Water Remediation Aided by a Graphene-Oxide

Research Article pubs.acs.org/journal/ascecg

Cite This: ACS Sustainable Chem. Eng. XXXX, XXX, XXX−XXX

Water Remediation Aided by a Graphene-Oxide-Anchored Metal Organic Framework through Pore- and Charge-Based Sieving of Ions Paresh Kumar Samantaray,† Giridhar Madras,‡ and Suryasarathi Bose*,§ †

Centre for BioSystems Science and Engineering, ‡Department of Chemical Engineering, and §Department of Materials Engineering, Indian Institute of Science, Bangalore, Karnataka 560012, India

Downloaded via INDIAN INST OF SCIENCE on December 11, 2018 at 14:12:05 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

S Supporting Information *

ABSTRACT: Herein, a unique reversible addition−fragmentation chain transfer (RAFT)-synthesized antibacterial copolymer was designed to target key requirements such as stringent and quick response toward bacteria and quick reversible response toward fouling using a multilayered assembly. In order to render the membrane assembly selective toward ions, a unique phosphonium-conjugated graphene oxide (P+GO)-anchored copper- and trimesic-acid-based metal organic framework (CuMOF) was sandwiched between the RAFT-synthesized polymer and a commercial reverse osmosis (RO) support. The sandwich architecture exhibited excellent antibacterial properties for both Gram-positive and Gram-negative bacterial cells. The membranes also retained an unimpeded flow of water even after longer continuous runs. The engineered active layer was excellent in rendering reversible antifouling against bovine serum albumin with 98.8% flux retention. The nanoexclusions/ channels offered by the P+GO-anchored CuMOF, sandwiched as an interlayer, though reduced the flux as compared to the support RO but manifested in an exemplary 99.9% salt removal for a formulation composed of monovalent and divalent ions through synergistic charge- and pore-based sieving. This multilayered assembly is bactericidal, is resistant to scaling unlike the base RO support, and shows excellent ion-sieving characteristics that makes it a potential candidate in water remediation. KEYWORDS: Sustainable multilayered membranes, RAFT polymerization, P+GO-anchored CuMOF, bacterial reduction and salt removal

■

INTRODUCTION With declining fresh water aquifers, the quest for providing access to safe and clean drinking water is the nexus of the 21st century.1 Since the 1950s, the consumption of water has astoundingly increased by approximately 300% globally.2 This is triggered mostly due to rising population, overburden on agriculture, and rapid industrialization. Conclusively, these factors contributed to the overexploitation and contamination of freshwater aquifers and the decline of the groundwater table.3 As nearly 71% of the surface of the Earth is covered with water and about 96.5% is oceans, harvesting seawater for commercial purpose in a low and sustainable way can only solve the water issue. Rapid technological advances took place in the field of membrane technology because the conventional methods were either inefficient (e.g., stand-alone filtration, boiling, and sieving) and/or required chemical treatment (e.g., ammonia treatment, chlorination, permanganate treatment, © XXXX American Chemical Society

coagulation aids such as alums, etc.) or involved multiple prefilters (large-scale reverse osmosis (RO) plants) that require high capital investment and time.4,5 Even if the membrane-based separation is facile and effective, it loses its performance due to many reasons with the most pronounced being fouling.6−9 The natural source of water is abundant with proteins and bacteria that generally adsorb on the membrane active layer and clog the pores. Further, the bacteria tend to multiply and form biofilms on the surfaces.10 This leads to a high pumping cost and subsequently membrane failure.11−14 This can be prevented by tailoring the membrane surface to reduce bacterial adhesion, arrest bacterial reproduction, and reverse the effect of protein-mediated Received: October 17, 2018 Revised: November 13, 2018 Published: November 27, 2018 A

DOI: 10.1021/acssuschemeng.8b05354 ACS Sustainable Chem. Eng. XXXX, XXX, XXX−XXX

Research Article

ACS Sustainable Chemistry & Engineering Scheme 1. (a) Antibacterial Copolymer Synthesis and (b) Schematics of Membrane Preparation

fouling.15 It has been reported in the literature that increased hydrophilicity or specific charges on the active surface of the membranes discourages bacterial attachment and promotes high flux retention after fouling attack.16−22 The surfaces can be tailored by either photochemical or chemical irradiation, redox modification, plasma treatment, or polymerization onto the membrane surface.23−27 It has been reported in the literature that polymers and macromolecules with a net positive charge are highly potent to exhibit a bactericidal response.28−30 In the recent literature, multilayered membranes with different interlayers show higher efficiency and excellent performance for antibacterial, antifouling, and desalination applications.5,31−36 Graphene derivatives have offered some promise in rendering faster transport of water and can be used in a wider range of operating conditions,37 yet it has a tendency to swell under the influence of hydration (i.e.,

absorbed water) and exhibit a dilated d spacing posthydration.38 This retards the efficiency of ion rejection in graphene oxide (GO)-based membranes. If the interlayer d spacing is stitched with some water-stable architectures such as metal organic frameworks (MOFs), the efficiency in terms of ion rejection and stability can be enhanced. To address this challenge, we stitched GO sheets using phosphonium moieties (P+GO) and anchored them onto a copper-based metal organic framework (CuMOF) to sieve ions. It is envisaged that the nanochannels offered by the GO and MOF can sieve ions based on their pore sizes and the P+ moieties can sieve the ions through either repulsion or via electrostatic interactions. The P+GO-anchored CuMOF has improved water stability and is sandwiched here between a reversible addition−fragmentation chain transfer (RAFT)synthesized copolymer, which offers excellent antibacterial B


