Anti-Fouling Behavior of Hyperbranched ... - ACS Publications

Jul 14, 2014 - ... generation with low organic fouling tendency. Ye Li , Saren Qi , Yining Wang , Laurentia Setiawan , Rong Wang. Desalination 2017 42...
0 downloads 0 Views 9MB Size
Download PDF
Article pubs.acs.org/est

Anti-Fouling Behavior of Hyperbranched Polyglycerol-Grafted Poly(ether sulfone) Hollow Fiber Membranes for Osmotic Power Generation Xue Li, Tao Cai, and Tai-Shung Chung* Department of Chemical & Biomolecular Engineering, National University of Singapore, Singapore 117585, Singapore S Supporting Information *

ABSTRACT: To sustain high performance of osmotic power generation by pressureretarded osmosis (PRO) processes, fouling on PRO membranes must be mitigated. This is especially true for the porous support of PRO membranes because its porous structure is very prone to fouling by feeding river water. For the first time, we have successfully designed antifouling PRO thin-film composite (TFC) membranes by synthesizing a dendritic hydrophilic polymer with well-controlled grafting sites, hyperbranched polyglycerol (HPG), and then grafting it on poly(ether sulfone) (PES) hollow fiber membrane supports. Compared to the pristine PES membranes, polydopamine modified membranes, and conventional poly(ethylene glycol) (PEG)grafted membranes, the HPG grafted membranes show much superior fouling resistance against bovine serum albumin (BSA) adsorption, E. coli adhesion, and S. aureus attachment. In high-pressure PRO tests, the PES TFC membranes are badly fouled by model protein foulants, causing a water flux decline of 31%. In comparison, the PES TFC membrane grafted by HPG not only has an inherently higher water flux and a higher power density but also exhibits better flux recovery up to 94% after cleaning and hydraulic pressure impulsion. Clearly, by grafting the properly designed dendritic polymers to the membrane support, one may substantially sustain PRO hollow fiber membranes for power generation.

1. INTRODUCTION Osmotic energy harvested from ocean and various sources of salty water is an attractive, renewable, and environmentally friendly energy alternative to conventional fossil fuel. Pressure retarded osmosis (PRO) offers the advantages of high scalability, accessibility, and cost efficiency to effectively harvest such energy into electricity.1−5 However, most of early PRO research were discontinued due to the absence of effective membranes.6−8 Statkraft of Norway was the first company initiating serious research on osmotic power4,9 and built the first prototype PRO plant in 2009. Since then, progress on PRO membranes10−20 and system modeling21−25 have been rapidly advanced. However, extensive pretreatment processes must be implemented in order to sustain the PRO process because fouling remains a big challenge for current PRO membranes. This is especially true for the thin-film composite (TFC) PRO membrane because both its polyamide layer and porous support are susceptible to fouling. Membrane fouling often occurs by the deposition and attachment of inorganic compounds, organic compounds, and biomolecular foulants onto the membrane surface, leading to pore clogging and permeability deterioration.26−37 Compared to conventional pressure-driven membrane processes, fouling on PRO membranes is much more complicated because fouling takes place on the outer surface of selective layers, outside and inside of support layers. If reverse osmosis (RO) retentate is used as the draw solution, it not only generates higher osmotic energy but © 2014 American Chemical Society

also saves expensive pretreatment cost because RO retentate is pretreated in its previous processes.3,38 Thus, the fouling on the selective layer of PRO membranes can be significantly reduced. In contrast, fouling on the support layer is unavoidable because water permeates from the feed solution (e.g., river water) to the draw solution (e.g., seawater or brine) that brings foulants into the porous support. As a result, the porous support of the PRO membranes must be modified to minimize its fouling propensity. Grafting hydrophilic linear polymers such as poly(ethylene glycol) on membrane surfaces has improved fouling resistance.39−41 However, such noncontrollable and randomly curly polymer chains can also block surface pores and reduce permeability.42 Unlike linear polymers, well-designed dendrimers show distinct advantages of properly planting on membrane surface with controllable shielding sizes and multiplied antifouling effects. Therefore, the aims of this study are to (1) explore highly effective fouling-control strategies and (2) to graft hyperbranched polymers on PRO membranes because these polymers are highly accessible in one-step reaction.43 A hyperbranched polyglycerol (HPG) was chosen because it can be synthesized with a controlled molecular weight, narrow Received: Revised: Accepted: Published: 9898

April 10, 2014 June 29, 2014 July 14, 2014 July 14, 2014 dx.doi.org/10.1021/es5017262 | Environ. Sci. Technol. 2014, 48, 9898−9907

Environmental Science & Technology

Article

Scheme 1. Synthesis Route for the Hyperbranched Polyglycerol

were introduced into a 50 mL single-necked round-bottom flask. The solution was purged with argon for 20 min to remove the dissolved oxygen. The reaction flask was sealed under an argon atmosphere, and the reaction was allowed to proceed at 100 °C with stirring for 2 h. Glycidol (10 mL, 151 mmol) in 1,4-dioxane (10 mL) was purged with argon for 30 min. It was added slowly with a syringe to the reaction mixture over a period of approximately 12 h. After complete addition of the solution, the reaction mixture was allowed to stir for an additional 12 h. At the end of the reaction, the reaction flask was quenched in cold water and diluted with methanol, followed by pouring into 300 mL of acetone. The adduct was purified twice by redissolving in methanol and reprecipitating in acetone. The HPG-S-S-HPG polymer was dried under vacuum at 80 °C to give a highly viscous liquid (yield ∼74%). The HPG-S-S-HPG (Mn,NMR = 12,200 g mol−1, 4.88 g, 0.4 mmol) polymer was dissolved in 50 mL of ethanol. The reaction mixture was purged with argon for 30 min. DTT (4 mmol) was then introduced into the reaction flask. The mixture was stirred under the protection of an argon stream for 24 h at 50 °C. The crude product was dialyzed against deionized water for 3 days, with the deionized water changed twice daily, using a cellulose acetate dialysis tubing (Sigma-Aldrich, MWCO 1000 g mol−1). Finally, the HPG-SH polymer was isolated via lyophilization and obtained a pale yellow liquid (yield ∼96%). 2.2. Preparation of the HPG-g-PES Hollow Fiber Membrane Supports. Prior to polymer grafting, polydopamine-coated hollow fiber membrane supports were prepared as reported previously.15 A 200 mg portion of dopamine-HCl in a 1 L Tris buffer solution (0.01 mol L−1, pH 8.5) was used to coat hollow fiber supports for 3 h. After polydopamine (PDA)

