ZIF-8

Research Article www.acsami.org

Layer-by-Layer Fabrication of High-Performance Polyamide/ZIF‑8 Nanocomposite Membrane for Nanofiltration Applications Luying Wang,†,‡ Manquan Fang,§ Jing Liu,†,‡ Jing He,†,‡ Jiding Li,§ and Jiandu Lei*,†,‡ †

Beijing Key Laboratory of Lignocellulosic Chemistry, ‡MOE Engineering Research Center of Forestry Biomass Materials and Bioenergy, Beijing Forestry University, Beijing 100083, PR China § Department of Chemical Engineering, Tsinghua University, Beijing 100084, PR China S Supporting Information *

ABSTRACT: The conventional blending fabrication for thin-film nanocomposite (TFN) membranes is to disperse porous fillers in aqueous/ organic phases prior to interfacial polymerization, and the aggregation of fillers may lead to the significant decrease in membrane performance. To overcome this limitation, we proposed a novel layer-by-layer (LBL) fabrication to prepare a polyamide (PA)/ZIF-8 nanocomposite membrane with a multilayer structure: a porous substrate, a ZIF-8 interlayer, and a PA coating layer. The PA/ZIF-8 (LBL) membrane for nanofiltration applications was prepared by growing an interlayer of ZIF-8 nanoparticles on an ultrafiltration membrane through in situ growth and then coating it with an ultrathin PA layer through interfacial polymerization. The obtained PA/ZIF-8 (LBL) membrane exhibited both better permeance and selectivity than did the conventional PA/ZIF-8 TFN membrane because of the ZIF-8 in situ growth producing a ZIF-8 interlayer with more ZIF-8 nanoparticles but fewer aggregates. Compared with the pure PA membrane (the flux of 11.2 kg/m2/h and rejection of 99.6%) for dye removal, the obtained PA/ZIF-8 (LBL) membranes achieved a significant improvement in membrane permeance and selectivity. (Flux was up to 27.1 kg/m2/h, and the rejection reaches 99.8%.) This LBL fabrication is a promising methodology for other polymer nanocomposite membranes simultaneously having high permeance and good selectivity. KEYWORDS: layer-by-layer, polyamide, ZIF-8, nanocomposite membrane, nanofiltration

1. INTRODUCTION Because of the rapid development of membrane technology, there is strong interest in developing high-performance membranes to meet the present and future requirements and challenges.1−5 Some microporous/mesoporous materials have gained considerable attention for having great potential to fabricate high-performance membranes such as, carbon nanotubes, 2,6−8 graphenes, 9,10 metal−organic frameworks (MOFs),11,12 covalent−organic materials,13,14 microporous organic polymers,15 zeolites,16−18 and double-layered hydroxide.16,19,20 MOF materials, a novel kind of porous materials, are basically composed of metal ions/clusters and organic struts and have been highly attractive for their applications in membrane separation because of the improvement in membrane performance based on the identical pore sizes and highly porous structures. It has been experimentally proven that MOF membranes can exhibit significantly improved performance in gas separation,4,21−26 pervaporation,11,27−30 nanofiltration (NF),12,31−36 and reverse osmosis (RO)37 processes. For a NF or RO process, it has been reported that polyamide (PA)/MOF TFN membranes exhibit higher permeance and better selectivity than pure PA thin-film composite (TFC) membranes.12,37,38 The presence of MOFs in the PA layers can significantly increase selective transport pathways to achieve © XXXX American Chemical Society

high performance, that is, a significant increase in permeance without considerable decrease in selectivity. In these reported works, the PA/MOF TFN membranes were all fabricated through interfacial polymerization by using MOF dispersions (presynthesized MOFs dispersed evenly in aqueous/organic phases). MOF aggregates may be formed in the dispersion of a high MOF concentration and then embedded in polymers during interfacial polymerization and/or solvent evaporation, even if the MOF dispersion is treated by sonication prior to interfacial polymerization. Although the excellent compatibility between polymers and MOF particles can eliminate the formation of nonselective interfacial defects in polymer/MOF membranes, the aggregation of MOFs attributes to the decrease in flux enhancement or in rejection.12,37,38 The development of a novel fabrication methodology for improving membrane performance of PA/MOF TFN membranes at high MOF concentrations is highly desirable. Recently, some novel methodologies have been reported for fabricating MOF membranes on porous inorganic or polymeric supports, including the in situ growth,24,32,39−41 seeded Received: August 3, 2015 Accepted: October 12, 2015

A

DOI: 10.1021/acsami.5b07128 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces

Figure 1. Schematic of membrane fabrication (ZIF-8 in blue, MPD in purple, TMC in orange, and PA in brown): ZIF-8 synthesis (a), LBL fabrication procedures (b), and cross-sectional structure of the PA/ZIF-8 (LBL) membrane (c).

growth,23,27,42−44 interfacial synthesis,21,31,36,45 contra-diffusion,30,46 self-assembly,29,33 and other methods.47−51 Derived from the in situ growth approaches,24,32 this study reports a novel layer-by-layer (LBL) fabrication for fabricating PA/ZIF-8 nanocomposite membranes via ZIF-8 in situ growth and interfacial polymerization. In this study, ZIF-8, a typical zeolitic MOF crystalline porous material, is selected to prepare PA/ ZIF-8 (LBL) membranes through this fabrication methodology because ZIF-8 can be easily synthesized at room temperature and has a microporous structure (shown in Figure 1a). The LBL fabrication includes two steps as shown in the Figure 1b: First, an interlayer of ZIF-8 nanoparticles grows on the porous surface of a support membrane through the in situ growth. Then, an ultrathin PA layer covers the ZIF-8 interlayer through interfacial polymerization and curing treatment, finally giving a layered membrane structure (Figure 1c). Because of the excellent compatibility between polymers and ZIF-8, the PA coating layer may seal gaps between the ZIF-8 nanoparticles, and then the ZIF-8 interlayer can be well-embedded in the PA coating layer resulting in a PA/ZIF-8 (LBL) nanocomposite layer. The introduction of the ZIF-8 and PA layers through this LBL fabrication need not disperse presynthesized ZIF-8 nanoparticles in aqueous/organic phases. Consequently, the formation of ZIF-8 aggregates in the ZIF-8 dispersions can be avoided, and the improvement in membrane performance of

the PA/ZIF-8 (LBL) nanocomposite membranes can be expected. Here, for the first time, the high-performance PA/ZIF-8 (LBL) membranes were successfully prepared through LBL fabrication. The particle number of the ZIF-8 interlayer can be increased by repeating several times the in situ growth procedure to increase the ZIF-8 loading in PA and then enhance membrane permeance. To study the effects of in situ growth on membrane (structure and separation) properties, a series of the PA/ZIF-8 (LBL) membranes were prepared by increasing the number of ZIF-8 in situ growth from one to four. The prepared PA/ZIF-8 (LBL) membranes were characterized with respect to the presence of ZIF-8 interlayers, chemical structure of PA/ZIF-8 (LBL) membranes, and membrane morphology; the separation performance was investigated by using NF experiments for water treatments (including desalination, phenol removal, and dye removal).

2. MATERIALS AND METHODS Materials. All reagents and solvents were commercially available for experiments and were used as received without further purification. Polysulfone (PSF) UF membranes (MWCO: 30 000 Da, dimensions: 0.28 m × 5.0 m) were supplied by the Pureach Tech. (Beijing) Co., Ltd. Zn(NO3)2·6H2O (99% purity) and 2-methylimidazole (99% purity) used for synthesizing of ZIF-8 were purchased from Tianjin Yongda Chemical Reagent Co., Ltd., and Sinopharm Chemical Reagent Co., Ltd., respectively. m-Phenylenediamine (MPD, 99% B


