumed to obtain more efficient grafting for membraneblended with acrylate-alumoxane due to the presence of carbon–carbon reactive double bonds as an active site for PAA chaingrowth. The AFM results (Sa value) for GBAA1 and BAA1 clearly rep-resented that grafting of PAA on the mixed matrix membrane re-sulted in reduction of surface roughness.

70 80 90 100

Rr

Rt

FR

Pure water flux (kg/m2.h)

eters for all fabricated membranes.

Fig. 12. AFM images of prepared support membranes and PAA grafted membranes (the membranes code are written on each photograph).

122 P. Daraei et al. / Chemical Engineering Journal 221 (2013) 111–123

The lowest flux recovery is attributed to BAA2 (Fig. 11). As dis-cussed before, high surface roughness allows more adhesion of thefoulants on the membrane surface. Accordingly, Rir is the highestfor BAA2. Both AFM (Table 2) and FE-SEM (Fig. 7) results confirmthe critical accumulation of nanoparticles on BAA2 surface. BAA0.5

has also rougher surface than PPES, and the amount of nanofillersis not enough to induce significant change in fouling resistance capa-bility relative to PPES. The different results obtained by previouswork [22] may be related to lower hydroxyl groups of newnanofillers (prepared in this study) compared to nanoboehmite. This

means that the hydroxyl groups on nanoparticles surface play a keyrole in improving antifouling properties of mixed matrixmembranes.

4. Conclusion

An acrylate modification was performed on nanoboehmite sur-face to introduce new properties on the nanoparticles and speciallyto form polymerization initiation sites for subsequent graftpolymerization. The FTIR techniques proved the successfulness of

P. Daraei et al. / Chemical Engineering Journal 221 (2013) 111–123 123

this modification. The novel acrylate-alumoxane nanoparticles andalso nanobohemit were added to PES membrane matrix. The per-meate flux reduction was occurred due to pore blockage by nanof-illers in membrane matrix prepared by the method in whichnanoparticles were added to the casting solution. PAA grafting ofprepared mixed matrix membranes was more efficient comparedto bare PES. It was revealed that PAA grafting of the membranecontaining acrylate-alumoxane (BAA1) resulted in higher dyeretention (around 99%) and higher water permeability (19 kg/m2 h bar).

Generally, addition of nanoboehmite induces a smooth surfaceto mixed matrix membranes which results in preferable antifoul-ing properties; however, sometimes, a lower pure water flux maybe obtained. Furthermore, blending of acrylate modified nano-boehmite may decrease surface roughness and not show improvedantifouling behavior whereas surface grafting of PAA surprisinglymakes the water flux much higher than pristine PES membrane.On the other hand, the increased hydrophilicity enhances the fluxrecovery of fouled membrane. Extra addition of nanofillers is notconvenient to improve the properties of the polymeric membraneswhile the proper amount of nanoboehmite or modified boehmiteparticles along with graft polymerization may be presented as anew modification method in order to obtain preferable antifoulingproperties and higher permeate flux as well as the efficient dye re-moval for the polymeric NF membranes.

References

[1] Y.L. Ji, Q.F. An, Q. Zhao, W.D. Sun, K.R. Lee, H.L. Chen, C.J. Gao, Novel compositenanofiltration membranes containing zwitterions with high permeate flux andimproved anti-fouling performance, J. Membr. Sci. 390–391 (2012) 243–253.

[2] J. Huang, K. Zhang, The high flux poly (m-phenylene isophthalamide)nanofiltration membrane for dye purification and desalination, Desalination282 (2011) 19–26.

[3] S. Cheng, D.L. Oatley, P.M. Williams, C.J. Wright, Characterisation andapplication of a novel positively charged nanofiltration membrane for thetreatment of textile industry wastewaters, Water Res. 46 (2012) 33–42.

[4] A. Akbari, S. Desclaux, J.C. Rouch, J.C. Remigy, Application of nanofiltrationhollow fibre membranes, developed by photografting, to treatment of anionicdye solutions, J. Membr. Sci. 297 (2007) 243–252.

[5] S. Roy, S.A. Ntim, S. Mitra, K.K. Sirkar, Facile fabrication of superiornanofiltration membranes from interfacially polymerized CNT-polymercomposites, J. Membr. Sci. 375 (2011) 81–87.

[6] A. Asatekin, E.A. Olivetti, A.M. Mayes, Fouling resistant, high flux nanofiltrationmembranes from polyacrylonitrile-graft-poly(ethylene oxide), J. Membr. Sci.332 (2009) 6–12.

