Desalination 333 (2014) 36-44

Contents lists available at ScienceDirect Desalination

journal homepage: www.elsevier.com/locateldesal

Studies on fouling by natural organic matter (NOM)on polysulfone membranes: Effect of polyethylene glycol (PEG)

Muharnad Zaini Yunos a.b. Zawati Harun "b.*, Hatijah Basri ', Ahmad Fauzi Ismail

" WGAW Depmhnent ofMatwink and Design Fa* ofMerhmdcnI and Manufacturing, UnivdTun Hwein Onn Malaysia. Rmit Rajk hluPahot 8&PW.]ohor, Malaysia Int- M-1 Proces, AdvmrcedMk ond Man- Cmtec Univerda' Tun Hlmein Onn MolaysirS Rui! Raja, Bofu PohotB6400,Johor.Wsia 'De~~echnologv ond Hen'kw. Frmlly ofsaence. Techndw andHuman Devebpnent Universili Tun Hunein Orm Malaysia Porit Raja Bntu Phhnt 86400,Johor. Malaysin Advonced Membmne Technology Rsmrrh Centre. Faculty of Cfmimland Natural Resourcer Engineer@, Universiti Teknologi Malaysia Skudot 81300JohorBahnr,]ohor, Malaysia

HIGHLIGHTS

- --- Fouling behaviour of PSf by NOM - Different concentrations and molecular weights of PEG were used as additives in membrane. - Detailed mwphology and membrane performance wmdmacterized and measured. Real river water which posseses hydrophobic and hydrophilic charackristia was wed to evaluate membrane fouling.

ARTICLE INFO ABSTRACT

ArtidehidoryI Polysulfone membranes were prepared via phase inversion technique by using polyethylene glycol with Received 19July2013 molecular weights of 400.1500 and 6000 Da as pore forming agent in dope formulation. The performance of Received in revised form 21 October 2013 membrane was characterized using humic acid and water sample taken from Sembrong River. Jobor. Malaysia Accepted 10 Novwber 2013 was used as natural organic- matter sources. Membrane properties- were also characterized in terms of mean Available online 14 December 2013 pore radius, pure water flux. humic acid rejection and fouling resistance. The results indicated that the pure flux PEG Keywordt: water and mean pore radius of membranes increased with the increase of content. Fourier transform PSf membrane inkued spechumpy results revealed the presence of hydrophilic component in PSf/PEG blend with the signif- PEG icant appearance of 0-H peak at3418.78 cm-'. Scanning electron miamcopy analysis revealed the presence of NOM finger-like stnicture for all membranes and the smhwintensified as PEG content was increased. The results Fouling obtained from the fouling study indicated that the membrane with the lowest PEG content and molecular weight has an excellent performance in mitigating fouling. Q, 2013 Elsevier B.V. All rights reserved.

remaining consists of neutral compounds. bases and contaminants. Although these compounds are relatively harmless, they are able to Natural organic matter (NOM) is one of the major pollutants in form carcinogenic disinfection byproducts such as hihalomethanes acidic and low turbidity water source. In Malaysia, most of the water [2,3].In order to remove these substances and render environmental sources are contaminated with NOM especially that from peat soil. remediation. ultrafiltration (UF) has been recognised as one of the Natural organic compounds such as humic acid and fulvic acid con* attractive approaches that has been highlighted in many studies due uted to the natural colour of water (brown to black) which becomes to its compactness, easy automation, high removal rate of organic more visible if the dissolved organic carbon (DOC) exceeds 5 mg/L matter and also capabiity to remove vim [4]. Therefore, the removal of NOM is usually known as colour removal. Polysulfone (PSf) is the most commonly used polymer in the fabrica- According to Thurman [I],surface water in average contains about tion of UF membrane. PSf is known for its resistance in exheme pH con- 45%fuhric acid. 5%hurnic acid. 25%low molecular weight acid. and the dition and high thermal stability [5]. However, one of the major problems of polysulfone is its hydrophobic characteristic which often causes hydrophobic particle to adsorb on the surface of the PSf membrane. This phenomenon had led to membrane fouling and drastically * Corresponding author atl integrated Material Process. Advanced Materials and decreased the membrane permeab'ity. Manufacturing Center. Universiti Tun Hussein OM Malaysia. Parit Raja Batu Pahat Fouling is desmias pore-blocking. solute aggregation or adsorp- 86400. Johor. Malaysia TeL: + 60 74537608: fax: + 60 74536080. E-mdadhsex [email protected] (MZYunos). [email protected] tion phenomenon. Irreversible membrane fouling by proteins. NOM id (2Harun). [email protected](H Basri), [email protected] (AF. ha]. other biomolecules' adsorption reduces the flux of membrane and

