Irreversible fouling control of PVDF ultrafiltration membrane with “fouled surface” for mimetic sewage treatment

Abstract

Polyvinylidene fluoride (PVDF) ultrafiltration membranes with sodium alginate (SA) as the “fouled surface” were prepared, in an effort to control irreversible membrane fouling in mimetic sewage treatment processes. The anchored SA layer improved the wetting ability of the PVDF membrane surface and used as a functional antifouling coating to improve the antifouling ability towards organic pollutants. The modified membrane (S-T-0.3) exhibited excellent fouling resistance, with the irreversible fouling ratio (IFR) values of 9%, 1%, 8% and 6% for the pollutant solutions of bovine serum albumin (BSA), SA, humic acid (HA) and mimetic sewage, respectively; while the IFR values of the neat PVDF membrane for all the pollutant solutions were higher than 30%. The fouled surface of the PVDF membrane suppressed the adsorption of hydrophobic pollutants due to the improved hydrophilicity, and prevented hydrophilic pollutants from entering the membrane pores due to hydrogen bonding, electrostatic and counterion effects. The higher water flux recovery ratios were observed because the pollutants were withheld on the modified membrane surface due to the cake layer feature of SA, and could be eliminated by simple flushing, leading to less irreversible fouling. The purpose of this article is to provide a novel and effective antifouling mechanism for the solution of the membrane fouling problem in sewage treatment.

---

1. Introduction

Membrane technology has been widely used in the fields of water treatment, industry, food and energy due to its high separation efficiency and energy-saving features. However, membrane fouling problems exist in almost all the membrane processes, which decreases the separation efficiency and increases the cost of membrane application.^1–4 Especially in the ultrafiltration (UF) process of sewage treatment, all the hydrophilic and hydrophobic pollutants existing in sewage water can induce serious irreversible fouling problems and shorten the service life of membrane modules during the membrane separation process.

To our knowledge, the dimensions of UF membrane pores vary from several to tens of nanometers, and nano-sized pollutants can accumulate on the membrane surface and clog the pores. The most of the pollutants accumulating on UF membrane surfaces can be cleaned by backwashing or flushing as reversible fouling. However, the non-removable pollutants may adsorb on the membrane surface and clog the membrane pores, resulting in the irreversible fouling problem.^5–7 Therefore, preventing pollutants from initially adsorbing on the membrane surface and entering the membrane pores to clog the separation channel are the key factors for the control of irreversible fouling during the UF membrane separation process.

Numerous studies have reported that the hydrophilicity, surface roughness and charge of membrane surfaces affect the antifouling property of the UF separation membranes.^8–10 Generally, the increased hydrophilicity of membrane materials prepared via blending and surface grafting of hydrophilic or amphiphilic polymers is conducive to improve the antifouling performance.^11–13 Almost all the membranes with the improved wetting ability exhibit excellent antifouling property towards the hydrophobic protein pollutants. However, few reports were published on the improvement of the antifouling ability towards the hydrophilic pollutants, which are one of the most important components causing membrane fouling problems. Professor Elimelech¹³ has demonstrated that the hydrophilic pollutants in feed were able to exhibit stronger interaction forces than hydrophobic pollutants when contacting the hydrophilic surfaces by using atomic force microscope (AFM). Hydrophilic membrane surfaces show weak resistance to the hydrophilic pollutants due to the interactions (e.g., hydrogen bonding) between pollutants and membrane surfaces during water-based separation process, leading to the persistent adsorption of pollutants on the membrane surface and severe irreversible fouling in membrane pores. To prepare low-fouling or non-fouling membranes, attentions should be paid to both the inhibition of adsorption of all kinds of organic pollutants on the membrane surface and the prevention of pollutants from entering membrane pores as irreversible fouling to block the separation channels during the separation process.

Sodium alginate (SA), a natural polysaccharide extracted from seaweed alginate with excellent hydrophilicity, colloidality and membrane forming properties,^14,15 is able to form a cake layer on the membrane surface, and as a reversible fouling pollutant, can be easily washed away during filtration process.^16–20 Therefore, due to the strong hydrophilicity, cake layer forming ability and various interactions with multiple pollutants, SA may serve as a potential antifouling material, inhibiting the adsorption of hydrophobic pollutants and preventing hydrophilic pollutants from entering membrane pores. The modified PVDF membranes with SA as fouled surface were prepared via the plasma treatment of the membrane surface and the subsequent surface crosslinking of SA. Pollutant solutions with single or mixed species of BSA, SA and HA were selected as the mimetic sewage to investigate the antifouling properties of the modified membranes. This article aims to provide a novel and efficient method to prepare antifouling PVDF membranes for sewage treatment application.

