Grafting polyzwitterions onto polyamide by click chemistry and nucleophilic substitution on nitrogen: A novel approach to enhance membrane fouling resistance

Journal of Membrane Science 449 (2014) 50–57

Contents lists available at ScienceDirect

Journal of Membrane Science journal homepage: www.elsevier.com/locate/memsci

Grafting polyzwitterions onto polyamide by click chemistry and nucleophilic substitution on nitrogen: A novel approach to enhance membrane fouling resistance Hai-Yin Yu a,b, Yan Kang a, Yaolin Liu a, Baoxia Mi a,n a b

Department of Civil and Environmental Engineering, University of Maryland, 1161 Glenn L. Martin Hall, College Park, MD 20742, USA College of Chemistry and Materials Science, Anhui Normal University, 1 East Beijing Road, Wuhu, Anhui 241000, China

art ic l e i nf o

a b s t r a c t

Article history: Received 27 June 2013 Received in revised form 13 August 2013 Accepted 15 August 2013 Available online 22 August 2013

We present a first-of-its-kind use of click chemistry to graft polyzwitterions (PZs) onto polyamide, the most widely used material to make semi-permeable membranes for desalination and water purification. We have also experimentally proven that SN2 nucleophilic substitution on nitrogen can occur on the polyamide polymer chain under mild reaction conditions, as opposed to harsh reaction conditions required by many traditional grafting approaches. To prepare the click reaction, we synthesized an alkyne-PZ via reversible addition-fragmentation chain-transfer radical polymerization, followed by functionalizing polyamide with azide functional groups through bromination and subsequently SN2 nucleophilic substitution of Br with azide. The alkyne-PZ was then grafted to azide-polyamide by an azide–alkyne cycloaddition click reaction. The PZ-grafted polyamide became much more hydrophilic than the virgin polyamide. Results of Fourier transform infrared spectroscopy and X-ray photoelectron spectroscopy indicated that a successful click reaction and almost full surface coverage by PZ were achieved under studied experimental conditions. Membrane flux testing in a forward osmosis mode showed that the PZ grafting did not significantly affect the water flux of a polyamide membrane, thereby demonstrating the new grafting approach as a safe route for the surface modification of polyamide membranes. Besides, the PZ-grafted polyamide membrane exhibited excellent antifouling capability, which can be attributed to the shielding of specific binding sites on membrane surface, strong hydrophilic repulsion caused by local charge-induced hydration forces, and steric repulsion introduced by the brush-like flexible PZ chains. Therefore, this study opens a new avenue to surface modification of polyamide with different functional polymers and hence paves the way to a next generation of high-performance polyamide membranes. & 2013 Elsevier B.V. All rights reserved.

Keywords: Polyamide Click chemistry Nucleophilic substitution on nitrogen Polyzwitterion Surface modification Antifouling

1. Introduction Polyamide is the most widely used material to make semipermeable membranes for highly effective desalination and water purification and hence plays a critical role in addressing the global water and energy issues [1]. The polyamide membrane, however, has a high propensity of fouling [2–4], a long-standing problem caused by the accumulation of foreign substances on the polyamide surface. Depending on the specific type of foulants, membrane fouling can be categorized into colloidal fouling, organic fouling, scaling, and biofouling [2–4]. Membrane performance, such as water flux and effluent quality, can be severely deteriorated by fouling [5]. Because membrane surface characteristics significantly affect the process of foulant deposition and fouling layer formation [2,3], developing fouling-resistant polyamide

n

Corresponding author. Tel.: þ 1 301 405 1262; fax: þ 1 301 405 2585. E-mail address: [email protected] (B. Mi).

0376-7388/$ - see front matter & 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.memsci.2013.08.022

membranes by surface modification is key to more efficient use of membrane technology [5]. Click chemistry offers an excellent route for membrane surface modification and has previously been used for various membrane materials (e.g., polypropylene [6], polysulfone [7,8]). A salient advantage of click chemistry is its ability to provide superior site selectivity and almost quantitative transformation under mild conditions, with nearly no side reactions or by-products [7,9–14]. Instead of referring to a single specific reaction, click chemistry provides a modular synthetic approach for producing substances by joining small units together [8]. In particular, the click reaction between azide and alkyne has extremely high yield (usually above 95%), high tolerance of functional groups, and moderate temperature requirement (25–70 1C) [15]. The resulting aromatic 1,2,3triazole ring is very stable and thus makes the grafted functional chain firmly and covalently bonded to the substrate [7,9]. Therefore, the azide–alkyne click reaction holds great promise for costeffective, sustainable surface modification [7,10–13]. Use of click reaction for surface modification of polyamide membranes, however, has not been reported in the literature.

