Remarkable Reduction of Irreversible Fouling and Improvement of the

11856

Langmuir 2005, 21, 11856-11862

Remarkable Reduction of Irreversible Fouling and Improvement of the Permeation Properties of Poly(ether sulfone) Ultrafiltration Membranes by Blending with Pluronic F127 Yan-qiang Wang, Ting Wang, Yan-lei Su, Fu-bing Peng, Hong Wu, and Zhong-yi Jiang* Key Laboratory for Green Chemical Technology, School of Chemical Engineering and Technology, Tianjin University, Tianjin 300072, China Received July 28, 2005. In Final Form: September 19, 2005 Hydrophilic modification of ultrafiltration membranes was achieved through blending of Pluronic F127 with poly(ether sulfone) (PES). The chemical composition and morphology changes of the membrane surface were confirmed by water contact angle, X-ray photoelectron spectroscopy, scanning electron microscopy, and protein adsorption measurements. The decreased static water contact angle with an increase in the Pluronic F127 content indicated an increase of surface hydrophilicity. XPS analysis revealed enrichment of PEO segments of Pluronic F127 at the membrane surface. The apparent protein adsorption amount decreased significantly from 56.2 to 0 µg/cm2 when the Pluronic F127 content varied from 0% to 10.5%, which indicated that the blend membrane had an excellent ability to resist protein adsorption. The ultrafiltration experiments revealed that the Pluronic F127 content had little influence on the protein rejection ratio and pure water flux. Most importantly, at a high Pluronic F127 content membrane fouling, especially irreversible fouling, has been remarkably reduced. The flux recoveries of blend membranes reached as high as 90% after periodic cleaning in three cycles.

Introduction Ultrafiltration is a novel and powerful technique that can be used to concentrate or fractionate protein solutions. It is gentler toward the delicate proteins than separation processes based on phase changes such as evaporation and more economical than gel chromatography. The main problem restricting its practical application is membrane fouling and the time-dependent flux and rejection behavior. Membrane fouling during filtration of protein solutions is mainly determined by adhesion or deposition of protein molecules on the membrane surface and entrapment or aggregates of proteins in the pores.1 Membrane fouling includes reversible and irreversible fouling. Reversible protein adsorption causes reversible fouling that can be removed by hydraulic cleaning, e.g., backwashing and cross-flushing. When membrane fouling can only be overcome by the use of chemical reagents, it is here defined as irreversible fouling.2 Chemical cleaning of a membrane to reduce or eliminate irreversible fouling should be limited to a minimum frequency because repeated chemical cleaning may affect the membrane life, and disposal of spent chemical reagents poses another problem. Thus, the control of membrane fouling, especially irreversible fouling, is of importance for more efficient use of membranes. Much research has revealed that increasing the hydrophilicity of the membrane surface and pore surface could reduce membrane fouling, especially irreversible fouling.3-5 * To whom correspondence should be addressed. Phone and fax: 86-22-27892143. E-mail: [email protected]. (1) Masahide, T.; Belfort, G. J. Membr. Sci. 2004, 231, 147. (2) Pieracci, J.; Crivello, J. V.; Belfort, Georges. J. Membr. Sci. 2002, 202, 1. (3) Musale, D. A.; Kulkarni, S. S. J. Membr. Sci. 1996, 111, 49. (4) Hester, J. F.; Banerjee, P. A.; Mayes, M. Macromolecules 1999, 32, 1643.