polydispersity index (PDI), and well-preserved end group functionality by ring-opening polymerization of glycidol with the aid of a strong base (sodium hydride, NaH) and a coinitiator (amine or hydroxyl group).44−47 Scheme 1 illustrates the synthesis route for a thiol-terminated hyperbranched polyglycerol (HPG-SH) of dendritic architecture. It involved ring-opening polymerization of glycidol using bis(2hydroxyethyl)disulfide (BHEDS) as the initiator, followed by cleavage of the disulfide bond using excess DL-1,4-dithiothreitol (DTT) as the reducing agent. Figure 1 illustrates the step-bystep procedure to fabricate the HPG-grafted polyether sulfone (PES) TFC hollow fiber membranes: (1) modules consisting of asymmetric PES hollow fiber membranes were fabricated; (2) a HPG-SH solution was circulated in the shell side so that HPGSH would react with the polydopamine-modified support via Schiff base reaction to induce steric/enthalpic barrier against fouling; and (3) an ultrathin polyamide was deposited on the inner surface of hollow fibers via thin-film interfacial polymerization. To the best of our best knowledge, grafting HPG-SH on polydopamine-modified PRO membranes has never been proposed. This pioneering work may provide useful insight on novel PRO membranes with fouling resistance for osmotic power generation.

2. EXPERIMENTAL SECTION 2.1. Synthesis of the HPG-SH. HPG-SH was synthesized in 1,4-dioxane from ring-opening polymerization of glycidol, using BHEDS as the initiator and NaH as the catalyst, followed by cleavage of the disulfide bond using excess DTT as the reducing agent.48,49 Briefly, BHEDS (92 μL, 0.755 mmol), NaH (1.8 mg, 0.076 mmol), and anhydrous 1,4-dioxane (10 mL) 9899

dx.doi.org/10.1021/es5017262 | Environ. Sci. Technol. 2014, 48, 9898−9907

Environmental Science & Technology

Article

Figure 1. Schematic procedure for the fabrication of HPG-g-TFC membranes.

soaked in the BSA-FITC solution (0.5 mg mL−1 PBS solution) at room temperature for 1 h. The adsorption was imaged with a Leica DMLM fluorescence microscope (Leica Microsystems, Wetzlar, Germany) equipped with an excitation filter of 495 nm and an emission filter of 525 nm. The fluorescence intensity, which was proportional to the surface density of the adsorbed BSA-FITC protein, was quantified using ImageJ software (National Institutes of Health, Bethesda, MD). Gram-negative Escherichia coli (E. coli, ATCC DH5α) and Gram-positive Staphylococcus aureus (S. aureus, ATCC 25923) were used for the evaluation of the antibacterial adhesion characteristics and bactericidal efficacy of the hollow fiber membranes. E. coli and S. aureus were cultured in the nutrient broth and tryptic soy broth, respectively, at 37 °C overnight. Overnight bacterial culture broths were centrifuged at 2700 rpm for 10 min to remove the supernatant. The bacteria were washed with PBS twice and resuspended in PBS (pH 7.4) at a concentration of 5 × 10 7 cells mL −1 . The bacterial concentrations were estimated from the optical density of the suspension based on a standard calibration from spread plate counting. An optical density of 0.1 at 540 nm was equivalent to 108 cells mL−1. Each membrane was then immersed in the bacterial suspension at 37 °C for 4 h. After fixing with 3% glutaraldehyde and dehydrating with serial ethanol, adhered bacterial cells were investigated under SEM. Quantification of bacteria adhesion and viability on membrane supports were carried out by the spread plate method. Briefly, after washing with PBS as described above, the substrates were put into 2 mL

coating, HPG-grafted hollow fiber supports were prepared by the Schiff base reaction between polydopamine and HPG-SH. PES hollow fiber supports were soaked in a 10 g L−1 HPG-SH water solution containing triethylamine (TEA) of 0.7%, v/v at room temperature. The amount of HPG was excessive and the reaction time was 12 h to ensure the completion of the coating. Reference experiments were conducted by using PEG-SH to replace HPG-SH in the same conditions. 2.3. Characterizations of Hollow Fiber Membranes and Antifouling Assays. The morphology of membranes was examined by a scanning electronic microscope (SEM JEOL JSM-5600LV) and a field emission scanning electronic microscope (FESEM, JEOL JSM-6700F). Before SEM/ FESEM tests, samples were prepared in liquid nitrogen followed by platinum coating using a Jeol JFC-1100E ion sputtering device. X-ray photoelectron spectroscopy (XPS) was used to measure the chemical composition of the functionalized membranes. XPS measurements were conducted on a Kratos AXIS Ultra HSA spectrometer equipped with a monochromatized Al Kα X-ray source (1468.6 eV photons). Elemental stoichiometries of the membrane surface were determined from peak-area ratios with the reliability of ±5%. Adsorption of fluorescein isothiocynateconjugated bovine serum albumin (BSA-FITC) on the pristine PES membrane supports, PDA-modified supports, PEG-g-PES supports, and HPG-g-PES supports were examined with a fluorescence microscope. The membrane supports were rinsed initially with a phosphate-buffered saline (PBS) solution and then 9900

dx.doi.org/10.1021/es5017262 | Environ. Sci. Technol. 2014, 48, 9898−9907

Environmental Science & Technology

Article

Figure 2. SEM images of inner surfaces and outer surfaces of four hollow fiber membrane supports.

