Contents lists available at ScienceDirect

Journal of Membrane Science

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

Highly permeable membranes enabled by film formation of blockcopolymers on water surface

Nina Yan1, Zhaogen Wang⁎, Yong Wang⁎

State Key Laboratory of Materials-Oriented Chemical Engineering, Jiangsu National Synergetic Innovation Center for Advanced Materials, College of Chemical Engineering,Nanjing Tech University, Nanjing 210009, Jiangsu, PR China

A R T I C L E I N F O

Keywords:Block copolymersSelective swellingWater surfaceSeparation membranes

A B S T R A C T

Separation membranes derived from block copolymers (BCPs) have attracted significant interest due to theirwell-defined pores and functional surfaces. However, it still remains a challenge to prepare ultrathin BCPcomposite membranes in an efficient way. Here we report a facile approach for the fabrication of ultrathin BCPcomposite membranes by forming thin films on water surface. The BCP layer as thin as ~ 17 nm can be preparedby spreading polystyrene-block-poly(2-vinyl pyridine) (PS-b-P2VP) solutions on water surface followed bycontrolled evaporation. The thin layer can be readily collected and composited with macroporous supports. Athickness of ~ 260 nm of the BCP layer is necessary to guarantee the integrity. Interconnected nanoporousstructures are then obtained in the BCP layer by the process of selective swelling induced pore generation. Onaccount of the ultrathin separation layer, the composite membranes exhibit excellent permselectivity, betterthan many other ultrafiltration membranes. Moreover, the separation performances can be facilely regulated bytuning the swelling durations. This method is featured as cost-effective and convenient, and is expected toprovide a higher chance for the production of BCP membranes at large scale.

1. Introduction

Membrane based separation has become one of the most frequentlyused strategies in purification/concentration of liquids and removing/recycling of by-products due to the advantages including low energyconsumption, good environmental friendliness, and the easiness to becoupled with other processes [1,2]. Nowadays, many newly emergedmaterials, such as carbon nanotubes [3–5], self-assembled polymers[6,7], and covalent organic frameworks [8,9], have been employed inmembrane fabrication toward the important goal of improving themembrane performances and expanding their applications. Amongthese materials, block copolymers (BCPs) are considered as one pro-mising candidate for preparing high-resolution separation membranes,although their large-scale preparation needs to be further explored[10–12]. The nanoporous structures derived from BCPs usually havenarrow pore size distributions and tunable chemical properties, whichare exactly desired for membranes with high permselectivity for spe-cific separations [10]. Meanwhile, a number of strategies for producingnanoporous structures from BCP precursors, such as selective etching ofthe minor blocks or the additives [11,12], nonsolvent-induced phase

separation [13,14] and selective swelling of amphiphilic BCPs [15,16],have emerged in recent years, which determine the feasibility and ef-ficiency of the mass production and real-world applications of BCP-based membranes.

However, most BCPs suffer from expensive cost [17] as well as lowmechanical strength, therefore they are usually employed in the form ofcomposite membranes with the support of macroporous substrates. Forexample, porous BCP composite membranes can be obtained by directlycasting concentrated polymer solutions on a nonwoven, followed bycoagulation in a water bath based on the nonsolvent-induced phaseseparation process [18,19]. However, the BCP solutions, whose con-centration is typically higher than 15wt%, are inevitably leading to athick polymer layer with the thickness over dozens of micrometers.Consequently, large BCP consumption still exists and the cost remainsat a high level. Moreover, the thick layers are usually accompanied withrelatively low permeability as the permeation rate is inversely pro-portional to thickness. A liquid pre-filling strategy has emerged recentlywith the employment of dilute BCP solutions [20–22]. In this strategy,the liquid, which is immiscible with the dissolving solvent, is firstwicked into the macropores of the substrate to avoid the filling of the

https://doi.org/10.1016/j.memsci.2018.09.047Received 6 August 2018; Received in revised form 21 September 2018; Accepted 22 September 2018

⁎ Corresponding authors.