2. Materials and methods

2.1 Reagents

Polyvinylidene fluoride (PVDF, MG15) was purchased from Arkema. Bovine serum albumin (BSA, 67 000 Da) and PEG (M_w ∼ 20 000) were purchased from Sinopharm Chemical Reagent Co. (China). Trimesoyl chloride (TMC, 98%), humic acid (HA, fulvic acid ≥ 90 wt%, M_w ∼ 2000) and sodium alginate (SA, M_w ∼ 230 000) were purchased from Aladdin Co. (China). Dimethylacetamide (DMAC) and calcium chloride were of analytical grade. All reagents were used without any purification. The selected BSA was chosen as hydrophobic pollutants due to the strong hydrophobic effect of its hydrophobic part. SA and HA were selected as hydrophilic pollutant and natural organic matter, respectively.

2.2 Membrane preparation and characterization

Neat PVDF UF membranes were prepared though non-solvent induced phase separation (NIPS) method. Briefly, the mass concentrations (w/w%) of PVDF and PEG in the casting solution were fixed at 16% and 12%, respectively, with DMAC as solvent and water as coagulation bath. The prepared PVDF membranes were freeze-dried, subsequently treated on one side using the dielectric barrier discharge plasma instrument (200 W, 60 s), oxidized in air atmosphere for 30 min, and then immersed in SA solution for 20 min to adsorb SA adequately. The excess SA solution on the membrane surface was scraped and the membrane was immersed in TMC (0.1 wt%) solution for 5 min to allow the crosslinking of SA on the membrane surface. Finally, the prepared PVDF membrane was washed with deionized water to remove the unreacted SA. The prepared PVDF membranes with SA layers were noted as S-T-0.1 and S-T-0.3, prepared with 0.1 wt% and 0.3 wt% SA solutions, respectively. Due to the limit of the solution viscosity, the SA solutions with higher concentration were not selected.

Attenuated total reflectance Fourier transform infrared spectroscopy (ATR-FTIR, Nicolet 8700, USA) was used to examined the surface composition of PVDF membranes. The samples were placed on a ZnSe crystal as the internal refection element with an aperture angle of 45°, and the spectra were recorded by the accumulated average of 32 scans at 4 cm⁻¹ resolution. The surface chemical composition of the membranes was examined using X-ray photoelectron spectroscopy (XPS, Kratos, AXIS UltraDLD). The Al K-α X-ray was used as radioactive source with the take-off angle at 90°. The wide scan survey and the high-resolution spectra were recorded with the resolution of 0.68 eV/(C 1s). The hydrophilicity of the membrane surface was evaluated by water contact angle (CA, OCA40Micro, Germany) test on the membrane surfaces at room temperature. A total of about 3 μl of deionized water was dropped onto the membrane surfaces using a micro syringe and the results were obtained using the drop shape image analysis system to evaluate the surface wetting ability. The surface zeta potential of neat and modified PVDF membranes were tested using the nanoparticle size analyzer (DelsaNano, Beckman Coulter, USA) by switching to the test mode of zeta potential. The surface morphology of membranes were viewed with the field emitting scanning electron microscope (FESEM, Hitachi SU8010, Japan). All the samples were coated with gold before observation. The average pore size distribution of the membranes was evaluated by membrane pore size analyzer (GQ/PSMA-10, China) by displacing the wetting liquid method.

2.3 Separation performance

The separation performances of the neat and modified PVDF membranes were investigated using a dead-end filtration system with the effective membrane area of 12.5 cm². The filtration experiments were carried out with the compacted pressure at 0.15 MPa for 30 min using pure water as the feed solution. Then the pressure was lowered to 0.10 MPa to obtain a stable flux by comparing the flux values every 5 minutes. The pure water flux was noted as J (L m⁻² h⁻¹) and calculated by the following eqn (1). The rejection ratio (R) of model pollutant (BSA) was calculated from the feed and permeate concentrations determined by UV spectrophotometer (UV-1800, Shimadzu) according to the following eqn (2), with BSA solution (1 g L⁻¹, pH 7.4) selected as the testing feed. Each test was repeated at least three timeswhere V (L) is the volume of the permeated water, t (h) is the permeation time and A (m²) is the effective area for filtration. C and C_p are the concentrations of BSA in the feed and permeate, respectively.