[7] H. Deng, Y. Xu, Q. Chen, X. Wei, B. Zhu, High flux positively chargednanofiltration membranes prepared by UV-initiated graft polymerization ofmethacrylatoethyl trimethyl ammonium chloride (DMC) onto polysulfonemembranes, J. Membr. Sci. 366 (2011) 363–372.

[8] K. Yoon, B.S. Hsiao, B. Chu, High flux nanofiltration membranes based oninterfacially polymerized polyamide barrier layer on polyacrylonitrilenanofibrous scaffolds, J. Membr. Sci. 326 (2009) 484–492.

[9] X.L. Li, L.P. Zhu, B.K. Zhu, Y.Y. Xu, High-flux and anti-fouling cellulosenanofiltration membranes prepared via phase inversion with ionic liquid assolvent, Sep. Purif. Technol. 83 (2011) 66–73.

[10] M.G. Buonomenna, L.C. Lopez, P. Favia, R. d’Agostino, A. Gordano, E. Drioli, NewPVDF membranes: the effect of plasma surface modification on retention innanofiltration of aqueous solution containing organic compounds, Water Res.41 (2007) 4309–4316.

[11] A. Akbari, S. Desclaux, J.C. Remigy, P. Aptel, Treatment of textile dye effluentsusing a new photografted nanofiltration membrane, Desalination 149 (2002)101–107.

[12] A. Akbari, S. Desclaux, J.C. Rouch, P. Aptel, J.C. Remigy, New UV-photograftednanofiltration membranes for the treatment of colored textile dye effluents, J.Membr. Sci. 286 (2006) 342–350.

[13] S. Chakraborty, M.K. Purkait, S. DasGupta, S. De, J.K. Basu, Nanofiltration oftextile plant effluent for color removal and reduction in COD, Sep. Purif.Technol. 31 (2003) 141–151.

[14] M. Amini, M. Arami, N.M. Mahmoodi, A. Akbari, Dye removal from coloredtextile wastewater using acrylic grafted nanomembrane, Desalination 267(2011) 107–113.

[15] M. Riera-Torres, C. Gutiérrez-Bouzán, M. Crespi, Combination of coagulation–flocculation and nanofiltration techniques for dye removal and water reuse intextile effluents, Desalination 252 (2010) 53–59.

[16] C.H. Liu, J.S. Wu, H.C. Chiu, S.Y. Suen, K.H. Chu, Removal of anionic reactivedyes from water using anion exchange membranes as adsorbers, Water Res. 41(2007) 1491–1500.

[17] N. Saffaj, M. Persin, S. Alami Younssi, A. Albizane, M. Bouhria, H. Loukili, H.Dach, A. Larbot, Removal of salts and dyes by low ZnAl2O4–TiO2 ultrafiltrationmembrane deposited on support made from raw clay, Sep. Purif. Technol. 47(2005) 36–42.

[18] J. Kim, B. Vander Bruggen, The use of nanoparticles in polymeric and ceramicmembrane structures: review of manufacturing procedures and performanceimprovement for water treatment, Environ. Pollut. 158 (2010) 2335–2349.

[19] A. Bhattacharya, B.N. Misra, Grafting: a versatile means to modify polymerstechniques, factors and applications, Prog. Polym. Sci. 29 (2004) 767–814.

[20] S.S. Madaeni, S. Zinadini, V. Vatanpour, A new approach to improve antifoulingproperty of PVDF membrane using in situ polymerization of PAAfunctionalized TiO2 nanoparticles, J. Membr. Sci. 380 (2011) 155–162.

[21] S.S. Madaeni, S. Zinadini, V. Vatanpour, Convective flow adsorption of nickelions in PVDF membrane embedded with multi-walled carbon nanotubes andPAA coating, Sep. Purif. Technol. 80 (2011) 155–162.

[22] V. Vatanpour, S.S. Madaeni, L. Rajabi, S. Zinadini, A.A. Derakhshan, Boehmitenanoparticles as a new nanofiller for preparation of antifouling mixed matrixmembranes, J. Membr. Sci. 401–402 (2012) 132–143.

[23] N. Ghaemi, S.S. Madaeni, A. Alizadeh, H. Rajabi, P. Daraei, Preparation,characterization and performance of polyethersulfone/organically modifiedmontmorillonite nanocomposite membranes in removal of pesticides, J.Membr. Sci. 382 (2011) 135–147.

[24] V. Vatanpour, S.S. Madaeni, A.R. Khataee, E. Salehi, S. Zinadini, H. Ahmadi, TiO2

embedded mixed matrix PES nanocomposite membranes: Influence ofdifferent sizes and types of nanoparticles on antifouling and performance,Desalination 292 (2012) 19–29.

