Amphiphilic PPESK-g-PEG graft copolymers for hydrophilic modification of PPESK microporous membranes

European Polymer Journal 43 (2007) 1383–1393

https://www.360docs.net/doc/aa7067367.html,/locate/europolj

0014-3057/$ - see front matter ? 2007 Elsevier Ltd. All rights reserved.doi:10.1016/j.eurpolymj.2007.01.037

Amphiphilic PPESK-g -PEG graft copolymers for hydrophilic

modi W cation of PPESK microporous membranes

Li-Ping Zhu, Chun-Hui Du, Li Xu, Yong-Xiang Feng, Bao-Ku Zhu, You-Yi Xu

¤

Institute of Polymer Science, Zhejiang University, Hangzhou Yuquan 310027, PR China Received 30 September 2006; received in revised form 6 December 2006; accepted 8 January 2007

Available online 30 January 2007

Abstract

Amphiphilic graft copolymers comprising poly(phthalazinone ether sulfone ketone) (PPESK) backbones and poly(eth-ylene glycol) (PEG) side chains were synthesized and blended into PPESK casting solutions to prepare hydrophilic and anti-fouling microporous membranes. The graft copolymer was prepared by a modi W ed Williamson etheri W cation method.Sodium alkoxide of methoxyl PEG (PEG–ONa) was used to react with chloromethylated PPESK (CMPPESK). FT-IR spectroscopy, 1H NMR and solid-state 13C CP-MAS NMR analysis con W rmed the covalent linking of PEG with PPESK backbones. The incorporation ratio of PEG calculated from 1H NMR was in agreement with that from TGA tests. The graft products were added into PPESK casting solutions to prepare composite porous membranes using phase inversion method. X-ray photoelectron spectroscopy (XPS) and water contact angle examinations indicated that the grafting copoly-mers were preferentially excluded to the membrane-coagulant interface during membrane forming, contributing the mem-branes with improved hydrophilicity and surface wettability. Compared with the neat membrane, the blend membranes exhibited a larger surface pore size and less susceptible to protein fouling.? 2007 Elsevier Ltd. All rights reserved.

Keywords:Amphiphilic graft copolymers; Poly(phthalazinone ether sulfone ketone); Membranes; Blending modi W cation

1. Introduction

Poly(phthalazinone ether sulfone ketone) (PPESK)is a newly-developed amorphous copolymer con-taining rigid phthalazinone moieties and aromatic structure. PPESK has shown excellent thermal sta-bility and chemical resistance [1,2]. This polymer is soluble in some polar aprotic solvents and thus can

be easily cast into porous membranes by traditional phase inversion technique. Jian and his co-workers have successfully used PPESK as membrane mate-rial for gas separation (GS), ultra W ltration (UF) and nano W ltration (NF), etc. [3,4]. However, the hydro-phobic nature of PPESK can lead to the adsorption of proteins or natural organic matters (NOM) which results in membrane fouling in water treatment.Many e V orts have been made to enhance the hydro-philicity and separation performance of the PPESK membrane. For example, Dai et al. used sulfonated PPESK to prepare NF membranes for the separa-tion of dyes and salts from water at a high operation

*

Corresponding author. Tel.: +86 571 87953011; fax: +86 57187953011.

E-mail address: opl-yyxu@https://www.360docs.net/doc/aa7067367.html, (Y.-Y. Xu).

1384L.-P. Zhu et al. / European Polymer Journal 43 (2007) 1383–1393

temperature. Sulfonated PPESK exhibits better

thermal stability and enhanced hydrophilicity than

the virgin PPESK [5,6]. Zhang et al. prepared thin W lm composite (TFC) membranes using sulfonated PPESK as the top layer coating on PPESK UF sup-

port membranes. The TFC membranes showed a

high water X ux at low pressures [7]. Su et al. made use of chloromethylated PPESK (CMPPESK) as the material to fabricate porous membranes by phase inversion method, and then the membranes were quaternized by immersing it into an aqueous tri-methylamine solution. The resultant membranes were used for the separation of dyes and divalent salts [8]. Xuan et al. synthesized a modi W ed PPESK contain-ing carboxylic groups in its side chains, this polymer showed high hydrophilicity and good solubility [9].In addition, several researchers fabricated TFC membranes via interfacial polymerization on PPESK UF membranes. The obtained membranes were used for desalination in high operation temperature [10,11]. However, little work was done to improve the properties of PPESK membrane through a chemical graft method.