Research Article


Figure 2. XRD pattern (a) and nitrogen adsorption/desorption isotherms (b) of the ZIF-8 nanoparticles. purity) and trimesoyl chloride (TMC, 99% purity) used for the fabrication of PA were purchased from Tianjin Guangfu Fine Chemical Research Institute and Zahn Chemical Technology (Shanghai) Co., Ltd., respectively. MgSO4 (analytical grade, 120.4 g/mol), phenol (analytical grade, 94.1 g/mol), and Congo red (certified grade, 696.7 g/mol, dye content ≥85%) were purchased from Tianjin Jinke Fine Chemical Research Institute, Beijing Chemical Works, and AMRESCO, respectively. Organic solvents (methanol and n-hexane, analytical grade) used for membrane preparation were obtained from Beijing Chemical Plant, and DI water was used in all experiments. Membrane Fabrication. The ZIF-8 suspension was prepared by mixing a Zn(NO3)2·6H2O solution (10.0 g of Zn(NO3)2·6H2O in 200 mL of methanol and 100 mL of DI water) and a 2-methylimidazole solution (30.0 g of 2-methylimidazole in 200 mL of methanol) together under stirring at room temperature for 0.5 h. As shown in Figures 1a,b, a PSF UF membrane with a desirable size was stuck on a polytetrafluoroethylene plate and immersed in the ZIF-8 suspension for 12 h. Then, the wet membrane was extensively washed with water and methanol to remove unreacted reagents. This was followed by drying the obtained membrane at 60 °C for 2 h to produce a modified supported membrane through the first in situ growth procedure; the obtained membrane is named ZIF-8@PSF (#1) membrane. The same ZIF-8 in situ growth procedure was repeated by using the ZIF-8@PSF (#1) membrane 2−4 additional times to produce ZIF-8@PSF (#2−4) membranes, respectively. The stability of the PSF membrane after immersion in methanol for 12 h was proven in literature.14 In addition, pure ZIF-8 nanoparticles were collected from the ZIF-8 dispersion after the in situ growth procedure in order to obtain ZIF-8 nanoparticles for characterization experiments. The synthesized ZIF8 was collected by repeated centrifugation (8000 rpm and 20 min) and was purified by washing with water and methanol. Then, the collected ZIF-8 nanoparticles were dried under vacuum at 120 °C overnight in a drying oven. Typically, a PA selective layer is fabricated on a UF membrane through interfacial polymerization between monomers with acyl chloride groups in an organic phase and monomers with amine groups in an aqueous phase. In this study, the organic phase was prepared by dissolved TMC in n-hexane to obtain the organic solution of 0.1% (w/v) TMC, and the aqueous phase was prepared by dissolved MPD in water to obtain an aqueous solution of 2% (w/v) MPD. Each of the PSF and ZIF-8@PSF membranes was stuck on a glass plate and immersed in the as-prepared aqueous phase for 2 min; after removal of the excess aqueous solution on the membrane surface, the wet membrane was immersed in the as-prepared organic phase for 20 s, as shown in Figure 1b. Then, the initial wet membranes obtained through the interfacial polymerization were dried at room temperature and then cured at 60 °C for 2 min to obtain the PA and PA/ZIF-8 (LBL) membranes. The PA and PA/ZIF-8 (LBL) membranes were rinsed with water several times to be used for characterization and separation experiments. The PA/ZIF-8 (LBL) membranes prepared

using different ZIF-8@PSF membranes were named PA/ZIF-8 (LBL#x), where x = 1−4 indicating the number of in situ growths for the corresponding ZIF-8@PSF membranes. Characterization Experiments. All samples including the ZIF-8 and prepared membranes were dried prior to the following characterization experiments. X-ray diffraction (XRD) analysis of the ZIF-8 nanoparticles was carried out via a Germany Bruker D8 X-ray diffractometer using a Cu Kα radiation source (λ = 1.5418 Å). The analysis was operated at 40 kV and 40 mA, and the 2θ was varied from 5 to 40° with a step of 1°/min. The morphologies of the ZIF-8 nanoparticles and membranes were observed using Hitachi S-4800 scanning electric microscopy (SEM) and MERLIN FESEM (Zeiss) instruments. The cross-sectional samples of the membranes were obtained by fracturing the membranes under liquid nitrogen condition, and all samples were coated by gold under vacuum. The nitrogen adsorption measurement of the ZIF-8 nanoparticles was carried out at 77 K using a Micro meritics ASAP 2020 gas sorption instrument to obtain the specific surface area and micropore volume. The ZIF-8 nanoparticles were degassed under vacuum at 200 °C for 4 h prior to the adsorption/desorption measurement. The specific surface area was calculated according to the Brunauer−Emmett−Teller (BET) method. The Fourier transform infrared (FTIR) spectra were recorded using a Nicolet IR560 FTIR spectrometer to characterize the chemical structures of the ZIF-8 nanoparticles and membranes. Each sample was scanned from 2800 to 800 cm−1 with a resolution of 4 cm−1. The surface roughness properties of the PA and PA/ZIF-8 (LBL) membranes were measured using a Shimadzu SPM-9600 atomic force microscope (AFM) with the tapping mode under ambient conditions. The contact angles of DI water on the membrane surfaces were determined using a DataPhysics OCA20 contact angle goniometer under ambient conditions. The water volume was set at 20 μL for each measurement, and at least 3−4 measurements were carried out to obtain an average value of the measured contact angles for each membrane. NF Experiments. The UF/NF apparatus from Hangzhou Watech Co. was used to investigate the membrane permeance of the ZIF-8@ PSF and PA/ZIF-8 (LBL) membranes. The membrane cell is a pumppressurized cross-flow unit, using a flat membrane, and the effective membrane area is 69.40 cm2. The UF experiments were carried out using pure water to measure the water permeance of the ZIF-8@PSF membranes, and the NF experiments were carried out using different aqueous feeds to obtain the separation performance of the PA/ZIF-8 (LBL) membranes. The solute concentrations of the feed mixtures were 1000 ppm MgSO4, 100 ppm phenol, and 200 ppm Congo red, respectively. During each experiment, the feed flow was continuously circulated from a feed tank to the membrane cell at a desired operating pressure of 1.0 MPa. Each permeate sample for evaluating the membrane performance was collected after reaching a steady-state condition (about 2−3 h). For each kind of support membrane and NF membrane, three membrane samples were prepared separately through C


Research Article

ACS Applied Materials & Interfaces the same fabrication conditions and then were measured respectively under the same operation conditions. The evaluated membrane performance of each kind of membrane was averaged using the three measurements obtained with different membrane samples, and the statistical error of each evaluation data was defined by the standard deviation. For each measurement, the permeate flux (J) was measured by weighing the permeate sample in a certain collecting time according to eq 1; the rejection (R) was calculated from eq 2.

J= R=

W At

(1)

Cf − C p Cf

× 100%

(2)

where W is the weight of permeate (kg), A the effective area of membrane (m2), t is the time (h), Cf and Cp represent the concentrations (ppm) in the feed and the permeate, respectively. Cf and Cp were analyzed by an electrical conductivity meter (Shanghai Yoke Instrument Co., DDS-307A) for MgSO4, a high-performance liquid chromatography (Beijing Tong Heng Innovation Technology Co., Ltd., LC3000) using UV detector at λ = 254.0 nm for phenol, and a UV−vis spectrophotometer (Shanghai Yoke Instrument Co., UV759) at λ = 496.0 nm for Congo red.

Figure 3. FTIR spectra of ZIF-8, PSF membrane, and ZIF-8@PSF (#4) membrane.

3. RESULTS AND DISCUSSION 3.1. Characterization Results. 3.1.1. Characterization of the ZIF-8 Nanoparticles and ZIF-8@PSF Membranes. The ZIF-8 nanoparticles were collected from the in situ growth procedure and characterized to obtain structure properties. The crystalline and porous structure of the ZIF-8 nanoparticles was confirmed by the XRD analysis and nitrogen adsorption measurement, respectively, as shown in Figure 2. The powder XRD pattern of ZIF-8 is well matched with the standard pattern simulated from the crystallographic data (Cambridge Crystallographic Data Centre (CCDC) no. 739161).52 The high similarity between the experimental and simulated XRD patterns indicates that the nanoparticles obtained from the in situ growth procedure were pure-phased ZIF-8 crystals. The adsorption/desorption isotherm (at 77 K) exhibits a rapid increase of nitrogen adsorption at very low relative pressure (P/ P0 < 0.03) and a nearly constant adsorption at high relative pressure, corresponding to the microporous structure of ZIF-8. Derived from the nitrogen adsorption isotherms, the BET surface area is 1168 m2/g, and the micropore volume is 0.642 cm3/g. The BET surface area and pore volume are both approximate to the corresponding values in literature.12,25,37 The FTIR analysis was carried out to compare the ZIF-8@ PSF (#4) membrane with the ZIF-8 nanoparticles and PSF membrane. As shown in Figure 3, the ZIF-8@PSF (#4) membrane exhibits an FTIR pattern similar to that of the PSF membrane, whereas stronger bands at 1400−1450 cm−1 and new characteristic bands at 1300 and 1000 cm−1 are observed from the spectra of the ZIF-8@PSF (#4) membrane. The difference between the spectra of the PSF and ZIF-8@PSF (#4) membranes is in accord with the ZIF-8 characteristic bands, indicating the presence of ZIF-8 on the membrane surface after the fourth in situ growth procedure. The morphologies of the ZIF-8 nanoparticles and the support membranes were observed by SEM to confirm the introduction of the ZIF-8 interlayer(s) through in situ growth. The SEM image of the obtained ZIF-8 nanoparticles (Figure 4a) reveals that the ZIF-8 crystals are nanoparticles with typical rhombic dodecahedral morphology and have an average particle size of about 150 nm. Figure 4b shows the surface SEM image of the PSF membrane, where the porous surface of the PSF