1 Present Address: Institute of Agricultural Facilities and Equipment, Key Laboratory of Agricultural Engineering in the Middle and Lower Reaches of Yangtze River,Jiangsu Academy of Agricultural Sciences, Nanjing 210014, Jiangsu PR China.

E-mail addresses: zhaogenwang@njtech.edu.cn (Z. Wang), yongwang@njtech.edu.cn (Y. Wang).

Journal of Membrane Science 568 (2018) 40–46

Available online 22 September 20180376-7388/ © 2018 Elsevier B.V. All rights reserved.

T

BCP solutions in the following direct- or spin-coating process. Subse-quently, the filled medium as well as the solvent for polymer dissolutionshould be evaporated completely in the drying step. With the pre-fillingstrategy, the thickness of the BCP layer can be dramatically reduced toseveral micrometers or even hundreds of nanometers. Nevertheless, thepreparation with the additional filling and evaporation of the pre-filledliquid not only complicates the preparation process but also bringsadditional uncertainties in controllability as care should be taken forboth the miscibility and volatility of the liquids. Films spin-coated on adense substrate followed by detaching from the dense substrate andattaching to a porous substrate are also recently used to prepare BCPcomposite membranes [23,24]. The BCP layer on the macroporoussubstrates can be as thin as hundreds or even tens of nanometers, whichminimizes the consumption of the BCPs and provides enhanced per-meance. However, the tedious detaching and attaching operations ac-tually limit this strategy to the small-scale fabrication for very specificapplications [25]. Therefore, it is highly demanding and remains achallenge for convenient and efficient strategies to build thin polymerlayers atop macroporous substrates.

In this work, we report on a new method for the preparation of BCPcomposite membranes by evaporating BCP dilute solutions on watersurface to form BCP thin layers. Dilute BCP solutions are spread onwater surface, followed by solvent evaporation to obtain thin BCPlayers covering on the water surface. The BCP layers are then easilycollected by macroporous substrates to form a composite structure.After soaking in hot ethanol, the BCP layers are cavitated as a result ofselective swelling-induced pore generation, thus producing highlypermeable composite membranes with the mesoporous BCP as the se-lective layer.

2. Experimental

2.1. Materials

The asymmetric diblock copolymer, polystyrene-block-poly (2-vinylpyridine) (PS-b-P2VP) with a polydispersity index of 1.09, was pur-chased from Polymer Source Inc., Canada. The number-average mole-cular weights of the PS and P2VP blocks are 50,000 gmol-1 and16,500 gmol-1, respectively. Organic solvents including o-xylene ofchemically pure, chloroform as well as ethanol of analytical grade wereobtained from local suppliers and used as received. Macroporouspolyvinylidene fluoride (PVDF) membranes with the diameter of 25mmand the nominal pore size of 0.22 µm were supplied by Millipore as thesupporting membranes. Silicon wafers with the size of 1.5 cm×1.5 cmwere also used as smooth substrates for the collection of the BCP layersin order to measure the thicknesses, porosities and the surface prop-erties of the layer accurately. Bovine serum albumin (BSA, 67 kDa) waspurchased from Aladdin and used as received. Phosphate buffer saline(PBS) solutions were prepared by dissolving the PBS tablets providedfrom MP Biomedicals, LLC. Deionized water was used in all experi-ments. All chemicals were used without further purification.