0011-91646 - see front matter O 2013 Elsorier B.V. AU righes reserved http://dx.doi.org/l0.1016/j.desa12013.11.019 hf.2 Yunos ef a1 I Dsah hinders the wide scale application of UF membranes 16.7). The current T*l technique to counter this problem is by cleaning the membrane with Compositionof tasting solution aggressive chemicals such as hypochlorite, nitric add. sodium hydrox- PSf NMP PFGW PU;15OOa PEG6000a ide and oxalic acid at extreme pH [8,9]. These chemicals can only remove the adsorbed foulants for some time and the membrane must be replaced when this cleaning method becomes ineffective. The mechanisms and causes of fouling that strongly depend on the membrane surface characteristic (hydrophilicity, charge and roughness). concentration polarization, cake layer formation, foulant properties and water chemistry have been well studied and reported i7.10-121. " Represent moldar weight (Da). Among an these causes. membrane surface seems to be the most mdal factor tobe looked into since the properties of membrane surface could potentially affect the overall performance of the resultant membranes. surface modifications of membrane to obtain high hydrophilicity. low surface roughness and positive charge haw been progressively microscopy (SEM) and Fourier transform-infrared spectroscopy investigated in order to produce a membrane with high antifouling (FIIR) to provide further insights in the fouling mechanism of PSI -properties - 113-151. membrane. A membrane with good rejection and water permeability is usually prepared by incorporating suitable additive Urnugh an efficient fabrication method. Various types of additive such as mineral fillers, polymer, salts. non-solvent and solvent were incorporated using different 2. I. Materials approachestoimprove the membrane properties [I6-1 91. Generally. inorganic and mineral fillers are used to enhance and create stability to Polymer solutions were prepared using polysulfone (UDEL P1700) the membrane performances. Hamid et al. reported that the addition of as polymeric material and N-methyl-2-pyrrolidone (NMP) (MERCK) titanium dioxide nanopartide in polysulfon~has reduced the fouling They as solvent. Meanwhile PEG 400.1500 and 6000 (Qrec) were used as resistance of the membrane 1201. also found that the additive an additive. Distilled water was used as a non-solvent bath for the has increased rejection of humic add up to 90%. ldris et al. studied the purposes of phase inversion. All chemicals purchased in this study effect of lithium bromide salts on the performance of a polyethersutfone were used without any further purification. hollow fibre membrane and found that lithium bromide enhanced membrane hydrophilicity and water permeabiity [21]. The capability of a strong non-solvent such as water as an additive in the membrane 22. Membrane pqmafion performance was studied by Yunos et al. [22]. They discovered that by adding water in PSf dope fornulation, pure water flux was greatly In this study. PSfpEG flat sheet membranes were prepared by enhanced despite the deterioration in the strength of the membrane. casting a polymer solution (18 wt%of PSf) with different PEG contents Typically, a low molecular weight polymer is used as an additive to im- and molecular weights on a glass plate. Table 1 shows the composition prove the pore forming of the membrane. Polyvinyl pyrmlidone (PVP) of various polymer solutions prepared in this study. Polymer solution and polyethylene glycol (PEG) arr the most common additives used in was cast on the glass plate with casting knife gap setting at 150 p membrane formulation 123-261. Investigation of PEG as an additive with an appropriatecasting shear. The cast solution was then immersed has attracted enormous attentions from many researchers due to the in water bath until the membrane thin film peels off naturally. The pro- ease of preparation as PEG can dissolve in water and organic solvent, cedures were conducted at constant temperature and relative humidity low toxicity and cost [26-291. PEG has shown promising potential (HR)(25 "C;HR 84%). in increased membrane pore size and permeability. Ma et al. have studied the effect of PEG on the resultant PSf membrane and they found that the presence of PEG has greatly improved the membrane 23. Smnning electron microscopy (SEMI hydrophilicity. porosity and water permeability 1271. The addition of PEG has also reduced the thermodynamic stability of dope formula- SEM JEOL GSM was used to examine the morphology of mem- tion which then led to finger-like formation in the membrane brane. The membrane was immersed in liquid nitrogen and fractured structure [26]. According to some previous reports [I6.30.311, surface-bound long-chain hydrophilic molecules in PEG were sufficient to form steric repulsion which prevents adsorption of protein molecules onto membrane. Zhu et al. 1321 found that the addition of PEG in membrane could reduce the total adsorption of bovine serum albu- min (BSA) on the membrane surface. Kim et al. [33] prepared a surface-coated membrane using PEG hydrogel on reverse osmosis (RO) membrane and demonstrated fouling reduction of salt water as foulant. Grafting of PEG on membrane surface was also found effective to reduce at least 96%BSA adhesion as studied by McCloskey et al. [34]. They found that grafted PEG created hydration shell that renders surface resistant to BSA adsorption. As far as we are concerned. the investigations in the effect of PEG concentration on the performance of membrane for NOM removal and fouling mitigation were rarely studied. Thus. in this work. the fouling behaviour of the PSf membrane with PEG as an additive Fi I. ul&aafiltrationp~meationtesting unit schematicdia~am, (1 ) feed tank; (2) control was studied by using NOM source from real river water. The Mhre:(3) pm-tmahnent W (4) osmoficpump: (5) flow meter: (6) pressure gauge: (7) prepared membranes were characterized via scanning electron filter holder. (8) beaker hr coUeciing permeate. T-Z prior to the weighing with electronic balance, and the initial weight Smface water characmklicafier preliltration pmcen was denoted as Ww.This wet membrane was dried in an oven at WW (on-') 0.178 f om 60 "C for 24 h and it was weighed again in dry state, denoted as DX Imw~1 987 f 1.42 WA. -~Membrane- porosity was determined by using the following equation:

Hydrophobic DOC % 42.3% where Ww is the weight of wet membranes (g). Wd is the weight of dry membranes (g), p.,, is the density of pure water at room temperature (g/cm3) and V is the volume of membrane in wet state (cm3). In order to minimize the experimental errors, each measurement carefully. The fractured samples were then gold sputtered prior to was repeated for five times and the average was calculated. the scanning. Mean pore radius was determined by using the filtration velocity method as stated in the Guerout-Elford-Ferry equation 1241: 2.4. Fourier Wansfom infrared (R7R) Mean pore radius, r,:

FllR (Perkin Elmer Spectrum 100) was employed to detect and analyse the functional groups within the molecules of the polymer based structure in the prepared membrane. The membranes were characterized using theattenuated total reflectance (ATR) technique where q is the water viscosity, l is the membrane thickness. Qis the vol- at a 4.0 cm-' resolution and the results of 32 scans were recorded. ume of permeate water per-unit time. A is the membrane area and AP is the operational membrane pressure. 25. Membrane mean pore mdius (r,) 2.6. Permeation flux, rejecton andfouling imedgation In order to determine the membrane mean pore radius, the porosity of the membrane was first being determined. The mem- The permeation flux and rejection of membrane were measured brane was first immersed in distilled water for 24 h at 25 "CThe based on the ultrafiltration experimental set-up. The schematic of membrane surface was then wiped carefully with tissue paper the experimental set up is presented in Fig. 1. The determination of

0.022 - 1584,43 cm.~cm-~,1495.9111;r- 1495.91 cm-' 0.020 - PSflPEGPS~IPEG ,,4,07,,.1314.07 cnl 1243.10 cm' / 0.016 - 0.014 - 0.010 - 0.008 -

0.004 - 0.0020,000o-EJ - uVYk 0,000 - 1493.89cm" 0.31 2

0.25 - 0.20 - q: 0.15 - 0.10 - PSf 2974.6 0.05 - 0.00 -- -0.02 - 0.09 -

2877.24 cm-I 1 1095.43 cm-' 0.07 - 1344.92 cm-' 0.06 - 0.05 - 0.04 - 1463.78cm.' PEGP 3418.78~3418.78 cr .' 948.73 cm-' 838.70 cm"

, 4000 3500 3000 2500 2000 1500 lo00 600 cm-'

Fii2. FllR multr Overall view Top crossection a)

Agl Scanning elebrca miaqmpb image at top and overall membrane aoss &on of PSfPEG at different PEG contents; a) 4 wt.% b)8 wt% c)10 wt% d)l6 wt% pure water flux by using distilled water as feed was conducted at pres- where PWF is the pure water flux (4rn2h). Q is the permeate volume sure 200 kPa. The flux was calculated using Eq. (3): (L). A is the membrane area (m2), and At is the permeate time (h). Rejection was charaderized using 100 mg/L hurnic add and Sembrong Q pw~=- (3) river water. Johor. Malaysia was used as feed solution. Membrane (A x At) was first filtered with distilled water until the flux is steady. The Fk'L SEM image of PSm at dilfgent PEG mtents andmddeweights: a)400 IXk b)1500 Dq c)Wk concentrations of feed and permeate solutions were determined by where %Risthe rejection percentage, C, is the permeate concentration using a UV spectrophotorneter (Thermo Scientific. Genesys 10s) and and 4is the feed concentration. total organic carbon (Shirnadzu. TOC-VCSH) which was calculated The capability of the prepared membrane to resist NOM water using Eq. (4): source was investigated. Fouling can be measured by membrane resistance during ultrafiltration process. This resistance is due to the cake layer formation on membrane surface and impurities ad- (41 sorption onto or within membrane pores. Therefore to determine

oh-, I-1- 400 1500 6000 4 6 8 10 12 14 16 PEG MW (da) PEG content (%) l3g 6. Mean pore radius of PSf/PEG membrane for 10 wt% PEG content at different PEG Fig 5. Mean pore radius of PSfAG 400 membrane at dfirent PEG concentrations molecular weights. fouling resistance of membrane. Darcy's Law was used as shown in Eq. (5) 1201:

where J is flux, pis the viscosity of permeate. AP is the transmembrane pressure. and & is the total filtration resistance of the membrane. Mean- while R, &. R, and % are intrinsic membrane resistance, membrane resistance due to concentrationpolarization, cake layer formation and adsorption.in partiah order. The membrane was first subjected to a pressure of 200 kPa and the feed temperature at 20 "CThe distilled water was then replaced with the river water sample at the same operating pressure. The stable water flux using distilled water was recognised as J-. The permeate readings were then periodically taken for every 5 min interval time throu&out- 120 min of filtration duration. After 120 rnin of filtration, -- - the permeate flux was recorded and known as J,, The membrane was 0 4 8 12 16 20 rinsed with distilled water to remove concentration polarization layer PEG content (%) for 10 min and pure water flux was measured again and labelled as J,. Fii7. Pllre water fhm of PSf at dierent PEG contents and molecular weigh&. Lastly* the membrane was washed with a 0.1 M Ha solution for 10 min and the PWF was measured and recorded as J,. HCI was used in this v an-'. study as a chemical cleaning agent in order to remove any foulant = 1150 Alkyl (R-CHI) stretching vibration is observed at adsorption on membrane surface. Individual resistant membrane was around 2850-3000 an-'. The spectrum also shows a broad absorp- v then calculated using the following equations: tion band for hydroxyl group at = 3200-3600 cm-'. Advanta- geously. PEG possesses both hydrophilic and hydrophobic properties thus giving them a unique ability to dissolve in both aqueous and organic solvents. Thus it could be homogenously mixed with PSF in NMP solution. The peak around 287720 an-' is assigned to N- CH3 (aliphatic) stretching vibration and this signifies the presence of PEG in PSf blend. The presence of OH vibration peak at 3418.78 cm-' for PSf/PEG membranes also indicates that they are able to interact AF' with water and perfom as a hydrophilic membrane. The findings R, = --R,-R, rJ, are in accordance with those reported elsewhere [I 7,361 and it is be- lieved that this characteristic has contributed to an increase of pure water flux of the PSf/PEG membrane. which will be further discussed in the following section.

32. Morphologicalpmperries of membranes 2.7. Sembrong surface water quality Fig. 3 represents the SEM micrographs of the cross section of the PSf The surface water used in this study was taken from Sembrong River membrane at different PEG concentrations. Observation on the mem- at Parit Raja. Johor. Malaysia. The river source water was found to be brane stmcture showed that all prepared membranes have asymmetric soft and rich in NOM. The characteristics of the surface water are porous substructure with a dense skin layer. This mernbrane structure is summarised in Table 2. The raw water was pre-filtered with 0.5 p ceramic microfiltration to remove the particulate materials. The UV254nm absorbance was measured using a Thermo Scientific. Genesys 10s W-Vis spectrophotometer at a wavelength of 254 nm Concentration of heavy metal was determined using Hach DR5000. The hydrophiliaty/hydrophobicily of NOM fraction characteristic was based on DOC and mass balance technique with DAX-8 and XAD-4 resins [35). c g so- 3. Remlts and dkuioms u 0 - 40 - 3.1. Fourier tmmJorm-infimed sjxctmcopy (FIIR) d 30- The FIRspectra for the prepared membranes are depicted in Fig. 2. The aromatic vibration ether band of PSf is detected at around 1241 cm- '. The S& groups' sh-etching vibration is around 1300 an-'. The absorptions at 1900 &'. 1775 an-'. 1605 an-'. 1488 cm-' 01 7-7 and 11 69 an-' are resulted from the aromatic ring of PSf. The vibration 4 6 8 10 12 14 16 at 2968 cm-I is attriito the two methyl groups of PSf. PEG with a PEG content (%) chemical shcture of (HO-CHr(CH2-O-CH~)n-CH2-OH)has resulted in significant peaks. The peaks at 1050 cm-' to 11 50 cm-' were Eig& Humic dad rejection of PSf membrane at different PEG contents and moledar due to the stretching of ether groups, with a maximum peak at weights 4 8 10 12 16 PEG content (%I

4 8 10 12 16 PEG content (%)