Modifying surfaces with poly(ethylene oxide) (PEO) is an effective method to resist protein adsorption and platelet adhesion.6 The protein-resistant character of PEO is attributed to the lake of ionic charge, high hydrophilicity, flexibility, and mobility in the aqueous environment.7-10 The mechanisms concerning fouling resistance of PEO chains have been investigated extensively, and the corresponding models have been proposed.9,11-18 The PEOrich surfaces have been prepared by adsorption of high molecular weigh PEO, adsorption of PEO-containing amphiphilic copolymers, and covalent grafting of PEO onto the surface.19-22 Adsorption of PEO-containing copolymers (5) Taniguchi, M.; Kilduff; J. E.; Belfort, G. J. Membr. Sci. 2003, 222, 59. (6) Harris, J. M. Poly(Ethylene Glycol) Chemistry: Biotechnical and Biomedical Applications; Plenum Press: New York, 1992. (7) Morra, M. J. Biomater. Sci., Polym. Ed. 2000, 11, 547. (8) Malmsten, M.; Emoto, K.; Van Alstine, J. M. J. Colloid Interface Sci. 1998, 202, 507. (9) Szleifer, I. Biophys. J. 1997, 72, 595. (10) Tosatti, S.; De Paul, S. M.; Askendal, A. Biomaterials 2003, 24, 4949. (11) Jeon, S. J.; Lee, J. H.; Andrade, J. D. Gennes, P. G. J. Colloid Interface Sci. 1991, 142, 149. (12) Jeon, S. J.; Andrade, J. D. J. Colloid Interface Sci. 1991, 142, 159. (13) Halperin, A. Langmuir 1999, 15, 2525. (14) Leckband, D. E.; Sheth, S. R.; Halperin, A. J. Biomater. Sci., Polym. Ed. 1999, 10, 1125. (15) Prime, K. L.; Whitesides, G. M. J. Am. Chem. Soc. 1993, 23, 10714. (16) Besseling, N. A. M. Langmuir 1997, 13, 2109. (17) Van Oss, C. J.; Good, R. J. Chaudhury, M. K. Langmuir 1988, 4, 884. (18) Harder, P.; Grunze, M.; Dahint, R.; Whitesides, G. M.; Laibinis, P. E. J. Phys. Chem. B 1998, 102, 426. (19) Kishida, A.; Mishima, K.; Corretge, E.; Konishi, H. Biomaterials 1992, 13, 113. (20) Llanos, G. R.; Sefton, M. V. Macromolecules 1991, 24, 6065. (21) Jeong, B. J.; Lee, J. H.; Lee, H. B. J. Colloid Interface Sci. 1996, 178, 757. (22) Lee, J. H.; Kopeckova, P.; Kopecek, P.; Andrade, J. D. Biomaterials 1990, 11, 455.

10.1021/la052052d CCC: $30.25 © 2005 American Chemical Society Published on Web 11/03/2005

