Contents lists available at ScienceDirect

Journal of Membrane Science

journal homepage: www.elsevier.com/locate/memsci

Unidirectional diffusion synthesis of covalent organic frameworks (COFs) onpolymeric substrates for dye separation

Rui Wang, Xiansong Shi, Zhe Zhang, Ankang Xiao, Shi-Peng Sun, Zhaoliang Cui, Yong Wang∗

State Key Laboratory of Materials-Oriented Chemical Engineering, College of Chemical Engineering, Jiangsu National Synergetic Innovation Center for Advanced Materials,Nanjing Tech University, Nanjing, 211816, PR China

A R T I C L E I N F O

Keywords:Covalent organic frameworksUnidirectional diffusion synthesisComposite membranesNanofiltrationDye separation

A B S T R A C T

Covalent organic frameworks (COFs) are regarded as important candidates for the preparation of advancedmembranes with remarkable performances on account of their uniform and tailorable channels and good che-mical stability. The development of COFs-based membranes is burgeoning, and there remains lacking of facileand mild methods to prepare to these membranes. Herein, we report on the unidirectional diffusion synthesis(UDS) of TpPa (an imine-based COFs) on commercial polyvinylidene fluoride (PVDF) microfiltration substratesfor the preparation of composite COFs-based membranes at room temperature. The p-phenylenediamine (Pa)aqueous solution and 1,3,5-triformylphloroglucinol (Tp) n-hexane solution were separately charged into the twosides of a diffusion cell. With the suitable solution pairs and molar ratio of two precursors, only Pa molecules canpass through the macropores of PVDF to react with Tp molecules at the phase interface formed by this solutionpairs, leading to the in situ synthesis of TpPa layer on the top side of PVDF substrates. The membrane fabricatedwith a synthesized duration of 24 h exhibits a superior selectivity for dyes (> 90%) and appreciable waterpermeance (60 L·m−2·h−1·bar−1). The permeance of the resultant membrane is ∼2–20 times higher than otherreported membranes with comparable dye rejections. The UDS offers a facile and mild process to directly growCOFs on porous polymeric substrates to construct composite membranes capable of efficient separations of dyemolecules.

1. Introduction

Covalent organic frameworks (COFs) are a new category of crys-talline porous polymers with predesigned skeletons and two or three-dimensional (2D or 3D) well-ordered nanopores, and constructed withmolecular building units via strong covalent bonds [1–6]. These uniquefeatures of facilely tailored functionalities, outstanding chemical sta-bility, permanent porosity as well as ordered and tunable pore structureendow COFs a broad range of potential applications including optoe-lectronics [7–10], semiconductors [11–13], gas storage and separation[14–18], energy conversion [19–21], catalysis [22–24]. In particular,imine-based COFs are superior to other COFs counterparts with regardto crystallinity and chemical stability (including water, acid and basestability) [25,26], making them greater possibility for water treatmentin harsh environment [27–30].

Recently, some significant progresses including fabrication of COFsto form high-performance separation membranes by solid-state mixingapproach [27,28] and in situ growth [31–34], as well as enhancement ofmembrane performances by introducing COFs interlayers [35] or

matrix [36–38] for gas separation and liquid purification have beenreported. In these studies, COFs usually synthesized under solvothermalconditions which require elevated temperatures, long durations andaggressive solvents, such as mesitylene, 1,4-dioxane, N, N-di-methylformamide (DMF), and N, N-dimethylacetamide (DMAc) [1,39].Clearly, it is extremely harsh for most of polymeric substrates to bearthe aggressive solvents and high temperature conditions for a longduration in the solvothermal synthesis procedure to form COFs-basedcomposite structure. As a result, some solvent-resistant and tempera-ture-resistant inorganic substrates such as anodic aluminum oxide(AAO) [40] and tubular/disk-like alumina substrates [31,32] areusually used for in situ growth to fabricate composite membranes de-spite the high costs of inorganic substrates [41]. Therefore, there stillremains a tremendous challenge for the preparation of COFs-basedmembranes via a facile and cost-effective approach under mild condi-tions. Especially, there are lack of appropriate strategies of combiningCOFs with polymeric substrates to construct composite structure for thereal-world application of COFs materials in membrane separation.

Therefore, room-temperature synthesis is very desirable for

https://doi.org/10.1016/j.memsci.2019.05.082Received 2 April 2019; Received in revised form 16 May 2019; Accepted 29 May 2019

∗ Corresponding author.,E-mail address: yongwang@njtech.edu.cn (Y. Wang).

Journal of Membrane Science 586 (2019) 274–280

Available online 31 May 20190376-7388/ © 2019 Elsevier B.V. All rights reserved.

T

fabricating COFs-based membranes with polymeric substrates.Progresses in synthesizing COFs thin films at room temperature areprosperous [42,43], but most of thus-synthesized thin films are hard toapply in membrane separation due to the problems of lateral sizes andstructural integrity. There are also some investigations have been re-ported for room-temperature interfacial synthesis of imine-based COFsmembranes. For example, Dey et al. [29], Matsumoto et al. [44] andValentino et al. [45] individually employed interfacial polymerizationto fabricate free-standing COFs films, and then transferred them toporous supports by various means to form nanofiltration membranes fordye separation. Lai and co-workers [46] adopted the Langmuir-Blodgett(LB) method to prepare ultrathin COFs films, which were then trans-ferred onto porous support to form multilayered composite membranesfor organic solvent nanofiltration (NF). However, aside from the issueof using aggressive solvents, the transfer process in these works makesthese preparation methods more tedious and complicated. Moreover,there might also exist the problem of the adhesion between the COFslayer and the substrates as well as the mechanical property of thus-prepared composite membranes. Therefore, room temperature synth-esis of COFs directly on polymeric substrates to form composite struc-tures should be a practicable way to deal with the aforementionedproblems.