variable-speed gear pump (Cole-Palmer, Vernon Hills, IL) was utilized to recirculate the feed solution through the shell side of the hollow fibers, and a high-pressure hydra cell pump was employed to recirculate the draw solution through the lumen side. Prior to tests, the TFC hollow fiber membranes were stabilized at 12.5 ± 1 bar for 1 h. To initiate the fouling test, 200 mg L−1 bovine serum albumin (BSA) was dosed into the feed solution. Baseline experiments were conducted under the same conditions in the absence of foulants. The purpose of baseline experiments is to study the flux decline due to the dilution of the draw solution. After fouling tests and before the next PRO test, membrane cleaning was performed. Countercurrent flows of deionized water at a flow rate of 0.1 L min−1 were applied to both the lumen and the shell sides of hollow fiber membranes without pressure for 6 h (24 ± 1 °C). The power density is calculated by eq 1

of PBS and subjected to ultrasonication for 7 min, followed by vortexing for 20 s to release the cells. The bacterial solution was then serially diluted, spread on the agar plate, and cultured overnight to quantify the number of bacterial cells. All experiments were performed in triplicate with three samples, and the mean values were calculated. The results were expressed as viable adherent fractions, defined as percentages of viable adherent bacteria cells on the modified membrane surface relative to those on the pristine PES surface. The cytotoxicity of membranes was evaluated by determining the viability of mouse 3T3 fibroblasts after incubation in the medium containing the membranes. The membranes were sterilized with 75% ethanol and recovered after drying under a reduced pressure before use. Control experiments were carried out using the growth culture medium without membranes. Cell viability testing was carried out via the reduction of the MTT reagent (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide). The 3T3 fibroblasts were first seeded in a 24-well culture plate and incubated at 37 °C for 24 h, and then the medium was replaced with a fresh medium containing different membranes of 1 cm length. The 3T3 fibroblasts were incubated at 37 °C for another 24 h in the medium. After that, the culture medium in each well was removed. Ninety milliliters of medium and 10 mL of MTT solutions (5 g L−1 in PBS) were mixed and distributed to each well. The volume of each well in the 24 well plate is ∼1 mL. After 4 h of incubation at 37 °C, the medium was removed, and the formazan crystals (a purple-color dye from reduction of MTT in living cells) were solubilized with 100 mL of dimethyl sulfoxide (DMSO) for 15 min. The optical absorbance was measured at 560 nm on a microplate reader (Tecan GENios). The results were expressed as percentages relative to that obtained in the control experiment. 2.4. Membrane Fouling Protocol in PRO Processes. The PRO fouling tests were conducted on a laboratory-scale PRO setup using membrane modules as described in previous publications.18,20,22 Synthetic seawater (3.5% NaCl) and deionized water were used as draw and feed solutions, respectively. TFC membranes were oriented in the PRO mode (i.e., selective layer faces the draw solution) for all tests. Countercurrent flows at a flow rate of 0.1 L min−1 were applied to both the draw solution and the feed solution at 24 ± 1 °C. A

W = Jw ΔP

(1)

where ΔP is the hydraulic pressure difference across the membrane and Jw is the water permeation flux.

3. RESULTS AND DISCUSSION 3.1. Membrane Characterization. In the present work, HPG-SH was first synthesized in 1,4-dioxane from ring-opening polymerization of glycidol, as shown in Scheme 1. The 1H NMR results as shown in the Supporting Information indicated that a HPG-SH polymer was successfully prepared. The HPGSH polymer used in preparing the HPG-g-PES hollow fiber membranes had a number-average molecular weight of 6100 g mol−1, calculated by 1H NMR results. Figure S1 (Supporting Information) displays morphologies of pristine PES hollow fiber membranes spun from dry-jet wet spinning via a dual-layer spinneret. The porous structure of outer surfaces is a result of the delayed demixing induced by the outer channel solvent NMP, while the relatively dense inner structure is ascribed to a faster demixing from the bore fluid water during the phase-inversion process.22,50 As shown in the cross-section image, a layer of finger-like macrovoids are formed between the two sponge-like layers beneath the inner and outer membrane surfaces. This structure is formed due to 9901

dx.doi.org/10.1021/es5017262 | Environ. Sci. Technol. 2014, 48, 9898−9907

Environmental Science & Technology

Article

Figure 3. XPS wide scan, C 1s, S 2p, and N 1s core-level spectra of four hollow fiber membrane supports.