Figure 4. SEM images of the ZIF-8 nanoparticles (a), surfaces of the PSF membrane (b), and ZIF-8@PSF membranes (c−f, #1−4, respectively), and cross section of the ZIF-8@PSF (#4) membrane (g and h).

membrane is observed and the pore size on the membrane surface is about 20−50 nm. The ZIF-8@PSF (#1) membrane still has a porous surface, and some ZIF-8 crystals showing particlelike spots are observed on the membrane surface, as seen in Figure 4c. After the second procedure of ZIF-8 in situ growth, Figure 4d shows a homogeneous surface with fewer D


Research Article


scale were proven to grow within the observed spongy pores.32 Therefore, the existence of ZIF-8 nanoscale crystals within the spongelike pores of the PSF membrane cannot be fully excluded even there are no visible ZIF-8 nanoparticles formed within pores. The contact angles of the PSF and ZIF-8@PSF (#1−4) membranes (Table 1) characterize the membrane hydrophilic

pore openings but more ZIF-8 crystals growing on the membrane surface. The ZIF-8 crystals from the first in situ growth may act as seeds on the membrane surface and then are helpful for the next in situ growth procedure to further grow ZIF-8; consequently, the ZIF-8@PSF (#2) membrane shows a continuous ZIF-8 interlayer. In addition to the seeding effect, the surface porosity of the support membrane affects the growth of ZIF-8 in the pore openings and on the membrane surface.24 During the first and second in situ growths, the ZIF-8 nanoparticles first grow in the pore openings and then gradually cover the porous support membrane. The pore size of the PSF membrane may restrict the growth of ZIF-8 particles on the porous surface, giving ZIF-8 interlayers with particle sizes smaller than that if the ZIF-8 nanoparticles collected from the dispersion (shown in Figure 4a). Because the porous surface of the PSF membrane is covered by the ZIF-8 interlayer after the second growth, the seeding effect becomes the significant factor for the surface morphology of the ZIF-8@PSF (#3 and 4) membranes; the ZIF-8 crystals from the previous in situ growth may help the ZIF-8 growth in the next in situ growth. Therefore, the ZIF-8 nanoparticles with obvious rhombic dodecahedral morphology and particle size similar to that of the collected ZIF-8 nanoparticles are observed for the ZIF-8@PSF (#3) membrane, and the following (fourth) in situ growth continually increases ZIF-8 nanoparticles that pack together resulting in the formation of ZIF-8 aggregates (about several hundred nanometers). The morphologies of the ZIF-8@PSF (#1−4) membranes shown in Figure 4 indicate that the ZIF-8 interlayer can successfully grow on the PSF membrane through repeated ZIF-8 in situ growth. The cross-sectional SEM images of the ZIF-8@PSF (#4) membrane are shown in Figure 4g,h. The membrane sample in Figure 4g, which is the ZIF-8@PSF (#4) membrane without the post-treatment procedure to remove unreacted reagents, shows an interlayer of ZIF-8 nanoparticles covered by unreacted reagents. As shown in Figure 4h, the ZIF-8 nanoparticles still exist on the membrane surface after the post-treatment procedure, showing the well-ordered arrangement and uniform particle size of the ZIF-8 nanoparticles. The thickness of the ZIF-8 layer ranges from a single layer of ZIF-8 to overlapped ZIF-8 nanoparticles corresponding to 150−300 nm (on the basis of the particle size of ZIF-8), which matches the characteristic thickness of the selective layer of the TFC or TFN membrane (50−200 nm).53 No matter the before or after the post-treatment procedure, the ZIF-8 nanoparticles are only observed on the membrane surface but not within the spongelike pores of the PSF membrane. The observation from the cross-sectional view confirms that the ZIF-8 nanoparticles having a particle size of about 150 nm can grow on the membrane surface to form a thin layer through four in situ growths. The porous surface of the PSF membrane contacting the ZIF-8 suspension provides enough spaces to grow ZIF-8 nanoparticles on the membrane surface, and the spongelike porous structure of the PSF membrane restricts ZIF-8 from growing within the pores throughout the PSF membrane, resulting in the ZIF-8 nanoparticles predominantly growing on the membrane surface. In reported works about MOFs growing in/on polymeric membranes, it was found that some MOF microparticles or nanoparticles can grow within the finger pores of support membranes through in situ growth and interfacial synthesis approaches.24,31,36 No individual MOF nanoparticles can be seen in the spongy structures of the support membrane, but the MOF crystals on the nanometer

Table 1. Contact Angles of the ZIF-8@PSF and PA/ZIF-8 (LBL) Membranes.a contact angle (°) number of in situ growths 0 1 2 3 4

ZIF-8@PSF 76.0 80.3 81.5 88.7 90.7

± ± ± ± ±

1.5 2.7 1.0 1.5 2.9

PA/ZIF-8 (LBL) 70.6 70.3 75.4 77.5 80.7

± ± ± ± ±

1.7 2.2 1.8 1.6 1.0

a

The statistical errors were calculated from the standard deviation of three or four measurements.

properties. The ZIF-8@PSF (#1 and 2) membranes both have slightly larger contact angles than that of the PSF membrane because the hydrophilic property of PSF is dominant over the hydrophobicity of ZIF-8. However, the third and fourth growths significantly increase the ZIF-8 loadings on the membrane surfaces, which is attributed to more hydrophobic surfaces of the ZIF-8@PSF (#3 and #4) membranes. In particular, the ZIF-8@PSF (#4) membrane exhibits contact angle values similar to the literature data of ZIF-8 and ZIF-71 membranes27,30,31 because the hydrophobic ZIF-type MOF materials growing on porous supports can produce slightly hydrophobic membrane surfaces. Moreover, the contact angle measurement is also affected by the surface roughness, which increases the measured value because of the air trapping between the considerably rough surface and the water droplet.54 The ZIF-8 nanoparticles packed together on the membrane surfaces of the ZIF-8@PSF (#3 and 4) membranes, as shown in Figure 4d,e, may increase the surface roughnesses and thus increase the measured contact angles. 3.1.2. Characterization of PA/ZIF-8 (LBL) Membranes. The FTIR spectra of the PA and PA/ZIF-8 (LBL) membranes are shown in Figure 5. The characteristic bands of typical amide structures at 1660 and 1550 cm−1, which correspond to the CO stretching and the N−H in-plane bending and C−N

Figure 5. FTIR spectra of the PA and PA/ZIF-8 (LBL) membranes. E


Research Article


Figure 6. Surface and cross-sectional SEM images of PA membrane (a and d), PA/ZIF-8 (LBL-#3) membrane (b and e), and PA/ZIF-8 (LBL-#4) membrane (c and f).