Fe a ww (a) and DOC (b) Sembmng water rejdon of PSf membrane at different PEG contents and molecular weights. generally governed by the interadion between the polymer solution, The membrane solidificationusually starts at the membrane's outer sur- non-solvent and kinetic process during phase inversion process 1371. face due to thermodynamic instability between dope solution and The addition of PEG in PSf meincreased the finger-like structure water. Then it is followed by diffusion of solvent into water and water at the top of the membrane. This could be due to dope solutions that be- into membrane film As water starts to penetrate into the membrane. come thennodynamidy less stable with the presence of PEG (27,291. droplets are formed and bind together to solidify at the inner surface Fig. 4 shows the cross section images of the PSf meat dser- of the membrane. In the meantime. solvent tends to diffuse towards ent PEG molecular weights and contenb.The figures show similar trend outer membrane. This mechanism leads to the formation of solidified for PEG 40D and PEG 1500, however as PEG 6000 content increased walls of finger-like structure (191. The size of the finger-like structure from 10 wt.% to 16 wt%. the dense layer of PSf becomes thicker and depends on the diffusion rate of discharge solvent towards the outer the sponge-like structure can be more clearly seen. Formation of surface of the membrane. Due to the low solubility of PEG in water membrane depends on the solubility and diffusivity between poly- at high PEG molecular weight. diffusion between solvent and non- mer. solvent and non-solvent (381.When membrane dope was cast solvent became slower and the wall created between PSf and water and immersed in water bath, the low molecular weight PEG and NMP became larger. In such situation, a larger finger-like structure is diffused in water instantaneously from dope solution and PSf solidified formed. to form a finger-like structure. However, as PEG molecular weight increased, the PEG diffusion to water becomes slower due to the low sol- 33. Mean pore radius, water permeability and rejection test ubility between PEG and water. hence resulted in thick sponge-like structure formation in the top layer of the membrane. A similar obser- The performance of the prepared membranes was characterized in vation has been reported by Li et al. (391 using PEG as additive and terms of mean pore radius, water permeability and rejection test. dimethyl acetamide (DMac) as sohrent Fig. 5 shows the mean pore radius of the PSfPEG membrane at different Fig. 4 shows that the size of the finger-like structure at the interme- PEG contents. As demonstrated from the figure. membrane mean pore diate layer becomes larger with the inaeasing PEG molecular weight. radius increased as the PEG content increased. This is due to the Q 0.9- - u"~~oo 3 0.8- q q q q V Y " 0000 q >( 0.7 - X 04% PEG OO~~oo 0.6- 0.6 - ALAb 00~~~~~~~ il A A A A 0400oa U 0.5 - iL 1 0.4- -3 0.4- A >> o15mOa - 010% PEG u""nLni, m 1 0.3 - i6M)ODa 012% PEG E 0.2- x 16% PEG 2 0.1 *o,,,,,,8,,,,1 07,~~~,~~~~~~~ 40 50 0 10 20 30 60 70 80 90 100110120 0 10 20 30 40 50 60 70 80 90 100110120 lime (min) Time (min)

400- g xx h E 350 < 250 dgo 04%PEG 3 200- 00?8f38ea 66086&a&0&aa nB%PEG E 200- OOOOOn 3 A \L - 8888888~~gg88888 oonoooo~~~~~~O~~O~ol(pkpEG I-- P 154- 1 015~)oa 3 loo- A L~ L n Afih~~h-'B)(IODa