H.-Y. Yu et al. / Journal of Membrane Science 449 (2014) 50–57

Lack of such research may be attributed to the absence of suitable cationic center (e.g., carbocation) on the polyamide polymer chain where clickable functional groups (e.g., azide and alkyne) can be introduced. To circumvent this problem, we propose a novel approach to introduce clickable azide functional groups onto polyamide using SN2 nucleophilic substitution on nitrogen atom, a reaction that is particularly important in the presence of abundant amide functional groups in the polyamide polymer chain. Despite being widely employed as a means to introduce azide functional groups onto a carbocation [16,17], nucleophilic substitution reaction has focused only on the reaction of highly activated nitrogen compounds (e.g., arylsulfonamides [18–33]) but rarely been used on nitrogen. Moreover, most nucleophilic substitution reaction on nitrogen atom has been associated with harsh experimental conditions and ultra-low reaction temperature [34,35]. It is interesting to point out that the nitrogen atom in polyamide has neighboring CQO and benzene groups, which are electron-withdrawing groups and thus may facilitate the nucleophilic substitution reaction without any requirement on extreme temperature conditions. Therefore, surface modification of polyamide by nucleophilic substitution not only serves as a rare example of nucleophilic substitution on nitrogen but also provides an efficient method for modifying polyamide via click reaction. As a new antifouling material, polyzwitterions (PZs) have received growing attention over recent years due to their high hydration capability and hence excellent anti-adhesive property against protein fouling and bacterial attachment [36]. Typically, PZ contains both positively and negatively charged functional groups within the same segment side chain while it maintains overall charge neutrality [37]. Compared with polyethylene glycol (PEG, an extensively studied hydrophilic, antifouling material) [38], PZ is able to strongly bind to water molecules with localized charges. Such electrostatically induced hydration can greatly inhibit the attachment of protein or bacteria to a polymer surface due to the physical swelling of the polymer [39,40]. This promising antifouling property makes PZs an ideal coating material for surface modification. So far, PZ-based surface modification efforts, including surfaceinitiated atom transfer radical polymerization and plasma-induced interfacial polymerization, have mostly targeted ultrafiltration membranes [41–47]. Only a very few attempts have been made to modify polyamide membranes by creating a PZ thin film on membrane surface via chemical vapor deposition [47] or concentration-polarization-enhanced radical graft polymerization [48]. These existing approaches, however, require specialized setup that is very difficult to scale up and/or involves harsh chemical conditions that are hard to control and lead to troublesome side reactions and by-products [47,48]. Therefore, an efficient, controllable, and easy to scale-up approach for grafting PZs onto the polyamide membrane is highly desirable. In the present study, alkyne-terminated PZ as a representative zwitterionic polymer was grafted onto a thin-film composite polyamide membrane using nucleophilic substitution on nitrogen followed by click chemistry. First, alkyne-terminated PZ was synthesized via reversible addition-fragmentation chain-transfer (RAFT) radical polymerization. Azide functional groups were then introduced onto the polyamide membrane using a two-step procedure: (i) bromination of polyamide with sodium hypobromite solution under high-pH conditions; and (ii) replacement of the bromine groups by azide functional groups using NaN3 via nucleophilic substitution. Finally, PZ was grafted onto the polyamide membrane by performing azide–alkyne click reaction between the alkyne-terminated PZ and azide-functionalized polyamide. Various characterization techniques were employed to confirm the success of the PZ grafting on the polyamide membrane surface. The permeate flux and fouling performance of the

51

polyamide membrane before and after PZ grafting were tested in a forward osmosis (FO) membrane system.

2. Materials and methods 2.1. Materials Unless noted otherwise all chemicals used in this study were obtained from Sigma-Aldrich (St. Louis, MO) and used as received. The polyamide membrane was synthesized in our lab by interfacial polymerization on a polysulfone membrane support, following a procedure similar to the one used in Ref. [5]. Briefly speaking, the polysulfone support was synthesized by a phase inversion process on a polyester fabric. The polysulfone (molecular weight of 22,000) was obtained from Solvay Advanced Polymers (Alpharetta, GA) and the polyester fabric sheets (40 μm, grade 3249) from Ahlstrom (Helsinki, Finland). To synthesize the polyamide membrane, the polysulfone support was fixed in a plastic frame and subsequently placed in an aqueous1,3-phenylenediamine (MPD) solution (3.4 wt%), surfacedried, and soaked in a 1,3,5-benzenetricarbonyl trichloride (TMC) solution (0.15 wt% in Isopar-G). The membrane was then treated in DI at 95 1C, a sodium hypochlorite (200 ppm) aqueous solution, a sodium bisulfite (1000 ppm) aqueous solution, and again in DI at 95 1C. Finally, the synthesized polyamide membrane was rinsed thoroughly with DI and ready for use. 2.2. Nucleophilic substitution on nitrogen A sodium hypobromite (NaBrO) solution was prepared by adding bromine dropwise to 0.5 M NaOH solution at 0 1C in dark environment (by using aluminum foil to shield the reaction from light) until a pH of 9.2 was reached [49]. A polyamide membrane coupon (15 cm in diameter) was fixed in a plastic frame, with the active side of the membrane in contact with NaBrO solution for 1–10 min [50]. The polyamide-Br membrane thus obtained was then submerged in 10 g/L NaN3 solution. The azide substitution reaction was allowed to continue for 24 h at 20 1C, with the solution continuously stirred in a nitrogen gas atmosphere. 2.3. Synthesis of PZ by RAFT polymerization [2-(methacryloyloxy)ethyl]dimethyl-(3-sulfopropyl) ammonium hydroxide (MEDSAH) was used to synthesize the model PZ via RAFT radical polymerization. In the polymerization, azodiisobutyronitrile (AIBN) was used as an initiator, and alkyne-terminated S-1-Dodecyl-S′-(R,R′-dimethyl-R″-acetic acid)trithiocarbonate (alkyneSDDAT) as a chain transfer agent. To perform the reaction, 5 g MEDSAH, 0.2 g alkyne-SDDAT, and 50 mg AIBN were mixed in 100 mL ethanol in a single-necked flask. Nitrogen was then bubbled through the solution for 30 min to remove oxygen from the flask. Next, RAFT polymerization was performed at 55 1C for 24 h to synthesize alkyne-polyMEDSAH, followed by the cooling of reactor in ice water. The synthesized polymer was then collected by filtering the solution via a vacuum filter (Cole Parmer, Vernon Hills, IL), rinsing with ethanol, and subsequently drying overnight in oven at 65 1C. The alkyne-polyMEDSAH thus collected was stored in a desiccator prior to use in the grafting experiments. 2.4. Surface grafting by click reaction The polymer solution for the click reaction was prepared by dissolving 0.50 g alkyne-polyMEDSAH and 0.011 g CuSO4 in 50 mL methanol/water. Then, the azide-functionalized polyamide was submerged in the polymer solution in a single-necked flask and nitrogen was bubbled into the flask for 30 min to remove oxygen.