through LBL fabrication may be related to the filler encapsulation mechanism and the increased hydrophobicity of the support membrane. Figure 6 shows the surface and cross-sectional morphologies of the PSF, PA, and PA/ZIF-8 (LBL-#3 and 4) membranes. The porous surface and ZIF-8 interlayers are seen from the SEM images of the PSF and ZIF-8@PSF membranes (Figure 4), respectively, whereas the PA and PA/ZIF-8 (LBL) membranes show dense surfaces without pores and nanoparticles, indicating that the PA layers formed on the support membranes. It is also observed from the high-resolution images of the PA/ZIF-8 (LBL-#4) membrane (Figure S1) that there are no interfacial defects/gaps showing on the membrane surface. The PA TFC membrane shows the typical ridge and valley surface morphology (Figure 6a), and the PA/ZIF-8 (LBL-#3 and 4) membranes both show rough surfaces with some bumps (Figure 6b,c). The presence of the ZIF-8 interlayer reduces the cross-linking extent of the selective layer, thus leading to different surface morphologies between the PA and PA/ZIF-8 (LBL-#3 and 4) membranes. As shown in Figure 4e,f, the ZIF-8@PSF (#4) from the fourth in situ growth has the ZIF-8 interlayer with more and larger ZIF-8 nanoparticles/aggregates than those from the third in situ growth. The PA/ZIF-8 (LBL-#4) membrane exhibits a rougher surface than the PA/ZIF-8 (LBL-#3) membrane, and the large bumps observed in Figure 6c due to the ZIF-8 aggregates are seen in Figure 4f. The cross-sectional SEM images in Figure 6 show that the thicknesses of the selective layers of the PA and PA/ZIF-8 (LBL) membranes are about 100−300 nm and that the selective layer of the PA membrane tends to be thicker than that of the PA/ZIF-8 (LBL-#3) membrane. As reported in literature, the thickness of the PA TFN membrane decreases with increasing loading of fillers in the PA layer because of the decrease in polymerization reactions to form PA.55 The uneven top surfaces of the PA and PA/ZIF-8 (LBL-#4) membranes are caused by the ridge and valley morphology for the PA membrane and the ZIF-8 nanoparticles/aggregates embedded in PA. Unlike the cross-sectional morphology of the ZIF-8@ PSF (#4) membrane, the ZIF-8 nanoparticles cannot be clearly seen from the dense layer of the PA/ZIF-8 (LBL-#4) membrane, indicating that the ZIF-8 interlayer is well-

stretching, respectively,55 are observed in the FTIR spectra of the PA TFC membrane, meaning that the cross-linked PA structure is formed through interfacial polymerization. The PA/ ZIF-8 (LBL) membranes show these characteristic bands of PA, but the bands become weaker with increasing in situ growths ZIF-8. It is also seen in Figure 5 that the PA/ZIF-8 (LBL) membranes exhibit new bands at 1755 and 1360 cm−1 compared with the PA membrane. These two peaks belong to the CO stretching and C−N stretching vibrations, respectively, corresponding to the amide structure of 1-acyl imidazole derivative.56 This additional amide structure in the nanocomposite selective layer is probably caused by a reaction between N−H groups of 2-methylimidazole in ZIF-8 and acyl chloride groups of TMC in the organic phase. The enhancement in these characteristic bands at 1755 and 1360 cm−1 indicates that more ZIF-8 nanoparticles may react with TMC during interfacial polymerization. The observations from Figure 5 mean that the ZIF-8 in situ growth prior to interfacial polymerization attributes to fewer cross-linking reactions between MPD and TMC but more reactions between ZIF-8 and TMC. It has been reported in literature that TFN membranes prepared through interfacial polymerization by dispersing fillers in organic phases were less cross-linked because of the increase in the viscosity of organic phases containing fillers, heat release from hydration of hydrophilic fillers, and the filler encapsulation mechanism.37,55 There is no distinct effect of the solution viscosity and filler hydration on the cross-linked PA structure in this study because the growth of the hydrophobic ZIF-8 nanoparticles on the membrane surface takes place prior to interfacial polymerization. However, the effect of filler encapsulation should be taken into account; the ZIF-8 nanoparticles encapsulated by PA during interfacial polymerization restricts the rate and pathway of MPD diffusing from the aqueous phase into the organic phase to react with TMC, resulting in the reduction in the degree of cross-linking.37 Moreover, the higher hydrophobicity of the ZIF-8@PSF membrane may attribute to lower adsorption of MPD monomers dissolved in the aqueous phase on the membrane surface, contributing to the lower degree of cross-linking of the PA/ZIF-8 (LBL) membrane. Therefore, the decreased crosslinking extent of the PA/ZIF-8 (LBL) membrane prepared F


Research Article


membranes. It has been found that the porous structure and hydrophilic property of the substrate has significant effects on the membrane properties and PA−substrate interaction, but the adhesion interaction between PA and porous substrate has not been well-documented and understood yet.59 For the PA/ZIF8 membranes with a LBL structure, the PA−substrate interaction may include the PA−(ZIF-8 interlayer), (ZIF-8 interlayer)−PSF, and PA−PSF interactions; to explore fundamental studies on these interactions is an interesting subject and should also be focused on in the future. 3.2. Nanofiltration Performance. 3.2.1. Water Permeance of the ZIF-8@PSF and PA/ZIF-8 (LBL) Membranes. Figure 7 shows the water fluxes of the ZIF-8@PSF and PA/

integrated with the PA layer because of the good compatibility between ZIF-8 and PA. However, the cross-sectional images of the PA/ZIF-8 (LBL-#3) membranes show a few chips having shapes and sizes similar to those of the ZIF-8 nanoparticles (featured by the red circle in Figure 6e), which may be ZIF-8 nanoparticles coated by PA. It indicates that the LBL fabrication affects the surface morphologies of the PA/ZIF-8 (LBL) membranes and that the ZIF-8 interlayer can be wellembedded in the PA layer. The surface roughness properties of the PA and PA/ZIF-8 (LBL) membranes were investigated by AFM analysis, and the roughness results are shown in Figure S2 and Table S1. The ridge and valley surface of the PA membrane and the rugged surfaces of the PA/ZIF-8 (LBL-#4) membrane in Figure S2 are consistent with the observations from the surface SEM images. The ZIF-8 in situ growth procedure leads to the increase in surface roughness, and the membrane roughness increases with increasing in situ growths. The PA membrane has the smoothest surface, exhibiting the Ra value of 11 nm, whereas the PA/ZIF-8 (LBL) membranes become rougher than the PA membrane. The Ra value of the PA/ZIF-8 (LBL-#4) membrane significantly increases to 45 nm, and the roughest surface is caused by the ZIF-8 aggregates on the membrane surface as seen in Figure 4f. The different roughness morphology and the increased surface roughness occur because of a combination of the reduction in interfacial polymerization reactions and the increase in the ZIF-8 nanoparticles/aggregates growing on the PSF membranes. The hydrophilic properties of the PA and PA/ZIF-8 (LBL) membranes characterized by contact angles are also summarized in Table 1. Compared with the contact angles of the PSF and ZIF-8@PSF membranes, the PA and PA/ZIF-8 (LBL) membranes are more hydrophilic than the corresponding support membranes because of the hydrophilic PA layers formed on the support membranes. As discussed above, the cross-linking extent of the PA/ZIF-8 (LBL) membrane decreases with increasing numbers of in situ growths. The lower degree of cross-linking of the PA (LBL) membrane should be attributed to a contact angle smaller than that of the PA TFC membrane because of more unreacted carboxylic acid groups of the PA layer.37,55,57 However, the PA/ZIF-8 (LBL) membranes in this study show larger contact angles than the PA membrane, which may be caused by the hydrophobic property of ZIF-8.12,38 As the number of in situ growth increases, the ZIF-8 loading on the membrane surface increases, and the PA layer coating on the ZIF-8 interlayer tends to be thinner than that of the pure PA membrane. Consequently, the hydrophobic ZIF-8 layer plays a more dominant role on the contact angle than the hydrophilic PA. The PA/ZIF-8 (LBL) membranes prepared in this work become rougher and more hydrophobic than the PA membrane. In general, the rough and hydrophobic surface easily causes membrane fouling owing to the large adsorption area to organic pollutants.58 Here the LBL fabrication produces the PA/ZIF-8 (LBL) membranes with rougher and more hydrophobic surfaces, which is not beneficial for antibiofouling applications. The possible ways to decrease the surface roughness may include growing hydrophilic MOFs on support membranes, decreasing MOF aggregates, and modifying membrane surfaces. These characterization results show that the presence of the ZIF-8 interlayer on the PSF membrane leads to different physicochemical properties between the PA/ZIF-8 and PA

Figure 7. Water permeance of the ZIF-8@PSF and PA/ZIF-8 (LBL) membranes.