2.4 Fouling test

The adsorption tests were carried out by immersing the neat and modified PVDF membranes in BSA solution (1 g L⁻¹, pH 7.4) for 12 h at room temperature to reach the adsorption equilibrium. The concentrations of BSA solutions before and after the BSA adsorption with the modified PVDF membranes were measured with UV-1800 and the mass (μg cm⁻²) of the adsorbed BSA on the membrane surface was calculated. It should be noted that only the active layer side of the modified PVDF membranes were allowed to contact with the BSA solutions.

BSA, SA and HA representing the proteins, polysaccharides and natural organic matter in sewage, respectively, were selected as model pollutants, which were commonly considered as the dominant pollutants in sewage causing membrane fouling. As shown in Table 1, the single pollutant feed and mimetic sewage were prepared according to the reported literatures,^21,22 which were the large-scale MBR supernatants from 10 selected wastewater treatment plants in China.

Table 1 The composition of prepared pollutants feed

	BSA	SA	HA	Mixture	Mimetic sewage
BSA (mg L^−1)	4	0	0	4	4
SA (mg L^−1)	0	15	0	15	15
HA (mg L^−1)	0	0	8	8	8
NaCl (mM)	9	9	9	0	9
NaHCO3 (mM)	2	2	2	0	2
MgCl2 (mM)	0.5	0.5	0.5	0	0.5
CaCl2 (mM)	1	1	1	0	1

---

The fouling experiments were carried out with circular mode using alternate feed solutions of pure water and pollutant solutions. Firstly, stable water flux was recorded as J₁. The feed was then replaced with pollutant solutions and the flux was recorded as J_p. The secondary water flux of the fouled membrane without cleaning is recorded as J_F. The water flux ratio (FR) of the fouled membrane was calculated by the followed eqn (3). Then the tested membranes were washed by pure water for 20 minutes (flushing process) and the secondary pure water flux was obtained as J₂ with the pure water feed. The water flux recover ratio (FRR) and the irreversible fouling resistance ratios (IFR) were calculated according to the followed eqn (4) and (5). Each fouling test was repeated at least three times.

IFR = 1 − FRR (5)

3. Results and discussion

3.1 Modification mechanism

Since PVDF membranes always exhibit a low surface energy due to the presence of stable C–F bonds, plasma technology was selected to activate the PVDF membrane to obtain a hydrophilic surface with a large number of polar groups without destruction to the matrix. The SA can be adsorbed onto the hydrophilic surface of the activated PVDF membrane for subsequent interface crosslinking reaction. As shown in Fig. 1, based on the convenient reaction of acyl chloride and hydroxyl group, the crosslinked SA by TMC as a functional coating covered the PVDF membrane surface. In addition, the generated polar groups including hydroxyl and amino groups by plasma treatment also acted as crosslinking sites for anchoring SA, as already proved by our previous study.²³

Fig. 1 Schematic diagram for the preparation process of the modified PVDF membrane.

3.2 Interface characteristics

To gain specific information about interfacial crosslinking of SA on PVDF membrane, the surface composition of the neat PVDF, S-T-0.1 and S-T-0.3 membranes were examined with ATR-FTIR, shown in Fig. 2. Two new adsorption peaks at 1725 cm⁻¹ and 1550 cm⁻¹ attributed to the C=O stretching vibration of the ester bond and the characteristic peak of carboxylate, respectively, appeared in the spectra of S-T-0.3 and S-T-0.1 membranes, and their intensity increased with the increase of SA concentration, indicating the occurrence of esterification reaction between the acid chloride groups of TMC and hydroxyl groups of SA, and the fixing of SA with carboxylate groups onto the PVDF membrane surface. Fig. 2B showed the surface element compositions of the neat and modified PVDF membranes from the results of XPS characterization. The atomic concentrations of O and F elements on the modified membrane surfaces exhibited the increase and decrease tendency, respectively, as compared to the neat PVDF membrane. O element belonged to the crosslinked SA layer and reduced F element indicated the change in membrane surface composition of the modified PVDF membrane. It could be known that the hydrophilic polar groups such as hydroxyl and carboxylate groups of SA were introduced onto the membrane interface by the surface crosslinking modification of TMC and SA, which was conducive to improve the wetting properties of the PVDF membranes surface.