molecules.53 For HPG-g-PES membranes, the presence of HPG grafting is ascertained by the XPS wide scan C 1s and S 2p spectra. The appearance of S 2s and S 2p signals in the wide scan is consistent with the sulfur site of the HPG polymer. The S 2p core-level spectrum consisting of the −C−S 2p3/2 species and −C−S 2p1/2 species also shows the covalent bonding of grafted HPG segments. Similar XPS results can be found for the PEG-g-PES membranes as they share similar chemical bonding and grafting on the membrane surfaces with HPG-gPES membranes. 3.2. Antiprotein Adsorption Behavior of the PES, PDA−PES, HPG-g-PES, and PEG-g-PES Membranes. The resistance against protein adsorption is an important indication of the antifouling performance of membranes. In this work, fluorescence microscopy of fluorescein-labeled bovine serum albumin (BSA) was chosen to sensitively and reliably measure protein adsorption onto membranes. Figure 4 displays the relative fluorescence intensities and fluorescence microscopy images of the pristine PES membrane surface and polymerfunctionalized membrane surfaces after exposure to a 0.5 mg mL−1 PBS solution of fluorescein isothiocynate-conjugated BSA (BSA-FITC). As expected, the pristine PES surface is covered with a large concentration of BSA molecules due to the hydrophobic−hydrophobic interactions between conjugated benzene rings of PES and protein molecules. The adsorption of BSA is slightly reduced on the PDA−PES membrane surface because of the repulsive forces arising from the hydrophilic PDA surface. However, the benzene rings and possible π−π

the fast phase inversion at the lumen side and strong effects of nonsolvent (i.e., water) intrusion. To graft HPG on the PES surface, a polydopamine treatment on PES hollow fiber membranes was first performed. Then, HPG was introduced to the PDA−PES membrane supports via Schiff base reaction.51 For comparison, a commercial poly(ethylene glycol) (PEG-SH, Mw = 6000 g mol−1) was also grafted on the PDA−PES supports to represent the linear antifouling agent. Figure 2 shows the SEM images of surfaces of four membrane supports. Apparently, all supports show similar surface morphologies of both inner and outer surfaces, indicating the PDA modification and grafting have no visible effects on the membrane structure. Figure 3 exhibits the respective XPS wide scan, C 1s, S 2p, and N 1s core-level spectra of outer surfaces of four membrane supports. In the case of the pristine PES support, C 1s, O 1s, S 2s, and S 2p signals are observed in the wide-scan spectrum. The C 1s core-level can be curve-fitted into two peaks with binding energies at about 284.6 and 286.1 eV, attributable to the C−H and C−O/C−SO2 species, respectively.52 The S 2p signal can be curve-fitted into peaks at 167.6 eV for the −SO2 2p3/2 species and 168.8 eV for the −SO2 2p1/2 species which belong to the strong spin-orbit split doublet of the sulfone group in PES molecules. After PDA treatments, S signals of PDA−PES membranes in the XPS wide scan spectrum disappear and the N 1s signal appears, confirming the successful covering of a PDA layer onto the PES support. The C 1s corelevel spectrum combines the C−N, C−H, C−O, CO, and O−CO species, which are components of polydopamine 9902

dx.doi.org/10.1021/es5017262 | Environ. Sci. Technol. 2014, 48, 9898−9907

Environmental Science & Technology

Article

fluorescence is barely visible on the HPG-g-PES surface. The fluorescence intensity is only 16% of that on the pristine PES surface. 3.3. Antibacterial Adhesion Behavior of the PES, PDA−PES, HPG-g-PES, and PEG-g-PES Membranes. Bacterial fouling on the pristine and polymer-functionalized membrane surfaces has been investigated. SEM images of the bacterial adhesion were taken after the hollow fiber membranes have been exposed to E. coli or S. aureus bacteria for 4 h (Figure 5A,B). The SEM images show that the pristine PES membrane and PDA−PES membrane seriously foul with both E. coli and S. aureus bacteria. A large amount of bacterial cells adhere readily to the membrane surfaces and stack in clusters. In comparison, the attachment on the HPG-g-PES membrane exhibits small amounts of bacterial cells, suggesting a significant antibacterial adhesion effect of HPG grafting. The PEG-g-PES membrane is also not highly susceptible to adhesion and colonization of bacteria. However, many small bacterial clusters are still observed on the PEG-g-PES surface. To further investigate the antimicrobial properties of the hollow fiber membranes, a quantitative antifouling assay was conducted by using the spread plate method. Figure 5C depicts the viable adherent fractions of E. coli and S. aureus bacteria after exposure to the pristine and modified membranes for 4 h. The amount of E. coli adhesion decreases monotonically after polymer modifications, with a viable adherent fraction of 98.2% for the PDA−PES, 10.7% for the PEG-g-PES, and 6.3% for the HPG-g-PES, respectively, compared to that of the pristine PES membrane. On the other hand, after S. aureus exposure, the PDA-modified membrane shows a similar quantitative adhesion (101.0%), in comparison to the pristine PES membrane, while the HPG-gPES and PEG-g-PES membranes reduce the adhesion to 11.3%

Figure 4. Relative fluorescence intensities and respective fluorescence microscopy images of the PES, PDA−PES, PEG-g-PES, and HPG-gPES hollow fiber membrane supports after exposure to 0.5 mg mL−1 BSA-FITC solution for 1 h.

stacking structure of PDA53,54 may attract proteins. Therefore, the fluorescence on the PDA−PES surface is also uniform and intense. The BSA adsorption significantly decreases on the PEG-g-PES surface. The fluorescence intensity is more than 70% lower than that on the pristine PES surface. The enhanced antiadsorption effect is attributed to the strong repulsive force between −OH groups of PEG and proteins.40,55,56 The dendritic structure of HPG achieves better antiadsorption behavior than the linear PEG. After HPG grafting, the