Research Article

ACS Sustainable Chemistry & Engineering

mixture. A 0.5 g amount of trimesic acid (BTC) was dissolved in 40 mL of ethanol. To the P+GO-retained suspension, the copper acetate mixture and trimesic acid solution were added dropwise, and the entire content was transferred into a round-bottom flask. This mixture was refluxed under N2 atmosphere for 6 h under mild stirring after which the contents were vacuum filtered and washed several times with a 1:1 (v/v) ethanol−water mixture and dried. Multilayer Membrane Preparation. Two sets of dope solution were prepared by dissolving 0.8 g of the RAFT-synthesized copolymer in 4 mL of DMF. One solution was casted using an automatic film applicator with a thickness of 100 μm. The membrane was then dried overnight. A 200 mg amount of the P+GO-anchored CuMOF was dispersed in 100 mL of acetone using bath sonication for 30 min and vacuum filtered on the synthesized membrane surface. The second dope solution was used to adhere the vacuum-filtered membrane and the RO membrane. This was achieved by coating the dope solution on the RO membrane and inverting the vacuum-filtered membrane onto it. This assembly was allowed to dry at 60 °C for 6 h and then overnight under vacuum. Scheme1b illustrates the preparation of the multilayer membrane. Characterization of Interlayer and Multilayered Membranes. The interlayer sorbent was characterized using Fourier transform IR (FTIR) spectroscopy on a PerkinElmer Frontier in the mid-IR range, 31P NMR spectroscopy using a JEOL ECX 500 highresolution NMR spectrometer at 500 MHz, X-ray diffraction (XRD) using an X-Pert PRO PANalytical having a copper target, scanning/ transmission electron microscopy (S/TEM) using a TEM-Titan Themis, and scanning electron microscopy (SEM) using an Ultra55 FE-SEM Karl Zeiss scanning electron microscope. The antibacterial copolymer was characterized using FTIR spectroscopy and 1H and 31 P NMR spectroscopy using a JEOL ECX 500 high-resolution NMR spectrometer at 500 MHz. The number-average molecular weight (Mn), polydispersity index (PDI), and weight-average molecular weight (Mw) were determined by GPC using a THF-based column and analyzed using Empower 3 software. The melting and crystallization temperatures were determined by differential scanning calorimetry (DSC) using a Q2000 TA Instruments at a heating and cooling rate of 10 °C/min. The surface and cross-section morphologies of the multilayered membrane were characterized by SEM using an Ultra55 FE-SEM Karl Zeiss scanning electron microscope. Membrane Uptake. Membrane uptake was assessed as per our previous reports.14,16 In brief, 10 multilayered membranes of equal size and area (0.63 cm2) were kept under vacuum overnight and weighed. The membranes were then immersed in double-distilled water for 24 h and weighed. The % uptake was calculated by

and antifouling properties, and a commercial RO support. The latter has a tendency to foul and is often not chlorine tolerant. This strategy of sandwiching the tailored interlayer between an antibacterial active layer and a support RO layer will offer some promise toward mitigating bacterial colonies and protein fouling, sieving ions, and rendering sustained permeation of pure water. This strategy also facilitated in controlling the swelling of the GO sheets and resulted in effective water remediation.

■

EXPERIMENTAL SECTION

Materials. Trihexyltetradecylphosphonium chloride (TPCl), triphenylphosphine, 4-vinylbenzyl chloride, polyethylenimine branched (∼25 000 Mw), azobis(isobutyronitrile) (AIBN), acrylic acid, 2-cyano-2-propyl dodecyl trithiocarbonate (RAFT reagent), silica gel (60−120 mesh size), copper(II) acetate (99.99%), 2-methyl2-propanol, Luria−Bertani (LB) broth, dithizone, (3-chloropropyl) triethoxysilane, magnesium chloride (≥98%), calcium nitrate tetrahydrate (≥99%), and sodium chloride (≥99) were procured from Sigma-Aldrich. (1-Ethyl-3-(3-(dimethylamino)propyl) carbodiimide hydrochloride (EDC) and N-hydroxysuccinimide (NHS) were procured from Thermo Scientific. GO was procured from BT Corp. Nutrient agar was purchased from Fluka. Methanol, toluene, ethanol, N, N′-Dimethylformamide (DMF), and hydrochloric acid (35.4%) were obtained from SD Fine Chemicals Limited. The RO membrane with application pressure of 72−120 psi was procured from K-Flow RO for this study. All the reagents used in the synthesis are analytical grade and are used directly without any further purification except AIBN that was purified and recrystallized using methanol. 4Vinylbenzyl chloride and acrylic acid were purified by passing through a column filled with silica gel and stored in a desiccator for synthesis. Polymer Synthesis. RAFT polymerization was used for this synthesis. The precursor monomer was synthesized using an earlier report.28 Precisely, 2 mL of 4-vinylbenzyl chloride and 5.26 g of triphenylphosphine were dissolved in 50 mL of toluene and refluxed for 4 h. The obtained mixture was then centrifuged, washed with toluene, and dried under vacuum. This monomer was then recrystallized in 2-methyl-2-propanol and used for synthesis. A 3 g amount of monomer and 2 mL of acrylic acid were taken in a 3:2 molar ratio in 1 mL of methanol in a glass vessel under nitrogen atmosphere. To this, 4 mg of the RAFT reagent and 4 mg of AIBN in 2 mL of methanol were injected, and subsequently, three freeze− pump−thaw cycles were carried out. Then, the polymerization was carried out for 48 h at 60 °C. An aliquot from the same was extracted with a syringe, washed with ethanol, and dried for gel permeation chromatography (GPC). After that, 2 g of polyethylenimine was dissolved in 2 mL of methanol, and to it, 3 mmol each of EDC and NHS were added; the pH of the solution was normalized to 4.5 by adding a few drops of HCl. This mixture was injected into the vessel, and the reaction was carried out further for 4 h at room temperature. The obtained product was washed with methanol and dried under vacuum. Scheme1a illustrates the polymerization schematics. P+GO-Anchored MOF Multilayered Preparation. GO was modified into phosphonium-conjugated GO (P+GO) using our previous report.16 Precisely, 200 mg of GO was dispersed in a 100 mL solution of water and ethanol (1:4 ratio of water and ethanol) using bath sonication for 30 min. To this solution, 1 g of trihexyltetradecylphosphonium chloride was added. The reaction was mechanically stirred for 24 h at 50 °C. The GO obtained was then washed with 1:1 ethanol−water solution to remove the unreacted compound, centrifuged, and vacuum filtered to obtain phosphonium-anchored GO. The obtained phosphonium-anchored GO was dried in a vacuum oven at 25 °C overnight. For making the P+GO-anchored CuMOF, 200 mg of P+GO was sonicated in 20 mL of ethanol for 30 min. The lighter nanoplatelets were isolated from the suspension and retained. A 0.86 g amount of copper acetate was dissolved in 60 mL of a 1:1 (v/v) DMF−water

ji W − Wd zyz z × 100 %uptake = jjj w j Ww zz k {

(1)

where Ww and Wd are weights after 24 h immersion and before immersion, respectively. Membrane Performance. Distilled Water Flux. An in-house cross-flow setup was used to quantify the distilled water flux.30 Token membranes, 45 mm in diameter, were mounted into a test cell and subjected to compaction at 0.34 MPa using a 300 GPD pump for 30 min. Postcompaction, the pressure was varied in sequence from 0.34 to 0.69 MPa, and the permeate flux (Jw) (L m−2 h−1) was recorded using the following formula: Jw =