0 10 20 30 40 50 60 70 80 90100110120 017,~m 8 8 Time (mln) 0 10 20 30 40 50 60 70 80 90 100110120 Time (mln) E?g 10. Fmbmame of PSfpEG400 ublibtion membrane at dierent PEG wntenh: a) Normalired flux, b) Absolute flux Fig 11. Performance of PWl0 wi%PEG uhliltratiw membrane at different PEG molecular weights: a) Normalized flu%b) Absolute tlux. formation of a finger-like and macrovoid structure that enhanced the membrane mean pore radius. The same trend is also obse~edat dier- pass through the membrane. The same phenomenon was observed for ent PEG moIecular weights with fixed PEG content (10 wt%)as shown addition of PEG 1500 and PEG 6000 in membrane. The result shows in Fig. 6. that addition of PEG 400 in membrane has excellent humic acid rejec- PWF of all membranes added with different PEG contents and mo- tion as compared to PEG 1500 and PEG 6000. A similar trend was lecular weights is presented in Fig. 7. These results demonstrated that found by Ma et al who used BSA and pepsin as rejection model 1271. water permeability increased as PEG content increased regardless of NOM rejection of the membranes is illustrated in Fig. 9. Fig. 9(a) the PEG molecular weight and content However for PEG 6000,the represents Ww rejection and Fig. 9(b) shows DOC rejection The PWF start to level off at 10% PEG content. The figure also shows that highest rejection of NOM is shown at 618% (UV254)and 44.4% (DOC) with addition of PEG at 12 wtX and 16 wtqthe membrane incorporated for Psiwith 4%PEG 400 content The overall results show that WZs4 with PEG 6000 shows lower flux as compared to PEG 1500 and PEG 400. rejection has higher rejection as compared to DOC. This suggests that In gened, The increment of water permeabilityi the prepared membrane was able to remove aromatic compound such to the increased membrane pore size and imprwed surface hydrophi- as hydrophobic acid which contributes 42.3% DOC in river water liaty. This unexpected trend for PEG 6000 might be due to the forma- (Table 2). The figure also indicates that NOM rejection decreased as tion of a thick sponge-like struchue at the top layer of membrane as PEG content and molecular weight increased. As compared to humic observed in SEM images. This is in line with the findings reported by acid rejection. the rejection was significa&ly affected by Hua et a1 who found that high molecular weight of PEG may contribute the increasing PEG content and molecular weight. This is due to the to the deterioration of the membrane pure water flw 1361. fact that NOM contains several compounds such as humic acid, tannic Humic add rejection exhibited by PSf membrane with different PEG acid, protein and carbohydrate with various sizes and charges. These contents and molecular weights is shown in Fig. 8. As shown in the various sizes and elecmstatic charges of solute molecule might affect figure. the humic acid rejection by PSf membrane was slightly decreased the performance of the membrane, in terms of declination of membrane as the PEG 400 content increased This is due to the undesired trade-off reiection. On the other hand, small sizes of the NOM molecule will pass effects, in which the rejection decreases as the flux increases. The mean through the membrane and the hydrophobic NOM molecules showed pore radius increment has also contributed to the decrease of humic high tendency to be adsorbed inside the membrane 140,411. acid rejection as larger pore size allowed more hurnic acid molecule to 3.4. Fouling investigation Table3 Individual filtmtiw resistant of membrane Fig. 10 presents the normalized and absolute flux for the PSf membrane with different PEG 400 contents. As shown in the figure. the %, (lo'* m-') & (10'~m-') & (10" m-') % (10'' m-'1 flux of the membrane reduced with time. It is noticed that the flux P.4.400 21388 (* 0221) 9.166(f OX8) OZ8 (*O-lS4) (* O-lQ2) significantlyreduced as the PEG content increases in the membrane. P.8.400 7575 (f 0.171) 0.161 (f0.141) 0.168 (f 0.104) 0.175 (f0.091) p.1a-loo 3.189 (*o.ng 0.1151 (fo.157) 0.m~(*om) o.;lsl (fo.101) Even thou& thememheincorporatedwith 4 PEG has the low- P.12400 2818 (-+0322) om1(-+om) 0.144 (~Iw)1.149 (f0.112) est flux the flw is more stable as comparrd to the membrane incorpo- P.16.400 1954 (f 0.125) 013212 (f0.321) 0.0439 (f0.077) 1.652 (f0373) rated with 16 wt% PEG. These results indicate that the addition of PEG ~.10-1500 2.693 (+0.175) 0336 (f0.164) 0.131 (f0JW 1.166 (f 0258) inaeased the membrane fouling The owedresults are in conkadic- P.10-rn 22% (f0.183) 0.743 [f0.129) 0.159 (f 0J-W 4385 (f0%) tion with the previous results reported by ~i~ et al. in which PEG 44 MZ Yunos eta1 / Desalination 333 (2014) 3W enhance membrane antifouling 1331. This might be due to variation of [I21 J.-y. Tian. Y.-p. Xu. 2-1. Chen, J. Nan. G.-b. ti. Air bubbling for alleviating membrane fouling of immersed hollow-fiber membrane for ultrafiltration of river water. the type offoulant used in different studies. Table 3 gives the value of Desalination 260 (2010) 225-230. individual filtration resistance of the membrane with different PEG (131 L-J. Mu, W.