In recent years, amphiphilic graft copolymers

having hydrophobic backbone and hydrophilic side

chains have absorbed many researchers’ attention

due to their particular properties and convenient synthesis process [12]. Poly(ethylene glycol) (PEG) is widely used as the hydrophilic side chains for amphiphilic graft copolymer due to its low cost and inherent biocompatibility. For example, Chiu et al. synthesized a series of amphiphilic copolymers con-taining monomeric units such as stearyl methacrylate, methyl acrylate, acrylic acid and PEG acrylate by free radical copolymerization. The obtained copolymers were applied in drug-delivery systems and exhibited a sustained release patten for pyrene in aqueous solu-tions [13]. Hou et al. prepared an amphiphilic co-polymer based on poly(styrene-co-maleic anhydride) (backbone copolymers) and methoxypolyethylene glycols (grafts) [14]. G uo et al. synthesized a blood compatible amphiphilic copolymer with uniform PEO side chains by the copolymerization of poly(eth-ylene oxide) (PEO) macromonomer with methacrylic acid [15]. Mihaylova et al. prepared the graft copoly-mers by the reaction between pentachlorophenyl fumarate-co-styrene and amino functionalized PEO. The supermolecular structure of obtained graft co-polymers was studied by WAXS [16]. Generally, there are two routes for the preparation of amphiphilic PEG-grafted copolymers: one is via copolymeriza-tion of vinyl monomers with PEG macromonomers,the other is the conjugation of polymeric precursors with the active functionalized group of PEG [13]. The products from the latter route produce comb-like PEG side chains that often show enhanced hydrophi-licity and biocompatibility compared to the poly-meric precursors.

Several authors synthesized graft copolymers having polysulfone backbones and PEG side chains via reaction of an alkoxide formed from PEG and sodium hydride with chloromethylated polysulfone. The resulting copolymers were added into casting solution of polysulfone to surface modify poly-sulfone membranes [17]. In the present work, analo-gous approach was employed for the chemical modi W cation of PPESK resin. Hydrophilic PEG chains were introduced to the rigid PPESK back-bones through a modi W ed Williamson reaction. The obtained PPESK-g-PEG graft copolymers were blended into PPESK to prepare hydrophilic porous membranes by immersion–precipitation technique. In comparison with the neat PPESK membrane, the blend membranes exhibited an elevated anti-fouling property.

2. Experimental

2.1. Materials and reagents

PPESK (sulfone/ketone D1:1) with an intrinsic viscosity of 0.65dL/g was provided by Dalian New Polymer Co. (PR China) and was vacuum dried at 100°C for 24h prior to use. Methoxyl poly(ethylene glycol) (PEG–OH) (average molecular weight of 350 and 750, respectively), sodium hydride (NaH) and bovine serum albumin (BSA) were purchased from Aldrich. Commercially available N-methyl-2-pyrrol-idone (NMP) (Shanghai Wulian chemical plant, PR China, reagent grade) served as solvent and was puri W ed by vacuum distillation from calcium hydride before use. Chloromethylether (CME) was obtained from Shanghai Haiqu Chemical Company. All other reagents were analytical grade and used without further puri W cation.