In our previous work, we reported interfacial polymerization ofCOFs directly on porous polysulfone (PSF) ultrafiltration substrates toprepare nanofiltration membranes within 1min [30]. We consider thatthe permselectivity of the COFs-based composite membranes can befurther promoted by reducing surface pores shrinkage of ultrafiltrationsubstrates during the heat treatment process and improving the crys-tallinity of COFs by the means of extending synthesized durations.Hence, we have designed a novel procedure for unidirectional diffusionsynthesis (UDS) of COFs selective layers directly on the commercialpolyvinylidene fluoride (PVDF) microfiltration membranes. To

demonstrate the feasibility of our proposed methodology, a common 2Dimine-based COF (TpPa) is selected as a model COF material in thiswork. For a typical process, the p-phenylenediamine (Pa) aqueous so-lution and 1,3,5-triformylphloroglucinol (Tp) n-hexane solution weresimultaneously charged into the corresponding side of a diffusion cell.During the reaction process, Pa molecules can pass through the PVDFsubstrates fixed on the diffusion cell and react with Tp molecules toproduce TpPa crystallites composited with substrates. It turns out thatthe TpPa/PVDF membranes fabricated by UDS method possess re-markable performances which are superior to other reported mem-branes with similar dye rejections.

2. Experiments

2.1. Materials

The substrates of macroporous polyvinylidene fluoride (PVDF)membranes (diameter: 25mm, nominal pore size: 0.22 μm) were or-dered from Millipore. 1,3,5-triformylphloroglucinol (Tp, 98 wt%) anddyes (acid fuchsin, chrome black T, Congo red, Evans blue, methyl blue,methyl orange and reactive blue 19) were supplied by ChengduTongchuangyuan Pharmaceutical Co. Ltd. (Sichuan, China) andInstitute of Chemical Reagent (Tianjin, China), respectively. Ethanol(99.8 wt%) and p-phenylenediamine (Pa, 97 wt%) were purchased fromAladdin. Hydrochloric acid (HCl, 37 wt%), sodium hydroxide (NaOH,96 wt%), n-hexane (99 wt%), p-toluene sulfonic acid monohydrate(PTSA·H2O, 99.5 wt%), deionized water (DI water, conductivity:8–20 μs cm−1, pH: 6.8, Wahaha Co.) were received from local suppliers.All the materials and chemicals used in this work have not been furtherpurified.

Fig. 1. Schematic of preparation process of the COFs-based membranes via the UDS method. (a) Schematic illustration of diffusion cell for TpPa growth by UDS. (b)Schematic illustration of the formation TpPa selective layer on the top sides of PVDF support via unidirectional diffusion of Pa and PTSA (catalyst) molecules throughthe pores of the PVDF support to react with Tp molecules. (c) Synthesis of TpPa with a keto form through the condensation of Tp (blue) and Pa (red). (d) Theschematic illustration of intergrowth appearance of the TpPa/PVDF membranes after synthesis. (For interpretation of the references to color in this figure legend, thereader is referred to the Web version of this article.)

R. Wang, et al. Journal of Membrane Science 586 (2019) 274–280

275

2.2. UDS of TpPa/PVDF membranes

To prepare precursors solutions for the UDS procedure, 0.03mmolof Tp (6.3 mg) was sufficiently dissolved in 40mL of n-hexane as well as0.90mmol of Pa (97.4mg) and 1.80mmol of PTSA·H2O (342.4 mg)were adequately dissolved in 40mL of water. The PVDF substrate wasfirstly fixed on a diffusion cell and then 10mL of the Pa aqueous so-lution and Tp n-hexane solution were simultaneously added into thecorresponding side of diffusion cell (Fig. 1a, Fig. S1). The apparatus waskept at room temperature for designed synthesized durations in un-disturbed condition. TpPa/PVDF membranes were collected by re-moving the two side solutions of the diffusion cell with a syringe andwashed with water and ethanol.

2.3. Characterization

The X-ray diffraction (XRD) patterns of the COFs composite mem-branes were recorded on an X-ray diffractometer (Rigaku SmartLab,Japan) with the 2θ range of 2–40° and scanning step of 0.02°. Thechemical structure of the monomers (including Tp and Pa) as well asPVDF substrates and TpPa/PVDF membranes surface was investigatedby a fourier transform infrared spectra (FT-IR, Nicolet 8700) with ascanning range of 4000–600 cm−1. A field emission scanning electronmicroscope (FE-SEM, Hitachi S-4800, Japan) was employed to char-acterize the microscopical morphologies (including surface, bottom andcross-section) of the PVDF substrates and TpPa/PVDF membranes. Thesamples for SEM characterization were treated by spraying a thin pla-tinum layer on their surface under vacuum condition to enhance theirconductivity. The PVDF substrates and TpPa/PVDF membranes surfacetopographies were examined by an atomic force microscopy (AFM, XE-100, Park Systems) at non-contact mode. The distribution of TpPacrystallites in the composite membrane was analyzed by the cross-sectional fluorescent images which obtained from the Leica TCS/SP2confocal microscope. The zeta potentials of the TpPa/PVDF membranessurface and dyes in solution were tested by the SurPASS (Anton Paar,Austria) and Zetasizer Nano ZS ZEN3600 (Malvern, UK) electrokineticanalyzer, respectively. The surface wettability of the resultant mem-branes was investigated by a contact angle goniometer (Krüss DSA100,Germany). A UV–vis absorption spectrophotometer (NanoDrop 2000c,Thermo) was employed to measure the dye solutions concentrationsduring the rejection test.