52

H.-Y. Yu et al. / Journal of Membrane Science 449 (2014) 50–57

Next, 0.024 g sodium ascorbate was added in the flask to initiate the click reaction. The flask was heated in an oil bath at 50 1C for 15 h to complete the click reaction. The PZ-grafted polyamide was taken out of the flask and thoroughly washed with DI water for 96 h, followed by overnight drying in oven at 65 1C. 2.5. Characterization of grafting density The grafting density (mmol/cm2) of PZ on polyamide is defined as Grafting density ¼

m2 m1  106 A  MW

ð1Þ

where m1 is the weight of virgin polyamide (g), m2 the weight of polyamide-PZ (g), A the surface area (cm2), and MW the molecular weight of alkyne-PZ (2500 g/mol, characterized by the gel permeation chromatography). Three membrane samples, each with a surface area of about 15.5 cm2, were analyzed in parallel for the grafting density analysis. The average weights of polyamide and polyamide-PZ membranes were 0.064 and 0.076 g, respectively.

2.7. Membrane performance tests The membrane flux and fouling performance were tested in a bench-scale, cross-flow FO membrane system [4]. This system comprises a custom-built cross-flow membrane cell, two variable speed gear pumps to form separate, closed loops for feed and draw solutions, respectively, a digital scale and computer to monitor permeate flux, and a water bath to maintain a constant system temperature at 20 71 1C. The membrane flux tests were conducted with a 4 M NaCl draw solution, a 20 mM NaCl feed solution, and a cross-flow rate of 8.5 cm/s. Aldrich humic acid (AHA) was used as model organic foulants to represent natural organic matters that are prevalent in natural waters. The experimental conditions for the fouling experiments were the same as for the flux tests, except that 200 mg/L AHA was added in the feed solution. Immediately following the fouling tests, membrane cleaning was performed using 20 mM NaCl cleaning solution at a cross-flow rate of 21 cm/s for 20 min. The membrane flux was tested after the cleaning experiment under the same conditions used for the flux tests.

2.6. Characterization of membrane surface properties 3. Results and discussion The effect of surface grafting on polyamide membrane morphology was studied using scanning electron microscopy (SEM). The water contact angles of the virgin and PZ-grafted polyamide membranes were analyzed using a sessile drop method with a Kruss G10 goniometer (Kruss, Matthews, NC). Changes in the chemical functionalities of polyamide after surface modification were characterized by Fourier transform infrared spectroscopy (FTIR) using a Thermo Nicolet Nexus 670 spectrometer with a Smart Golden Gate accessory and a diamond crystal window (Thermo Fisher, Madison, WI). Data collection and analysis was performed using the OMNIC software 6.2 provided with the instrument. X-ray photoelectron spectroscopy (XPS) analysis was used to characterize the elemental composition of the virgin polyamide and PZ-grafted polyamide.

3.1. Synthesis of azide-polyamide membrane by nucleophilic substitution on nitrogen As shown in Fig. 1, a two-step procedure was used to introduce clickable azide (N3) functional groups onto polyamide. In the first step, polyamide is activated by NaBrO via bromination to introduce Br groups onto the polyamide polymer chain [50]. It is expected that hypobromite quickly attacks the amide group in polyamide to produce an N–Br bond, which is used in the subsequent nucleophile substitution reaction. Note that N–Br bond may gradually rearrange to form a ring-brominated product via irreversible Orton rearrangement and thus eventually cause polyamide to degrade [51]. Such rearrangement, however, has slow

Fig. 1. Azide functionalization of polyamide (PA) membrane: Step 1 – bromination of PA to form N–Br, and Step 2 – SN2 nucleophilic substitution of Br on nitrogen atoms with azide (N3) groups following Route 1.

H.-Y. Yu et al. / Journal of Membrane Science 449 (2014) 50–57

reaction kinetics and hence can be inhibited by the nucleophilic substitution reaction right after bromination. In the second step, Br in N–Br is replaced by azide groups (in NaN3 solution) through SN2 nucleophilic substitution. It has been widely reported [16,17] that the nucleophilic substitution reaction can take place between C–Br and NaN3 to produce C–N3, where the carbon in C–Br is the cationic center, Br the leaving group, and NaN3 the nucleophilic reactant. The nitrogen cation (N–Br), however, is used as the cationic center in the present nucleophilic substitution, as Br has been primarily attached to the nitrogen in the amide groups during the polyamide bromination. The nucleophilic substitution reaction can effectively occur on nitrogen because the primary N in the amide group is linked to an electron-withdrawing group (CQO) and unsaturated carbon (benzene ring), greatly enhancing the reaction rate of SN2. Note that such nucleophilic substitution reaction most likely follows Route 1 instead of Route 2, as indicated in Fig. 1. This is

53

because Br in Route 2 would be more strongly bonded to the unsaturated benzene ring, making it difficult to replace Br by azide groups. 3.2. Synthesis of alkyne-PZ by RAFT polymerization In order to graft PZ onto polyamide by azide–alkyne click reaction, we synthesized an alkyne-terminated PZ (alkyne-PZ) by RAFT radical polymerization, a very useful method for operating on the largest range of vinyl monomers under diverse experimental conditions [11–13]. RAFT also allows the synthesis of polymers with predefined molecular weight, low polydispersity, and well-controlled composition and functionality under reaction conditions that are very similar to those of traditional free radical polymerization [7]. In this study, we used MEDSAH to synthesize the model PZ. As schematically shown in Fig. 2, MEDSAH contains a sulfonic group as the negatively charged center and a quaternary amine as the positively charged center. The two charged centers interact to maintain overall neutrality for the MEDSAH molecule as a whole. Note that the synthesized polyMEDSAH has a molecular weight of 2500, as characterized by gel permeation chromatography [6]. 3.3. Surface grafting of PZ onto polyamide membrane by azide–alkyne click reaction

Fig. 2. Synthesis of alkyne-PZ (i.e., polyMEDSAH) using RAFT polymerization.