Fig. 2 Surface property of PVDF, S-T-0.1 and S-T-0.3 membranes (A: ATR-FTIR spectra, B: surface element composition, C: water contact angles, D: surface zeta potential).

As shown in Fig. 2C, S-T-0.3 and S-T-0.1 membranes exhibited the sharp decrease of the water contact angle (CA) with the values decreasing to 51° and 36°, respectively, while the CA of the neat PVDF membrane was 92°. Fig. 2D showed the surface zeta potential of modified PVDF membranes, which was an important parameter affecting membrane fouling property. Obviously, the modified membranes always showed a negative charge (pH < 10) due to the presence of the carboxylate groups of SA, which enhanced the hydrophilicity and electronegativity of the modified PVDF membrane surface.

3.3 Surface topography

In addition, the surface morphologies of PVDF membranes before and after interface crosslinking were shown in Fig. 3A. The modified membrane surface was observed to be covered with a film-like media. The average pore size of the modified membranes (as shown in Fig. 3B) decreased as compared with the neat membrane, and showed a decrease with the increase of the SA concentration, indicating the coverage of the SA layer. These results were consistent with the SEM observation.

Fig. 3 Surface topography (A) and average pores size (B) of the neat and modified PVDF membranes.

3.4 Separation performance

The separation performances of the neat and modified PVDF membranes were shown in Fig. 4. Although the wetting ability of the modified membrane surface was improved by the anchored SA layer, the modified membranes of S-T-0.1 and S-T-0.3 did not show the expected water flux with an increasing trend, which should be attributed to the reduced pore sizes of the membrane surface. As a contrast, the rejection of the modified membranes exhibited an obvious increase, with the value of S-T-0.3 membrane increasing to 88%. In general, the separation performance of the modified PVDF membrane was improved. For ultrafiltration membranes, the membrane surface hydrophilicity and average membrane pores size show positive and negative effect to the separation flux, respectively. But both contribute to the enhancement of rejection, which will help to reduce the probability of pollutants into the membrane pores and decrease membrane fouling.

Fig. 4 Separation performance of the neat and modified PVDF membranes.

3.5 Antifouling performance

In order to investigate the antifouling performance of modified PVDF membranes, single and mimetic sewage feed were employed to quantitatively study the antifouling ability of the neat and modified PVDF membranes by comparing the FR, FRR and IFR values. The high FR value indicated the better anti-adsorption ability, the lower IFR value indicated the less irreversible fouling performance, and the difference of FR and FRR values represented the reversible fouling ratio.

For the single protein pollutant of BSA based on its hydrophobic part, the fouling test results were illustrated in Fig. 5. Fig. 5A showed the time dependent normalized flux variation. The absolute flux values of the membranes in Fig. 5 can be calculated according to Fig. 4. The J_p values of PVDF, S-T-0.1 and S-T-0.3 membranes were 25, 26 and 25 L m⁻² h⁻¹ respectively. The J_F values of PVDF, S-T-0.1 and S-T-0.3 membranes were 35, 65 and 64 L m⁻² h⁻¹ respectively. The J₂ values of PVDF, S-T-0.1 and S-T-0.3 membranes were 46, 66 and 67 L m⁻² h⁻¹, respectively. As shown in Fig. 5B, the FR values of S-T-0.1 and S-T-0.3 membranes without flushing were 80% and 88%, respectively, which were significantly higher than 45% of the neat PVDF membrane. After flushing, The FRR values of S-T-0.1 and S-T-0.3 membranes increased to 82% and 91%, respectively, but the FRR value of the neat PVDF membrane was only 55%. Meanwhile, IFR values of PVDF, S-T-0.1 and S-T-0.3 membranes were 45% and 18% and 9%, respectively. The changed flux suggested that the adsorbed and accumulated pollutants in the PVDF membranes during the separation process were responsible for the decrease of water flux, and the SA layer was conducive to reduce the adsorption of pollutants. On the one hand, the improved FR values of modified PVDF membranes indicated the enhanced anti-adsorption ability of the surface SA layer with less BSA adsorbed on membranes. Adsorption results (Fig. 5C) also demonstrated that the BSA adsorption capacity was less on the hydrophilic S-T-0.1 and S-T-0.3 membrane surfaces. On the other hand, the enhanced FRR values indicated that the amount of irremovable pollutants in membrane pores were reduced, leading to the reduction of irreversible fouling due to the increased hydrophilicity of the membrane surface, which provided hydration layer as a protective layer to reduce the attachment of the hydrophobic pollutants based on hydrophobic effect, and prevent BSA pollutants from entering the membrane pores as irremovable pollutants. At the same time, the membrane surface with higher electronegativity would show strong repulsive forces to negatively charged BSA molecules (pH 7.4). It demonstrated that enhanced wetting ability and charge of hydrophobic membranes provided by SA layer improved the fouling resistance for protein pollutants mainly due to the hydrophobic effect.