2.2. Preparation of PPESK-g-PEG copolymers

The synthesis route for PPESK-g-PEG copoly-mers was illustrated in Fig.1. The polymeric precur-sor, chloromethylated PPESK (CMPPESK), was synthesized as an intermediate to prepare PPESK–PEG graft copolymer. The chloromethylation of PPESK followed the procedures described by previ-

L.-P. Zhu et al. / European Polymer Journal 43 (2007) 1383–13931385

ous reports [8,18,19]. In a typical reaction, 10g PPESK resin was dissolved in 150ml of 98% con-centrated H 2SO 4 under vigorous stirring at room temperature. About 40ml of CME was added drop-wise into the PPESK solution, and the reaction was performed under stirring for 5h. Then, the reaction mixture was precipitated in cool de-ionized water under mechanical agitation, and was repeatedly washed with de-ioned water until the washed water was neutral. After W ltering and completely drying,the obtained CMPPESK was used for the synthesis of PPESK-g -PEG graft copolymer.

A modi W ed Williamson etheri W cation was employed for the synthesis of PPESK-g -PEG [17].In a typical procedure, an alkoxide solution was pre-pared by adding dropwise 40ml of 10.0 mmol PEG–OH solution into 40ml of 10.0mmol NaH solution using NMP as the solvent. The PEG –OH with molecular weight of 350 and 750g/mol were used,respectively. After stirring for 2h at room tempera-ture, the mixture was added dropwise to a solution of 3g CMPPESK dissolved in 60ml NMP. The reac-tion was performed at room temperature under an argon atmosphere for 72h and was terminated through neutralizing the mixture with acetic acid. At last, the reaction mixture was precipitated into an adequate amount of de-ioned water. After W ltrated and dried, the product was puri W ed by repeatedly dissolving in NMP and precipitating in ethanol. The

resultant product was extracted for 48h by metha-nol using a Soxhlet extractor and vacuum dried at 60°C overnight.

2.3. Characterization of PPESK-g-PEG

Fourier transform infrared (FT-IR) spectra were recorded using a Vector22 (Bruker Co. Germany )instrument. The 1H NMR spectra of CMPPESKs and PPESK-g -PEG s in deutrated chloroform were determined using a Bruker Avance DMX 500 MHz instrument. Solid-state 13C CP-MAS NMR spectra for virgin PPESK and PPESK-g -PEG compolymers were acquired on a Bruker Avance AV 400MHz spectrometer using a standard Bruker broad-band MAS 4mm probe. All samples were examined with the MAS spinning speed of 5kHz, the CP contact time of 1ms and the recycle delay of 3s. The thermal weight loss of the raw material and products were evaluated at a heating rate of 10°C/min under a pro-tective nitrogen atmosphere using a Pyris 6 TG A (Perkin–Elmer, Germany) instrument.2.4. Blend membrane preparation

Casting solution for membrane formation was prepared by dissolving PPESK and PPESK-g -PEO in NMP. The weight ratio of PPESK/PPESK-g -PEO was 4/1 and the total solid content in the solution

Fig.

1. Schematic diagram for the synthesis of PPESK-g -PEG graft copolymers.

1386L.-P. Zhu et al. / European Polymer Journal 43 (2007) 1383–1393

was 15wt%. 10 weight percent of ethylene glycol methyl ether was used as a non-solvent additive.After W ltered and degassed, the solution was cast onto a horizontal glass plate at room temperature using a glass blade. After exposed in air for 30s, the nascent membrane was immersed into an 80°C deioned water bath for gelation. The resulting mem-brane was stored in de-ioned water for at least 2 days before characterization.2.5. Membrane characterization

XPS spectra for the surface of the neat and blend PPESK membranes were recorded on a PHI 5000C ESCA System (PHI co., America) employing Al K excitation radiation (1486.6eV). A W eld emitting scanning electronic microscopy (FESEM SIRION-100, FEI CO., LTD., Netherlands) was used to observe the surface morphologies and cross-section structure of the membranes. The changes of water contact angle with drop age on the top surface of the membranes were acquired in the movie mode using an OCA20 contact angle system (Dataphysics Instru-ments with G mbH, G ermany). Membrane fouling was evaluated through the UF experiments of BSA aqueous solution in a dead-end ultra W ltration cell.