Alkyne-PZ was grafted onto azide-polyamide by performing a click reaction that involved a CuI-catalyzed Huisgen 1,3-dipolar cycloaddition of azides and alkynes to form an aromatic 1,2,3triazole ring. The detailed reaction mechanism between alkyne-PZ and azide-polyamide is illustrated in Fig. 3. The satisfactory stability of the resulting aromatic triazole ring enabled the grafted PZ chain to be firmly and covalently bonded to the polyamide membrane. The average grafting density of PZ on the polyamide surface is plotted in Fig. 4 as a function of bromination time. It is seen that a relatively high grafting density was achieved within the first minute, indicating very fast bromination reaction. From 1 to 10 min, PZ on the polyamide surface in general increased slowly but steadily with the bromination time, suggesting that the introduction of more bromine atoms resulted in the grafting of more polymers on the polyamide surface. Therefore, it is expected that surface modification of polyamide can be well modulated by varying the experimental conditions in the bromination reaction. The grafting densities of Br and azide groups on polyamide were below the precision limit (0.0001 g) of the digital balance

Fig. 3. Click reaction between alkyne-PZ and the azide-polyamide (PA) membrane.

54

H.-Y. Yu et al. / Journal of Membrane Science 449 (2014) 50–57

(Denver Instrument, Bohemia, NY) used in the experiments and hence could not be measured. Therefore, it would be inconvenient to directly characterize the SN2 nucleophilic exchange of Br atoms (in polyamide-Br) with azide groups (in NaN3 solution). The grafting density of PZ, however, indirectly confirmed the effectiveness of azide functionalization of polyamide. This is because if the SN2 nucleophilic substitution had not occurred, the click reaction would not have taken place and accordingly the polyamide weight after click reaction would not have increased.

3.4. Surface properties of the PZ-grafted polyamide membrane The effect of surface grafting on polyamide morphology was studied using SEM. Comparison of SEM images of the virgin polyamide (Fig. 5(a)) and the PZ-grafted polyamide (Fig. 5(b)) demonstrates that PZ grafting dramatically changed the polyamide surface morphology. As shown in Fig. 5(c), the water contact angle decreases on average from 75 to 261, indicating that a PZ-grafted polyamide is much more hydrophilic than a virgin polyamide. Changes in the chemical functionalities of polyamide at different stages of surface modification were characterized by FTIR. Fig. 6 (a) shows that the virgin polyamide has signature absorption peaks at 1540 cm  1 corresponding to the N–H in-plane bending in secondary amines, at 1610 cm  1 associated with the hydrogen

Grafting Density (µmol/cm2)

0.5

0.4

0.3

0.2 0

2

4

6

8

10

12

Bromination Time (min) Fig. 4. Variation of the grafting density of PZ on polyamide membrane with bromination time.

bonding between CQO and N–H in the amide functional groups, and at 1667 cm  1 related to the CQO stretching in the amide group [51]. After polyamide bromination, peaks at both 1540 cm  1 and 1610 cm  1 are significantly reduced, indicating that bromination affected both N–H bonds and hydrogen bonds between CQO and N–H. In contrast, the peak at 1667 cm  1 remains almost the same after bromination, suggesting that CQO group was not attacked by hypobromite. These observations confirm that bromine was succesfully attached to the nitrogen atom in the polyamide chains, verifying the reaction mechanism proposed in Fig. 1. Subsequent replacement of Br by the azide functional group does not cause observable changes in the FTIR spectra. However, three new peaks as the characteristic absorption bands of PZ appear after the click coupling of alkyne-PZ to azide-polyamide: the SO3  vibration band at 1035 cm  1, the CQO stretching band at 1730 cm  1, and the hydrogen bond at 3450 cm  1, indicating that PZ was indeed successfully grafted onto polyamide [52]. To assess the effectiveness of PZ grafting, XPS analysis was used to characterize the elemental composition of the virgin polyamide and PZ-grafted polyamide. As shown in Fig. 6(b), the virgin polyamide has peaks for C1s, N1s, and O1s. In contrast, polyamide after bromination exhibits additional peaks for Br3d, Br3p, and Br3s, all being the characteristics of covalently bonded bromine [53]. As listed in Fig. 6(c), the bromine concentration in the brominated polyamide is 4.2%, confirming successful polyamide bromination. Furthermore, after SN2 nucleophilic substitution reaction, the bromine concentration decreases from 4.2% to 3.2%, indicating that about 25% of bromine was replaced by azide functional groups. Note that the introduction of nitrogen atoms during the azide functionalization of polyamide only slightly changed the nitrogen concentration from 15% to 14%, a difference within the experimental error of an XPS analysis. The bromination of polyamide also led to the occurrence of a weak S2p peak, possibly a consequence of damage to polyamide due to a combined action of bromination and vigorous washing during the PZ grafting process [54]. Such damage can probably be lessened, if not completely avoidable, by using milder bromination and/or adjusting other experimental conditions (e.g., NaBrO solution with lower NaOH concentration, shorter bromination time). Nevertheless, we emphasize that treatment of polyamide with sodium hypobromite could potentially lead to polyamide degradation unless careful control of experimental conditions is exercised. Therefore, future grafting experiments may be advantageously performed with chlorination or with shorter bromination time and under milder conditions.