ZIF-8 (LBL) membranes measured at 0.4 MPa. For the ZIF8@PSF membranes, the pure water flux decreases from 89.7 to 60.8 kg/m2/h after the first ZIF-8 in situ growth, and then considerably decreases after further in situ growths. The ZIF8@PSF (#4) membrane has the lowest water flux of 24.6 kg/ m2/h, about 27% of the water flux of the PSF membrane. The water permeance primarily depends on the membrane structure of a UF membrane; the water permeance increases with increasing pore size or decreasing membrane thickness. For the PSF and ZIF-8@PSF membranes, ZIF-8 in situ growth has little effect on the membrane thickness but significantly changes the surface morphology. The porous surface of the PSF membrane is gradually coved by the ZIF-8 layer as the number of ZIF-8 in situ growth increases as shown in Figure 4. Therefore, the decrease in the pore openings on the membrane surface gives rise to a resistance of water transport through the membrane resulting in the reduction in water flux. The organic solvent (methanol) used in the in situ growth procedure may lead to membrane swelling that enlarges pores of PSF; oppositely, the heat treatment after the in situ growth procedure may shrink pores of PSF. Overall, the ZIF-8@PSF membranes exhibit reduced water permeance because of the ZIF-8 layer on the membrane surfaces blocking transport pathways. The PA and PA/ZIF-8 (LBL) membranes exhibit significantly lower water fluxes than do the PSF and ZIF-8@PSF membranes because the dense PA layer obtained through interfacial polymerization is highly cross-linked and has small free-volume pore size. Unlike the reduction in water permeance of the PSF and ZIF-8@PSF membranes, the water permeance G


Research Article


Figure 8. Fluxes (a) and rejections (b) of the PA, PA/ZIF-8 (O), and PA/ZIF-8 (LBL-#4) membranes.

interfacial polymerization using the same recipe and methodology. The water-soluble solutes were inorganic salt (MgSO4, 1000 ppm), organic pollutant with low molecular weight (phenol, 100 ppm), and dye (Congo red, 200 ppm). As shown in Figure 8, the presence of ZIF-8 in PA attributes to the enhancement in flux and the reduction in rejection, and the PA/ZIF-8 (LBL-#4) membrane exhibits better performance in both flux and rejection than does the PA/ZIF-8(O) membrane. As mentioned above and in literature,12,37,38 the permeance of the nanocomposite membranes is improved because of a combination of the porous structure of ZIF-8, the hydrophobicity of ZIF-8, and the decrease in the cross-linking degree of the nanocomposite layer. The good compatibility between ZIF-8 and PA is good for avoiding the interfacial defects formed in the nanocomposite layer and then avoiding the significant reduction in membrane selectivity. Although it was proven that the inner voids of ZIF-8 aggregates were nonselective mesopores, they may exist in membranes at a high ZIF-8 loading.25 The ZIF-8 aggregates embedded in PA contribute to the nanocomposite layer with nonselective voids; consequently, the rejections of the PA/ZIF-8(O) and PA/ZIF-8 (LBL-#4) membrane are both lower than that of the pure PA membrane. Moreover, the thinner and looser PA layers of the PA/ZIF-8(O) and PA/ZIF-8 (LBL-#4) membranes also may give rise to higher fluxes but lower rejection. The PA/ZIF-8 (LBL-#4) membrane has higher flux and rejection than does the PA/ZIF-8(O) membrane, indicating that the PA nanocomposite membrane has a higher ZIF-8 loading and fewer nonselective voids. Therefore, the LBL fabrication in this study may increase the ZIF-8 loading but decrease the ZIF-8 aggregates and is more favorable than the traditional interfacial polymerization using the ZIF-8 dispersion. With respect to the performance of the PA/ZIF-8 (LBL-#4) membrane for removing salt, phenol, and dye from water, the flux enhancements are 182, 198, and 242%, respectively, primarily caused by the increase in the porosity of the nanocomposite selective layer. Compared with the permeance for desalination and phenol removal, the lower fluxes for dye removal are due to the effects of concentration polarization and membrane fouling on the membrane performance;60 the high concentration of Congo red (200 ppm) in this study may enhance these effects, contributing to the flux decline. Unlike the significantly enhanced fluxes, the selectivity of the PA/ZIF8 (LBL-#4) membrane decreases except for the dye removal result; the rejections for desalination and phenol removal are about 65 and 75% of the rejections of the PA membrane, respectively. The mesopores formed by the ZIF-8 aggregates

of the PA and PA/ZIF-8 membranes tends to increase with increasing numbers of ZIF-8 in situ growths. Compared with the flux of the PA membrane, the PA/ZIF-8 (LBL-#1 and 2) membranes show slightly lower and similar fluxes, respectively, which may be caused by the reduction in the pore openings of the support membrane as explained above and fewer ZIF-8 nanoparticles embedded in PA. After the further procedures of ZIF-8 in situ growth, the greater number of ZIF-8 nanoparticles embedded in PA increase water permeance because of the microporous and hydrophobic structure of ZIF-8. Moreover, the degree of cross-linking decreases with increasing numbers of ZIF-8 in situ growths, which also results in faster water permeance. Therefore, the membrane with the fourth in situ growth produces the highest water flux among the PA/ZIF-8 (LBL) membranes of 17.2 kg/m2/h, and the flux enhancements (defined by the flux ratio of the PA/ZIF-8 (LBL) membrane to the PA membrane) reaches about 221%. We note that the permeance difference between the support and composite membranes decreases with increasing numbers of ZIF-8 in situ growths. It can be expected that the water permeance of a ZIF-8@PSF membrane having a denser ZIF-8 layer on the support membrane may be close to that of the corresponding PA/ZIF-8 (LBL) membrane if the number of ZIF-8 in situ growths continues to increase, and the dense ZIF8@PSF membrane may be directly used as a NF membrane. It has been reported that MOF membranes prepared on polymeric UF membranes for NF processes can exhibit much higher rejections but much lower fluxes than the porous UF membranes,31,32,36 whereas the NF performance of these MOF membrane cannot reach the higher performance of the PA TFC membranes, especially for the high selectivity. This study focuses on the improvement in NF performance of the aromatic PA membrane based on the microporous structure of ZIF-8 to prepare high-performance membranes with enhanced permeance without deteriorating selectivity. The separation properties of the PA/ZIF-8 (LBL) membranes will be further investigated by removing solutes from water by NF processes. 3.2.2. Effect of Membrane Fabrication on NF Performance. To investigate NF performance for separating aqueous mixtures, the pure PA membrane, the PA/ZIF-8(O) membrane in our previous work,38 and the PA/ZIF-8 (LBL-#4) membrane in this study were tested at the operating pressure of 1.0 MPa using three different solutes. The PA/ZIF-8(O) membrane was prepared by dispersing presynthesized ZIF-8 nanoparticles in the organic phase (0.20% (w/v) ZIF-8 concentration) prior to interfacial polymerization. The PA, PA/ZIF-8 (O), and PA/ ZIF-8 (LBL-#4) membranes were all fabricated through H