2.4. Filtration and separation tests

The pure water permeances (PWP) and dye rejections of the re-sultant membranes with various synthesized durations were measuredat room temperature in a stirred filtration cell (Amicon 8003, Millipore)with a testing pressure of 2 bar. In order to obtain a stable water per-meance, there was a half hour prepressing process at 2.5 bar prior to thePWP test. The PWP (L·m−2·h−1·bar−1) was calculated by the followingequation:

PWP = V / (S t ΔP)

Where V (L), S (m2), t (h), and ΔP (bar) are volume of the permeate,effective filtration area, testing duration, and testing pressure, respec-tively.

For the rejection test, dyes including acid fuchsin, chrome black T,Congo red, Evans blue, methyl blue, methyl orange and reactive blue 19were separately dissolved in DI water at the concentration of 50 μmol/Land then the dye solutions were used as feed solutions to test the na-nofiltration performance of TpPa/PVDF membranes. In order to obtaina stabilized rejection, the initial filtrate (around 1mL) was abandonedand the next filtrate was collected for the analysis of separation per-formance. The concentrations of the feed (CF, μmol/L), permeate (CP,μmol/L) and retention (CR, μmol/L) solutions were obtained from the

UV absorbance at the maximum absorption wavelength of each dye.The detailed information of various dyes is given in Table S1. The re-jection (R, %) was calculated by the following equation:

R = (CF− CP) / CF× 100%

2.5. Long-term stability and chemical stability tests

To evaluate the long-term performance stability of the TpPa/PVDFmembranes in the separation process, a 12 h continuous separation testwas conducted at room temperature with a trans-membrane pressure of4 bar and the 50 μmol/L acid fuchsin aqueous solution was used as feedsolutions. The chemical stability was evaluated by testing the perfor-mances of the membranes which were separately soaked in the strongacid (2M HCl) and strong base (2M NaOH) aqueous solution for a weekand comparing with those of the membrane without treatment.

3. Results and discussion

The fabrication of TpPa/PVDF membranes was realized in the dif-fusion cell by the UDS method (Fig. 1a). In order to realize the uni-directional diffusion of one kind of monomer through the macropores ofthe PVDF support (Fig. 1b), we elaborately designed the experimentalscheme from two aspects. On the one hand, immiscible water and n-hexane were selected as the solution pairs to create a phase interface.Here Tp can be dissolved in n-hexane with a solubility at least largerthan 0.15 wt% and Pa is highly soluble in water. However, Tp and Paexhibit neglectable solubility in water and hexane, respectively. Thediscrepancy of solubility of the two monomers in water and n-hexaneenables the UDS of COFs across the substrate. Therefore, Tp moleculescan hardly pass through the created phase interface, and the reaction(Fig. 1c) between the two precursors of Tp and Pa only occurs near then-hexane side (top side). On the other hand, the molar ratio of Pa to Tpis 30, which is far higher than the stoichiometric proportions of 1.5.That is to say, Pa molecules are excessive near the phase interface aswell as the condensation reaction between the two precursors will occurpromptly. As a result, Tp molecules will be consumed immediately oncethe Tp molecules diffuse to the phase interface from the bulk solution.

The reaction, occurring at the phase interface, between the twoprecursors of Tp and Pa is illustrated in Fig. 1c. In general, the totalreaction involves two steps. A reversible Schiff-base condensation re-action occurring between the precursors result in the formation of un-stable enol-imine structure in the first step, which provides a self-cor-rection mechanism to improve the crystallinity. The irreversibletautomerization from unstable enol-imine structure to stable β-ketoe-namine structure in the second step enhances the chemical stability ofTpPa crystallites [47]. The highly chemically stable COFs synthesizedby the method of combining a reversible and an irreversible organicsynthetic route make it possible to apply COFs in water treatment. Weexpect that the TpPa crystallites primarily grew along the pore walls onthe top side of PVDF substrates by the way of conformal growth to forma selective layer (Fig. 1d) [30].

The synthesis of TpPa on PVDF substrates was first examined by FT-IR (Fig. 2a). The stretching bands of the aldehyde groups (–CHO,2892 cm−1 and –C=O, 1643 cm−1) of Tp and amino groups(3100–3300 cm−1) of Pa appeared in the monomers and vanished en-tirely in the TpPa/PVDF membranes, which indicates the entire con-sumption of the monomers. The simultaneous appearance of new peaksof –C=C (1578 cm−1) and –C–N (1248 cm−1) in the resultant mem-branes indicates the successful formation of TpPa with a keto form onthe surface of PVDF supports [30,37,48,49]. Similar to other works[31–33,45], the XRD characterization of COFs grown on substrates is atricky issue although the crystallinity of COFs powder synthesizedunder the same or similar conditions of membrane fabrication is ap-preciable. As shown in Fig. S2, XRD analysis reveals that the

R. Wang, et al. Journal of Membrane Science 586 (2019) 274–280

276

characteristic diffraction peaks of TpPa located at 4.7° (100) and 27°(001) still appear although the strong peak intensity of the PVDF sub-strates obscures the peaks originated from TpPa crystallites, which is inline with other works [25,50]. In order to eliminate the effect of sub-strates to the XRD analysis, we selected AAO substrates with nominalpore size of 100 nm to replace PVDF substrates to fabricate TpPa/AAOmembranes for XRD characterization because the AAO substrates areamorphous [40]. The XRD pattern of the TpPa/AAO membrane pre-sents an appreciable crystallinity evidenced by peaks at around 4.7° and27° (Fig. 2b), which match well with those of TpPa powders. Conse-quently, we draw a conclusion that the desired crystalline TpPa hasbeen successfully generated on the PVDF substrates.