Fig. 5. Polyamide (PA) membrane surface morphology (a) before and (b) after PZ grafting. (c) Changes in membrane hydrophilicity..

H.-Y. Yu et al. / Journal of Membrane Science 449 (2014) 50–57

1000

1500

3000

Wavenumbers

3500

4000

0

100

(cm-1)

Grafting stage

55

200 300 400 Binding Energy (eV)

Elemental concentration (%)

500

Atomic ratio

C

N

O

S

Br

N/C

O/C

S/N

O/N

Virgin PA

73.1

10.2

14.8

0.0

0.0

0.17

0.17

0.0

1.0

0.14

0.20

-

1.4

PA-Br

70.4

10.4

14.6

0.4

4.2

0.15

0.20

-

-

PA-N3

70.6

9.8

15.5

0.8

3.2

0.14

0.22

-

-

PA-PZ

65.1

3.8

22.0

3.9

1.1

0.06

0.34

1.0

5.8

0.09

0.45

1.0

5.0

Theoretical PA*

Theoretical PZ*

* Atomic ratio was calculated per the theoretical molecular formula. Fig. 6. Characterization of polyamide (PA) at different grafting stages. (a) FTIR spectra, (b) XPS spectra, and (c) elemental composition.

3.5. Membrane performance To examine whether the PZ grafting procedure causes any damage to the polyamide membrane, permeate fluxes of the membrane before and after surface modification were tested in an FO membrane system. The virgin polyamide membrane synthesized in the lab had a permeate flux of 570.4 mm/s with 4 M NaCl draw and 20 mM NaCl feed solution. The permeate flux of the PZ-grafted polyamide membranes with different durations of bromination time were tested under the same conditions as those for the virgin polyamide membrane. Plotted in Fig. 7 are the flux data normalized by the virgin polyamide membrane flux. Overall, a slightly decreasing trend with increasing bromination time is observed. This is because longer bromination time increased the grafting degree (Fig. 4) and hence the density of PZ on the polyamide membrane, thereby enhancing its hydraulic resistance. Note that the PZ-grafted polyamide membrane with very short bromination time (2 min) had a slightly higher flux than the virgin polyamide membrane. This difference is possibly attributed to the enhanced membrane hydrophilicity introduced by the PZ grafting and also in part to the slight breaking of polyamide polymer chains during

1.50 Normalized Water Flux

After PZ was grafted onto polyamide through click reaction, the nitrogen concentration decreased while the concentration of both oxygen and sulfur increased significantly. These changes are expected because, compared with the virgin polyamide, PZ has lower nitrogen concentration and higher oxygen and sulfur contents (Fig. 6(c)). Note that an aromatic polyamide has a theoretical formula of C18O3N3 and the atomic ratio is calculated thereof. The theoretical atomic ratio of PZ was calculated based on the zwitterionic structure used (i.e., MEDSAH in Fig. 2). As listed in Fig. 6(c), the atomic ratios of the PZ-grafted polyamide are very close to those of PZ, indicating that the click reaction led to an almost full polyamide surface coverage by PZ.

1.25 1.00 0.75 0.50 0.25 0.00

PA 2 8 10 4 6 Bromination Time (min)

Fig. 7. Effect of PZ grafting on membrane flux as a function of bromination time. The membrane flux data are normalized by the virgin polyamide (PA) membrane flux.

the bromination process, the latter factor being minimizable by the optimization of bromination conditions. Nevertheless, the PZ grafting did not significantly affect the polyamide membrane flux. It was further found that the NaCl rejection of the membrane was not considerably affected by the PZ grafting. Therefore, the proposed grafting approach provides a safe route to the surface modification of polyamide membranes. The effect of PZ grafting on membrane fouling resistance was tested in the FO membrane system using AHA as a model foulant. The flux declines for the virgin and PZ-grafted polyamide membranes during the AHA fouling experiments are plotted in Fig. 8(a). The results clearly demonstrate that the flux decline for the PZ-grafted polyamide membrane is almost negligible, while an obvious flux decline is observed for the virgin polyamide membrane. At the end of the fouling experiment, a cleaning procedure was conducted to examine if

Normalized Water Flux

H.-Y. Yu et al. / Journal of Membrane Science 449 (2014) 50–57

Normalized Water Flux

56

1.0 0.8 0.6

PA-PZ PA

0.4 0.2 0.0

0

100

200

300

400

500

1.0 0.8 0.6 0.4 0.2 0.0 PA

PA-PZ

Time (min)

Fig. 8. Effects of PZ grafting on the performance of polyamide (PA) membranes in a forward osmosis system. (a) Flux decline in AHA fouling experiments. The flux data are normalized by the initial membrane flux and corrected by a baseline flux using a procedure described in Refs. [2,55]. (b) Flux recovery after membrane cleaning. The flux data are normalized by the membrane flux prior to the fouling experiments. (c) Schematic illustration of the antifouling mechanisms of the PZ-grafted membrane.