3. Results and discussion

3.1. Chloromethylation of PPESK

Chloromethylation provides PPESK with active group, –CH 2Cl, for the subsequent graft reaction, as shown in Fig.1. The chloromethylation of aromatic polymers is an electrophilic Friedel–Crafts substitu-tion reaction, and the substitution position depends on the type of activating substituents linked to the phenyl ring. In the chemical structure of PPESK, the ether bond is an electron-donating substituent lead-ing to selective ortho - and para -substitution, while phthalazinyl and sulfone groups are electron-with-drawing substituents which favor the meta -substitu-tion. Thus, the chloromethyl substitution occurs at the carbon atom adjacent to the ether bond, as shown in Fig.2 [5,20]. PPESK resin used in this work is a random copolymer, in which the mole ratio of the sulfone and ketone is 1:1. Fig.2 presents the 1H NMR spectrum of CMPPESK. For chlorom-ethylated PPESK, the signal of –CH 2Cl protons was observed at about 4.8ppm (a and b ) and only a slight di V erence was between the ones adjacent to the sulfone unit and ketone unit. The incorporation of chloromethyl group was also veri W ed by the FT-IR analysis. As shown in Fig.3, the FT-IR spectrum

Fig.

2. 1H NMR spectrum of chloromethylated PPESK.

L.-P. Zhu et al. / European Polymer Journal 43 (2007) 1383–13931387

of CMPPESK presents a new strong absorbance at 754cm ?1, which is attributed to the stretching vibra-tion of C–Cl bond.

A more detailed discussion about the synthesis,puri W cation and characterization of CMPPESK was reported elsewhere [19]. The only di V erence is that,in this work, CME, instead of chloromethyl butyl ether, was employed as chloromethylation agent. In Fig.2, the peak at 8.63 ppm is assigned to the proton adjacent to the C

B O group of phthalazinone. Since its relative intensity remains unchanged after chlo-romethylation, the peak at 8.63ppm was selected as the reference peak. The number of chloromethyl group per PPESK repeat unit (DS, degree of substi-tution) was calculated based on the intensity of the 4.8 ppm peak vs the 8.63ppm peak. A series of CMPPESKs were prepared in various reaction con-ditions. The DS values of di V erent samples are listed in Table 1. The DS of CMPPESK strongly depends on the reaction conditions, including reaction tem-perature, reaction time, the amount of CME, con-centration of reactants, etc. Obviously, the ratio of CME to PPESK is the main factor that determines the DS. The value of DS increased observably with the increasing amount of CME. In this work, the DS of CMPPESK varied from 0.81 to 3.22, which was controlled by the dosage of CME. The DS value increases steadily as the reaction progresses. Increas-ing reaction temperature speeds up the chlorome-thylation process, but the ultimate DS declines because of sulfonation and chain degradation of PPESK at elevated temperatures. Thus, the reaction was performed at the temperature range of 20–30 °C.3.2. Synthesis of PPESK-g-PEG copolymers After PPESK was chloromethylated, the resultant CMPPESK reacted with a sodium alkoxide to pre-pare PPESK-g -PEG graft copolymers. The sodium alkoxide was formed by reacting PEG –OH with NaH. Because the sodium alkoxide was highly sensi-tive to moisture and water, the reactants and the sol-vent needed to be su Y ciently dry. The reaction was performed under a protective inert gas atmosphere.The purpose of selecting mono hydroxyl PEG as the graft agent is to avoid the cross-linking of reactants and the gelation of reaction system. The molecular weights of PEG–OH used in this work are 350 and 750 g/mol.