V A×t

(2)

where V is the permeate volume, A is the cross-sectional area of the token, and t is the elution time. A run-time flux experiment was carried out to ensure sustained permeation of water. An experiment for 15 days was done on the membranes at 0.69 MPa, and its performance was compared with the commercial RO membrane. Synthetic Water Rejection. A combination of monovalent and divalent salts was used to formulate 250, 400, 1000, and 2000 ppm C


Research Article

ACS Sustainable Chemistry & Engineering solution. For the 250 ppm solution, 1 g of MgCl2, 1 g of Ca(NO3)2· 4H2O, and 0.5 g of NaCl were dissolved in 1 L of double-distilled water, and then, the solution was diluted accordingly with 9 L of double-distilled water. For the 400 ppm solution, 1.6 g of MgCl2, 1.75 g of Ca(NO3)2·4H2O, and 0.65 g of NaCl were dissolved in 1 L of double-distilled water, and then, the solution was diluted in 9 L of double-distilled water. For the 1000 ppm solution, 0.4 g of MgCl2, 0.4 g of Ca(NO3)2·4H2O, and 0.2 g of NaCl were dissolved in 1 L of double-distilled water. For the 2000 ppm solution, 0.8 g of MgCl2, 0.8 g of Ca(NO3)2·4H2O, and 0.4 g of NaCl were dissolved in 1 L of double-distilled water. Experiments were performed at 0.69 MPa poststabilization at 0.34 MPa for 30 min. The desalination capabilities were evaluated using eq 3. For this, the concentrations were obtained using inductively coupled plasma mass spectrometry (ICP-MS) for the feed and permeate. %rejection = 100 − ((permeate in ppm)/(feed in ppm))

The water-soluble tetrazolium salt (WST-1) assay, a colorimetric assay, was used to assess the cell proliferation using the manufacturer’s protocol. In brief, 3000 SVEC cells/well were seeded in a 96-well plate and allowed to grow for 24 h. Meanwhile, the permeate water was collected, and UV sterilization was performed for 30 min. A 1:10 (v/v) feed/media ratio was prepared in complete media, and the cells were incubated with the same for 24 h in the humidified incubator. After this, the wells were gently washed with 1× PBS, supplemented with WST-1 assay reagent (10 μL/100 μL media) in each well, and incubated for 3 h in the dark in the humidified incubator at 37 °C with 5% CO2. The 96-well plate was read under a plate reader at an excitation wavelength of 450 nm. The data was presented as mean ± standard error mean (n = 5). Significance was evaluated via Tukey’s posthoc analysis and set at 95% confidence (p < 0.05).

■

RESULTS Characterization of the P+GO-Anchored CuMOF. Figure 1 shows the FTIR spectra of GO, P+GO, and the

(3)

These performance values were compared with the commercial RO membrane without any modification. Antifouling Resistance. The antifouling property of the membranes was evaluated using the dynamic fouling test taking bovine serum albumin (BSA) as a model biofoulant. The protocol used was similar to our previous report.14 In short, the multilayered membrane was compacted at 0.34 MPa, and a stable distilled water flux (Jw) was obtained at 0.69 MPa pressure. Then, BSA was charged into the feed stream at a concentration of 1 g/L, and a stable flux (JP) was obtained at the same pressure. After this, the membranes were backflushed for 30 min with phosphate-buffered saline (PBS). A second cycle of water flux (J) was done after backflushing. The flux recovery ratio (FRR) and the irreversible flux decline ratio (IFR) were calculated using the following formulas: ij J yz FRR(%) = jjjj zzzz × 100 j Jw z k {

(4)

IFR(%) = 100 − FRR(%)

(5)

These were compared with the commercial RO membrane. Antibacterial Performance. ATCC25922 Escherichia coli (Gram-negative bacteria) and ATCC25923 Staphylococcus aureus (Gram-positive bacteria) were taken as model bacterial strains. The master culture was subcultured in LB broth at 37 °C in a shaking incubator. These cultures were subsequently harvested in log phase. A pellet of bacteria was obtained through centrifugation from the same, washed three times with PBS, and resuspended in PBS. This was then normalized to 109 CFU/mL for carrying out all experiments. Standard Plate Count. Disk tokens (4.5 mm) were taken in a 96well plate, and 100 μL amounts of resuspended cultures of E. coli or S. aureus were added to it. They was incubated in a shaking incubator for different incubation times (15, 30, and 60 min). Suspensions were periodically taken, and 100 μL amounts of 7-fold serial dilutions were cultured on freshly made nutrient agar. The agar plates were then incubated at 37 °C for 24 h. Colonies were counted after 24 h of incubation. After the test, the tokens were then fixed with 3.7% (v/v) formaldehyde and desiccated for SEM imaging. Intracellular Reactive Oxygen Species (ROS) Generation. Aliquots (100 μL) of E. coli and S. aureus cultures were added to the 4.5 mm disk samples for different incubation periods (15, 30, and 60 min) in a 96-well plate in a shaking incubator. After incubation, 50 μL of dichlorodihydrofluorescein diacetate (DCFH-DA) dye was added to the system, and the samples were kept in the dark for 30 min. Fluorescence intensity measurements were taken at the excitation wavelength of 485 nm and the emission wavelength of 528 nm using a plate reader. Permeate Cytotoxicity. The permeate cytotoxicity was assessed using seminal vesicle epithelial cells (SVEC) cells in complete Dulbecco’s modified eagle media containing 10% fetal bovine serum and 1% antibiotic and antimycotic. SVEC cells were incubated at 37 °C in a humidified atmosphere of purged 5% CO2. The fifth cell passage was taken for the experiment.

Figure 1. FTIR spectra for GO, P+GO, and the P+GO-anchored CuMOF used in this study.