Figure 5. Antibacterial adhesion behavior of four membrane supports: (A) E. coli ( ATCC DH5α); (B) S. aureus (ATCC 25923); (C) viable adherent fractions of E. coli and S. aureus cells in PBS (5 × 107 cells mL−1) in contact with four hollow fiber membrane supports at 37 °C for 4 h. 9903

dx.doi.org/10.1021/es5017262 | Environ. Sci. Technol. 2014, 48, 9898−9907

Environmental Science & Technology

Article

3.4. PRO Performance and PRO Fouling. Figure 7 shows the membrane performance in high-pressure PRO tests in terms of water flux and power density as a function of pressure. Synthetic seawater (3.5% NaCl in water) and deionized water were used as the stream pair. The PRO tests were conducted in a safe pressure range (0−16 bar) to avoid the “bursting” phenomenon where the water flux reversely flows across the membrane from the draw solution into the fresh water. As shown in Figure 7, the TFC−HPG membrane possesses an enhanced initial water flux at 0 bar than the TFC−PES and TFC−PEG membranes. This is because of the higher wettability of the HPG-grafted membrane surface. All TFC membranes demonstrate a water flux decline with increasing ΔP because of the reduction in effective driving force. The TFC−PES membrane exhibits a maximum power density of 5.0 W m−2, while the TFC−PEG membrane fabricated from the PEG-g-PES support shows an improved maximum power density of 6.0 W m−2. In contrast, the TFC−HPG membrane further increases the maximum power density to 6.7 W m−2 due to the higher water flux induced by HPG grafting. Clearly, the HPG-functionalized membranes produce a better PRO performance. Prior to the fouling experiments, baseline experiments were performed to quantify the flux decline due to the dilution of draw solution, concentration of feed solution, and the resultant decrease in osmotic pressure difference. As shown in Figure 8A, the baseline experiments were conducted at an identical effective hydraulic pressure difference of 12.5 ± 1 bar for the TFC−PES, TFC−PEG, and TFC−HPG membranes. The normalized water flux is plotted against the normalized accumulative permeate volume (i.e., normalizing it to the same membrane surface area). The baseline curves of TFC− PES (rectangles) and TFC−PEG (circles) membranes have trends similar to those of their initial water fluxes, and the final water fluxes are about 96% of the initial ones. The TFC−HPG membrane exhibits the most serious dilution effect as shown by the triangles because it has the highest initial water flux among three membranes. Its final water flux is 92% of the initial value. To study the fouling in PRO processes, BSA was chosen to dose the feed solution and model the protein foulants during PRO tests. Before the foulant dosage, fresh TFC hollow fiber membranes were conditioned at 12.5 ± 1 bar for 1 h. Figure 8B demonstrates the fouling and recovering behaviors of the TFC membranes. After fouling tests and before the next PRO test, membrane cleaning by deionized water was performed on both sides of the membrane for 6 h. The rectangular symbols indicate the normalized water flux of the TFC−PES membrane during the fouling test as a function of normalized cumulative

and 14.9%, respectively. The HPG-g-PES and PEG-g-PES membranes became much more hydrophilic after modification. These hydrophilic polymer brushes can form highly hydrated ultrathin coatings that provide an effective enthalpic and entropic barrier to nonspecific bacteria adsorption and impart the membrane surface with improved antifouling property. It has been predicted theoretically that, at an equivalent area per molecules, a branched polymer will be more efficient in protein/bacteria rejection than its linear counterpart.57 In summary, the bacterial adhesion is essentially inhibited by the grafting of hyperbranched polyglycerol, and the antibacterial performance of surface-grafted HPG membranes is higher than that of membranes with conventional linear PEG of a similar molecular weight. From a safety perspective, the saline stream pair (e.g., seawater and river water) used in PRO processes does not allow passage through a toxic membrane surface. Therefore, the cytotoxicity of the newly modified membranes was evaluated by the in vitro experiments of the MTT (3-(4,5-dimethylthiazol-2yl)-2,5-diphenyltetrazolium bromide) cell-viability assay. The 3T3 fibroblasts were incubated with membranes for 24 h. The results show that both the HPG-g-PES and PEG-g-PES membranes exceed 96% for the viability of 3T3 fibroblasts cells in comparison of the control experiment (Figure 6),

Figure 6. Cytotoxicity assays of different hollow fiber supports in 3T3 fibroblasts culture medium after 24 h of incubation. Error bars represent the standard deviation of four measurements.

indicating that the introduction of HPG branches onto the PES membrane surface has negligible cytotoxicity effect to viable cells. Meanwhile, the viability of cells on the PDA−PES membrane surface is slightly lower (92%), which may be due to the toxicity of imine groups of the PDA structure.

Figure 7. Membrane performance in PRO tests using deionized water as the feed solution and 3.5% NaCl as the draw solution. The flow rates for both feed and draw solutions were 0.1 L min−1. 9904

dx.doi.org/10.1021/es5017262 | Environ. Sci. Technol. 2014, 48, 9898−9907

Environmental Science & Technology

Article

Figure 8. (A) Baseline tests of the TFC membranes in PRO and (B) membrane fouling performance by dosing BSA as the foulants. The normalized water flux decline is plotted against normalized permeate volume. The draw solution initially contained 3.5% NaCl, and deionized water was used as the feed solution. Hydraulic pressure difference between feed and draw solutions was 12.5 ± 1 bar. The flow rates for both feed and draw solutions were 0.1 L min−1.