The incorporation of PEG chains to the PPESK backbone was qualitatively con W rmed by FT-IR spectroscopy, as shown in Fig.3. In comparison to CMPPESK, the intensity of C–Cl absorbance peak (754cm ?1) for PPESK-g -PEG decreases remarkably,whereas that of ether bond peak (1098cm ?1) exhibits a great increase. Moreover, a new peak occurs at 2863cm ?1, which is attributed to the stretching vibration of C–H in the PEG chains. FT-IR analysis cannot provide su Y cient evidence for the covalent linking of PEG with the PPESK backbone, thus fur-ther NMR measurements were performed.

Fig.4 shows a typical 1H NMR spectrum of PPESK-g -PEG. The signals of methylene protons (denoted in Fig.4 as f ) and terminal methyl protons (denoted in Fig.4 as g ) were observed at about 3.4and 3.7ppm, respectively. They are the characteris-tics of the PEG chains. The peaks at about 4.35ppm are attributed to the protons between phenyl and PEG chains, as denoted in Fig.4 as d and e . Com-parison of Fig.4 with Fig.3, the peak at about 4.8ppm corresponding to the protons of –CH 2Cl (a and b ) was greatly reduced after the grafting of

Table 1

The DS of CMPPESK in various reaction conditions a a

In all chloromethylation experiments, the amounts of concen-trated H 2SO 4 were 15mL/g PPESK.

Sample label

Reaction conditions

DS

Mole ratio of CME/PPESK repeat unit Temperature (°C)

Time (h)CMPPESK-111.52020.81CMPPESK-234.5202 1.65CMPPESK-357.5202 3.03CMPPESK-457.5205 3.22CMPPESK-557.52010 2.96CMPPESK-6

57.530

10

2.47

1388L.-P. Zhu et al. / European Polymer Journal 43 (2007) 1383–1393

PEG. Solid-state 13C CP-MAS NMR spectra of the virgin PPESK and PPESK-g -PEG are shown in Fig.5. In the spectrum of PPESK-g -PEG 350, a series of new peaks appeared at 38–70 ppm, which are attributed to the incorporation of PEG chains. The

peaks for the carbon of methyl and methylene of PEG occurred at about 58 and 69ppm, respectively.The peak at 64.6 ppm is corresponding to the carbon linking PPESK with PEG. Based on these data, it was concluded that PEG chain is covalently linked to PPESK backbones.

The grafting degree (GD) of PEG on PPESK, de W-ned as the number of grafted PEG chains per PPESK repeat unit, was determined according to the ratio of peak intensity at 4.35ppm (d and e ) to that at 8.6ppm (c ) in Fig.4. The grafting e Y ciency (namely the per-centage of –CH 2Cl conversion into –CH 2O–PEG)was calculated according to the ratio of PEG grafting degree to the corresponding DS of CMPPESK. The results are listed in Table 2. The PEG content of the graft copolymers for PEG –OH 350 and PEG –OH 750are 29.2 and 37.2wt%, respectively. With same synthe-sis conditions, the PEG grafted degree of PPESK-g -PEG 350 (1.18) is higher than that of PPESK-g -PEG 750(0.79), and so is the grafting e Y ciency. This phenome-non might be attributed to the higher chemical activ-ity of hydroxyl group of PEG –OH 350 than that of PEG–OH 750. In addition, the shorter molecular chain has less steric hindrance for the reaction. It is worth

Fig.

4. The typical 1H NMR spectrum of PPESK-g -PEG.

L.-P. Zhu et al. / European Polymer Journal 43 (2007) 1383–1393

1389

noting that the grafting e Y ciencies of both PPESK-g -PEG 350 and PPESK-g -PEG 750 were lower than 36.6%.This suggests that only a small portion of –CH 2Cl were converted into –CH 2O–PEG. This phenomenon is mainly owed to the steric hindrance of molecular chains to PEG grafting.