P+GO-anchored CuMOF. The −OH characteristic band for GO was obtained at 3373 cm−1 while P+GO showed a characteristic band at 3319 cm−1 corresponding to the −OH functional group of GO. The alkyl chains in trihexyltetradecylphosphonium chloride showed corresponding bands at 2854, 2923, and 2955 cm−1. For GO and P+GO, the bands at 1606 and 1626 cm−1, respectively, indicate the presence of CC due to the aromatic rings of GO, while the bands at 1715 and 1726 cm−1, respectively, suggest the presence of C O due to the carboxylic groups of GO. For the P+GOanchored CuMOF, the CO band due to GO and trimesic acid was seen at 1733 cm−1 and the CC band due to GO was seen at 1638 cm−1. 31 P NMR was done on TPCl (the P+GO precursor), P+GO, and the P+GO-anchored CuMOF. TPCl, P+GO, and the P+GO-anchored CuMOF showed chemical shifts at 33.33, 37.97, and 37.96 ppm, respectively (see Figure S1). From here, it can be clearly understood that the phosphorus environment changed for TPCl and P+GO while it was same for P+GO and the P+GO-anchored CuMOF. Figure 2a shows the high-resolution TEM (HRTEM) image of the P+GO-anchored CuMOF. It can be clearly seen that the MOF is well dispersed on the planar sheet of P+GO. The selected area diffraction (SAED) confirms that the crystal D


Research Article


Figure 2. (a) HRTEM image for the P+GO-anchored CuMOF; (b) HAADF elemental maps of the P+GO-anchored CuMOF used in this study.

architecture possesses crystallinity. To further support our hypothesis, we obtained the high angle annular dark field (HAADF) image of the same area as shown in Figure 2b. It is clearly evident from the HAADF elemental maps that the P+GO-anchored CuMOF was successfully fabricated. To further support the same, SEM was done after drop-casting the suspension of the P+GO-anchored CuMOF. Figure 3 shows the electron micrograph of the same. It can be clearly seen that the organic framework is anchored on the sheet of the P+GO sheets supporting the TEM images. Figure 4 shows the powder XRD patterns of the P+GOanchored CuMOF before and after hydration. The CuMOF did show the (001) basal plane of P+GO captured at 2θ of 9.6° (d spacing of 0.69 nm) indicating the presence of GO in the framework. We performed a water stability analysis of the P+GO-anchored CuMOF, and it was observed that the architecture was structurally stable even after hydration for 24 h, and it was also noted that the d spacing of the GO sheets

Figure 3. Electron micrograph of the P+GO-anchored CuMOF.

remains unaltered after hydration, which proves our claims about stitching the GO d spacing: restricting the undesirable swelling for water permeation application. The Brunauer− Emmett−Teller method was used to assess the surface area, E


Research Article


the bulky and pendant groups in the side chains of the polymer chain. Finally, GPC was done on the copolymer to determine the molecular weight and PDI. The number-average molecular weight was found to be 1.54 × 105 g/mol whereas the weightaverage molecular weight was found to be 1.82 × 105 g/mol. The PDI was found to be 1.18. Characterization of the Multilayered Membrane. Figure 6a, b shows the surface and the cross-sectional

Figure 4. Powder XRD patterns of the P+GO-anchored CuMOF.

and the Barret−Joyner−Halenda analysis was used for the average pore radius of the P+GO-anchored CuMOF (see Table 1 in the Supporting Information). Characterization of the Antibacterial Copolymer. Cationic compounds such as amines and phosphonium have been shown to exhibit a robust bactericidal response and reversible fouling characteristics.14,16,30,39 This inspired us to fabricate polymers based on these compounds rather than grafting it on to the membrane surface. Figure 5 shows the FTIR spectrum of the RAFT-synthesized copolymer. The characteristic NH stretching at 3386 cm−1

Figure 6. (a) Surface morphology of the multilayered membrane; (b) cross-sectional morphology of the multilayered membrane (arrangement from left to right: RO support−P+GO-anchored CuMOF− copolymer active layer).

morphologies of the multilayered membrane. The pore size was in the range 180−350 μm (estimated through imageJ). The cross-sectional cryo-fractured section shows the interlayer P+GO-anchored CuMOF stitched effectively between the RO support and the copolymer active layer. The “gluing” of the P+GO-anchored CuMOF with the RO support is also evident from Figure 6b. Membrane Uptake. The bulk hydrophilicity of the multilayered membrane was assessed using water uptake capacity. It was observed that the % uptake in the commercial RO membrane was 34.2 ± 5.8% whereas the multilayered membrane had an uptake of 72.2 ± 5.4%. This can be attributed to the high uptake capacity of the porous interlayer P+GO-anchored CuMOF. Membrane Performance. Distilled Water Flux. Figure 7a shows the transmembrane flux versus the transmembrane pressure for the multilayered assembly. It was observed that there was a significant decline in the flux for the multilayered membrane when compared with that of the control. This might be due to the increased resistance of the water molecules as they transit through the interlayered architecture. Figure 7b shows the stability of the membranes. The flux values for the control membranes attain saturation at ∼19.5 L m−2 h−1 while the multilayered membrane attains saturation at 16.97 L m−2 h−1. Additionally, it can be seen from the long-term experiments that the multilayered membranes are capable of rendering unimpeded water flow even after a continuous run for 15 days. Synthetic Water Rejection. In this work, the rejection capabilities of the membranes were assessed using a formulation of synthetic water mimicking the test environment (250, 400, 1000, and 2000 ppm). Figure 8 shows the salt rejection values for the commercial RO membrane and the designed multilayered membrane. It was seen that the commercial RO membrane rejected 97%, 96.2%, 94.3%, and 92.7% of the salts at 250, 400, 1000, and 2000 ppm, respectively, while the multilayered membrane showed average rejections of 99.9%, 98.6%, 97.9%, and 96.8% of the salts at 250, 400, 1000, and 2000 ppm, respectively, of the synthetic water formulation.