permeate volume. Immediately after the foulant dosage, the water flux substantially declines over the first 20−40 L m−2 of the normalized permeate volume, and then the water flux drops slowly and gradually to less than 70% of the initial flux at the end of the fouling test when the normalized permeate volume is about 150 L m−2. The significant flux drop is consistent with the fouling assay of the PES membrane support, attributable to the fouling susceptible nature of PES and resultant adhesion of foulants onto the porous membrane support. After water cleaning for 6 h, the cleaned membranes were retested under PRO tests at 12.5 ± 1 bar. A sudden flux increase to 80% of the initial value is observed for the fouled and recleaned TFC-PES membrane. This phenomenon may be due to the “back pulse” effect from the hydraulic pressure in the draw solution side (Figure S2, Supporting Information). Since foulants have accumulated within the porous support and beneath the dense layer during the fouling test, some may be washed away during the cleaning, while some remain. In the subsequent high-pressure PRO test, the pressurized draw solution suddenly impulses the dense layer from the lumen side, expands and stretches the hollow fiber radially outward, and loosens the foulants adhered or trapped within the membrane. Therefore, the water flux goes up to some extent. Similar trends can be observed for both TFC−PEG and TFC−HPG membranes. In the case of the TFC−PEG membrane (circles), a steep water flux decrease is found in the early stage of the fouling test. Then the fouling progressively deteriorates the membrane performance, with a normalized flux dropping by ∼23% at the end of the fouling test. The water flux of the TFC−PEG membrane can be eventually recovered to ∼84% at the subsequent PRO test, suggesting an antifouling effect of the surface-grafted TFC− PEG membrane. For the TFC−HPG membrane, the antifouling effect of HPG branches is more evident. The water flux attrition significantly eases off in the fouling test. The reduction in water flux due to fouling is only ∼11%. After cleaning and hydraulic pressure impulsion, the eventual flux decline is ∼6%, which is impressively low. Clearly, the hyperbranched polyglycerol grafting onto the hollow fiber membrane support achieves a good antifouling performance. It is worthwhile to note once again that the antifouling effects of

the HPG membrane are much better than its linear representative PEG membrane. In this work, we have successfully designed antifouling PRO membranes by synthesizing hyperbranched polyglycerols (HPG) with one thiol site (HPG-SH) from ring-opening polymerization of glycidol and grafting them onto the poly(ether sulfone) (PES) hollow fiber membranes. Assays of SEM, XPS, fluorescence microscopy, and spread plate method have verified that the introduction of hydrophilic HPG branches to PES membrane surfaces imparts good antiprotein adsorption and antibacterial adhesion properties without changing the morphology and bulk properties of the hollow fiber membranes. Cytotoxicity experiments further confirm the nontoxic nature of the HPG grafting. In high pressure PRO tests, the HPG-grafted TFC membrane achieves higher performance in terms of water flux and power density in comparison of the pristine TFC and PEG grafted TFC membranes. Furthermore, the protein fouling on the HPG grafted TFC membrane has been largely alleviated. After cleaning, the recovered flux in PRO tests is as high as ∼94% of the initial value from a fresh TFC−HPG membrane. The present approach provides an effective means for the molecular design of antifouling PRO membranes.

■

ASSOCIATED CONTENT

S Supporting Information *

Materials and chemicals; characterizations of thiol-terminated HPG polymer; fabrication and morphology of PES hollow fiber supports; fabrication of thin-film composite membranes. This material is available free of charge via the Internet at http:// pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Tel: (65) 6516-6645. Fax: (65) 6779-1936. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This research was funded under the project entitled “Membrane development for osmotic power generation, Part 9905

dx.doi.org/10.1021/es5017262 | Environ. Sci. Technol. 2014, 48, 9898−9907

Environmental Science & Technology

Article

(17) Bui, N. N.; McCutcheon, J. R. Hydrophilic Nanofibers as new supports for thin film composite membranes for engineered osmosis. Environ. Sci. Technol. 2013, 47, 1761−1769. (18) Li, X.; Chung, T. S. Effects of free volume in thin-film composite membranes on osmotic power generation. AIChE J. 2013, 59, 4749−4761. (19) Alsvik, I.; Hägg, M. B. Pressure retarded osmosis and forward osmosis membranes: materials and methods. Polymers 2013, 5, 303− 327. (20) Zhang, S.; Sukitpaneenit, P.; Chung, T. S. Design of robust hollow fiber membranes with high power density for osmotic energy production. Chem. Eng. J. 2014, 241, 457−465. (21) Kim, J.; Park, M.; Snyder, S. A.; Kim, J. H. Reverse osmosis (RO) and pressure retarded osmosis (PRO) hybrid processes: modelbased scenario study. Desalination 2013, 322, 121−130. (22) Li, X.; Chung, T. S. Thin-film composite P84 co-polyimide hollow fiber membranes for osmotic power generation. Appl. Energy 2014, 114, 600−610. (23) Sivertsen, E.; Holt, T.; Thelin, W.; Brekke, G. Modelling mass transport in hollow fibre membranes used for pressure retarded osmosis. J. Membr. Sci. 2012, 417−418, 69−79. (24) Sivertsen, E.; Holt, T.; Thelin, W.; Brekke, G. Pressure retarded osmosis efficiency for different hollow fibre membrane module flow configurations. Desalination 2013, 312, 107−123. (25) Park, M.; Kim, J. H. Numerical analysis of spacer impacts on forward osmosis membrane process using concentration polarization index. J. Membr. Sci. 2013, 427, 10−20. (26) Mi, B. X.; Elimelech, M. Chemical and physical aspects of organic fouling of forward osmosis membranes. J. Membr. Sci. 2008, 320, 292−302. (27) Mi, B. X.; Elimelech, M. Organic fouling of forward osmosis membranes: Fouling reversibility and cleaning without chemical reagents. J. Membr. Sci. 2010, 348, 337−345. (28) Zaky, A. M.; Escobar, I. C.; Gruden, C. L. Application of atomic force microscopy for characterizing membrane biofouling in the micrometer and nanometer scales. Environ. Prog. Sustain. Energy 2013, 32, 449−457. (29) Li, Z. Y.; Yangali-Quintanilla, V.; Valladares-Linares, R.; Li, Q. Y.; Zhan, T.; Amy, G. Flux patterns and membrane fouling propensity during desalination of seawater by forward osmosis. Water Res. 2012, 46, 195−204. (30) Li, Z. Y.; Linares, R. V.; Abu-Ghdaib, M.; Zhan, T.; YangaliQuintanilla, V.; Amy, G. Osmotically driven membrane process for the management of urban runoff in coastal regions. Water Res. 2014, 48, 200−209. (31) Hausman, R.; Escobar, I. C. A comparison of silver- and coppercharged polypropylene feed spacers for biofouling control. J. Appl. Polym. Sci. 2013, 128, 1706−1714. (32) Thelin, W. R.; Sivertsen, E.; Holt, T.; Brekke, G. Natural organic matter fouling in pressure retarded osmosis. J. Membr. Sci. 2013, 438, 46−56. (33) Chen, S. C.; Fu, X. Z.; Chung, T. S. Fouling behaviors of polybenzimidazole (PBI)−polyhedral oligomeric silsesquioxane (POSS)/polyacrylonitrile (PAN) hollow fiber membranes for engineering osmosis processes. Desalination 2014, 335, 17−26. (34) Flemming, H.-C.; Schaule, G.; Griebe, T.; Schmitt, J.; Tamachkiarowa, A. Biofouling-the achilles heel of membrane processes. Desalination 1997, 113, 215−225. (35) Zhang, S.; Qiu, G.; Ting, Y. P.; Chung, T. S. Silver−PEGylated dendrimer nanocomposite coating for anti-foulingthin film composite membranes for water treatment. Colloids Surf., A 2013, 436, 207−214. (36) Yip, N. Y.; Elimelech, M. Influence of natural organic matter fouling and osmotic backwash on pressure retarded osmosis energy production from natural salinity gradients. Environ. Sci. Technol. 2013, 47, 12607−12616. (37) She, Q. H.; Wong, Y. K. W.; Zhao, S. F.; Tang, C. Y. Organic fouling in pressure retarded osmosis: experiments, mechanisms and implications. J. Membr. Sci. 2013, 428, 181−189.