2.2. Preparation of PS-b-P2VP composite membranes

The BCP was first dissolved in the mixture of o-xylene/chloroformwith the volume ratio of 4/1 and a concentration of 10mgmL-1. Aftermild stirring for over 3 h, the as-obtained solutions were filtratedthrough a 0.22 µm polytetrafluoroethylene (PTFE) filter for three timesto remove any big aggregates. A clean weighing bottle with the dia-meter of 75mm was filled with ca. 30mL deionized water and then apiece of PVDF membrane or cleaned silicon wafer was placed at thebottom of the weighing bottle. After that, the bottle was transferred intoa container with the diameter of ca. 210mm and PS-b-P2VP solutionswere dropped on the water surface immediately. The droplet spread outand covered the surface upon contacting water. After that, the con-tainer was closed at once to allow the evaporation of solvent inside the

container. 5 min later, the lid of the container was removed to allow thesolvent to evaporate in open air for another 30min. A thin PS-b-P2VPlayer was then gradually formed on the water surface. The as-formedfilms were gently collected by the PVDF membranes or silicon waferswhich were preset at the bottom of bottle, followed by natural drying atroom temperature for over 1 h in the fume hood.

For preparing the PVDF composite membranes, an extra thermaltreatment was employed. Briefly, the as-prepared membranes with BCPlayers collected by PVDF substrates were transferred to an oven pre-heated to 30 °C and then warmed up to 110 °C. The membranes werevacuum-dried at this temperature for 20min and subsequently cooleddown to 30 °C in order to avoid cracking. We then used the selectiveswelling-induced pore generation process to make pores in the PS-b-P2VP layer. The as-prepared composite membranes were soaked inethanol at the temperature of 55 °C for different periods followed by airdrying at room temperature for over 0.5 h.

2.3. Characterizations

The surface and cross-sectional morphologies of the pristine com-posite membranes as well as membranes with various swelling condi-tions were examined with a field emission scanning electron micro-scope (FESEM, Hitachi S-4800) at the accelerating voltage of 5 kV. Thesamples were first vacuum coated with a thin layer of platinum forimproving their conductivity. To measure the static water contact an-gles (WCAs), thicknesses as well as the porosities of the PS-b-P2VPlayers, clean silicon wafers were used as the substrates to collect thefilms on water surface. A contact angle goniometer (Dropmeter A-100,Maist) was employed to quantitatively probe the WCAs of both the topand the bottom surfaces of the as-prepared layers. Each sample wastested on at least 5 positions and the average value was used as thereported one. The thicknesses of the PS-b-P2VP films prepared withdifferent conditions were determined using a spectroscopic ellipsometer(Complete EASE M-2000U, J. A. Woollam) with the incidence angle of70°, while the porosities after swelling were calculated based on thethickness change. A transmission electron spectroscope (TEM, Tecnai12) was used to examine the microphase separation behaviors of thefilms with the thickness of ~60 nm at the operating voltage of 120 kV.The sample was directly collected by the TEM copper grids and sub-sequently stained with I2 at room temperature for ~5 h before char-acterizations.

2.4. Filtration tests

The characterizations of water permeabilities and BSA retentions ofthese composite membranes with different swelling degrees were car-ried out at room temperature in a stirred cell module (Amicon 8010,Millipore) under the pressure of 0.1 bar. A 10-min prepressing ofmembranes was employed for ensuring a stable flux prior to the ex-periments. For the retention tests, BSA was first dissolved in the pre-pared PBS solutions at the concentration of 0.5 g L-1 and the BSA so-lution was then used as the feed solution to characterize the retentionproperties of the membranes. The concentrations of the feed as well asthe filtrate solutions were determined from the UV absorbance at thewavelength of 280 nm with a UV–vis absorption spectrophotometer(NanoDrop 2000c, Thermo).