-Z. Zhao. Hydrophilic modification of polyethersulfone porous mem- contents it was noticed that PEG addition reduced intrinsic membrane branes via a thermal-induced surface crosslinking approach. Appl. Surf. Sci. 255 (2009) 7273-7278. resistance but inaeased adsorption resistance of the membrane. The 1141 A.V.R Reddy. D.J. Mohan, A Bhattacharya. V.J. Shah. P.K. Ghosh. Surface modification figure also shows that contribution of fouling resistance due to conm- of ultrafiltration membranes by preadsorption of a negatively charged polymer: I. tration polarization and cake layer formation is not significant as com- Permeation of water soluble polymers and inorganic salt solutions and fouling resistance properties. J. Membr. Sci. 214 (2003) 211-221. pared to resistance due to fouling caused by membrane adsorption. 1151 D. Rana, T. Matsuura, Surface modifications for antifouling membranes. Chem. Rev. This phenomenon is due to the existence of hydrophilic and hydrophw 110 (2010) 2448-2471. bic substances in the river water sample. These substances tend to 1161 Z Peng. L.X. Kong. A thermal degradation mechanism of polyvinyl alcohol/silica adsorb easily on the membrane surface and subsequently foul the mem- nanocomposites, Polym Degrad. Stab. 92 (2007) 1061-1071. 1171 E. Yuliwati. AF. Ismail. Effect of additives concentration on the surface properties brane. As the pore size increased m the membrane,tendency for adsorp- and perfonnance of WDF ultraliltration membranes for refinery produced waste- tion of this substance also inarased. This aspect is dearly observed for water treatment. Desalination 273 (2011) 226-234. fouling behaviour at different PEG molecular weights as shown in 1181 M.A. Aroon, AF. Ismail. M.M. Montazer-Rahmati. T. Matsuura, Morphology and permeation properties of polysulfone membranes for gas separation: effects of Fig. 1 1 and Table 3. Normalized flux for PEG 6000 is drastically reduced non-solvent additives and co-solvent. Sep. Purif. Technol. 72 (2010) 194-202. to 63%for 20 min due to the membrane fouling. Meanwhile PEG 400 1191 A. Mansourizadeh. A.F. Ismail. Preparation and characterization of porous WDF shows excellence normalized flux as compared to PEG 1500 and PEG hollow fiber membranes for C02 absorption: effect of different non-solvent additives in the polymer dope. Int J. Greenhouse Gas Control 5 (2011) 640-648. 6000. As shown in Table 3. even though PEG 6000 shows the lowest 1201 N.A.A. Hamid. A.F. Ismail, T. Matsuura. A.W. Zularisam. W.J. Lau. E. Yuliwati. M.S. intrinsic membrane resistance. it has the highest resistance towards Abdullah. Morphological and separation performance study of polysulfonejtitanium the fouling that was due to adsorption, cake fonnation and concentra- dioxide (PSFlTiO2) ultrafiltration membranes for humic add removal. Desalination 273 (2011 ) 85-92. tion polarization 1211 A. Idris. I. Ahmed. M. Misran. Novel high performance hollow fiber ultrafiltration membranes spun from LiBr doped solutions. Desalination 249 (2009) 541-548. 1221 M.Z Yunos. Z. Harun, H. Basri. A.F. Ismail, Effects ofwater as non-solvent additive on performance of polysulfone ultrafiltration membrane. Adv. Mater. Res. 488-489 (2012) 46-50. The effects of different PEG contents and molecular weights to the 1231 A. Xu. A Yang. S. Young. D. deMontigny, P. Tontiwachwuthikul. Effect of internal PSf membrane in respect to morghology. flu. rejection and fouling coagulant on effectiveness of polyvinylidene fluoride membrane for carbon dioxide separation and absorption. J. Membr. Sci. 311 (2008) 153-158. were investigated. It was found that PEG enhanced pore fonnation of 1241 H. Basri. A.F. Ismail. M. Aziz Polyethersulfone (PES)-silver composite UF membrane: the membrane. As PEG molecular weight increased. a finger-like and effect of silver loading and WP molecular weight on membrane morphology and dense layer structure also enhanced. The same effect was also noticed antibacterial activiw. Desalination 273 (201 1) 72-80. (251 Y. Zhang, 2 Jin. X. Shan, J. Sunarso. P. Cui. Preparation and characterization of phos- for the membrane water flux In the case of PEG 6000. the flux started phorylated Zr-doped hybrid silica/PSF composite membrane, J. Hazard. Mater. 186 to level off at 10 wt.% PEG content. Overall rejection of membrane (2011) 390-395. decreased as PEG content and molecular weight increased for both 1261 S. Wongchitphimon. R Wang. R. Jiraratananon. L Shi. C.H. Loh. Effect of polyethylene glycol (PEG) as an additive on the fabrication of polyvinylidene fluoride-co- humic add and Sembrong River water sample. PEG enhand pore for- hexafluropropylene (PVDF-HFP) asymmetic microporous hollow fiber mem- mation of the membrane but increased the fouling properties of the branes. J. Membr. Sci. 369 (2011) 329-338. membrane. [Dl Y. Ma. F. Shi.J. Ma, M W&J. ZhangC Gao. Effect ofPEC additive on themorphology and performance of polysulfone ulhafilhation membranes Desalination 272 (201 1) 51-58. [XI] J.