It is clear that TpPa crystallites are uniformly generated on themembrane surface because the resultant membrane surface displays auniform orange color after synthesis (Fig. S3). The change of micro-scopic morphology was then observed by SEM. The SEM top viewshows a continuous and dense TpPa layer (Fig. 2d), which is distinctlydifferent from the pristine PVDF substrate having a macroporousmorphology (Fig. 2c). The cross-sectional morphologies also show tre-mendous change after synthesis comparing with the pristine PVDFsubstrate (Fig. 2e). Clearly, the cross-sectional SEM image of TpPa/PVDF membrane (Fig. 2f) reveals the intergrowth appearance betweenTpPa crystallites and PVDF substrates. From the megascopic appear-ance of the resultant membrane in Fig. S3, we can easily find that TpPacrystallites just grow on the top side of the PVDF substrate and hardly

exist on bottom side, which is consistent with our previous experimentdesign. As shown in the cross-sectional fluorescent images (Fig. 2g, Fig.S4), there exists an intensive distribution of TpPa crystallites near thetop side of the membrane with a fluorescent layer thickness of∼6.5 μm, which further demonstrated the single-side growth of TpPacrystallites on the substrates.

We prepared a series of TpPa/PVDF membranes by the UDS methodat room temperature with different synthesized durations. The influ-ence of synthesized durations for UDS on the nanofiltration membranesperformances is shown in Fig. 3 and the permselectivity of the mem-branes presents a typical trade-off effect [51]. The PWP is progressivelydeclined from 86 to 16 L·m−2·h−1·bar−1 while the rejection of the re-sultant membranes to acid fuchsin is gradually increased from 78.0% to96.2% with extended synthesized durations. We speculate that thechange of permselectivity could be attributed to the effect of reducingeffective pore size of the membrane surface by increasingly morecrystalline TpPa grown along the substrates pore wall [30]. Althoughthe intrinsic permanent aperture of TpPa is able to allow water to pass[52,53], the excess TpPa crystallites still leads to the surface of mem-branes much denser and consequently brings increased water resistanceduring filtrations, which results in decreasing PWP.

To verify the aforementioned interpretation, we systematicallycharacterized the TpPa/PVDF membranes prepared under differentsynthesized durations by FT-IR, AFM and SEM. As shown in Fig. S5, theintensity of TpPa characteristic peaks (including –C=C at 1578 cm−1,

Fig. 2. Structural characterization. (a) FT-IR spectra of PVDF substrate (black), precursors of Tp (blue) and Pa (red), as well as TpPa/PVDF membrane (orange) (b)XRD patterns of simulated TpPa with an eclipsed stacking model (black) and TpPa/AAO membranes (red). (c–f) SEM images of the surface and cross-sectionalmorphologies of the pristine PVDF substrates (c, e) and TpPa/PVDF membranes (d, f). (g) Fluorescent image of the cross-section of TpPa/PVDF membranes. (Forinterpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

R. Wang, et al. Journal of Membrane Science 586 (2019) 274–280

277

–C=C(Ar) at 1455 cm−1 and –C–N at 1248 cm−1) presents a slight butgradual increase with prolonging synthesized durations, revealing theincreasing number of TpPa crystallites on the membrane surface[29,46]. The AFM topography images in Fig. S6 further exhibit thequantity change of TpPa crystallites formed on the PVDF substratesbecause increasing TpPa crystallites can slightly reduce the surfaceroughness (Fig. S6f). As shown in Fig. S7, no pinholes, cracks, or otherdefects are observable in the magnified SEM images of the top surfaceof the resultant membranes prepared after TpPa growth for variousdurations. The bottom surface maintains the topography of the pristinePVDF substrates, which well meets the expectation of single-sidegrowth on the substrates. Besides, there also exists an intensive dis-tribution of TpPa crystallites near the top side of the membrane with athickness of 6–7 μm in all of the cross-sectional SEM images, which isconsistent with the fluorescent characterization (Fig. 2g, Fig. S4). Inaddition, from the water contact angle (WCA) tests (Fig. S8), we canfind the membranes prepared under various synthesized durations areweakly hydrophilic, implying the suitability of their application inwater treatment.

Considering the trade-off effect, we believe that the optimal syn-thesized duration is 24 h. Thus, the TpPa/PVDF membrane preparedunder this condition was selected to evaluate its sieving performancetowards various negatively charged dyes including acid fuchsin,chrome black T, Congo red, Evans blue, methyl blue, methyl orange andreactive blue 19. As shown in Fig. 4, the resultant membranes with aPWP of 60 L·m−2·h−1·bar−1 (Fig. 3) exhibit rejection values of 82.8%(methyl orange), 84.7% (chrome black T), 90.4% (acid fuchsin), 92.7%(reactive blue 19), 98.7% (Congo red), 96.5% (methyl blue) and 98.8%(Evans blue), respectively. The UV–vis absorption spectra in Fig. S9indicates that the size-sieving plays a leading role comparing with theadsorption because there is a significant increase of the retention so-lutions concentrations for all separation tests. The 24 h static adsorptionof TpPa/PVDF membranes to the acid fuchsin aqueous solution alsodemonstrates the adsorption is negligible (Fig. S10).