Modification of PES Ultrafiltration Membranes

is the most simple way, but the main disadvantage is that the adsorbed polymers may not remain on the surface permanently due to the weak hydrophobic adsorption forces between amphiphilic polymers and the substrate.23 Covalent coupling of PEO or PEO derivatives to substrates can effectively inhibit the nonspecific adsorption, but the coupling process is quite complicated and needs activated PEO and a functional group on the surface.24-26 Similar PEO-enriched surfaces with greater stability could be achieved by blending small amounts of PEO-containing block copolymers into the polymer matrix.27 The copolymers are anchored in the polymer matrix through the hydrophobic segments, while the hydrophilic segments have been shown to accumulate at the surfaces due to the incompatibility between the membrane matrix and PEO. Poly(ethylene oxide)-poly(propylene oxide)-poly(ethylene oxide) (PEO-PPO-PEO) triblock copolymer, commercially available as Poloxamer or Pluronic, is a frequently used amphiphilic copolymer whose properties have been studied extensively.28-33 Since the PEO segment of Pluronic is highly hydrophilic, Pluronic has been widely investigated in biomedical applications for reducing protein adsorption and cell adhesion.34-42 However, no report can be found in the ultrafiltration area which concerns both the effect of Pluronic on the permeation properties and the ability to resist membrane fouling. The aim of this work is to fabricate antifouling ultrafiltration membranes by blending amphiphilic Pluronic with PES and to investigate the properties of blend membranes. Pluronic F127 is selected for its better structure property: the long PEO segments give Pluronic F127 the excellent ability to reduce protein adsorption and the PPO segment with the appropriate length makes the Pluronic F127 located in the membrane matrix stable.43 The surface properties of the blending membranes were investigated by water contact angle measurements and X-ray photoelectron spectroscopy (XPS). The pore morphologies of the blend membranes were observed by scanning electron microscopy (SEM). Bovine serum albumin (BSA) was used as a model protein to investigate the protein adsorption resistivity and permeation proper(23) Lee, J. H.; Young, M. J.; Lee, W. K. J. Biomed. Mater. Res. 1998, 40, 314. (24) Qiang, Y.; Zeng, F.; Shiping, Z. Macromolecules 2001, 34, 1612. (25) Yu W. H.; Kang, E. T.; Neoh, K. G.; Shiping Z. J. Phys. Chem. B 2003, 107, 10198. (26) Zhu, X. Y.; Jun, Y.; Staarup, D. R.; Major, R. C. Langmuir 2001, 17, 7798. (27) Lee, J. H.; Young, M. J.; Dong, M. K. Biomaterials 2000, 21, 683. (28) Alexandridis, P.; Hatton, T. A. Colloid Surf., A 1995, 96, 1. (29) Alexandridis, P.; Holzwarth, J. F.; Hatton, T. A. Macromolecles 1994, 27, 2414. (30) Alexandridis, P.; Nivaggioli, T.; Hatton, T. A. Langmuir 1995, 11, 2442. (31) Nivaggioli, T.; Tsao, B.; Alexandridis, P.; Hatton, T. A. Langmuir 1994, 10, 2604. (32) Su, Y. L.; Wang, J.; Liu, H. Z. Macromolecules 2002, 35, 6426. (33) Su, Y. L.; Wang, J.; Liu, H. Z. J. Phys. Chem. B 2002, 106, 11823. (34) Lee, J.; Kopecek, J. H.; Andrade, J. D. J. Biomed. Mater. Res. 1989, 23, 351. (35) Freij, L. C.; Wessle´n, B. J. Appl. Polym. Sci. 1993, 50, 345. (36) Amiji, M.; Park, K. Biomaterials 1992, 13, 682. (37) Lee, J.; Kopeckova, P.; Kopecek, J.; Andrade, J. D. Biomaterials 1990, 11, 455. (38) Lee, J.; Martic, P. A.; Tan, J. S. J. Colloid Interface Sci. 1989, 131, 252. (39) Norman, M. E.; Williams, P.; Illum, L. Biomaterials 1992, 13, 841. (40) Akon, H.; Kaichiro, S.; Yoona, B. O.; Masaru, S.; Mariko, H.; Takashi, S. Biomaterials 2003, 24, 3235. (41) Christina, F. L.; Patric, J.; Bengt, W. Biomaterials 2000, 21, 307. (42) Li, Z. F.; Ruckenstein, E. J. Colloid Interface Sci. 2003, 264, 362. (43) Karl, D. P.; Cedric, J. O. Biomaterials 1999, 20, 885.