the fouling was reversible by rinsing. As shown in Fig. 8(b), the water flux of the cleaned virgin polyamide membrane is only about 77% of its original flux, indicating AHA fouling on the virgin polyamide membrane was almost irreversible under the tested conditions. In contrast, the water flux of the PZ-grafted polyamide membrane after cleaning is comparable to its original flux prior to fouling, indicating there is no irreversible fouling taking place on the PZ-grafted polyamide membrane. Therefore, the PZ grafting greatly enhances the fouling resistance of a polyamide membrane by minimizing irreversible fouling. As illustrated in Fig. 8(c), the excellent antifouling properties of the PZ-grafted polyamide membrane can be attributed to three major effects introduced by PZ: (1) the reduction of specific bindings and charge interactions between foulants and membrane by shielding the binding sites and charged functional groups (e.g., carboxylate functional groups) on the polyamide membrane, (2) the strong hydration capability of the localized charges on PZ to create large hydrophilic repulsion forces between foulants and membrane, and (3) the steric repulsions introduced by the long, flexible, brush-like PZ chains. 4. Conclusions We have demonstrated that the click chemistry and nucleophilic substitution reaction can be successfully integrated to graft PZ onto an aromatic polyamide membrane. Key steps include synthesis of an alkyne-PZ via RAFT polymerization, functionalization of polyamide with azide functional groups by bromination and subsequently SN2 nucleophilic substitution of Br with azide, and finally grafting of alkyne-PZ onto azide-polyamide by an azide–alkyne cycloaddition click reaction. Characterization data have shown that the PZ-grafted polyamide was much more hydrophilic than the virgin polyamide, and that almost full surface coverage by PZ was achieved under studied experimental conditions. Flux tests have verified that the polyamide membrane did not experience significant flux change after PZ grafting, suggesting

that the present grafting approach provides a safe route to the surface modification of polyamide membranes. Using AHA as a model foulant, we have also shown that the PZ-grafting improved the antifouling performance of the polyamide membrane, most likely due to a combined effect of reduced specific binding, strong hydrophilic repulsion, and steric repulsions. The present study represents a novel route of using click chemistry to graft PZ onto polyamide, and proves that SN2 nucleophilic substitution on nitrogen can occur on the polyamide polymer chain under mild reaction conditions. Therefore, the proposed grafting method opens a new avenue to surface modification of the polyamide membrane with different functional polymers and hence paves the way to next-generation, highly fouling-resistant polyamide membranes for sustainable desalination and water purification. Acknowledgments This material is based upon work supported by the National Science Foundation under Grant nos. CBET-1154572 and CBET1158601. These supports are gratefully acknowledged. We also acknowledge the facility support of the Maryland NanoCenter and its NispLab. The NispLab is supported in part by the NSF as a Materials Research Science and Engineering Center shared experimental facility. The opinions expressed herein are those of the authors and do not necessarily reflect those of the sponsors. The work was done while the first author was a postdoctoral research associate at the University of Maryland at College Park.

References [1] M.A. Shannon, P.W. Bohn, M. Elimelech, J.G. Georgiadis, B.J. Marinas, A. M. Mayes, Science and technology for water purification in the coming decades, Nature 452 (2008) 301–310.