Research Article


membranes and tend to be located at the bottom region of the PA layer after interfacial polymerization. The decrease in pore openings of the support membrane increases the resistance to water transport, and the ZIF-8 location leads to fewer ZIF-8 nanoparticles districted from the top to middle region of the dense PA layer. Consequently, the permeance of the PA/ZIF-8 (LBL-#1 and 2) membranes cannot be significantly enhanced. As proven through the FTIR results, the PA/ZIF-8 (LBL-#1 and 2) membranes are less cross-linked than is the pure PA membrane. Although the lower degree of cross-linking cannot considerably enlarge the free-volume pores of the PA layer to allow Congo red molecules to pass through the PA/ZIF-8 (LBL-#1 and 2) membranes, there is no enhanced water transport for the PA/ZIF-8 (LBL-#1 and 2) membranes as shown in Figure 7. Therefore, the rejections of these PA/ZIF-8 (LBL) membranes are nearly the same as that of the PA membrane, meaning that the first and second ZIF-8 in situ growths have little influence on the improvement in membrane performance of the PA/ZIF-8 (LBL) membrane. The fluxes of the PA/ZIF-8 (LBL-#3 and 4) membranes (20.1 and 27.1 kg/m2/h, respectively) are considerably enhanced compared with that of the PA membrane (11.2 kg/ m2/h). The ZIF-8 nanoparticles growing on the ZIF-8@PSF (LBL-#3 and 4) membranes become much greater than the first and second in situ growths as shown in Figure 4 and are well-embedded in the PA layers through interfacial polymerization as shown in Figure 6. Consequently, the ZIF-8 loadings in the PA/ZIF-8 (LBL-#3 and 4) membranes increase. The particle size of the ZIF-8 nanoparticles is comparable to the thickness of the PA/ZIF-8 (LBL-#3 and 4) layers, which leads to the ZIF-8 nanoparticles locating in the PA/ZIF-8 (LBL-#3 and 4) layer from top to bottom. The ZIF-8 loading and distribution both significantly increase the porosity of the dense PA/ZIF-8 (LBL) layer, resulting in the increase in the ability of water transport of the PA/ZIF-8 (LBL-#3 and 4) membranes. Moreover, the enhancement in flux is also caused by the hydrophobic structure of ZIF-8, the decrease in the crosslinking degree and thickness of the PA layer, and/or the existence of nonselective voids/interfacial nanogaps. The third in situ growth results in the PA/ZIF-8 (LBL-#3) membrane exhibiting faster water permeance, and the pore size of PA/ZIF8 rejects Congo red molecules from passing through the membrane. Therefore, the PA/ZIF-8 (LBL-#3) membrane has a slightly higher rejection (99.8%) than that of the PA membrane (99.6%), whereas the fourth in situ growth results in a decrease in the rejection (99.2%) with respect to that of the PA membrane, owing to the increase in the nonselective voids and/or interfacial nanogaps at the higher ZIF-8 loading in PA. The PA/ZIF-8 (LBL-3) membrane with the enhanced permeance and best selectivity is suitable for removing solutes with a molecular size comparable to or smaller than that of Congo red; the membrane performance may be further enhance at higher operating pressures. For removing larger solutes, the PA/ZIF-8 (LBL-#4) membrane is better because of the enhanced flux and good rejection. The separation results indicate that the regulation of the ZIF8 interlayer is important to the PA/ZIF-8 membrane performance, which is affected by the ZIF-8 loading and nonselective voids in the PA/ZIF-8 nanocomposite layer. The well-ordered distribution of ZIF-8 nanoparticles in PA that improves both the permeance and selectivity is related to the particle size, number of particles, particle aggregates, and PA− particles interface which can be regulated through the recipe of

and the larger free-volume pores of the PA layer having the lower degree of cross-linkage both provide nonselective transport pathways to solutes with relative small molecular sizes. In addition, the existence of interfacial nanogaps in the PA/ZIF-8 (LBL-#4) membrane cannot be fully excluded because of the ultrathin PA layer coating. Because the ion/ molecular sizes of MgSO4 and phenol are both much smaller than that of Congo red, the nonselective pathways and/or interfacial nanogaps allow many more MgSO4 ions and phenol molecules to transport through the PA/ZIF-8 (LBL-#4) membrane, resulting in significantly lower rejections for desalination and phenol removal. The differences between the rejections for removing MgSO4 and phenol may be caused by the electrostatic repulsion mechanism on differently charged solutes and the different solute concentration, which both play important roles on NF performance. On the basis of reported works about PA TFN membranes, adjusting the membrane fabrication can improve the membrane selectivity for salts or small organic solutes, such as the lower ZIF-8 loading, longer reaction time of interfacial polymerization, higher monomer concentration, and so on.37,61 Note that the PA/ZIF-8 (LBL-#4) membrane shows the significantly improved permeance without deteriorating selectivity for removing Congo red (flux enhancement of 242%, rejection >99%). The high performance of the PA/ZIF-8 (LBL#4) membrane for dye removal indicates that the PA/ZIF-8 (LBL) membrane prepared through the LBL fabrication has a great potential to be applied in water treatments for removing large organic solutes. 3.2.3. Effect of in Situ Growth on Dye-Removal Performance. The effect of ZIF-8 in situ growth on the NF performance of the PA/ZIF-8 (LBL) membrane was studied by carrying out NF experiments for dye removal at 1.0 MPa operating pressure, and the separation results are shown in Figure 9. Similar to the water permeance of the PA and PA/

Figure 9. NF performance of the PA/ZIF-8 (LBL) membranes for dye removal.

ZIF-8 (LBL) membranes shown in Figure 7, the flux first decreases and then increases with increasing numbers of ZIF-8 in situ growths from 0 to 4. Considering the statistical errors of the measure values, the first and second in situ growths cannot enhance membrane permeance. As seen in the surface SEM images of the ZIF-8@PSF membranes (Figure 4), the ZIF-8 layers on the membrane surface introduced from the first and second in situ growths fill the pore openings of the PSF I


Research Article


m2/h) without deterioration of the selectivity. This study proposed a novel LBL fabrication for PA/ZIF-8 (LBL) nanocomposite membranes with high performance for NF and RO processes, and the LBL fabrication may be further applied in the preparation of other PA/MOF nanocomposite membranes.

ZIF-8 synthesis, the duration of in situ growth, the posttreatment of in situ growth, the number of in situ growth, the properties of the support membrane, and the interfacial polymerization fabrication. This study shows that the membrane permeance can be regulated by ZIF-8 in situ growth, but the membrane selectivity may decrease because of the ZIF-8 aggregation. It can be expected to improve membrane performance further by optimizing the regulation of the ZIF-8 interlayer, growing other MOF materials with different pore sizes, functionalizing MOFs with functional groups, and so on. Table S2 compares the NF performance between the PA/ ZIF-8 (LBL) membranes prepared in this study and the PA/ MOF TFN membranes reported in literature. The PA/ZIF-8 (LBL-#3 and 4) membranes have higher permeance values or flux enhancements than those of the PA/ZIF-8 TFN membrane, and the rejections of all PA/ZIF-8 membranes are on the same level (from 99.2 to 99.8%). The smaller pore sizes of NH2-MIL-53 and MIL-53 lead to perfect selectivity with the rejections of nearly 100%, but the permeance of the PA/NH2MIL-53 and PA/MIL-53 membranes cannot been significantly enhanced. Although the PA TFN membrane prepared using MIL-101 of the largest micropores exhibits the highest permeance, the flux enhancement and rejection are both lower than those of the PA/ZIF-8 (LBL-3) and PA/ZIF-8 (LBL-#4) membranes. The PA/ZIF-8 (LBL-3) membrane has a rejection of 99.8% and a flux enhancement of up to 179%, whereas the PA/ZIF-8 (LBL-#4) membrane shows the highest flux enhancement and a rejection still higher than 99%. The NF results of the PA/MOF membranes indicated that the MOF materials embedded in PA can improve membrane performance, and the LBL fabrication has good potential to be a feasible methodology for the preparation of PA/MOF (LBL) membranes with better performance than that of the conventional fabrication.

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsami.5b07128. HR-SEM images of the PA/ZIF-8 (LBL-#4) membrane; AFM images of the PA membrane and PA/ZIF-8 (LBL) membranes (LBL-#1−4), roughness analysis of the PA and PA/ZIF-8 (LBL) membranes, and membrane performance of PA/MOF TFN membranes. (PDF)

■

AUTHOR INFORMATION

Corresponding Author

*Tel.: +86-10-62337251. E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We greatly appreciate the financial support of the Fundamental Research Funds for the Central Universities (no. BLX2013007), the National Natural Science Foundation of China (no. 21406013, 21576029), the Beijing Natural Science Foundation (no. 2154054), and the Scientific Research Foundation for the Returned Overseas Chinese Scholars, State Education Ministry (no. 14JLX-03).