Langmuir, Vol. 21, No. 25, 2005 11857

ties of the blend membranes through static BSA adsorption experiments and ultrafiltration experiments of BSA solutions. Experimental Section Materials. PES 6020P was purchased from BASF Co. (Germany) and was dried at 110 °C for 12 h before use. N,NDimethylformamide (DMF) was purchased from Kewei Chemicals Co. (Tianjin, China). PEO-PPO-PEO triblock copolymer, Pluronic F127, was purchased from Sigma. Pluronic F127 has a molecular weight of 12600 and a PEO content of 70%. On the basis of molecular weight and chemical composition, Pluronic F127 can be represented by the formula EO100-PO65-EO100. Poly(ethylene glycol) 2000 (PEG 2000) was purchased from the Damao Chemical Reagent Co. (Tianjin, China). Bovine serum albumin (BSA) was purchased from the Institute of Hematology, Chinese Academic of Medical Sciences (Tianjin, China). All other chemicals were of commercial analytical grade. Preparation of Membranes. Casting solutions were prepared by dissolving PES, Pluronic F127, and PEG 2000 in DMF. PES (18 wt % in casting solutions) is a membrane matrix, PEG 2000 (15 wt % in casting solutions) is a pore-forming agent, and Pluronic F127 (Pluronic F127/PES weight ratios are 0, 1.5, 3.0, 4.5, 6.0, 7.5, 9.0, and 10.5 wt %) is a membrane-modified agent. After the homogeneous solution was obtained, the casting solutions were left for 4 h to allow complete release of bubbles. The solutions were cast on a glass plate with a steel knife at a wet thickness of 250 µm and then immersed in a coagulation bath of deionized water. The formed membranes were peeled off and subsequently washed thoroughly with deionized water to remove residual solvent and pore-forming agent and kept in deionized water before testing. Characterization of Blend Membranes. The static contact angle was measured at room temperature with a contact angle goniometer (Erma Contact Angle Meter, Japan). The chemical composition of the blend membrane surfaces was analyzed by X-ray photoelectron spectroscopy (PHI-1600) using Mg KR (1486.6 eV) as the radiation source. Survey spectra were collected over a range of 0-1100 eV, and high-resolution spectra of C1s were also collected. The takeoff angle of the photoelectron was set at 90°. The cross-section morphologies of control and blend membranes were observed by SEM using a Philips XL30E scanning microscope. The membranes frozen in liquid nitrogen were broken and sputtered with gold before SEM analysis. BSA Adsorption. The membranes were cut into a round shape with about 20 cm2 of external surface. The membranes were put into vials containing 5 mL of 1.0 mg/mL BSA solution whose pH was adjusted to 7.0 with 0.1 M phosphate buffer solution and incubated at 25 °C for 24 h to reach equilibrium. The concentrations of BSA in the solution before and after contact with the membranes were measured with a UV-vis spectrophotometer (Hitach UV-2800); then the apparent adsorbed BSA amount was calculated. The reported data were the mean value of triplicate samples for each polymer membrane. Ultrafiltration Experiments. A dead-end stirred cell filtration system connected with a N2 gas cylinder and solution reservoir was designed to characterize the filtration performance of the membranes. The system consists of a filtration cell (model 8200, Millipore Co.) with a volume capacity of 200 mL and an inner diameter of 62 mm. The effective area of the membrane was 28.7 cm2. The feed side of the system was pressed by extra nitrogen gas. All the ultrafiltration experiments were carried out at a stirring speed of 400 rpm and a temperature of 25 ( 1 °C. After the membrane was fixed in the cell, the stirred cell and solution reservoir were filled with deionized water. Each membrane was initially pressurized for 30 min at 150 kPa; then the pressure was reduced to the operating pressure of 100 kPa, and the water flux (Jw1) was measured by weighing the permeate solution. The cell and solution reservoir were emptied and refilled rapidly with 1.0 mg/mL BSA solution and 0.1 M buffer solution, respectively; the flux was recorded (Jp). The BSA rejection ratio was calculated by the following equation:

rejection ratio (%) ) (1 - Cp/Cf) × 100

(1)

11858

Langmuir, Vol. 21, No. 25, 2005

Wang et al.

where Cp and Cf (mg/mL) are the protein concentrations of the permeate and feed solutions, respectively. After 2 h of ultrafiltration, the membranes were cleaned with deionized water for 20 min, and then the water flux was measured (Jw2).

Results and Discussion Characterization of PES/Pluronic F127 Blend Membranes. PES membranes have excellent temperature and pH stabilities and good selectivity and mechanical strength and are capable of withstanding rigorous cleaning. However, their hydrophobic nature makes them susceptible to protein fouling. In the present work, amphiphilic polymer Pluronic F127 was blended with PES to endow fouling-resistant capacity. The hydrophilicity of the blend membranes was evaluated by the water contact angle. Figure 1 presents the water contact angles of the control PES membrane and PES/Pluronic F127 blend membranes. The control PES membrane has the highest contact angle, indicating the lowest hydrophilicity. The decreasing water contact angles of the blend membranes with an increase in the Pluronic F127 content indicates that the membrane surface is more hydrophilic at higher Pluronic F127 content. The surface compositions of PES/Pluronic F127 blend membranes were characterized by XPS analysis. Figure 2 presents the high-resolution XPS spectra of C1s. The peak of C1s is composed of an alkyl carbon (-C-C-, binding energy ∼285.0 eV) peak and an ether carbon (-C-O- and -C-S-, ∼286.6 eV) peak. The blend membranes show a significant increase of the ether carbon peak intensity with increasing Pluronic F127 content. The theoretical and experimental values of the ether carbon molar ratios for the control membrane are 33.3 and 34.8, respectively; the slightly higher experimental value may be due to the remaining pore-forming agent. The theoretical values of the ether carbon molar ratios are 41.2, 43.5, and 47.4 for 4.5%, 6.0%, and 9.0% Pluronic F127 content in the membranes; the experimental results of the ether carbon molar ratios for those membranes are 46.6, 52.3, and 55.9, respectively. The theoretical values of the ether carbon molar ratios are all lower than the experimental results of XPS analysis, which indicates that Pluronic F127 is enriched at the membrane surface. The enrichment of Pluronic F127 at the membrane surface may due to the thermodynamic incompatibility between PES and Pluronic F127. The PPO segment anchors Pluronic F127 at the membrane surface, and the PEO segment stretches out of the membrane surface and forms a hydrated polymer layer. The XPS results also indicate that the PEO segment density on the membrane surface increases with increasing F127 content. Asymmetric PES/Pluronic F127 blend membranes were prepared by a phase inversion method in a wet process. Figure 3 shows the cross-section morphologies of control and blend membranes. SEM micrographs of the membrane structures are very similar, consisting of a skin layer on top, an intermediate layer with a fingerlike structure, and a bottom layer of fully developed macropores. There are no appreciable morphological variations between the membranes without and with different contents of Pluronic F127 in the blend membranes. Reduction of Protein Adsorption on PES/Pluronic F127 Blend Membranes. The data in Figure 4 show that the apparent protein adsorption amount decreases with an increase in the Pluronic F127 content; when the Pluronic F127 content reaches 10.5%, the apparent protein adsorption amount decreases to nearly zero. The effective reduction in the protein amount is attributed to the PEO