The molecular weight cut-off (MWCO) [31,54] is often used toevaluate separation capability of a membrane. From Fig. 4, MWCO isread approximately as 600 Da. We can also find that the rejection rateto small dyes, such as methyl orange, acid fuchsin and reactive blue 19,with a relatively small size of 1.1–1.6 nm is also considerably high eventhough the inherent aperture size of TpPa crystallite is simulated fromcrystallographic data to be ∼1.8 nm [47]. A possible reason for thisresult could be the staggered pores which results from the intergrowthstructure of TpPa [31]. Another reason may be the Donnan effect due tothe same charged nature of TpPa/PVDF membranes (Fig. S11) and dye

molecules (Fig. S12). That is to say, the electrostatic repulsion pays apositive role as well. The negatively charged nature of TpPa/PVDFmembranes could be ascribed to the existence of the structure of α, β-unsaturated ketones in the backbone of TpPa because the strong effectof electron withdrawing of double bond makes oxygen atoms in car-bonyl groups display negative electricity. Furthermore, the TpPa/PVDFmembranes display very stable dye rejections (Fig. 5a) benefiting fromthe outstanding stability of TpPa [25,26]. Meanwhile, there is no no-ticeable change on the membranes performances after being treatedwith strong acid (2M HCl) and strong base (2M NaOH) aqueous so-lution for a week (Fig. 5b), which displays its outstanding acid and baseresistance.

As mentioned above, the TpPa/PVDF membranes prepared with asynthesized duration of 24 h presents a stable rejection rates to dyes(such as, acid fuchsin of 90.4%, Congo red of 98.7%) and an appreci-able PWP of up to 60 L·m−2·h−1·bar−1. In order to highlight the su-perior performances, we conducted a comparison of performances be-tween the TpPa/PVDF membranes and other reported membraneswhere the polymer used as substrates or matrix (Fig. 6). In spite ofappreciable rejection rates (e.g.>75%), majority of the reportedmembranes exhibit a moderate PWP of less than 25 L·m−2·h−1·bar−1.That is, the PWP of our membranes are∼2–20 times higher than that ofother reported membranes including those prepared using some newmaterials [55,56]. Such a remarkable PWP of the membranes synthe-sized in this work should be attributed to the synergistic effect of theapplications of macroporous PVDF substrates with less mass transferresistance and ordered channel structure of TpPa crystallites for waterpermeation.

4. Conclusions

In summary, continuous and compact imine-based COFs selectivelayers have been successfully synthesized on the polymeric substratesby a unidirectional diffusion synthesis (UDS) method. Porous andcrystalline TpPa separation layers with a thickness up to 6–7 μm on thetop side of PVDF substrates were fabricated. The UDS method offers afacile and cost-effective way to directly grow COFs on the one side ofpolymeric substrates under mild conditions due to the elaborate se-lection of suitable solution pairs and appropriate molar ratio of twoprecursors during the synthesis process. The performances of TpPa/PVDF membranes can be effectively governed by the synthesizeddurations and the composite membrane prepared with a synthesized

Fig. 3. Influence of synthesized durations for UDS on the TpPa/PVDF mem-branes performances.

Fig. 4. Rejection of TpPa/PVDF membranes to various dyes (the embeddedphotograph shows the color change of various dye solutions before (Fe) andafter (Fi) filtration). (For interpretation of the references to color in this figurelegend, the reader is referred to the Web version of this article.)

R. Wang, et al. Journal of Membrane Science 586 (2019) 274–280

278

duration of 24 h achieves considerable pure water permeance (PWP) ofup to 60 L·m−2·h−1·bar−1 and desired dyes (molecule size > 1.1 nm)rejection higher than 90%. The resultant membranes also display ex-cellent long-term performance stability and chemical stability (treatedby 2M acid/base for a week). The composite membranes are expectedto apply in many fields including the treatment of dyeing wastewater,recovery of high value-added species and removal of pharmaceuticalwastes in harsh conditions. Moreover, we expect that the UDS methodcan be further extended to the synthesis of other COFs-based mem-branes with small apertures for desalination as well.

Acknowledgements

Financial support from the National Basic Research Program ofChina (2015CB655301), the Jiangsu Natural Science Foundation(BK20150063), and the National Science Found for DistinguishedYoung Scholars (21852503) is gratefully acknowledged. We also thankthe Program of Excellent Innovation Teams of Jiangsu HigherEducation Institutions and the Project of Priority Academic ProgramDevelopment of Jiangsu Higher Education Institutions (PAPD).

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi.org/10.1016/j.memsci.2019.05.082.

References

[1] X. Feng, X. Ding, D. Jiang, Covalent organic frameworks, Chem. Soc. Rev. 41 (2012)6010–6022.

[2] C.S. Diercks, O.M. Yaghi, The atom, the molecule, and the covalent organic fra-mework, Science 355 (2017) eaal1585.

[3] P.J. Waller, F. Gandara, O.M. Yaghi, Chemistry of covalent organic frameworks,Acc. Chem. Res. 48 (2015) 3053–3063.