■

REFERENCES

(1) Qin, F.; Li, S.; Qin, P.; Karim, M. N.; Tan, T. A PDMS membrane with high pervaporation performance for the separation of furfural and its potential in industrial application. Green Chem. 2014, 16, 1262−1273. (2) Inukai, S.; Cruz-Silva, R.; Ortiz-Medina, J.; Morelos-Gomez, A.; Takeuchi, K.; Hayashi, T.; Tanioka, A.; Araki, T.; Tejima, S.; Noguchi, T.; Terrones, M.; Endo, M. High-performance multifunctional reverse osmosis membranes obtained by carbon nanotube·polyamide nanocomposite. Sci. Rep. 2015, 5, 13562. (3) Gu, J.-E.; Lee, J. S.; Park, S.-H.; Kim, I. T.; Chan, E. P.; Kwon, Y.N.; Lee, J.-H. Tailoring interlayer structure of molecular layer-by-layer assembled polyamide membranes for high separation performance. Appl. Surf. Sci. 2015, 356, 659−667. (4) Zhang, C.; Dai, Y.; Johnson, J. R.; Karvan, O.; Koros, W. J. High performance ZIF-8/6FDA-DAM mixed matrix membrane for propylene/propane separations. J. Membr. Sci. 2012, 389, 34−42. (5) Tao, K.; Kong, C.; Chen, L. High performance ZIF-8 molecular sieve membrane on hollow ceramic fiber via crystallizing-rubbing seed deposition. Chem. Eng. J. 2013, 220, 1−5. (6) Qin, X.; Yuan, Q.; Zhao, Y.; Xie, S.; Liu, Z. Measurement of the rate of water translocation through carbon nanotubes. Nano Lett. 2011, 11, 2173−2177. (7) Holt, J. K.; Park, H. G.; Wang, Y.; Stadermann, M.; Artyukhin, A. B.; Grigoropoulos, C. P.; Noy, A.; Bakajin, O. Fast Mass Transport Through Sub-2-Nanometer Carbon Nanotubes. Science 2006, 312, 1034−1037. (8) Majumder, M.; Chopra, N.; Andrews, R.; Hinds, B. J. Nanoscale hydrodynamics: Enhanced flow in carbon nanotubes. Nature 2005, 438, 44. (9) Cohen-Tanugi, D.; Grossman, J. C. Water desalination across nanoporous graphene. Nano Lett. 2012, 12, 3602−3608.

4. CONCLUSIONS A novel fabrication for PA/ZIF-8 (LBL) membranes was developed by combining ZIF-8 in situ growth and interfacial polymerization: First, an ZIF-8 interlayer grew on the porous surface of a support membrane through ZIF-8 in situ growth. Then, a PA layer was coated on the ZIF-8 interlayer to obtain a PA/ZIF-8 selective layer through interfacial polymerization. A series of PA/ZIF-8 (LBL) membranes were prepared by increasing the number of in situ growths from one to four, and the membrane properties were investigated. Increasing the numbers of ZIF-8 in situ growth procedures led to more ZIF-8 nanoparticles growing on the PSF membrane and produced the PA/ZIF-8 (LBL) membrane with higher ZIF-8 loading but lower degree of cross-linking. After the second in situ growth, the hydrophobicity and roughness of the PA/ZIF-8 (LBL) membrane were both considerably enhanced with respect to that of the PA membrane, and the typical ridge and valley surface morphology of the PA TFC membrane disappeared. Because of the micropores and hydrophobic property of ZIF-8, the flux of the PA/ZIF-8 (LBL-#4) membrane was at least 182% of that of the PA membrane for pure water transport or separating aqueous mixtures by NF. The prepared PA/ZIF-8 (LBL) membranes exhibited favorable selectivity (rejection >99%) for removing dye (Congo red) from water: The PA/ ZIF-8 (LBL-#3) membrane with a flux of 20.1 kg/m2/h and a rejection of 99.8% exhibited the best performance, and the PA/ ZIF-8 (LBL-#4) membrane showed the highest flux (27.1 kg/ J


Research Article

ACS Applied Materials & Interfaces (10) Huang, K.; Liu, G.; Shen, J.; Chu, Z.; Zhou, H.; Gu, X.; Jin, W.; Xu, N. High-efficiency water-transport channels using the synergistic effect of a hydrophilic polymer and graphene oxide laminates. Adv. Funct. Mater. 2015, 25, 5809−5815. (11) Liu, X.; Li, Y.; Zhu, G.; Ban, Y.; Xu, L.; Yang, W. An organophilic pervaporation membrane derived from metal-organic framework nanoparticles for efficient recovery of bio-alcohols. Angew. Chem., Int. Ed. 2011, 50, 10636−10639. (12) Sorribas, S.; Gorgojo, P.; Téllez, C.; Coronas, J.; Livingston, A. G. High flux thin film nanocomposite membranes based on metal− organic frameworks for organic solvent nanofiltration. J. Am. Chem. Soc. 2013, 135, 15201−15208. (13) Xiang, Z.; Cao, D. Porous covalent−organic materials, synthesis, clean energy application and design. J. Mater. Chem. A 2013, 1, 2691− 2718. (14) Xiang, Z.; Cao, D. Synthesis of luminescent covalent-organic polymers for detecting nitroaromatic explosives and small organic molecules. Macromol. Rapid Commun. 2012, 33, 1184−1190. (15) Du, N.; Park, H. B.; Robertson, G. P.; Dal-Cin, M. M.; Visser, T.; Scoles, L.; Guiver, M. D. Polymer nanosieve membranes for CO2capture applications. Nat. Mater. 2011, 10, 372−375. (16) Friebe, S.; Wang, N.; Diestel, L.; Liu, Y.; Schulz, A.; Mundstock, A.; Caro, J. Deuterium/hydrogen permeation through different molecular sieve membranes: ZIF, LDH, zeolite. Microporous Mesoporous Mater. 2015, 216, 127−132. (17) Zornoza, B.; Seoane, B.; Zamaro, J. M.; Téllez, C.; Coronas, J. Combination of MOFs and zeolites for mixed-matrix membranes. ChemPhysChem 2011, 12, 2781−2785. (18) Wang, J.; Li, Y.; Zhang, Z.; Hao, Z. Mesoporous KIT-6 silica− polydimethylsiloxane (PDMS) mixed matrix membranes for gas separation. J. Mater. Chem. A 2015, 3, 8650−8658. (19) Liu, Y.; Wang, N.; Pan, J. H.; Steinbach, F.; Caro, J. In-situ synthesis of MOF membranes on ZnAl-CO3 LDH buffer layermodified substrates. J. Am. Chem. Soc. 2014, 136 (41), 14353−14356. (20) Liao, J.; Wang, Z.; Gao, C.; Wang, M.; Yan, K.; Xie, X.; Zhao, S.; Wang, J.; Wang, S. A high performance PVAm−HT membrane containing high-speed facilitated transport channels for CO2 separation. J. Mater. Chem. A 2015, 3, 16746−16761. (21) Kwon, H. T.; Jeong, H.-K. In situ synthesis of thin zeolitic− imidazolate framework ZIF-8 membranes exhibiting exceptionally high propylene/propane separation. J. Am. Chem. Soc. 2013, 135, 10763− 10768. (22) Liu, D.; Ma, X.; Xi, H.; Lin, Y. S. Gas transport properties and propylene/propane separation characteristics of ZIF-8 membranes. J. Membr. Sci. 2014, 451, 85−93. (23) Dumée, L.; He, L.; Hill, M.; Zhu, B.; Duke, M.; Schütz, J.; She, F.; Wang, H.; Gray, S.; Hodgson, P.; Kong, L. Seeded growth of ZIF-8 on the surface of carbon nanotubes towards self-supporting gas separation membranes. J. Mater. Chem. A 2013, 1, 9208−9214. (24) Nagaraju, D.; Bhagat, D. G.; Banerjee, R.; Kharul, U. K. In situ growth of metal-organic frameworks on a porous ultrafiltration membrane for gas separation. J. Mater. Chem. A 2013, 1, 8828−8835. (25) Fang, M.; Wu, C.; Yang, Z.; Wang, T.; Xia, Y.; Li, J. ZIF-8/ PDMS mixed matrix membranes for propane/nitrogen mixture separation, Experimental result and permeation model validation. J. Membr. Sci. 2015, 474, 103−113. (26) Lin, R.; Ge, L.; Hou, L.; Strounina, E.; Rudolph, V.; Zhu, Z. Mixed matrix membranes with strengthened MOFs/polymer interfacial interaction and improved membrane performance. ACS Appl. Mater. Interfaces 2014, 6, 5609−5618. (27) Dong, X.; Lin, Y. Synthesis of an organophilic ZIF-71 membrane for pervaporation solvent separation. Chem. Commun. 2013, 49, 1196−1198. (28) Kang, C.; Lin, Y.; Huang, Y.; Tung, K.; Chang, K.; Chen, J.; Hung, W.; Lee, K.; Lai, J. Synthesis of ZIF-7/chitosan mixed-matrix membranes with improved separation performance of water/ethanol mixtures. J. Membr. Sci. 2013, 438, 105−111. (29) Fan, H.; Shi, Q.; Yan, H.; Ji, S.; Dong, J.; Zhang, G. Simultaneous spray self-assembly of highly loaded ZIF-8−PDMS