Figure 5. FTIR spectrum of the RAFT-synthesized copolymer.

indicates the presence of the free amine of the polymer that also begins to suggest that PEI is successfully grafted to the polymer. The CH stretching band at 2925 cm−1 indicates the presence of alkane groups while the the CO stretching band at 1717 cm−1 is an indication that an amide is formed in the process of polymerization. To verify the change in the chemical environment upon polymerization, 31P NMR spectroscopy was done on the polymer sample and compared with respect to triphenylphosphine (the precursor for synthesizing the phosphonium monomer). The shift of triphenylphosphine was found to be −4.85 ppm whereas it was 23.48 ppm for the polymer. 1H NMR spectroscopy was also performed to support the structure (see Figure S2). Further, from the DSC thermogram, it was observed that the polymer had diffused and had less crystallinity wherein the crystallization temperature was 91.5 °C and the melting temperature was 112 °C (see Figure S3). The plausible explanation for this behavior can be attributed to F


Research Article


Figure 7. (a) Transmembrane flux versus transmembrane pressure for the multilayered membrane and the commercial RO membrane; (b) longterm flux studies for the multilayered membrane and the commercial RO membrane.

nanocapillaries that can harvest pure water and be selective barriers to heavy metal and salts. These sheets acts as molecular sieves to selectively reject ions based on size.40,41 However, GO sheets have a tendency to swell upon hydration,38 and the molecular sieving efficiency is reduced due to this swelling. The CuMOF is highly crystalline with less water stability.42 In situ growth of the CuMOF on GO has been reported to impart water stability to the MOF architecture.43,44 Further, in the work of Gutiérrez-Sevillano et al., the hydro-stability of the CuMOF was improved by the presence of cationic moieties in the vicinity of the CuMOF.45 Hence, when P+GO sheets were used to grow the CuMOF, the CuMOF became water stable, and in turn, these MOFs made the GO sheets resistant to swelling. This was the rationale for designing the P+GO-anchored CuMOF architecture. This arrangement was indeed efficient in sieving salts. To support this, we performed synthetic water rejection of multilayered membranes in the absence of P+GO and in the absence of the CuMOF (see Figure S4). It was observed that the sieving efficiency of these membranes was better than the commercial RO but less efficient than the P+GO-anchored CuMOF architecture. Permeation of water in the interlayer P+GO−MOF gallery will be mediated through the d spacing of the P+GO sheets retained to be 0.69 nm (inferred from XRD data). While the feed has a mixture of water molecules and di/monovalent cations, water molecules can pass through the d spacing easily. Some cations from the mixture may pass through the slit width between adjacent P+GO sheets (see Scheme2). Those monovalent Na+ and divalent Ca2+ and Mg2+ ions will be laterally entrapped in the porous architecture of the CuMOF due to the size-exclusion effect. (The hydrated radii of Na+, Ca2+, and Mg2+ are 0.358, 0.42, and 0.44 nm, respectively, and the average pore size of the MOF is 0.198 nm.) The escaped salt cations from the interlayer will be confined by the support RO membrane. Hence, the three layers work in tandem for removal of salts. This module can be backflushed and reused for many cycles via making the RO as the active layer and the copolymer as the support, recovering all the rejected salts and making the membranes reusable. This is completely different from the current backflushing technique for most of the RO membranes. As the feed concentration increases, the salts start to accumulate on the membrane surface and eventually crystallize, leading to mineral salt scaling of the RO membranes surface.46 This scale formation reduces flux drastically, increases input power in terms of pumping cost, and shortens

Figure 8. Synthetic water rejection for the membranes.

The plausible mechanism of enhanced rejection lies in the multilayered architecture of the membranes. The active layer being positively charged renders some electrostatic repulsion to the cations; however, being microporous, most of the cations pass through the pores (see Scheme2). The interlayer P+GO−MOF gallery has a pivotal role in the multilayered architecture. It has been established in the literature that GObased membranes have extraordinary performance capabilities in terms of salt and heavy metal rejection that can be attributed to their ability to form a layered structure consisting of Scheme 2. Specific Roles of Each Layer towards the Complete Decontamination of Water

G


Research Article


Figure 9. Standard plate count results for (a) E. coli and (b) S. aureus.

Figure 10. (a) Multilayered membrane with E. coli; (b) multilayered membrane with S. aureus; (c) commercial RO with E. coli; and (d) commercial RO with S.aureus.

of the material. It was seen that the multilayered membranes had lower IFR values compared to that of the commercial RO membrane. Antibacterial Performance. The cell wall of bacteria carry an overall negative charge due to the presence of peptidoglycan, the structural backbone of Gram-positive bacteria, and lipopolysaccharides, the structural backbone of Gram-negative bacteria.47 For our antibacterial copolymeric membranes, the phosphonium cation and the free NH2 groups (in the presence of water form NH3+ ion) are the cationic agents to manifest antibacterial action. The effect of these agents to arrest bacterial cell growth was quantified by a standard plate count method. Standard Plate Count. An incubation of 60 min for Grampositive and Gram-negative strains (E. coli and S. aureus, respectively) showed a 7-log reduction in visible bacterial colonies indicating the specific and targeted response of these membranes toward the bacteria. Figure 9a, b shows the cell viability in terms of colony-forming units per milliliter (CFU/ mL) for E. coli and S. aureus, respectively.

the life of the membranes. To reduce scaling, antiscalant is introduced in the feed in addition to backflushing. Hence, our multilayered approach might be the next step toward antiscalant-free sustainable membranes for water decontamination. Dynamic Antifouling Studies. A biofoulant such as a protein tends to adsorb on free surfaces such as membranes. Upon adsorption, these proteins block the pores causing the membrane to lose its property and resulting in a drastic reduction in the performance of the membrane followed by membrane failure. To be resistant to protein fouling attack, a surface which is subjected to a foulant should recover its purification property after backflushing. This is called flux retention in membranes. The higher the retention value, the closer is its efficiency in removal of deposited surface foulants. Figure S5 shows the dynamic fouling test results for the multilayered membranes. It was seen that the multilayered membrane surface allowed average flux retention (FRR) of 98.8% as compared to the commercial RO membrane. Additionally, the lower the value of the irreversible fouled flux ratio (IFR), the better is the reversible antifouling behavior H