[4] S. Kandambeth, K. Dey, R. Banerjee, Covalent organic frameworks: chemistry be-yond the structure, J. Am. Chem. Soc. 141 (2019) 1807–1822.

[5] N. Huang, P. Wang, D.L. Jiang, Covalent organic frameworks: a materials platformfor structural and functional designs, Nat. Rev. Mater. 1 (2016) 16068.

[6] P.B. Ryan, R.D. William, Covalent organic frameworks as a platform for multi-dimensional polymerization, ACS Cent. Sci. 3 (2017) 533–543.

[7] M. Calik, F. Auras, L.M. Salonen, K. Bader, I. Grill, M. Handloser, D.D. Medina,M. Dogru, F. Lobermann, D. Trauner, A. Hartschuh, T. Bein, Extraction of photo-generated electrons and holes from a covalent organic framework integrated het-erojunction, J. Am. Chem. Soc. 136 (2014) 17802–17807.

[8] L. Chen, K. Furukawa, J. Gao, A. Nagai, T. Nakamura, Y.P. Dong, D.L. Jiang,Photoelectric covalent organic frameworks: converting open lattices into ordereddonor-acceptor heterojunctions, J. Am. Chem. Soc. 136 (2014) 9806–9809.

[9] D.D. Medina, M.L. Petrus, A.N. Jumabekov, J.T. Margraf, S. Weinberger,J.M. Rotter, T. Clark, T. Bein, Directional charge-carrier transport in orientedbenzodithiophene covalent organic framework thin films, ACS Nano 11 (2017)2706–2713.

[10] D.D. Medina, T. Sick, T. Bein, Photoactive and conducting covalent organic fra-meworks, Adv. Energy Mater. 7 (2017) 1700387.

[11] M. Dogru, M. Handloser, F. Auras, T. Kunz, D. Medina, A. Hartschuh, P. Knochel,T. Bein, A photoconductive thienothiophene-based covalent organic frameworkshowing charge transfer towards included fullerene, Angew. Chem. Int. Ed. 52(2013) 2920–2924.

[12] S.L. Cai, Y.B. Zhang, A.B. Pun, B. He, J.H. Yang, F.M. Toma, I.D. Sharp, O.M. Yaghi,J. Fan, S.R. Zheng, W.G. Zhang, Y. Liu, Tunable electrical conductivity in orientedthin films of tetrathiafulvalene-based covalent organic framework, Chem. Sci. 5(2014) 4693–4700.

[13] S. Dalapati, M. Addicoat, S.B. Jin, T. Sakurai, J. Gao, H. Xu, S. Irle, S. Seki,D.L. Jiang, Rational design of crystalline supermicroporous covalent organic fra-meworks with triangular topologies, Nat. Commun. 6 (2015) 7786.

[14] Y.F. Zeng, R.Q. Zou, Y.L. Zhao, Covalent organic frameworks for CO2 capture, Adv.Mater. 28 (2016) 2855–2873.

[15] Y. Pramudya, J.L. Mendoza-Cortes, Design principles for high H2 storage usingchelation of abundant transition metals in covalent organic frameworks for 0-700bar at 298 K, J. Am. Chem. Soc. 138 (2016) 15204–15213.

[16] Y. Du, H.S. Yang, J.M. Whiteley, S. Wan, Y.H. Jin, S.H. Lee, W. Zhang, Ioniccovalent organic frameworks with spiroborate linkage, Angew. Chem. Int. Ed. 55(2016) 1737–1741.

[17] C.J. Doonan, D.J. Tranchemontagne, T.G. Glover, J.R. Hunt, O.M. Yaghi,

Fig. 5. Performance stabilities. (a) Long-term performance stability of the re-sultant membranes for acid fuchsin separation. The left inset is the basic in-formation of acid fuchsin, and the right inset is the UV–vis absorption spectra ofacid fuchsin solutions under various testing durations. (b) Chemical stability ofthe TpPa/PVDF membranes: separation performances of the pristine membraneand composite membranes after treating with 2M HCl and NaOH aqueous so-lution.

Fig. 6. Comparison of membrane performances between this study and otherreported in the literature (Congo red: red symbols, acid fuchsin: purple sym-bols), details are given in Table S2. (For interpretation of the references to colorin this figure legend, the reader is referred to the Web version of this article.)

R. Wang, et al. Journal of Membrane Science 586 (2019) 274–280

279

Exceptional ammonia uptake by a covalent organic framework, Nat. Chem. 2(2010) 235–238.

[18] Y.P. Ying, D.H. Liu, J. Ma, M.M. Tong, W.X. Zhang, H.L. Huang, Q.Y. Yang,C.L. Zhong, A GO-assisted method for the preparation of ultrathin covalent organicframework membranes for gas separation, J. Mater. Chem. A 4 (2016)13444–13449.

[19] L. Stegbauer, K. Schwinghammer, B.V. Lotsch, A hydrazone-based covalent organicframework for photocatalytic hydrogen production, Chem. Sci. 5 (2014)2789–2793.

[20] V.S. Vyas, F. Haase, L. Stegbauer, G. Savasci, F. Podjaski, C. Ochsenfeld, B.V. Lotsch,A tunable azine covalent organic framework platform for visible light-inducedhydrogen generation, Nat. Commun. 6 (2015) 8508.