Figure 1. Contact angles for PES/Pluronic F127 blend membranes as a function of the Pluronic F127 content.

segments that stretch out from the membrane surface into the surrounding aqueous environment, which can prevent protein adsorption from the solution onto the surfaces. It is seen in Figure 4 that the content of Pluronic F127 has a great effect on the protein-adsorption resistance of blend membranes. According to Halperin’s model,13,14 BSA molecules may penetrate the PEO brush in an end-on orientation and deposit on the surface at a low PEO chain density and thus cause significant adsorption. Thus, to obtain effective reduction of protein adsorption, the brush density must be high enough to make the distance between the neighboring PEO chains smaller than the dimensions of the indwelling BSA molecules. In our experiment, we imagine that, at a higher Pluronic F127 content, a dense PEO polymer layer can be formed at the surfaces of PES/ Pluronic F127 blend membranes. Permeation Properties of PES/Pluronic F127 Blend Membranes. Ultrafiltration experiments were carried out to study the permeability of the blend membranes. The results of the pure water flux and BSA rejection ratio are shown in Figure 5; the control membrane and the blend membranes have nearly similar pure water flux, which indicates that the blending of Pluronic F127 has little influence on the pure water flux of membranes. While the rejection ratio of the blend membrane decreases slightly with an increase in the Pluronic F127 content, all rejection ratios are higher than 98%, which means the membranes have better separation performance. To investigate membrane fouling, the membranes were cleaned after BSA solution ultrafiltration, and the pure water flux of the cleaned membranes was measured. Figure 6 presents the time-dependent flux of membranes in operation. It can be seen that the pure water fluxes (Jw1) before ultrafiltration of the BSA solution change little during ultrafiltration. The flux decreased dramatically at the initial operation of BSA solution ultrafiltration due to protein adsorption or convective deposition. It is proposed that some protein molecules in the feed will deposit or adsorb on the membrane surface (cake formation), causing a drop in flux in the first few minutes of operation. Under constant pressure, the effects of membrane fouling and concentration polarization are usually observed by a considerable decline in the permeate flux with time. In the present work, the concentration polarization was minimized because of the high molecular weight of BSA molecules and rigorous stirring near the membrane surface. Therefore, we would conjecture that membrane fouling mostly caused the flux decline of the membranes. The vigorous stirring in the ultrafiltration cell and the deposition and diffusion of the protein may

Modification of PES Ultrafiltration Membranes

Langmuir, Vol. 21, No. 25, 2005 11859

Figure 2. High-resolution XPS spectra of C1s for PES/Pluronic F127 blend membranes with Pluronic F127 contents of 0% (a), 4.5% (b), 6.0% (c), and 9.0% (d).

reach equilibrium within our experimental time, so that a relatively steady flux (Jp) is retained at the final operation of BSA solution ultrafiltration. After 2 h of ultrafiltration, the membranes are cleaned with fresh deionized water under vigorous stirring to recover the flux to some extent. Irreversible and Reversible Fouling. Membrane fouling causes flux loss (Jw1 - Jp). To study the antifouling property, we defined several ratios to describe the fouling process. The first ratio is Rt in eq 2. Here, Rt is the degree