Research Article

ACS Sustainable Chemistry & Engineering After the plate count, 60 min incubated samples were fixed with 3.7% (v/v) formaldehyde for SEM. Figure 10a−d shows the SEM images of the membranes with E. coli and S. aureus. The SEM images clearly showed irreversible cell damage (indicated by arrow marks) as compared to the control. In the control samples, commercial RO, bacteria remain unperturbed and tend to colonize (encircled). The results are in agreement with the plate count results to show the high antibacterial property of the multilayered membrane. ROS. DCFH-DA dye is nonfluorescent in nature, but it triggers a fluorescence response after getting converted into dichlorofluorescein (DCF) in the presence of cellular esterase and upon oxidation in bacterial cells. This fluorescence acts as an indicator of the intracellular stress by generation of ROS, such as hydroxyl radicals and peroxide radicals, and hence is directly correlated with the extent of ROS generated. The value for the ROS (in counts per second) is then divided by the live cells in each case to obtain ROS/CFU. This gives a quantified analysis about the viable cells under oxidative stress. Figure S6a, b shows the ROS/CFU for the multilayered membrane with E. coli and S. aureus, respectively. It is clearly evident from the ROS/CFU values that the copolymeric active layer was intrinsically efficient enough to induce significant intracellular oxidative stress in the bacterial cells. This was due to cationmediated interaction of the copolymer that has the phosphonium cation and free NH2 groups (in the presence of water form NH3+ ion) present in the macromolecular chains with the bacteria cells. It was also observed that the ROS mediated by the control (commercial RO membrane) is negligible in both the cases. Permeate Cytotoxicity. WST-1 is a colorimetric marker that cleaves to form a soluble formazan by mitochondrial dehydrogenases in viable cells. The higher the number of viable and metabolically active cells, the more intense the fluorescence absorbance will be. From Figure S7a, it is clearly evident that the permeate from the multilayered membrane showed no visible cytotoxicity because the absorbance by the permeate was not much different from the control; the control for this experiment was sterilized permeate from the commercial RO membrane. To further corroborate the statement, copper ions in the permeate were analyzed using ICP-MS (Figure S7b). The copper ions were found to be in the range 0.01−0.015 ppm, which is safe for human consumption based on permissible limits of copper in drinking water.9

cell growth by inducing intracellular leakage and rapid cell death, quantified by plate count and SEM studies. This copolymer was also highly effective in mitigating protein fouling because the active layer was highly polar, which increased the hydrophilicity, decreased the BSA foulant deposition, and improved the backflushing action. The active layer also rendered some electrostatic repulsion to mono/ divalent cations helping to improve the salt rejection. The interlayer P+GO−MOF gallery had a major role in harvesting water and sieving ions. As ordinary GO sheets only acted as molecular sieves to selectively reject ions based on size and upon hydration, the molecular sieving efficiency gets reduced. We tried to solve this issue by first modifying it with phosphonium salt to impart positive charge and stitched the layers by growing a Cu-BTC-based MOF. Anchoring this MOF into the gallery of GO made the GO resistant to swelling, and the cationic modification intensified the electrostatic repulsion that further aided in enhanced rejection. This module can be backflushed and reused for many cycles via making the RO membrane as the active layer and the copolymer as the support, recovering all the rejected salts and making the membranes reusable, unlike the commercial RO membrane that fails under scaling attack. Hence, our multilayered approach offers some potential toward sustainable water remediation.

■

CONCLUSIONS A novel antibacterial copolymer was designed via RAFT polymerization, and multilayered membranes were prepared using this polymer as the active surface layer, commercial RO membrane as the support, and the P+GO-anchored CuMOF as the interlayer. These membranes had quick bacterial reduction capabilities by reducing 7-log fold bacterial colonies for both Gram-positive and Gram-negative strains and exhibited a stringent and targeted response in terms of ROS generation. These membranes were capable enough to exhibit stable permeation of water even after 15 days of continuous run and had a flux value of 16.97 L m−2 h−1. In addition, the multilayered membranes exhibited excellent tailored and reversible antifouling with BSA as the model foulant with 98.8% FRR. The nanoexclusions of the P+GO-anchored CuMOF present as the interlayer though reduced the flux as compared to the control RO membranes, yet they were exemplary in rejecting 99.9% of salts at 250 ppm and 98.6% of salts for 400 ppm for a formulation composed of monovalent and divalent salts. As the current RO membranes are not bactericidal, require prefiltration and UV treatment a priory, and require antiscalant in the feed in addition to backflushing for scaling remediation due to high salt inputs, our multilayered approach might be the next step toward antiscalant-free sustainable membranes for water decontamination. Overall these membranes can be the next generation membranes seeking potential applications in purification applications without the need of a prefiltration kit.

■

DISCUSSION The porous multilayered architecture was radically designed to target all the key requirements for remediation of water such as bactericidal response, antifouling, and salt removal. Cationic agents such as amines and phosphonium tailored on the surface of the membranes as grafts have shown to exhibit an extensive bactericidal response and antifouling behavior. Inspired from these properties, polycationic amine and phosphonium groups were introduced onto the main chain using the RAFT-based strategy and used as the active layer. When this active layer was in contact with bacteria, it exhibited evasive bactericidal action. Because the bacterial cell wall carried an overall negative charge, upon being contacted with the positively charged RAFT copolymer, the outer cell wall integrity of the bacteria was lost, and the bacterial cells experienced intense oxidative stress that was recorded using ROS studies. The RAFT copolymer also arrested the bacterial

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acssuschemeng.8b05354. 31

P and 1H NMR spectra; DSC thermogram; synthetic water rejection assessment; dynamic fouling test; ROS/