Fig. 5 Fouling test results of PVDF, S-T-0.1 and S-T-0.3 membranes using BSA solution as pollutant feed (A: time dependent normalized flux variation. B: FR, FRR and IFR values. C: Adsorption mass of BSA).

For the single hydrophilic pollutant solution of SA, the fouling test results were shown in Fig. 6A and a similar trend with that of BSA was observed. SA was more likely to be trapped on the membrane surface due to the larger molecular weight. It was worthy noted that the water flux (FR value) of S-T-0.1 and S-T-0.3 membranes decreased to 53% and 61% after fouling, respectively, which were significantly lower than the FR values of BSA. These results were attributed to the feature of SA which was more easily aggregated on membranes surface to form a cake layer and declined the separation efficiency due to its gel feature. However, the cake layer type SA as the removable pollutants could be eliminated though simple water flushing,¹³ leading to the higher FRR values. Humus is the major component of the natural sewage including humic acid (HA), fulvic acid (FA) and humin (HM). HA is more likely to enter the membrane pores and cause membrane fouling due to the smaller molecular weight. The fouling test results of HA (FA > 90%) in Fig. 6B showed that the FRR values and IFR values of the modified membrane were significantly improved compared to those of the neat PVDF membrane, demonstrating that the modified PVDF membranes with the SA layer exhibited excellent antifouling ability for the natural sewage humus.

Fig. 6 Fouling test results of PVDF, S-T-0.1 and S-T-0.3 membranes using SA (A), HA (B), mixed pollutants (C) and mimetic sewage (D) as pollutant feed.

To further study the membrane fouling problem of modified PVDF membranes used in the filtration of mimetic sewage solution, the mixed pollutants solution of BSA, SA and HA were used to examine the antifouling ability. As shown in Fig. 6C, the modified S-T-0.1 and S-T-0.3 membranes exhibit higher FR values compared to the neat PVDF membrane, suggesting that the existence of SA layer was conducive to reduce the flux decline caused by mixed pollutants and improve the anti-adsorption ability. In addition, the FRR value of S-T-0.3 membrane was as high as 94%, exhibiting excellent antifouling effect in dynamic filtration. It could be speculated that the SA layer of modified PVDF membranes surface played a key role for enhancing the antifouling performance. According to the results of Fig. 5 and 6, the modified membrane surface may show different interaction mechanisms with hydrophilic and hydrophobic pollutants in the feed solutions due to the different type of interactions, including the selective adsorption and selective exclusion during the separation process. Hydrophilic SA layer exhibited repulsion effect for hydrophobic pollutants (hydrophobic part of BSA) based on different Gibbs free energy and steric effects. On the contrary, hydrophilic pollutants (SA and FA) could be bonded to the fouled membrane surface through the interaction of hydrogen bond and static electricity, which prevented the pollutants from entering the membranes pores to cause more irreversible fouling. Finally, the pollutants form a filter cake layer on the membrane surface, which can be washed away though flushing, leading to the reduction of IFR values.^24,25 Based on this principle, fouled surface with SA layer could exhibit improved antifouling effect for numerous hydrophilic and hydrophobic pollutants, which may induce serious membrane fouling problems.