[21] H.W. Chen, H.Y. Tu, C.J. Hu, Y. Liu, D.R. Dong, Y.F. Sun, Y.F. Dai, S.L. Wang,H. Qian, Z.Y. Lin, L.W. Chen, Cationic covalent organic framework nanosheets forfast Li-ion conduction, J. Am. Chem. Soc. 140 (2018) 896–899.

[22] S.Y. Ding, J. Gao, Q. Wang, Y. Zhang, W.G. Song, C.Y. Su, W. Wang, Construction ofcovalent organic framework for catalysis: Pd/COF-LZU1 in Suzuki-Miyaura cou-pling reaction, J. Am. Chem. Soc. 133 (2011) 19816–19822.

[23] C.S. Diercks, S. Lin, N. Komienko, E.A. Kapustin, E.M. Nichols, C.H. Zhu, Y.B. Zhao,C.J. Chang, O.M. Yaghi, Reticular electronic tuning of porphyrin active sites incovalent organic frameworks for electrocatalytic carbon dioxide reduction, J. Am.Chem. Soc. 140 (2018) 1116–1122.

[24] P.F. Wei, M.Z. Qi, Z.P. Wang, S.Y. Ding, W. Yu, Q. Liu, L.K. Wang, H.Z. Wang,W.K. An, W. Wang, Benzoxazole-linked ultrastable covalent organic frameworks forphotocatalysis, J. Am. Chem. Soc. 140 (2018) 4623–4631.

[25] B.P. Biswal, S. Chandra, S. Kandambeth, B. Lukose, T. Heine, R. Banerjeet,Mechanochemical synthesis of chemically stable isoreticular covalent organic fra-meworks, J. Am. Chem. Soc. 135 (2013) 5328–5331.

[26] S. Chandra, S. Kandambeth, B.P. Biswal, B. Lukose, S.M. Kunjir, M. Chaudhary,R. Babarao, T. Heine, R. Banerjee, Chemically stable multilayered covalent organicnanosheets from covalent organic frameworks via mechanical delamination, J. Am.Chem. Soc. 135 (2013) 17853–17861.

[27] S. Kandambeth, B.P. Biswal, H.D. Chaudhari, K.C. Rout, H.S. Kunjattu, S. Mitra,S. Karak, A. Das, R. Mukherjee, U.K. Kharul, R. Banerjee, Selective molecularsieving in self-standing porous covalent-organic-framework membranes, Adv.Mater. 29 (2017) 1603945.

[28] R. Banerjee, A. Halder, S. Karak, M. Addicoat, S. Bera, A. Chakraborty, S. HKunjattu, P. Pachfule, T. Heine, Ultrastable imine-based covalent organic frame-works for sulfuric acid recovery: an effect of interlayer hydrogen bonding, Angew.Chem. Int. Ed. 130 (2018) 5899–5904.

[29] K. Dey, M. Pal, K.C. Rout, S. Kunjattu H, A. Das, R. Mukherjee, U.K. Kharul,R. Banerjee, Selective molecular separation by interfacially crystallized covalentorganic framework thin films, J. Am. Chem. Soc. 139 (2017) 13083–13091.

[30] R. Wang, X.S. Shi, A.K. Xiao, W. Zhou, Y. Wang, Interfacial polymerization ofcovalent organic frameworks (COFs) on polymeric substrates for molecular se-parations, J. Membr. Sci. 566 (2018) 197–204.

[31] H. Fan, J. Gu, H. Meng, A. Knebel, J. Caro, High-flux membranes based on thecovalent organic framework COF-LZU1 for selective dye separation by nanofiltra-tion, Angew. Chem. Int. Ed. 57 (2018) 4083–4087.

[32] H.W. Fan, A. Mundstock, A. Feldhoff, A. Knebel, J.H. Gu, H. Meng, J. Caro, Covalentorganic framework-covalent organic framework bilayer membranes for highly se-lective gas separation, J. Am. Chem. Soc. 140 (2018) 10094–10098.

[33] H.W. Fan, A. Mundstock, J.H. Gu, H. Meng, J. Caro, An azine-linked covalent or-ganic framework ACOF-1 membrane for highly selective CO2/CH4 separation, J.Mater. Chem. A 6 (2018) 16849–16853.

[34] H.W. Fan, Y.F. Xie, J.C. Li, L. Zhang, Q.Y. Zheng, G.J. Zhang, Ultra-high selectivityCOF-based membranes for biobutanol production, J. Mater. Chem. A 6 (2018)17602–17611.

[35] M.Y. Wu, J.Q. Yuan, H. Wu, Y.L. Su, H. Yang, X.D. You, R.N. Zhang, X.Y. He,N.A. Khan, R. Kasher, Z.Y. Jiang, Ultrathin nanofiltration membrane with poly-dopamine-covalent organic framework interlayer for enhanced permeability andstructural stability, J. Membr. Sci. 576 (2019) 131–141.

[36] H. Yang, H. Wu, Z. Xu, B.W. Mu, Z.X. Lin, X.X. Cheng, G.H. Liu, F.S. Pan, X.Z. Cao,Z.Y. Jiang, Hierarchical pore architectures from 2D covalent organic nanosheets for

efficient water/alcohol separation, J. Membr. Sci. 561 (2018) 79–88.[37] B.P. Biswal, H.D. Chaudhari, R. Banerjee, U.K. Kharul, Chemically stable covalent

organic framework (COF)-polybenzimidazole hybrid membranes: enhanced gasseparation through pore modulation, Chem. Eur J. 22 (2016) 4695–4699.