H.-Y. Yu et al. / Journal of Membrane Science 449 (2014) 50–57

[2] B. Mi, M. Elimelech, Organic fouling of forward osmosis membranes: fouling reversibility and cleaning without chemical reagents, J. Membr. Sci. 348 (2010) 337–345. [3] B. Mi, M. Elimelech, Gypsum scaling and cleaning in forward osmosis: measurements and mechanisms, Environ. Sci. Technol. 44 (2010) 2022–2028. [4] Y. Liu, B. Mi, Combined fouling of forward osmosis membranes: synergistic foulant interaction and direct observation of fouling layer formation, J. Membr. Sci. 407 (2012) 136–144. [5] N.Y. Yip, A. Tiraferri, W.A. Phillip, J.D. Schiffman, M. Elimelech, High performance thin-film composite forward osmosis membrane, Environ. Sci. Technol. 44 (2010) 3812–3818. [6] X.-M. Wu, L.-L. Wang, Y. Wang, J.-S. Gu, H.-Y. Yu, Surface modification of polypropylene macroporous membrane by marrying RAFT polymerization with click chemistry, J. Membr. Sci. 421–422 (2012) 60–68. [7] R. Ranjan, W.J. Brittain, Combination of living radical polymerization and click chemistry for surface modification, Macromolecules 40 (2007) 6217–6223. [8] J.E. Moses, A.D. Moorhouse, The growing applications of click chemistry, Chem. Soc. Rev. 36 (2007) 1249–1262. [9] C.J. Huang, F.C. Chang, Using click chemistry to fabricate ultrathin thermoresponsive microcapsules through direct covalent layer-by-layer assembly, Macromolecules 42 (2009) 5155–5166. [10] J.F. Lutz, H.G. Borner, K. Weichenhan, Combining atom transfer radical polymerization and click chemistry: a versatile method for the preparation of end-functional polymers, Macromol. Rap. Commun. 26 (2005) 514–518. [11] T. Zhang, Z. Zheng, X. Ding, Y. Peng, Smart surface of gold nanoparticles fabricated by combination of RAFT and click chemistry, Macromol. Rap. Commun. 29 (2008) 1716–1720. [12] D.X. Wu, X.H. Song, T. Tang, H.Y. Zhao, Macromolecular brushes synthesized by grafting from approach based on click chemistry and RAFT polymerization, J. Polym. Sci. Part A Polym. Chem. 48 (2010) 443–453. [13] G. Gody, C. Rossner, J. Moraes, P. Vana, T. Maschmeyer, S. Perrier, One-pot RAFT/click chemistry via isocyanates: efficient synthesis of α-endfunctionalized polymers, J. Am. Chem. Soc. 134 (2012) 12596–12603. [14] H.-L. Jiang, D. Feng, T.-F. Liu, J.-R. Li, H.-C. Zhou, Pore surface engineering with controlled loadings of functional groups via click chemistry in highly stable metal-organic frameworks, J. Am. Chem. Soc. 134 (2012) 14690–14693. [15] W.H. Binder, C. Kluger, Azide/alkyne-click reactions: applications in material science and organic synthesis, Curr. Org. Chem. 10 (2006) 1791–1815. [16] J.A. Johnson, D.R. Lewis, D.D. Díaz, M.G. Finn, J.T. Koberstein, N.J. Turro, Synthesis of degradable model networks via ATRP and click chemistry, J. Am. Chem. Soc. 128 (2006) 6564–6565. [17] N.V. Tsarevsky, S.A. Bencherif, K. Matyjaszewski, Graft copolymers by a combination of ATRP and two different consecutive click reactions, Macromolecules 40 (2007) 4439–4445. [18] S. Yoshida, K. Igawa, K. Tomooka, Nucleophilic substitution reaction at the nitrogen of arylsulfonamides with phosphide anion, J. Am. Chem. Soc. 134 (2012) 19358–19361. [19] S. Hirota, T. Sakai, N. Kitamura, K. Kubokawa, N. Kutsumura, T. Otani, T. Saito, Synthesis of nitrogen heterocycle-fused 1,2,4-benzothiadiazine-1,1-dioxide, quinazolinone, and pyrrolidinone derivatives with a guanidine joint via sequential aza-Wittig reaction/intramolecular NH-addition cyclization/ nucleophilic substitution ring closure methodology, using functionalyzed carbodiimides as key intermediates, Tetrahedron 66 (2010) 653–662. [20] T. Hayashi, W. Peng, Y.Y. Nakai, K. Yamada, M. Somei, Nucleophilic substitution reaction on the nitrogen of indole nucleus: a novel synthesis of 1aryltryptamines, Heterocycles 57 (2002) 421–424. [21] M. Somei, F. Yamada, T. Hayashi, A. Goto, Y. Saga, Nucleophilic substitution reaction on the nitrogen of indole nucleus: formation of 1-(indol-3-yl)indoles upon reaction of 1-hydroxyindoles with indole in formic acid, Heterocycles 55 (2001) 457–460. [22] R. Gareyev, S. Kato, V.M. Bierbaum, Gas phase reactions of NH2Cl with anionic nucleophiles: nucleophilic substitution at neutral nitrogen, J. Am. Soc. Mass Spectrom. 12 (2001) 139–143. [23] I. Bitter, G. Tóth, I. Hermecz, Z. Mészŕos, Nitrogen bridgehead compounds part 59. Nucleophilic substitution reactions of 9-bromo-6,7,8,9-tetrahydro-4Hpyrido [1,2-a] pyrimidin-4-ones, Heterocycles 26 (1987) 869–873. [24] R. Roy, M. Chaudhury, S.K. Mondal, K. Nag, Metal complexes of sulphur– nitrogen chelating agents. Part 11. Synthesis, characterisation, and thermodynamics of nucleophilic substitution reactions of monohalogeno nickel(II) complexes of tridentate ligands of the type SNN in solution and the solid state, J. Chem. Soc.: Dalton Trans. (1984) 1681–1686. [25] D.R. Carver, T.D. Greenwood, J.S. Hubbard, A.P. Komin, Y.P. Sachdeva, J.F. Wolfe, SRN1 mechanism in heteroaromatic nucleophilic substitution. Reactions involving certain dihalogenated π-deficient nitrogen heterocycles, J. Org. Chem. 48 (1983) 1180–1185. [26] R.D. Chambers, P.A. Martin, J.S. Waterhouse, D.L.H. Williams, B. Anderson, Mechanisms for reactions of halogenated compounds. Part 4.[1] activating influences of ring-nitrogen and trifluoromethyl in nucleophilic aromatic substitution, J. Fluorine Chem. 20 (1982) 507–514. [27] J.E. Johnson, E.A. Nalley, C. Weidig, M. Arfan, Mechanisms of alkoxide substitution reactions at the carbon-nitrogen double bond. Stereoelectronic control during nucleophilic substitution, J. Org. Chem. 46 (1981) 3623–3629. [28] B.A. Sudbury, J.H. Krueger, Nucleophilic substitution on nitrogen. Kinetics of reactions of hydroxylamine-O-sulfonic acid in dimethyl sulfoxide–water solvents, Inorg. Chem. 13 (1974) 1736–1741.