The TG A curves for PPESK, CMPPESK and PPESK-g -PEGs are shown in Fig.6. The weight loss temperature (5% loss) of virgin PPESK was 476°C,exhibiting an outstanding thermal stability. Two decomposition events were observed for CMPPESK and PPESK-g -PEG s. The W rst decomposition occurs, from 250 to 320°C. This is attributed to the decomposition of chloromethyl groups and PEG chains. The second decomposition step, beginning at about 480°C, corresponds to the degradation of PPESK backbones. From the weight loss ratios at the W rst decomposition step, the grafting degree of PEG on the PPESK was also calculated. Due to the graft products containing unreacted CMPPESK, the weight loss ratio of CMPPESK must be considered when calculating the G D of PPESK-g -PEG . As shown in Table 2, the results from TG A were approximately in agreement with the data calculated

1acted PEG–OH was completely been removed from the products by precipitation and extraction.3.3. Surface characterizations of the blend membranes

The Neat and blend PPESK asymmetrical membranes were prepared by the immersion precipi-tation method. The elevated coagulant bath temperature would lead to the exchange of solvent and non-solvent more rapidly, thus hydrophilic components are excluded to the top surface of the membrane [21,22]. The chemical compositions for both surfaces of the neat and blend PPESK mem-branes were evaluated by XPS analysis. The X-ray source was run at a power of 250W (14.0kV,93.9eV). Binding energies were calibrated by using the containment carbon (C 1s D 284.6eV). Surface elemental stoichiometry, with the accuracy of was §5%, was determined from peak area ratios after correcting with experimentally determined instru-mental sensitivity factors. The XPS atomic percents for both surfaces of PPESK control membrane (M-1) and PPESK/PPESK-g -PEG blend membranes (M-2, M-3) are shown in Table 3. It can be seen clearly that the O contents for the blend membranes are higher than those of the neat PPESK membrane.On the contrary, the contents of S and N decreased after blending with the graft copolymers. Moreover,the element content variations on the top surface are bigger than those on the bottom surface. For exam-ple, the O content on the top surface of M-3increases from 13.73% to 18.15%, while that on the bottom surface of M-3 increases from 13.78% to only 15.98%. The XPS O 1s core level spectra for both surfaces of these membranes are shown in Fig.7.The peak intensity of –C–O–C– bond for the top surfaces of M-2 and M-3 increases remarkably when compared to that of M-1, while that for the bottom surfaces increases only moderately. These results indicate that the hydrophilic PEG chains were pref-erentially excluded to the membrane-coagulant

Table 2

The grafting degree (GD) and the grafting e Y ciency of PPESK-g -PEG a PEG content calculated according to the grafting degree from 1H NMR analysis.

b

Grafting e Y ciency denotes the conversion percentage of –CH 2Cl and calculated from 1H NMR analysis.

Sample label DS of used CMPPESK GD of PEG

PEG content (%)a

Grafting e Y ciency (%)b 1

H NMR

TGA PPESK-g -PEG350 3.22 1.18 1.0929.236.6PPESK-g -PEG750

3.22

0.79

0.71

37.2

24.5

1390L.-P. Zhu et al. / European Polymer Journal 43 (2007) 1383–1393

a hydrophilic surface modi W cation was achieved for the PPESK membranes. This conclusion was further veri W ed by the following water contact angle mea-surements.neat PPESK membrane and the blend membranes was performed through a contact angle measure-ment. The curves of contact angle as a function of the drop age was plotted to evaluate the membranes’

Table 3

XPS atomic percentage for the neat and blend PPESK membranes

Membrane label Composition (wt%)XPS atomic percent

Carbon Oxygen Sulfur Nitrogen M-1 (bottom surface)PPESK 100%78.5113.78 1.69 6.02

M-1 (top surface)78.5813.73 1.64 6.05

M-2 (bottom surface)PPESK-g-PEG350 20%, PPESK 80%78.7314.15 1.40 5.72

M-2 (top surface)79.1415.980.91 3.97

M-3 (bottom surface)PPESK-g-PEG750 20%, PPESK 80%77.1415.89 1.36 5.61

M-2 (top surface)76.7418.15 1.08 4.03

L.-P. Zhu et al. / European Polymer Journal 43 (2007) 1383–13931391

hydrophilicity and surface wettability, as shown in Fig.8. Due to the hydrophobic nature of PPESK, the contact angle declined very slowly with drop age for the neat membrane (M-1). After adding the graft polymers, the initial value of contact angle for the blend membranes (M-2 and M-3) was smaller, and contact angle decreased with drop age more rapidly. This phenomenon indicates that both surface and bulk of the blend membranes held an enhanced hydrophilic compared with the membrane made from pure PPESK.