1. Materials development and membrane fabrication” (1102IRIS-11-01) and NUS Grant No. of R-279-000-381-279. This research grant is supported by the National Research Foundation, Prime Minister’s Office, Singapore, under its Environmental & Water Technologies Strategic Research Programme and administered by the Environment & Water Industry Programme Office (EWI) of the PUB. Special thanks are due to Prof. E.T. Kang, Dr. S. Zhang, Mr. C.F. Wan, Ms. S.C. Chen, Mr. Z.L. Cheng, and Miss J. Gao for their useful comments and assistance. Dr. Xue Li also acknowledges the World Future Foundation (WFF) for her Ph.D. Prize in Environmental and Sustainability Research 2014.

■

REFERENCES

(1) Logan, B. E.; Elimelech, M. Membrane-based processes for sustainable power generation using water. Nature 2012, 488, 313−319. (2) Ramon, G. Z.; Feinberg, B. J.; Hoek, E. M. V. Membrane-based production of salinity-gradient power. Energy Environ. Sci. 2011, 4, 4423−4434. (3) Chung, T. S.; Li, X.; Ong, R. C.; Ge, Q. C.; Wang, H. L.; Han, G. Emerging forward osmosis (FO) technologies and challenges ahead for clean water and clean energy applications. Curr. Opin. Chem. Eng. 2012, 1, 246−257. (4) Gerstandt, K.; Peinemann, K.-V.; Skilhagen, S. E.; Thorsen, T.; Holt, T. Membrane processes in energy supply for an osmotic power plant. Desalination 2008, 224, 64−70. (5) Achilli, A.; Childress, A. E. Pressure retarded osmosis: from the vision of Sidney Loeb to the first prototype installation  review. Desalination 2010, 261, 205−211. (6) Loeb, S. Production of energy from concentrated brines by pressure-retarded osmosis. I. Preliminary technical and economic correlations. J. Membr. Sci. 1976, 1, 49−63. (7) Loeb, S.; Van Hessen, F.; Shahaf, D. Production of energy from concentrated brines by pressure-retarded osmosis, II. Experimental results and projected energy costs. J. Membr. Sci. 1976, 1, 249−269. (8) Lee, K.; Baker, R.; Lonsdale, H. Membranes for power generation by pressure-retarded osmosis. J. Membr. Sci. 1981, 8, 141−171. (9) Skilhagen, S. E.; Dugstad, J. E.; Aaberg, R. J. Osmotic power power production based on the osmotic pressure difference between waters with varying salt gradients. Desalination 2008, 220, 476−482. (10) Yip, N. Y.; Tiraferri, A.; Phillip, W. A.; Schiffman, J. D.; Hoover, L. A.; Kim, Y. C.; Elimelech, M. Thin-film composite pressure retarded osmosis membranes for sustainable power generation from salinity gradients. Environ. Sci. Technol. 2011, 45, 4360−4369. (11) Chou, S.; Wang, R.; Shi, L.; She, Q.; Tang, C. Y.; Fane, A. G. Thin-film composite hollow fiber membranes for pressure retarded osmosis (PRO) process with high power density. J. Membr. Sci. 2012, 389, 25−33. (12) Li, X.; Wang, K. Y.; Helmer, B.; Chung, T. S. Thin-film composite membranes and formation mechanism of thin-film layers on hydrophilic cellulose acetate propionate substrates for forward osmosis processes. Ind. Eng. Chem. Res. 2012, 51, 10039−10050. (13) Han, G.; Zhang, S.; Li, X.; Chung, T. S. High performance thin film composite pressure retarded osmosis (PRO) membranes for renewable salinity-gradient energy generation. J. Membr. Sci. 2013, 440, 108−121. (14) She, Q.; Jin, X.; Tang, C. Y. Osmotic power production from salinity gradient resource by pressure retarded osmosis: effects of operating conditions and reverse solute diffusion. J. Membr. Sci. 2012, 401−402, 262−273. (15) Li, X.; Zhang, S.; Fu, F. J.; Chung, T. S. Deformation and reinforcement of thin-film composite (TFC) polyamide-imide (PAI) membranes for osmotic power generation. J. Membr. Sci. 2013, 434, 204−217. (16) Zhang, S.; Chung, T. S. Minimizing the instant and accumulative effects of salt permeability to sustain ultrahigh osmotic power density. Environ. Sci. Technol. 2013, 47, 10085−10092. 9906