57

[29] P.F. Blanchet, J.H. Krueger, Nucleophilic substitution on nitrogen. Reactions of hydroxylamine-O-sulfonate with thiosulfate and thiourea, Inorg. Chem. 13 (1974) 719–723. [30] S. Oae, F. Yamamoto, Kinetic study of nucleophilic substitution reaction on nitrogen atom, Tetrahedron Lett. 14 (1973) 5143–5146. [31] J.H. Krueger, P.F. Blanchet, A.P. Lee, B.A. Sudbury, Nucleophilic substitution on nitrogen. Kinetics of reactions of hydroxylamine-O-sulfonate ion, Inorg. Chem. 12 (1973) 2714–2718. [32] G. Ayrey, W.A. Wylie, A nitrogen isotope effect study of an aromatic nucleophilic substitution reaction, J. Chem. Soc. B: Phys. Org. (1970) 738–739. [33] E.A. Mistryukov, N.I. Aronova, V.F. Kucherov, Stereochmeistry of heterocyclic compounds – Communication 11. Effect of substitution on the nitrogen on the configuration of octahydro-4(1H)-quinolones, and the stereochemistry of some nucleophilic reactions at the keto group, Bull. Acad. Sci. USSR Div. Chem. Sci. 11 (1963) 1514–1518. [34] N. Matsuda, K. Hirano, T. Satoh, M. Miura, Copper-catalyzed amination of arylboronates with N,N-dialkylhydroxylamines, Angew. Chem. Int. Ed. 51 (2012) 3642–3645. [35] T. Hatakeyama, Y. Yoshimoto, S.K. Ghorai, M. Nakamura, Transition-metal-free electrophilic amination between aryl grignard reagents and N-chloroamines, Org. Lett. 12 (2010) 1516–1519. [36] R.E. Holmlin, X. Chen, R.G. Chapman, S. Takayama, G.M. Whitesides, Zwitterionic SAMs that resist nonspecific adsorption of protein from aqueous buffer, Langmuir 17 (2001) 2841–2850. [37] J.S. Liu, Y. Zhan, T.W. Xu, G.Q. Shao, Preparation and characterizations of novel zwitterionic membranes, J. Membr. Sci. 325 (2008) 495–502. [38] A. Sagle, E. Van Wagner, H. Ju, B. McCloskey, B. Freeman, M. Sharma, PEGcoated reverse osmosis membranes: desalination properties and fouling resistance, J. Membr. Sci. 340 (2009) 92–108. [39] I. Eshet, V. Freger, R. Kasher, M. Herzberg, J. Lei, M. Ulbricht, Chemical and physical factors in design of antibiofouling polymer coatings, Biomacromolecules 12 (2011) 2681–2685. [40] M.Y. Zhou, H.W. Liu, J.E. Kilduff, R. Langer, D.G. Anderson, G. Belfort, Highthroughput membrane surface modification to control NOM fouling, Environ. Sci. Technol. 43 (2009) 3865–3871. [41] Q. Shi, Y.L. Su, W. Zhao, C. Li, Y.H. Hu, Z.Y. Jiang, S.P. Zhu, Zwitterionic polyethersulfone ultrafiltration membrane with superior antifouling property, J. Membr. Sci. 319 (2008) 271–278. [42] M. Okabe, G. Szakacs, M.A. Reimers, T. Suzuki, M.D. Hall, T. Abe, J.N. Weinstein, M.M. Gottesman, Profiling SLCO and SLC22 genes in the NCI-60 cancer cell lines to identify drug uptake transporters, Mol. Cancer Ther. 7 (2008) 3081–3091. [43] L.J. Wang, Y.L. Su, L.L. Zheng, W.J. Chen, Z.Y. Jiang, Highly efficient antifouling ultrafiltration membranes incorporating zwitterionic poly(3-(methacryloylamino)propyl-dimethyl(3-sulfopropyl) ammonium hydroxide), J. Membr. Sci. 340 (2009) 164–170. [44] Y.F. Yang, Y. Li, Q.L. Li, L.S. Wan, Z.K. Xu, Surface hydrophilization of microporous polypropylene membrane by grafting zwitterionic polymer for antibiofouling, J. Membr. Sci. 362 (2010) 255–264. [45] H.J. Yu, Y.M. Cao, G.D. Kang, J.H. Liu, M. Li, Q. Yuan, Enhancing antifouling property of polysulfone ultrafiltration membrane by grafting zwitterionic copolymer via UV-initiated polymerization, J. Membr. Sci. 342 (2009) 6–13. [46] J. Zhao, Q.A. Shi, S.F. Luan, L.J. Song, H.W. Yang, H.C. Shi, J. Jin, X.L. Li, J.H. Yin, P. Stagnaro, Improved biocompatibility and antifouling property of polypropylene non-woven fabric membrane by surface grafting zwitterionic polymer, J. Membr. Sci. 369 (2011) 5–12. [47] R. Yang, J.J. Xu, G. Ozaydin-Ince, S.Y. Wong, K.K. Gleason, Surface-tethered zwitterionic ultrathin antifouling coatings on reverse osmosis membranes by initiated chemical vapor deposition, Chem. Mater. 23 (2011) 1263–1272. [48] R. Bernstein, S. Belfer, V. Freger, Bacterial attachment to RO membranes surface-modified by concentration-polarization-enhanced graft polymerization, Environ. Sci. Technol. 45 (2011) 5973–5980. [49] J. Tang, N. He, L. Nie, P. Xiao, H. Chen, Hydrolysis of microporous polyamide-6 membranes as substrate for in situ synthesis of oligonucleotides, Surf. Sci. 550 (2004) 26–34. [50] C.W. Carr, H.P. Gregor, K. Sollner, The structure of the collodionmembrane and its electrical behavior 12: the preparation and properties of "megapermselective" protamine collodion membranes combining high ionic selectivity with high permeability. J. Gen. Physiol. 28 (1945) 179–185. [51] G.-D. Kang, C.-J. Gao, W.-D. Chen, X.-M. Jie, Y.-M. Cao, Q. Yuan, Study on hypochlorite degradation of aromatic polyamide reverse osmosis membrane, J. Membr. Sci. 300 (2007) 165–171. [52] M.G. Santonicola, M. Memesa, A. Meszyńska, Y. Ma, G.J. Vancso, Surfacegrafted zwitterionic polymers as platforms for functional supported phospholipid membranes, Soft Matter 8 (2012) 1556–1562. [53] F.J. Xu, J.P. Zhao, E.T. Kang, K.G. Neoh, J. Li, Functionalization of nylon membranes via surface-initiated atom-transfer radical polymerization, Langmuir 23 (2007) 8585–8592. [54] Z.B. Zhang, X.L. Zhu, F.J. Xu, K.G. Neoh, E.T. Kang, Temperature- and pHsensitive nylon membranes prepared via consecutive surface-initiated atom transfer radical graft polymerizations, J. Membr. Sci. 342 (2009) 300–306. [55] B. Mi, M. Elimelech, Chemical and physical aspects of organic fouling of forward osmosis membranes, J. Membr. Sci. 320 (2008) 292–302.