Fig.9 presents the SEM images of the bottom sur-face, the top surface and the cross-section for the membranes M1, M-2 and M-3, respectively. In com-parison with pure PPESK membrane (M-1), the blend membranes (M-2, M-3) have rougher surfaces with larger pore size. In addition, the cross-sections of M-2 and M-3 exhibit a porous sponge-like structure, while that of M-1 had a structure of a dense skin-layer linked by W nger-like pores. Meanwhile, some nodular aggregates were observed in the bottom sur-face of M-2 and M-3. This phenomenon is also observed in other membrane-forming systems where hydrophilic polymers are used as additive [23–25]. The changes of membrane morphologies are mainly owed to a change of precipitation kinetics during the membrane formation. On one hand, the addition of PPESK-g-PEG results in an increase of casting dope viscosity and slowed down the exchange of solvent–non-solvent [26]. On the other hand, inclusion of the more hydrophilic additive hastens the exchange of non-solvent with solvent, causing the membrane to precipitate more rapidly [17]. The ultimate membrane morphologies depend on the trade-o V between the two factors. As a result, the macrovoid formation is suppressed and the sponge-like pore is favoured.

Fig.9. SEM images of the surface and cross-section morphologies: (A) pure PPESK membrane (M-1); blend membranes: (B) M-2 and (C) M-3. 1, 2 and 3 represent the bottom surface, top surface and cross-section of the asymmetrical membranes, respectively.

1392L.-P. Zhu et al. / European Polymer Journal 43 (2007) 1383–1393

3.4. Evaluation of anti-fouling property

To compare the anti-fouling property of the neat and blend membranes, UF experiments for protein solution were performed using these membranes. After the membrane sample was pressurized with de-ioned water under 0.15MPa for 30min, a 1g/L of BSA aqueous solution was W ltrated under the trans-membrane pressure of 0.05MPa. The permeate

X uxes at di V erent W ltration time were measured to evaluate the anti-fouling properties of membrane. Fig.10 shows the permeate X ux declines with the

W ltration time. In the initial period, the X ux showed a drastic decline due to the protein adsorption and deposition on the membrane. After running for about 60 min, the X ux reached a stable value. From the initial X ux, J w0, and the steady protein solution X ux, J w1, the relative X ux decline, RFD, was de W ned and determined as: RFD D(1?J w1/J w0)￡100%. The RFD of the neat membrane (M-1) was 68.8%, while that of M-2 and M-3 was 60.5% and 54.2%, respectively. The RFD value re X ects the membrane fouling degree. The smaller RFD of M-2 and M-3 than M-1 indicates a lower fouling due protein adsorption and deposition on the membrane. From these data, it is concluded that the grafting with PEG allows PPESK membrane to be less susceptible to protein fouling. The enrichment and arrangement of PEG chains on the separation side (top surface) are responsible for the improvement of anti-fouling property. The repulsion force coming from the change of conformation and desolvation of PEG chains depresses the adsorption and deposition of protein molecules on membrane surface, and thus enhanced permeability and fouling resistance are achieved after PPESK-g-PEG addition [27].