[25] N. Maximous, G. Nakhla, W. Wan, K. Wong, Preparation, characterization andperformance of Al2O3/PES membrane for wastewater filtration, J. Membr. Sci.341 (2009) 67–75.

[26] V. Vatanpour, S.S. Madaeni, R. Moradian, S. Zinadini, B. Astinchap, Fabricationand characterization of novel antifouling nanofiltration membrane preparedfrom oxidized multiwalled carbon nanotube/polyethersulfone nanocomposite,J. Membr. Sci. 375 (2011) 284–294.

[27] 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.

[28] P. Daraei, S.S. Madaeni, N. Ghaemi, E. Salehi, M.A. Khadivi, R. Moradian, B.Astinchap, Novel polyethersulfone nanocomposite membrane prepared byPANI/Fe3O4 nanoparticles with enhanced performance for Cu(II) removal fromwater, J. Membr. Sci. 415–416 (2012) 250–259.

[29] C. Ozdilek, K. Kazimierczak, D. van der Beek, S.J. Picken, Preparation andproperties of polyamide-6-boehmite nanocomposites, Polymer 45 (2004)5207–5214.

[30] C. Ozdilek, K. Kazimierczak, S.J. Picken, Preparation and characterization oftitanate-modified boehmite-polyamide-6 nanocomposites, Polymer 46 (2005)6025–6034.

[31] A.A. Derakhshan, L. Rajabi, Review on applications of carboxylate-alumoxanenanostructures, Powder Technol. 226 (2012) 117–129.

[32] C.D. Jones, M. Fidalgo, M.R. Wiesner, A.R. Barron, Alumina ultrafiltrationmembranes derived from carboxylate–alumoxane nanoparticles, J. Membr.Sci. 193 (2001) 175–184.

[33] S.J. Maguire-Boyle, A.R. Barron, A new functionalization strategy for oil/waterseparation membranes, J. Membr. Sci. 382 (2011) 107–115.

[34] W. Chen, S. Wu, Y. Lei, Z. Liao, B. Guo, X. Liang, D. Jia, Interfacial structure andperformance of rubber/boehmite nanocomposites modified by methacrylicacid, Polymer 52 (2011) 4387–4395.

[35] A.A. Derakhshan, L. Rajabi, H. Karimnezhad, Morphology and productionmechanism of the functionalized carboxylate alumoxane micro andnanostructures, Powder Technol. 225 (2012) 156–166.

[36] C.C. Landry, N. Pappe, M.R. Mason, A.W. Apblett, A.N. Tyler, A.N. Maclnnes, A.R.Barron, From minerals to materials: synthesis of alumoxanes from the reactionof boehmite with carboxylic acids, J. Mater. Chem. 5 (2) (1995) 331–341.

[37] L. Rajabi, A.A. Derakhshan, Room temperature synthesis of boehmite andcrystallization of nanoparticles: effect of concentration and ultrasound, Sci.Adv. Mater. 2 (2010) 163–172.

[38] G. Arthanareeswaran, T.K. Sriyamuna Devi, M. Raajenthiren, Effect of silicaparticles on cellulose acetate blend ultrafiltration membranes: Part I, Sep.Purif. Technol. 64 (2008) 38–47.

[39] X. Zhao, J. Ma, Z. Wang, G. Wen, J. Jiang, F. Shi, L. Sheng, Hyperbranched-polymer functionalized multi-walled carbon nanotubes for poly (vinylidenefluoride) membranes: From dispersion to blended fouling-control membrane,Desalination 303 (2012) 29–38.

[40] Y. Mansourpanah, S.S. Madaeni, A. Rahimpour, M. Adeli, M.Y. Hashemi, M.R.Moradian, Fabrication new PES-based mixed matrix nanocompositemembranes using polycaprolactone modified carbon nanotubes as theadditive: property changes and morphological studies, Desalination 277(2011) 171–177.

[41] J. Kim, B. Van der Bruggen, The use of nanoparticles in polymeric and ceramicmembrane structures: review of manufacturing procedures and performanceimprovement for water treatment, Environ. Pollut. 158 (2010) 2335–2349.

[42] R. Field, Fundamentals of fouling, in: K.V. Peinemann, S.P. Nunes (Eds.),Membranes for Water Treatment, WILEY-VCH Verlag GmbH & Co. KGaA,Weinheim, 2010, pp. 1–24.

[43] A. Saxena, B.P. Tripathi, M. Kumar, V.K. Shahi, Membrane-based techniques forthe separation and purification of proteins: an overview, Adv. Colloid InterfaceSci. 145 (2009) 1–22.