nanohybrid membranes exhibiting exceptionally high biobutanolpermselective pervaporation. Angew. Chem., Int. Ed. 2014, 53, 5578. (30) Huang, K.; Li, Q.; Liu, G.; Shen, J.; Guan, K.; Jin, W. A ZIF-71 hollow fiber membrane fabricated by contra-diffusion. ACS Appl. Mater. Interfaces 2015, 7, 16157. (31) Li, Y.; Wee, L. H.; Volodin, A.; Martens, J. A.; Vankelecom, I. F. J. Polymer supported ZIF-8 membranes prepared via an interfacial synthesis method. Chem. Commun. 2015, 51, 918−920. (32) Campbell, J.; Szekely, G.; Davies, R.; Braddock, D. C.; Livingston, A. G. Fabrication of hybrid polymer/metal organic framework membranes, Mixed matrix membranes versus in-situ growth. J. Mater. Chem. A 2014, 2, 9260−9271. (33) Zhang, R.; Ji, S.; Wang, N.; Wang, L.; Zhang, G.; Li, J. Coordination-driven in situ self-assembly strategy for the preparation of metal-organic framework hybrid membranes. Angew. Chem. 2014, 126, 9933−9937. (34) Basu, S.; Maes, M.; Cano-Odena, A.; Alaerts, L.; DeVos, D. E.; Vankelecom, I. F. J. Solvent resistant nanofiltration (SRNF) membranes based on metal-organic frameworks. J. Membr. Sci. 2009, 344, 190−198. (35) Sani, N. A. A.; Lau, W. J.; Ismail, A. F. Polyphenylsulfone (PPSU)-based solvent resistant nanofiltration (SRNF) membrane incorporated with Cu-BTC nanoparticles for methanol separation. RSC Adv. 2015, 5, 13000−13010. (36) Campbell, J.; Davies, R. P.; Braddock, D. C.; Livingston, A. G. Improving the permeance of hybrid polymer/metal−organic framework (MOF) membranes for organic solvent nanofiltration (OSN) − development of MOF thin films via interfacial synthesis. J. Mater. Chem. A 2015, 3, 9668−9674. (37) Duan, J.; Pan, Y.; Pacheco, F.; Litwiller, E.; Lai, Z.; Pinnau, I. High-performance polyamide thin-film-nanocomposite reverse osmosis membranes containing hydrophobic zeolitic imidazolate framework-8. J. Membr. Sci. 2015, 476, 303−310. (38) Wang, L.; Fang, M.; Liu, J.; Deng, L.; He, J.; Li, J.; Lei, J. Effect of dispersed phases on polyamide/ZIF-8 nanofiltration membranes for the removal of dye from water. RSC Adv. 2015, 5, 50942−50954. (39) Bux, H.; Liang, F.; Li, Y. S.; Cravillon, J.; Wiebcke, M.; Caro, J. Zeolitic imidazolate framework membrane with molecular sieving properties by microwave-assisted solvothermal synthesis. J. Am. Chem. Soc. 2009, 131, 16000−16001. (40) Xu, G.; Yao, J.; Wang, K.; He, L.; Webley, P. A.; Chen, C.-S.; Wang, H. Preparation of ZIF-8 membranes supported on ceramic hollow fibers from a concentrated synthesis gel. J. Membr. Sci. 2011, 385−386, 187−193. (41) Liu, Y.; Hu, E.; Khan, E. A.; Lai, Z. P. Synthesis and characterization of ZIF-69 membranes and separation for CO2/CO mixture. J. Membr. Sci. 2010, 353, 36−40. (42) Venna, S. R.; Carreon, M. A. Highly permeable zeolite imidazolate framework-8 membranes for CO2/CH4 separation. J. Am. Chem. Soc. 2010, 132, 76−78. (43) Li, Y. S.; Bux, H.; Feldhoff, A.; Li, G.; Yang, W.; Caro, J. Controllable synthesis of metal-organic frameworks: from MOF nanorods to oriented MOF membranes. Adv. Mater. 2010, 22, 3322−3326. (44) Pan, Y.; Wang, B.; Lai, Z. P. Synthesis of ceramic hollow fiber supported zeolitic imidazolate framework-8 (ZIF-8) membranes with high hydrogen permeability. J. Membr. Sci. 2012, 421−422, 292−298. (45) Biswal, B. P.; Bhaskar, A.; Banerjee, R.; Kharul, U. K. Selective interfacial synthesis of metal−organic frameworks on a polybenzimidazole hollow fiber membrane for gas separation. Nanoscale 2015, 7, 7291. (46) Brown, A. J.; Brunelli, N. A.; Eum, K.; Rashidi, F.; Johnson, J. R.; Koros, W. J.; Jones, C. W.; Nair, S. Interfacial microfluidic processing of metal-organic framework hollow fiber membrane. Science 2014, 345, 72−75. (47) Tao, K.; Kong, C.; Chen, L. High performance ZIF-8 molecular sieve membrane on hollow ceramic fiber via crystallizing-rubbing seed deposition. Chem. Eng. J. 2013, 220, 1−5. K


Research Article

ACS Applied Materials & Interfaces (48) Kong, L.; Zhang, X.; Liu, H.; Wang, T.; Qiu, J. Preparation of ZIF-8 membranes supported on macroporous carbon tubes via a dipcoating−rubbing method. J. Phys. Chem. Solids 2015, 77, 23−29. (49) Cacho-Bailo, F.; Caro, G.; Etxeberría-Benavides, M.; Karvan, O.; Téllez, C.; Coronas, J. High selectivity ZIF-93 hollow fiber membranes for gas separation. Chem. Commun. 2015, 51, 11283−11285. (50) Zhang, X.; Liu, Y.; Kong, L.; Liu, H.; Qiu, J.; Han, W.; Weng, L.T.; Yeung, K. L.; Zhu, W. A simple and scalable method for preparing low-defect ZIF-8 tubular membranes. J. Mater. Chem. A 2013, 1, 10635−10638. (51) Cacho-Bailo, F.; Catalán-Aguirre, S.; Etxeberría-Benavides, M.; Karvan, O.; Sebastian, V.; Téllez, C.; Coronas, J. Metal-organic framework membranes on the inner-side of a polymeric hollow fiber by microfluidic synthesis. J. Membr. Sci. 2015, 476, 277−285. (52) Moggach, S. A.; Bennett, T. D.; Cheetham, A. K. The effect of pressure on ZIF-8: Increasing pore size with pressure and the formation of a high-pressure phase at 1.47 GPa. Angew. Chem. 2009, 121, 7221−7223. (53) Lind, M. L.; Ghosh, A. K.; Jawor, A.; Huang, X.; Hou, W.; Yang, Y.; Hoek, E. M. V. Influence of zeolite crystal size on zeolite-polyamide thin film nanocomposite membranes. Langmuir 2009, 25, 10139− 10145. (54) Quéré, D. Rough ideas on wetting. Phys. A 2002, 313, 32−46. (55) Namvar-Mahboub, M.; Pakizeh, M.; Davari, S. Preparation and characterization of UZM-5/polyamide thin film nanocomposite membrane for dewaxing solvent recovery. J. Membr. Sci. 2014, 459, 22−32. (56) Chen, R.; Liu, T.; Zuo, X. Synthesis of some 1-acyl benzimidazole and 1-acyl imidazole derivatives. Chem. J. Chin. Univ. 1984, 5, 361−365. (57) Vrijenhoek, E. M.; Hong, S. K.; Elimelech, M. Influence of membrane surface properties on initial rate of colloidal fouling of reverse osmosis and nanofiltration membranes. J. Membr. Sci. 2001, 188, 115−128. (58) Yin, J.; Deng, B. Polymer-matrix nanocomposite membranes for water treatment. J. Membr. Sci. 2015, 479, 256−275. (59) Lau, W. J.; Ismail, A. F.; Misdan, N.; Kassim, M. A. A recent progress in thin film composite membrane: A review. Desalination 2012, 287, 190−199. (60) Zhang, X.; Yuan, Q.; Fang, L.; Yang, X. Experimental studies on transport flux and rejection for nanofiltration in organic solvents. Beijing Huagong Daxue Xuebao 2004, 31, 5−8. (61) Kim, H. J.; Lim, M.-Y.; Jung, K. H.; Kim, D.-G.; Lee, J.-C. Highperformance reverse osmosis nanocomposite membranes containing the mixture of carbon nanotubes and graphene oxides. J. Mater. Chem. A 2015, 3, 6798−6809.

L


ZIF-8

Recommend Documents