Rt ) 1 - Jp/Jw1

(2)

of total flux loss caused by total fouling. A high value of Rt corresponds to a large reduction in flux. We also define two ratios, Rr and Rir, to distinguish reversible fouling and irreversible fouling. Rr is defined by eq 3, which is the

Rr ) (Jw2 - Jp)/Jw1

(3)

degree of reversible flux loss caused by reversible fouling. The reversible BSA adsorption on the membranes causes reversible fouling, which can be eliminated by hydraulic cleaning. Rir is defined by eq 4, which is the degree of

Rir ) (Jw1 - Jw2)/Jw1

(4)

irreversible flux loss caused by reversible fouling. Irreversible fouling is caused by irreversible BSA adsorption or deposition, which cannot be avoided by hydraulic washing. Thus, Rt is the sum of Rir and Rr:

Rt ) Rr + Rir

(5)

A summary of Rt, Rr, and Rir of PES/Pluronic F127 blend membranes as a function of the Pluronic F127 content is shown in Figure 7. It can be seen that Rt decreases from

0.74 to 0.42 with an increase of the Pluronic F127 content from 0% to 9.0%. The lower Rt indicates lower total flux loss, corresponding to less protein adsorption or deposition on the membrane surfaces. Figure 7 also shows that the blend membranes have not only lower Rt but also lower Rir than the control PES membrane. When the Pluronic F127 content increases from 0% to 9.0%, Rir decreases from 0.52 to 0.09. The introduction of Pluronic F127 reduces total membrane fouling, especially irreversible membrane fouling. In other words, at a high Pluronic F127 content, reversible fouling is the dominant factor which is responsible for the flux loss of the blend membranes. From the protein adsorption experimental results, it can be seen that when the Pluronic F127 content reaches 10.5%, the protein adsorption amount decreases to nearly zero. However, membrane fouling cannot be inhibited completely in ultrafiltration. This can be explained as follows: there is a great difference of protein adsorption between static conditions and convective conditions. In static protein adsorption, protein molecules diffuse from the solution bulk to the membrane surface; thus, the interaction, including an attraction force and a repulsion force, between protein molecules and membranes plays an important role, and the adsorbed protein molecules form a protein molecular layer on the membrane surface. In the dynamic operation of ultrafiltration, the protein molecules are driven to the membrane surface by fluid flow and may then deposit on the membrane surface and/ or become entrapped in the membrane pores. Thus, though the protein adsorption amount decreases to nearly zero in the static experiment, membrane fouling still exists during dynamic ultrafiltration. Introduction of Pluronic F127 cannot inhibit membrane fouling completely; however, the values of Rr and Rir change greatly with an increase of the Pluronic F127 content. Fouling may be largely attributed to irreversible

11860

Langmuir, Vol. 21, No. 25, 2005

Wang et al.

Figure 3. Cross-sectional SEM morphology of PES/Pluronic F127 blend membranes with Pluronic F127 contents of 0.0% (a), 3.0% (b), 4.5% (c), 6.0% (d), and 9.0% (e).

Figure 4. BSA adsorption amount on PES/Pluronic F127 blend membranes as a function of the Pluronic F127 content.

membrane fouling in the case of a lower Pluronic F127 content. At a low Pluronic F127 content, the PEO segments appearing at the membrane surface cannot form a compact polymer layer; thus, the protein molecule can easily penetrate the PEO layer and reach the membrane surface (hydrophobic PES), causing irreversible fouling.44,45 Increasing the Pluronic F127 content, that is, increasing (44) Szleifer, I. M.; Carignano, A. Macromol. Rapid Commun. 2000, 21, 423.

Figure 5. Pure water flux and BSA rejection ratio of PES/ Pluronic F127 blend membranes as a function of the Pluronic F127 content.

the PEO segment density on the membrane surface, makes the PEO layer more compact; therefore, a highly hydrated and dense PEO polymer layer on the membrane surface prevents the protein molecule from contacting the membrane surface directly. Thus, most of the protein molecules (45) Unsworth, L. D.; Sheardown, H.; Brash, J. L. Langmuir 2005, 21, 1036.

Modification of PES Ultrafiltration Membranes

Figure 6. Time-dependent flux of membranes during the ultrafiltration process for PES/Pluronic F127 blend membranes. The ultrafiltration process includes four steps: pure water flux measurement (1), BSA solution ultrafiltration (2), 20 min of water washing (3), and pure water flux measurement of the cleaned membranes (4). Ultrafiltration was carried out at a temperature of 25 °C and a pressure of 100 kPa. The BSA concentration is 1.0 mg/mL in PBS solution.