I


Research Article


■

(12) Park, S. Y.; Choi, S. H.; Chung, J. W.; Kwak, S.-Y. Anti-scaling ultrafiltration/microfiltration (UF/MF) polyvinylidene fluoride (PVDF) membranes with positive surface charges for Ca 2+/silicarich wastewater treatment. J. Membr. Sci. 2015, 480, 122−128. (13) Ridgway, H. F.; Rigby, M. G.; Argo, D. G. Bacterial Adhesion and Fouling of Reverse Osmosis Membranes (PDF). J. - Am. Water Works Assoc. 1985, 77 (7), 97−106. (14) Samantaray, P. K.; Madras, G.; Bose, S. Antibacterial and Antibiofouling Polymeric Membranes through Immobilization of Pyridine Derivative Leading to ROS Generation and Loss in Bacterial Membrane Integrity. ChemistrySelect 2017, 2 (26), 7965−7974. (15) Kang, S.-T.; Subramani, A.; Hoek, E. M.; Deshusses, M. A.; Matsumoto, M. R. Direct observation of biofouling in cross-flow microfiltration: mechanisms of deposition and release. J. Membr. Sci. 2004, 244 (1−2), 151−165. (16) Samantaray, P. K.; Bose, S. Cationic biocide anchored graphene oxide based membranes for water purification. Proc. Indian Natl. Sci. Acad., Part A 2018, 84 (3), 669−679. (17) Combe, C.; Molis, E.; Lucas, P.; Riley, R.; Clark, M. The effect of CA membrane properties on adsorptive fouling by humic acid. J. Membr. Sci. 1999, 154 (1), 73−87. (18) Knoell, T.; Safarik, J.; Cormack, T.; Riley, R.; Lin, S. W.; Ridgway, H. Biofouling potentials of microporous polysulfone membranes containing a sulfonated polyether-ethersulfone/polyethersulfone block copolymer: correlation of membrane surface properties with bacterial attachment. J. Membr. Sci. 1999, 157 (1), 117−138. (19) Kouwonou, Y.; Malaisamy, R.; Jones, K. L. Modification of PES membrane: reduction of biofouling and improved flux recovery. Sep. Sci. Technol. 2008, 43 (16), 4099−4112. (20) Rana, D.; Matsuura, T. Surface modifications for antifouling membranes. Chem. Rev. 2010, 110 (4), 2448−2471. (21) Vatanpour, V.; Madaeni, S. S.; Moradian, R.; Zinadini, S.; Astinchap, B. Fabrication and characterization of novel antifouling nanofiltration membrane prepared from oxidized multiwalled carbon nanotube/polyethersulfone nanocomposite. J. Membr. Sci. 2011, 375 (1−2), 284−294. (22) Park, J.; Park, J.; Kim, S. H.; Cho, J.; Bang, J. Desalination membranes from pH-controlled and thermally-crosslinked layer-bylayer assembled multilayers. J. Mater. Chem. 2010, 20 (11), 2085− 2091. (23) Louie, J. S.; Pinnau, I.; Ciobanu, I.; Ishida, K. P.; Ng, A.; Reinhard, M. Effects of polyether−polyamide block copolymer coating on performance and fouling of reverse osmosis membranes. J. Membr. Sci. 2006, 280 (1), 762−770. (24) Ma, H.; Hakim, L. F.; Bowman, C. N.; Davis, R. H. Factors affecting membrane fouling reduction by surface modification and backpulsing. J. Membr. Sci. 2001, 189 (2), 255−270. (25) Hilal, N.; Kochkodan, V.; Al-Khatib, L.; Levadna, T. Surface modified polymeric membranes to reduce (bio) fouling: a microbiological study using E. coli. Desalination 2004, 167, 293−300. (26) Malaisamy, R.; Bruening, M. L. High-flux nanofiltration membranes prepared by adsorption of multilayer polyelectrolyte membranes on polymeric supports. Langmuir 2005, 21 (23), 10587− 10592. (27) Kull, K. R.; Steen, M. L.; Fisher, E. R. Surface modification with nitrogen-containing plasmas to produce hydrophilic, low-fouling membranes. J. Membr. Sci. 2005, 246 (2), 203−215. (28) Kanazawa, A.; Ikeda, T.; Endo, T. Novel polycationic biocides: synthesis and antibacterial activity of polymeric phosphonium salts. J. Polym. Sci., Part A: Polym. Chem. 1993, 31 (2), 335−343. (29) Kanazawa, A.; Ikeda, T.; Endo, T. Antibacterial activity of polymeric sulfonium salts. J. Polym. Sci., Part A: Polym. Chem. 1993, 31 (11), 2873−2876. (30) Samantaray, P. K.; Madras, G.; Bose, S. PVDF/PBSA membranes with strongly coupled phosphonium derivatives and graphene oxide on the surface towards antibacterial and antifouling activities. J. Membr. Sci. 2018, 548, 203−214.

CFU as a function of incubation time; permeate cytotoxicity assessment and copper ion release analysis; and table of the MOF surface and pore sizes (PDF)

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID

Paresh Kumar Samantaray: 0000-0003-2533-929X Giridhar Madras: 0000-0003-2211-5055 Suryasarathi Bose: 0000-0001-8043-9192 Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS The authors would like to extend their acknowledgement towards DST for financial support, NMR research Centre for NMR facility, Prof. Satish Patil for GPC experiments, Dr. Kaushik Chatterjee for extending the facilities for bacterial studies, Mr. Ambresh M. from CeNSE for TEM imaging, and CeNSE IISc for various characterization facilities.

■

REFERENCES

(1) Qadir, D.; Mukhtar, H.; Keong, L. K. Mixed Matrix Membranes for Water Purification Applications. Sep. Purif. Rev. 2017, 46 (1), 62− 80. (2) Döll, P.; Hoffmann-Dobrev, H.; Portmann, F. T.; Siebert, S.; Eicker, A.; Rodell, M.; Strassberg, G.; Scanlon, B. Impact of water withdrawals from groundwater and surface water on continental water storage variations. J. Geodyn. 2012, 59-60, 143−156. (3) Famiglietti, J. S. The global groundwater crisis. Nat. Clim. Change 2014, 4 (11), 945−948. (4) Shannon, M. A.; Bohn, P. W.; Elimelech, M.; Georgiadis, J. G.; Marinas, B. J.; Mayes, A. M. Science and technology for water purification in the coming decades. Nature 2008, 452 (7185), 301− 310. (5) Tripathi, B. P.; Dubey, N. C.; Stamm, M. Functional polyelectrolyte multilayer membranes for water purification applications. J. Hazard. Mater. 2013, 252-253, 401−412. (6) Chan, W.-F.; Chen, H.-y.; Surapathi, A.; Taylor, M. G.; Shao, X.; Marand, E.; Johnson, J. K. Zwitterion functionalized carbon nanotube/polyamide nanocomposite membranes for water desalination. ACS Nano 2013, 7 (6), 5308−5319. (7) Gunawan, P.; Guan, C.; Song, X.; Zhang, Q.; Leong, S. S. J.; Tang, C.; Chen, Y.; Chan-Park, M. B.; Chang, M. W.; Wang, K.; Xu, R. Hollow fiber membrane decorated with Ag/MWNTs: toward effective water disinfection and biofouling control. ACS Nano 2011, 5 (12), 10033−10040. (8) Lv, Y.; Zhang, C.; He, A.; Yang, S. J.; Wu, G. P.; Darling, S. B.; Xu, Z. K. Photocatalytic Nanofiltration Membranes with SelfCleaning Property for Wastewater Treatment. Adv. Funct. Mater. 2017, 27 (27), 1700251. (9) Padmavathy, N.; Samantaray, P. K.; Ghosh, L. D.; Madras, G.; Bose, S. Selective Cleavage of Polyphosphoester in Crosslinked Copper based Nanogels: Enhanced Antibacterial Performance Through Controlled release of Copper. Nanoscale 2017, 9 (34), 12664−12676. (10) Diagne, F.; Malaisamy, R.; Boddie, V.; Holbrook, R. D.; Eribo, B.; Jones, K. L. Polyelectrolyte and silver nanoparticle modification of microfiltration membranes to mitigate organic and bacterial fouling. Environ. Sci. Technol. 2012, 46 (7), 4025−4033. (11) Brink, L.; Romijn, D. Reducing the protein fouling of polysulfone surfaces and polysulfone ultrafiltration membranes: Optimization of the type of presorbed layer. Desalination 1990, 78 (2), 209−233. J