There were a variety of organic matter and salt ions in sewage system, acting as important components of sewage to induce membrane fouling. As shown in Fig. 6D, the S-T-0.3 membrane exhibited 97% water flux recovery rate after the addition of several kinds of ions in the mixed pollutants feed. Obviously, the antifouling properties of modified PVDF membranes were further improved. The reason may be contributed to the role of Ca²⁺ in pollutants feed. Ca²⁺ as multivalent ions could exhibit crosslinking effect with two carboxylate anion simultaneously, promoting the aggregation effect of pollutants^26,27 and leading to the increase of pollutant sizes and effectively hinder the pollutants from entering the membrane pores. In addition, Ca²⁺ ions as positive ion also show pairing effect with the carboxylic acid anion on the PVDF membrane surface to provide bridging molecule for the formation of cake layer, and increase the traction and fixed forces for the pollutants contacting the membrane surface in feed solution. Gathered hydrophilic pollutants enhanced the wetting ability of the membrane surface, resulting in the stronger repulsive force against the hydrophobic pollutants. Therefore, the addition of calcium ions in the feed solution led to a sharp decrease of irreversible membrane fouling extent to a very low level, and the modified PVDF membrane could exhibit better antifouling properties for sewage. However, it should be worth noted that the addition of Ca²⁺ ion might accelerate the fouling process and requires higher cleaning frequency.

3.6 Antifouling mechanism

According to the above described fouling test results, it was precisely learned that the PVDF UF membrane with the SA layer exhibited excellent antifouling effect for hydrophobic, hydrophilic and natural pollutants. A possible antifouling mechanism for the fouled PVDF membranes was shown in Fig. 7. According to the reported interaction mechanism of pollutants with different membrane surfaces by Professor Elimelech,¹³ the control principle of the fouled hydrophilic surface for the fouling problem may be explained as follows: the possibility of hydrophobic pollutants adhered to the hydrophilic membrane surface was greatly reduced due to the differences in surface energy. Meanwhile, the protective space barrier of SA hydration layer would expel hydrophobic pollutants directly. The hydrophilic pollutants with typically hydrophilic polar or ionic hydrophilic groups were restricted and withheld on the membrane surfaces to form cake layer based on various interactions such as hydrogen bonding, electrostatic and ionic bond between hydrophilic SA layer and hydrophilic pollutants. Compared with the pollutants entering the membrane pores, the cake-layer type pollutants can be washed off easily.²⁸ To sum up, the main function of the fouled layer of the modified PVDF membranes was to reduce the irreversible fouling, preventing all kinds of organic pollutants from adsorbing on the membrane surface and entering the membrane pores.^29–31

Fig. 7 Schematic diagram for the antifouling mechanism of fouled PVDF membranes.

4. Conclusions

The novel modified UF PVDF membranes with SA as the fouled and functional layer were prepared via plasma treatment and surface crosslinking, which were applied to the separation of mimetic sewage system and exhibited excellent antifouling property. For single and mixed sewage feed, the modified PVDF membrane showed high FRR values with significant reduction of irreversible fouling, and the IFR values were no more than 6% for the mimetic sewage solution due to the improved hydrophilicity, screening ability and gather ability of SA. The presence of SA layer reduced the possibility of various kinds of hydrophilic and hydrophobic pollutants to enter the UF membrane pores. In addition, the SA layer enhanced the cake layer formation of hydrophilic pollutants, which can be cleaned by backwashing to decrease irreversible fouling. It was demonstrated that the fouled surface of the modified PVDF membrane was able to control the membrane fouling problem and enhance the antifouling ability of hydrophobic PVDF membranes. This paper provides a new design method for the preparation of antifouling UF membrane for sewage treatment applications.

Acknowledgements

This work was financially supported by the Research Foundations of Shenzhen (JCYJ20140418095735550, JCYJ20160308105200725, KQCX20140519103908550, ZDSYS20160606153007978), and the National Natural Science Foundation of China (21307083).