[38] C.B. Wang, Z.Y. Li, J.X. Chen, Z. Li, Y.H. Yin, L. Cao, Y.L. Zhong, H. Wu, Covalentorganic framework modified polyamide nanofiltration membrane with enhancedperformance for desalination, J. Membr. Sci. 523 (2017) 273–281.

[39] J.L. Segura, M.J. Mancheno, F. Zamora, Covalent organic frameworks based onSchiff-base chemistry: synthesis, properties and potential applications, Chem. Soc.Rev. 45 (2016) 5635–5671.

[40] X.S. Shi, A.K. Xiao, C.X. Zhang, Y. Wang, Growing covalent organic frameworks onporous substrates for molecule-sieving membranes with pores tunable from ultra- tonanofiltration, J. Membr. Sci. 576 (2019) 116–122.

[41] E.V. Perez, K.J. Balkus, J.P. Ferraris, I.H. Musselman, Mixed-matrix membranescontaining MOF-5 for gas separations, J. Membr. Sci. 328 (2009) 165–173.

[42] D.D. Medina, J.M. Rotter, Y.H. Hu, M. Dogru, V. Werner, F. Auras, J.T. Markiewicz,P. Knochel, T. Bein, Room temperature synthesis of covalent-organic frameworkfilms through vapor-assisted conversion, J. Am. Chem. Soc. 137 (2015) 1016–1019.

[43] W.Y. Dai, F. Shao, J. Szczerbinski, R. McCaffrey, R. Zenobi, Y.H. Jin, A.D. Schluter,W. Zhang, Synthesis of a two-dimensional covalent organic monolayer throughdynamic imine chemistry at the air/water interface, Angew. Chem. Int. Ed. 55(2016) 213–217.

[44] M. Matsumoto, L. Valentino, G.M. Stiehl, H.B. Balch, A.R. Corcos, F. Wang,D.C. Ralph, B.J. Mariñas, W.R. Dichtel, Lewis-acid-catalyzed interfacial poly-merization of covalent organic framework films, Chem 4 (2018) 308–317.

[45] L. Valentino, M. Matsumoto, W.R. Dichtel, B.J. Marinas, Development and perfor-mance characterization of a polyimine covalent organic framework thin-filmcomposite nanofiltration membrane, Environ. Sci. Technol. 51 (2017)14352–14359.

[46] D.B. Shinde, G. Sheng, X. Li, M. Ostwal, A.H. Emwas, K.W. Huang, Z.P. Lai,Crystalline 2D covalent organic framework membranes for high-flux organic sol-vent nanofiltration, J. Am. Chem. Soc. 140 (2018) 14342–14349.

[47] S. Kandambeth, A. Mallick, B. Lukose, M.V. Mane, T. Heine, R. Banerjee, Crystalline2D covalent organic frameworks with remarkable chemical (acid/base) stability viaa combined reversible and irreversible route, J. Am. Chem. Soc. 134 (2012)19524–19527.

[48] P. Pachfule, S. Kandmabeth, A. Mallick, R. Banerjee, Hollow tubular porous cova-lent organic framework (COF) nanostructures, Chem. Commun. 51 (2015)11717–11720.

[49] H. Wei, S.Z. Chai, N.T. Hu, Z. Yang, L.M. Wei, L. Wang, The microwave-assistedsolvothermal synthesis of a crystalline two-dimensional covalent organic frame-work with high CO2 capacity, Chem. Commun. 51 (2015) 12178–12181.

[50] C.C. Zou, Q.Q. Li, Y.Y. Hua, B.H. Zhou, J.G. Duan, W.Q. Jin, Mechanical synthesis ofCOF nanosheet cluster and its mixed matrix membrane for efficient CO2 removal,ACS Appl. Mater. Interfaces 9 (2017) 29093–29100.

[51] H.B. Park, J. Kamcev, L.M. Robeson, M. Elimelech, B.D. Freeman, Maximizing theright stuff: the trade-off between membrane permeability and selectivity, Science(2017) 356 eaab0530.

[52] L.C. Lin, J.W. Choi, J.C. Grossman, Two-dimensional covalent triazine frameworkas an ultrathin-film nanoporous membrane for desalination, Chem. Commun. 51(2015) 14921–14924.

[53] K. Zhang, Z.J. He, K.M. Gupta, J.W. Jiang, Computational design of 2D functionalcovalent-organic framework membranes for water desalination, Environ. Sci.:Water Res. Technol. 3 (2017) 735–743.

[54] F. Rashidi, N.S. Kevlich, S.A. Sinquefield, M.L. Shofner, S. Nair, Graphene oxidemembranes in extreme operating environments: concentration of kraft black liquorby lignin retention, ACS Sustain. Chem. Eng. 5 (2017) 1002–1009.

[55] C. Zhang, K. Wei, W. Zhang, Y. Bai, Y. Sun, J. Gu, Graphene oxide quantum dotsincorporated into a thin film nanocomposite membrane with high flux and anti-fouling properties for low-pressure nanofiltration, ACS Appl. Mater. Interfaces 9(2017) 11082–11094.

[56] L. Wang, M. Fang, J. Liu, J. He, L. Deng, J. Li, J. Lei, The influence of dispersedphases on polyamide/ZIF-8 nanofiltration membranes for dye removal from water,RSC Adv. 5 (2015) 50942–50954.

R. Wang, et al. Journal of Membrane Science 586 (2019) 274–280

280