Research Article

ACS Sustainable Chemistry & Engineering (31) Liu, X.; Qi, S.; Li, Y.; Yang, L.; Cao, B.; Tang, C. Y. Synthesis and characterization of novel antibacterial silver nanocomposite nanofiltration and forward osmosis membranes based on layer-bylayer assembly. Water Res. 2013, 47 (9), 3081−3092. (32) Xu, G.-R.; Wang, S.-H.; Zhao, H.-L.; Wu, S.-B.; Xu, J.-M.; Li, L.; Liu, X.-Y. Layer-by-layer (LBL) assembly technology as promising strategy for tailoring pressure-driven desalination membranes. J. Membr. Sci. 2015, 493, 428−443. (33) Tang, L.; Gu, W.; Yi, P.; Bitter, J. L.; Hong, J. Y.; Fairbrother, D. H.; Chen, K. L. Bacterial anti-adhesive properties of polysulfone membranes modified with polyelectrolyte multilayers. J. Membr. Sci. 2013, 446, 201−211. (34) Ashraf, M.; Maah, M.; Qureshi, A.; Gharibreza, M.; Yusoff, I. Synthetic polymer composite membrane for the desalination of saline water. Desalin. Water Treat. 2013, 51 (16−18), 3650−3661. (35) Jiang, J.; Zhu, L.; Zhu, L.; Zhang, H.; Zhu, B.; Xu, Y. Antifouling and antimicrobial polymer membranes based on bioinspired polydopamine and strong hydrogen-bonded poly (Nvinyl pyrrolidone). ACS Appl. Mater. Interfaces 2013, 5 (24), 12895− 12904. (36) Wang, J.; Wang, Y.; Zhang, Y.; Uliana, A.; Zhu, J.; Liu, J.; Van der Bruggen, B. Zeolitic imidazolate framework/graphene oxide hybrid nanosheets functionalized thin film nanocomposite membrane for enhanced antimicrobial performance. ACS Appl. Mater. Interfaces 2016, 8 (38), 25508−25519. (37) Cohen-Tanugi, D.; Grossman, J. C. Water desalination across nanoporous graphene. Nano Lett. 2012, 12 (7), 3602−3608. (38) Zheng, S.; Tu, Q.; Urban, J. J.; Li, S.; Mi, B. Swelling of graphene oxide membranes in aqueous solution: characterization of interlayer spacing and insight into water transport mechanisms. ACS Nano 2017, 11 (6), 6440−6450. (39) Mural, P. K. S.; Banerjee, A.; Rana, M. S.; Shukla, A.; Padmanabhan, B.; Bhadra, S.; Madras, G.; Bose, S. Polyolefin based antibacterial membranes derived from PE/PEO blends compatibilized with amine terminated graphene oxide and maleated PE. J. Mater. Chem. A 2014, 2 (41), 17635−17648. (40) Hu, M.; Mi, B. Enabling graphene oxide nanosheets as water separation membranes. Environ. Sci. Technol. 2013, 47 (8), 3715− 3723. (41) Joshi, R.; Carbone, P.; Wang, F.-C.; Kravets, V. G.; Su, Y.; Grigorieva, I. V.; Wu, H.; Geim, A. K.; Nair, R. R. Precise and ultrafast molecular sieving through graphene oxide membranes. Science 2014, 343 (6172), 752−754. (42) Burtch, N. C.; Jasuja, H.; Walton, K. S. Water stability and adsorption in metal−organic frameworks. Chem. Rev. 2014, 114 (20), 10575−10612. (43) Li, L.; Liu, X. L.; Geng, H. Y.; Hu, B.; Song, G. W.; Xu, Z. S. A MOF/graphite oxide hybrid (MOF: HKUST-1) material for the adsorption of methylene blue from aqueous solution. J. Mater. Chem. A 2013, 1 (35), 10292−10299. (44) Petit, C.; Burress, J.; Bandosz, T. J. The synthesis and characterization of copper-based metal−organic framework/graphite oxide composites. Carbon 2011, 49 (2), 563−572. (45) Gutiérrez-Sevillano, J. J.; Dubbeldam, D.; Bellarosa, L.; López, N.; Liu, X.; Vlugt, T. J.; Calero, S. Strategies to Simultaneously Enhance the Hydrostability and the Alcohol−Water Separation Behavior of Cu-BTC. J. Phys. Chem. C 2013, 117 (40), 20706−20714. (46) Shih, W.-Y.; Rahardianto, A.; Lee, R.-W.; Cohen, Y. Morphometric characterization of calcium sulfate dihydrate (gypsum) scale on reverse osmosis membranes. J. Membr. Sci. 2005, 252 (1−2), 253−263. (47) Samantaray, P. K.; Baloda, S.; Madras, G.; Bose, S. A designer membrane tool-box with a mixed metal organic framework and RAFT-synthesized antibacterial polymer perform in tandem towards desalination, antifouling and heavy metal exclusion. J. Mater. Chem. A 2018, 6 (34), 16664−16679.

K