-H Kim. K.-H. Lee. Effect of PEG additive on membrane formation by phase inversion. J. Membr. Sci. 138 (1998) 153-163. 1291 8. Chakrabarty, AK Ghoshal. M.K Pwkait Effect of molecular weight of PEG on membrane morphology and transport properties. J. Membr. Sci. 309 (2008) 209-221. The authors thank the financial support for the LRGSVot A022 grant (301 D.Y. Koseoglu-lmer. B. Kose. M. Altinbas. I. Koyuncu. The production of polysulfone and the FRGS Vot 0754 grant that were sponsored by the Ministry of (PS) membrane with silver nanoparticles (AgNP):physical propertis. filtration per- Higher Education Malaysia (MOHE). formances, and biofouling resistances of membranes, J. Membr. Sci. 428 (2013) 620-628. [31] F.-Q. Nie. Z:IC Xu. P. Ye. J. Wu. P. Seta. Acrylonitrile-based copolymer membranes References containing reactive groups: effects of surface-immobilized poly(ethy1ene glycol)^ on anti-fouling properties and blood compatibility. Polymer 45 (2004) 399-407. [I] E.M. Thurman. Preparative isolation of aquatic humic substance, Environ. Sci. 132) L.-P. Zhu. L Xu. B.-K Zhu. Y.-X. Feng. Y.-Y. Xu. Preparation and characterization of Technol. 15 (1981) 468-466. improved fouling-resistant PPESK ultrafiltration membranes with amphiphilic 121 A. Matilainen. E.T. Gjessing. T. Lahtinen. L Hed. A Bhatnagar. M. Sillanpaa. An over- PPESK-grafi-PEG copolymers as additives,J. Membr. Sci. 294 (2007) 19f5-206. view of the methods used in the characterisation of natural organic matter (NOM) 1331 D.H. Kim. S.-H. Moon. J. Cho, Investigation of the adsorption and transport of natural in relation to drinking water treatment. Chemosphere 83 (2011) 1431-1442. organic matter (NOM) in ion-exchange membranes. Desalination 151 (2003) [3] H. Wong KM. Mok. XJ. Fan. Natural organic matter and formation of tihalometh- 11-20. anes in two water treatment processes. Desalination 210 (2007) 44-51. 1341 B.D. McCloskey. H.B. Park. H. Ju. B.W. Rowe. D.J. Miller. B.J. Chun. K. Kin, B.D. 141 M. Mulder. Basic Principles of Membrane Technology. 2nd ed. Kluwer Academic. Freeman, Influence of polydopamine deposition conditions on pure water flux London, 1996. and foulant adhesion resistance of reverse osmosis, ultrafiltration,and microfiltration [5] S.P. Nunes. K.V. Peinemam. Membrane Technology in the Chemical Industry. 2006. membranes. Polymer 51 (2010) 3472-3485. 161 S. Lee, C.-H. Lee. Effect of membrane properties and pretreatment on flux and NOM 1351 J. Labanowski. G. Feuillade. Dissolved organic matter: precautions for the study of rejection in surface water nanofiltration. Sep. Purif. Technol. 56 (2007) 1-8. hydrophilic substances using XAD resins. Water Res. 45 (2011) 315-327. 171 X. Wei. R Wang. Z. Li. A.G. Fane. Development of a novel electrophoresis-UV 136) H. Hua, N. I.L Wu. H. Zhong G. Wu. Z. Yuan, X tin, L Tang, Anti-fouling ultrafiltra- grafting technique to modify PES UF membranes used for NOM removal. J. tion membrane prepared from polysulfone-graft-methyl acrylate copolymers by Membr. Sci. 273 (2006) 47-57. W-induced grafting method. J. Environ. Sci. 20 (2008) 565-570. 181 S.KS. Al-Obeidani. H. Al-Hinai. M.FA Goosen. S. Sablani. Y. Taniguchi. H. Okamura. 1371 Q.2. Zheng. P. Wang. Y.-N. Yang. Rheological and thermodynamic variation in Chemical cleaning of oil contaminated polyethylene hollow fiber microfiltration polysulfone solution by PEG introduction and its effect on kinetics of membrane membranes. J. Membr. Sd. 307 (2008) 299-308. formation via phase-inversion process. J. Membr. Sci. 279 (2006) 230-237. [91 W.S. Ang. N.Y. Yip. A. Tirafem. M. Elimelech. Chemical cleaning of RO membranes (381 GR Guillen, Y. Pan. M. ti. E.M.V. Hoek. Preparation and characterization of mem- fouled by wastewater effluent: achieving higher efficiency with dual-step cleaning branes formed by nonsolvent induced phase separation: a review. Ind. Eng. Chem. J. Membr. Sci. 382 (2011) 100-106. Res. 50 (2011) 3798-3817. [lo] A.W. Zularisamn. AF. Ismail. M.R Salim. M. Sakinah. T. Matsuurq Application of coag- I391 J.-f. ti. Z.4. Xu. H. Yang. C.-p. Feng. J.-h. Shi. Hydrophilic Microporous PES Mem- ulation-ultrafiltration hybrid process for drinking water treatment: optimization of branes Prepared by PES/PEG/DMAc Casting Solutions. 2007. operating conditions using experimental design. Sep. Purif. Technol. 65 (2009) 1401 AW. Zularisam. AF. Ismail. R Salim. Behaviours of IIdhlral organic matter in membrane 193-210. filtration for surface water treatment - a review. Desalination 194 (2006) 21 1-237. [Ill W. Shi. M.M. Benjamin, Membrane interactions with NOM and an adsorbent in a vi- [41] A.W. Zularisam. A.F. Ismail. M.R. Salim. M. Sakinah. H. Ozaki. The effects of natural bratory shear enhanced filtration process (VSEP) system. J. Membr. Sci. 312 (2008) organic matter (NOM) fractions on fouling characteristics and flux recovery of ultra- 23-33. filtration me~nbranes.Desalination 212 (2007) 191-208.