dx.doi.org/10.1021/es5017262 | Environ. Sci. Technol. 2014, 48, 9898−9907

Environmental Science & Technology

Article

(38) Chung, T. S.; Zhang, S.; Wang, K. Y.; Su, J. C.; Ling, M. M. Forward osmosis processes: yesterday, today and tomorrow. Desalination 2012, 287, 78−81. (39) Dong, B.; Jiang, H.; Manolache, S.; Wong, A. C. L.; Denes, F. S. Plasma-mediated grafting of poly(ethylene glycol) on polyamide and polyester surfaces and evaluation of antifouling ability of modified substrates. Langmuir 2007, 23, 7306−7313. (40) Nie, F. Q.; Xu, Z. K; Yang, Q.; Wu, J.; Wan, L. S. Surface modification of poly(acrylonitrile-co-maleic acid) membranes by the immobilization of poly(ethylene glycol). J. Membr. Sci. 2004, 235, 147−155. (41) Sofia, S. J.; Premnath, V.; Merrill, E. W. Poly(ethylene oxide) grafted to silicon surfaces: grafting density and protein adsorption. Macromolecules 1998, 31, 5059−5070. (42) Asatekin, A.; Menniti, A.; Kang, S.; Elimelech, M.; Morgenroth, E.; Mayes, A. M. Antifouling nanofiltration membranes for membrane bioreactors from self-assembling graft copolymers. J. Membr. Sci. 2006, 285, 81−89. (43) Kainthan, R. K.; Muliawan, E. B.; Hatzikiriakos, S. G.; Brooks, D. E. Synthesis, characterization, and viscoelastic properties of high molecular weight hyperbranched polyglycerols. Macromolecules 2006, 39, 7708−7717. (44) Dechy-Cabaret, O.; Martin-Vaca, B.; Bourissou, D. Controlled ring-opening polymerization of lactide and glycolide. Chem. Rev. 2004, 104, 6147−6176. (45) Cameron, D. J. A.; Shaver, M. P. Aliphatic polyester polymer stars: synthesis, properties and applications in biomedicine and nanotechnology. Chem. Soc. Rev. 2011, 40, 1761−1776. (46) Kim, H.; Olsson, J. V.; Hedrick, J. L.; Waymouth, R. M. Facile synthesis of functionalized lactones and organocatalytic ring-opening polymerization. ACS Macro Lett. 2012, 1, 845−847. (47) Hans, M.; Gasteier, P.; Keul, H.; Moeller, M. Ring-opening polymerization of ε-caprolactone by means of mono- and multifunctional initiators: comparison of chemical and enzymatic catalysis. Macromolecules 2006, 39, 3184−3193. (48) Wilms, D.; Stiriba, S. E.; Frey, H. Hyperbranched polyglycerols: from the controlled synthesis of biocompatible polyether polyols to multipurpose applications. Acc. Chem. Res. 2010, 43, 129−141. (49) Yeh, P.-Y. J.; Kainthan, R. K.; Zou, Y.; Chiao, M.; Kizhakkedathu, J. N. Self-assembled monothiol-terminated hyperbranched polyglycerols on a gold surface: a comparative study on the structure, morphology, and protein adsorption characteristics with linear poly(ethylene glycol)s. Langmuir 2008, 24, 4907−4916. (50) Sukitpaneenit, P.; Chung, T. S. High performance thin-film composite forward osmosis hollow fiber membranes with macrovoidfree and highly porous structure for sustainable water production. Environ. Sci. Technol. 2012, 46, 7358−7365. (51) Lee, H.; Dellatore, S. M.; Miller, W. M.; Messersmith, P. B. Mussel-inspired surface chemistry for multifunctional coatings. Science 2007, 318, 426−430. (52) Moulder, J. F.; Stickle, W. F.; Sobol, P. E.; Bomben, K. D. Handbook of X-ray Photoelectron Spectroscopy; Perkin-Elmer Corp.: Eden Prairie, MN, 1992. (53) Dreyer, D. R.; Miller, D. J.; Freeman, B. D.; Paul, D. R.; Bielawski, C. W. Elucidating the Structure of Poly(dopamine). Langmuir 2012, 28, 6428−6435. (54) Liebscher, J.; Mrówczyński, R.; Scheidt, H. A.; Filip, C.; Hădade, N. D.; Turcu, R.; Bende, A.; Beck, S. Structure of polydopamine: a never-ending story? Langmuir 2013, 29, 10539−10548. (55) Zhang, M.; Desai, T.; Ferrari, M. Proteins and cells on PEG immobilized silicon surfaces. Biomaterials 1998, 19, 953−960. (56) Malmsten, M.; Emoto, K.; Van Alstine, J. M. Effect of chain density on inhibition of protein adsorption by poly(ethylene glycol) based coatings. J. Colloid Interface Sci. 1998, 202, 507−517. (57) Xu, Y.; Takai, M.; Ishihara, K. Suppression of protein adsorption on a charged phospholipid polymer interface. Biomacromolecules 2009, 10, 267−274.

9907

dx.doi.org/10.1021/es5017262 | Environ. Sci. Technol. 2014, 48, 9898−9907