4. Conclusion

Amphiphilic PPESK-g-PEG graft copolymer were prepared by a modi W ed Williamson ether reac-tion, in which sodium alkoxide of poly(ethylene glycol) methyl ether was covalently coupled with chloromethylated PPESK. The grafting degree of PEG calculated from 1H NMR was in approxi-mately agreement with that from TG A tests. The PEG content of the graft copolymers was in the range of 29.2–37.2wt%. The graft products were blended with PPESK to prepare porous membranes by immersing precipitation method. The graft copolymers were excluded to membrane surface dur-ing membrane formation and a hydrophilic surface modi W cation was achieved for the PPESK mem-branes. The modi W ed membrane exhibited enhanced hydrophilicity, surface wettability and ant-fouling properties than the neat one. Acknowledgements

The authors are grateful to the W nancial support of the National Basic Research Program of China (Grant No. 2003CB615705).

References

[1]Meng YZ, Hay AS, Jian XG, Tjong SC. J Appl Polym Sci

1997;66:1425–32.

[2]Jian XG, Chen P, Liao GX, Zhu XL, Zhang SH, Wang JY.

Acta Polym Sin(4) 2003:469–75.

[3]Jian XG, Dai Y, Zeng L, Xu RX. J Appl Polym Sci 1999;

71:2385–90.

[4]Jian XG, Dai Y, He G H, Chen G H. J Membr Sci 1999;

161:185–91.

[5]Dai Y, Jian XG, Liu XM, G uiver MD. J Appl Polym Sci

2001;79:1685–92.

[6]Dai Y, Jian XG, Zhang SH, Guiver MD. J Membr Sci 2001;

188:195–203.

[7]Zhang SH, Jian XG, Dai Y. J Membr Sci 2005;246:121–6.

[8]Su Y, Jian XG, Zhang SH, Wang G Q. J Membr Sci 2004;

241:225–33.

[9]Xuan YN, G ao Y, Huang Y, Jian XG. J Appl Polym Sci

2003;88:1111–4.

[10]Wei J, Jian XG, Wu CR, Zhang SH, Yan Chun. J Membr Sci

2005;256:116–21.

[11]Wu CR, Zhang SH, Yang DL, Wei J, Yan C, Jian XG. J

Membr Sci 2006;279:238–45.

[12]Horgan A, Saunders B, Vincent B, Heenan RK. J Colloid

Interf Sci 2003;262:548–59.

L.-P. Zhu et al. / European Polymer Journal 43 (2007) 1383–13931393

[13]Chiu HC, Chern CS, Lee CK, Chang HF. Polymer 1998;39:

1609–16.

[14]Hou SS, Kuo PL. Polymer 2001;42:2387–94.

[15]Guo SR, Shen LJ, Feng LX. Polymer 2001;42:1017–22.

[16]Mihaylova MD, Krestev VP, Kresteva MN, Amzil A,

Berlinova IV. Eurp Polym J 2001;37:233–9.

[17]Park JY, Acar MH, Akthakul A, Kuhlman W, Mayes AM.

Biomaterials 2006;27:856–65.

[18]Li L, Wang YX. J Membr Sci 2005;262:1–4.

[19]Su Y, Jian XG, Zhang SH. J Funct Mater Contents, China

2004;35:385–8.

[20]Zhang AQ, Li XD, Nah CW, Hwang KJ, Lee MH. J Polym

Sci: Part A: Polym Chem 2003;41:22–9.[21]Hester JF, Banerjee P, Mayes AM. Macromolecules 1999;

32:1643–50.

[22]Hester JF, Mayes AM. J Membr Sci 2002;202:119–35.

[23]Jung B, Yoon JK, Kima B, Rhee H-W. J Membr Sci 2004;

243:45–57.

[24]Yoo JE, Kim JH, Kim Y, Kim CK. J Membr Sci 2003; 216:

95–106.

[25]Jung B. J Membr Sci 2004;229:129–36.

[26]Wang M, Wu LG, Mo JX, G ao CJ. J Membr Sci 2006;

274:200–8.

[27]Wang YQ, Su YL, Ma XL, Sun Q, Jiang ZY. J Membr Sci

2006;283:440–7.