Langmuir, Vol. 21, No. 25, 2005 11861

Figure 8. Flux recovery ratio (Jw2/Jw1) for PES/Pluronic blend membranes as a function of the Pluronic F127 content.

Figure 9. The cleaned PES/Pluronic F127 membranes were reused in BSA solution ultrafiltration. The fluxes were plotted as a function of time.

Figure 7. A summary of Rt, Rr, and Rir of the PES/Pluronic F127 blend membranes as a function of the Pluronic F127 content.

deposit on the PEO layer and cause reversible fouling which can be removed easily by water washing. Therefore, fouling may largely attribute to reversible membrane fouling at higher Pluronic F127 contents. When a protein molecule is adsorbed on a polymer surface, water molecules between the protein and polymer need to be replaced in the aqueous environment.46 Protein adsorbed on the surface lost bound water at the surfacecontacting position. This phenomenon induces a conformational change in the protein, causing protein adsorption, especially irreversible protein adsorption. Much research has revealed that if the water state at the surface is similar to an aqueous solution or the free water fraction is high, the protein adsorption, especially irreversible protein adsorption, could be reduced effectively.47,48 In this study, the highly hydrated PEO polymer layer contains much free water; thus, protein molecules have little or no conformational change when they approach the PEO layer. Thus, irreversible protein adsorption or deposition is decreased greatly, and the flux can be sufficiently recovered after water washing. Recycling of the PES/Pluronic F127 Membranes. During ultrafiltration, protein molecules deposited or adsorbed on the surface and inside the membrane pores, (46) Lu, R. D.; Lee, S. J.; Park, K. J. Biomater. Sci., Polym. Ed. 1991, 3, 127. (47) Erwin, A. V. Adv. Colloid Interface Sci. 1998, 74, 69. (48) Ishihara, K.; Nomura, H.; Mihara, T.; Kurita, K. J. Biomed. Mater. Res. 1998, 39, 323.

causing the permeate flux to decrease. The values of the flux recovery ratio (Jw2/Jw1) were calculated for the blend membranes and are presented in Figure 8. The value of Jw2/Jw1 is 0.47 for the control membrane, while all the blend membranes have higher Jw2/Jw1 values which increase with an increase of the Pluronic F127 content. The blend membrane with a Pluronic F127 content of 9.0% has a Jw2/Jw1 value of 0.90. PES/Pluronic F127 blend membranes give higher flux recoveries after cleaning, suggesting that protein fouling is reversible because of the introduction of a dense hydrophilic PEO layer. The excellent flux recovery property of blend membranes indicates that the blend membranes can be reused for several runs. It can be seen in Figure 9 that the pure water flux of the PES/Pluronic F127 blend membrane containing 9.0% Pluronic F127 can be retained at 110 L/(m2 h) after three BSA solution ultrafiltrations, while the pure water flux of the control PES membrane decreases to nearly zero after only two BSA solution ultrafiltrations. Comparison of all these data allows us to conclude that blending of Pluronic F127 with PES is an appropriate method to prepare antifouling ultrafiltration membranes. Conclusion Blending Pluronic F127 with PES can create highly hydrophilic PES blend membranes with superior fouling resistance. Though the apparent protein adsorption amount decreases to nearly zero for blend membranes at higher Pluronic F127 contents, membrane fouling cannot be inhibited completely during ultrafiltration. The total membrane fouling and irreversible membrane fouling decrease with an increase of the Pluronic F127 content, and excellent flux recovery renders the blend membranes

11862

Langmuir, Vol. 21, No. 25, 2005

with good recycling utilization. It is proposed that a highly hydrated and dense PEO polymer layer formed on the membrane surface prevents protein molecules from contacting the membrane surface directly, and the protein molecules deposited on the PEO layer can be removed easily by water washing due to the reversible fouling characteristics.

Wang et al.

Acknowledgment. This research was financially supported by the Tianjin Natural Science Foundation (Grant No. 05YFJZJC 00100) and Ministry of Education of China. We also thank Prof. Fei. He for her help in the XPS measurements. LA052052D