References

1. M. A. Shannon, P. W. Bohn, M. Elimelech, J. G. Georgiadis, B. J. Marinas and A. M. Mayes, Nature, 2008, 452, 301.
2. M. Elimelech and W. A. Phillip, Science, 2011, 333, 712.
3. D. Rana and T. Matsuura, Chem. Rev., 2010, 110, 2448.
4. W. Chen, Y. Su, J. Peng, X. Zhao, Z. Jiang, Y. Dong, Y. Zhang, Y. Liang and J. Liu, Environ. Sci. Technol., 2011, 45, 6545.
5. A. Asatekin and A. M. Mayes, Environ. Sci. Technol., 2009, 43, 4487.
6. M. Zhou, H. Liu, J. E. Kilduff, R. Langer, D. G. Anderson and G. Belfort, Environ. Sci. Technol., 2009, 43, 3865.
7. H. Yamamura, K. Kimura, T. Okajima, H. Tokumoto and Y. Watanabe, Environ. Sci. Technol., 2008, 42, 5310.
8. Y. Chang, C. Y. Ko, Y. J. Shih, D. Quemener, A. Deratani, T. C. Wei, D. M. Wang and J. Y. Lai, J. Membr. Sci., 2009, 345, 160.
9. M. Kobayashi, Y. Terayama, H. Yamaguchi, M. Terada, D. Murakami, K. Ishihara and A. Takahara, Langmuir, 2012, 28, 7212.
10. E. M. Van Wagner, A. C. Sagle, M. M. Sharma, Y. H. La and B. D. Freeman, J. Membr. Sci., 2011, 367, 273.
11. Y. Cho, H. S. Sundaram, C. J. Weinman, M. Y. Paik, M. D. Dimitriou, J. A. Finlay, M. E. Callow, J. A. Callow, E. J. Kramer and C. K. Ober, Macromolecules, 2011, 44, 4783.
12. M. Kobayashi, Y. Terayama, M. Kikuchi and A. Takahara, Soft Matter, 2013, 9, 5138.
13. A. Tiraferri, Y. Kang, E. P. Giannelis and M. Elimelech, Environ. Sci. Technol., 2012, 46, 11135.
14. J. Zlopasa, B. Norder, A. B. Koenders and S. J. Picken, Macromolecules, 2015, 48, 1204.
15. Y. F. Li, H. P. Jia, Q. L. Cheng, F. S. Pan and Z. Y. Jiang, J. Membr. Sci., 2011, 375, 304.
16. Q. L. Li and M. Elimelech, Environ. Sci. Technol., 2004, 38, 4683.
17. A. M. Mayes, A. Asatekin, K. Seoktae and M. Elimelech, J. Membr. Sci., 2007, 298, 136.
18. K. Listiarini, W. Chun, D. D. Sun and J. O. Leckie, J. Membr. Sci., 2009, 344, 244.
19. A. Achilli, T. Y. Cath, E. A. Marchand and A. E. Childress, Desalination, 2009, 239, 10.
20. C. Jarusutthirak and G. Amy, Environ. Sci. Technol., 2006, 40, 969.
21. Y. X. Shen, K. Xiao, P. Liang, J. Y. Sun, S. J. Sai and X. Huang, J. Membr. Sci., 2012, 415, 336.
22. S. Liang, G. Qi, K. Xiao, J. Sun, E. P. Giannelis, X. Huang and M. Elimelech, J. Membr. Sci., 2014, 463, 94.
23. L. Li, N. Y. Gao, Y. Deng, J. J. Yao and K. J. Zhang, Water Res., 2012, 46, 1233.
24. F. S. Qu, H. Liang, Z. Z. Wang, H. Wang, H. R. Yu and G. B. Li, Water Res., 2012, 46, 1490.
25. A. W. Zularisam, A. F. Ismail and R. Salim, Desalination, 2006, 194, 211.
26. B. X. Mi and M. Elimelech, J. Membr. Sci., 2010, 348, 337.
27. B. Mi and M. Elimelech, J. Membr. Sci., 2008, 320, 292.
28. L. Li, Z. M. Wang, L. C. Rietveld, N. Y. Gao, J. Y. Hu, D. Q. Yin and S. L. Yu, Environ. Sci. Technol., 2014, 48, 14549.
29. H. Huang, T. A. Young and J. G. Jacangelo, Environ. Sci. Technol., 2008, 42, 714.
30. Q. Wang, Z. Wang, C. Zhu, X. Mei and Z. Wu, J. Membr. Sci., 2013, 446, 154.
31. C. J. van Oss, Colloids Surf., B, 2007, 54, 2.

---