3. Results and discussion

3.1. Film formation on water surface

Typically, polymer solutions are spread on a smooth solid surface toprepare thin films. However, it would be technically difficult and te-dious to remove the polymer thin films from the solid surface withoutdestroying their structural integrity. Alternatively, liquid surface canalso serve as a substrate to prepare polymer thin films, facilitating the

N. Yan et al. Journal of Membrane Science 568 (2018) 40–46

41

subsequent transferring operations as films floating on the liquid sur-face can be readily collected by various solid substrates. As the mostinexpensive and ubiquitous liquid in the world, water is utilized herefor providing a smooth liquid surface for film preparation. The lowviscosity and the high surface tension of water also make it possible forsome specific solvents to spread on its surface. We first take an in-vestigation on the BCP layer prepared on water surface. The film pre-paration process was shown in Fig. 1. The PS-b-P2VP solutions werefirst obtained by dissolving the polymer in the mixture of o-xylene andchloroform with a volume ratio of 4 and a concentration of 10mgmL-1.The utilization of small amount of chloroform was aimed to adjust thesurface tension of the BCP solutions for better spreading on the watersurface. Deionized water was added to a clean weighing bottle with apiece of substrate such as the PVDF macroporous membrane placed atthe bottom of the bottle. After that, the bottle was transferred to a largecontainer and dosages of PS-b-P2VP solutions were added dropwise tothe water surface (Fig. 1a). After slower evaporation of the solvent(Fig. 1b), another 30-min evaporation in open air was carried out forcomplete removal of the solvent (Fig. 1c). The first slower evaporationwas employed because it can provide efficient time for the solution tospread on water evenly, thus eventually leading to a more uniform film.With the evaporation of the solvent, a thin layer of PS-b-P2VP film wasgradually formed on the water surface and it can be collected directlyby the preset substrate for further characterizations (Fig. 1d).

To investigate the thicknesses as well as the surface properties of theas-obtained PS-b-P2VP films, silicon wafers were utilized instead of thePVDF macroporous membranes as the substrates. The correlation be-tween the thickness and the dosages for film preparation was first in-vestigated and the results were shown in Fig. 2. Interestingly, a film asthin as ~17 nm can be produced on water surface by applying BCPsolutions of 10 μL. As the solutions were increased to 100 μL, a film withthe thickness of ~261 nm was obtained. Meanwhile, it is also worthnoting that the thickness is nearly linear increased with the solutiondosages ranging from 0 to 100 μL. That is, the thicknesses of the as-obtained PS-b-P2VP films can be easily tuned with the dosages ofpolymer solutions. For comparison, the theoretical thicknesses of thesefilms were also calculated through the following equation with theassumption that the PS-b-P2VP solutions were spread out on watersurface evenly:

=C V ρ S Ts s m m m (1)

where Cs, Vs represent the concentration and the dosage of the PS-b-P2VP solutions used for film preparation. Sm and Tm are the surface areaand the calculated thickness of the as-prepared films. The surface area

can be associated with the diameter of the weighing bottle for assumingthat the solutions, which are transformed into film with evaporation,cover the whole water surface completely. ρm is the density of the filmwhich can be counted as the density of the polymer in ideal conditionsas the film should be in dense state. The density of the polymer is es-timated to be 1.07 g cm-3 with those of the corresponding blocks whichare 1.05 g cm-3 for PS and 1.11 g cm-3 for P2VP [26]. According to Eq.(1), the calculated thicknesses of the as-prepared films with differentdosages were depicted in Fig. 2. The measured thicknesses agree wellwith the calculated values. For example, the calculated thickness of thefilms should be 21 nm as prepared by 10 μL BCP solutions, and it in-creases to 212 nm with the dosage of 100 μL. This implies that the BCPfilms produced on the water surface are uniform in thickness, and thethickness of the BCP films can be predictably tuned by changing thedosages of the spreading solutions if we fix the concentration.

To investigate the surface properties of the as-prepared BCP layer,silicon wafers were also employed. Briefly, the wafers were used tosupport the films from the bottom side to characterize the WCAs of thetop surfaces of the films. The WCAs of the bottom surfaces facing towater in the film-forming course were measured after compositing thefilms by wafers from the top surface of the films to make the bottomface up. The results of the WCAs were plotted in Fig. 3.

We note that the WCAs of the top surfaces of the PS-b-P2VP films arevaried in the range from 80° to 90° regardless of the thickness, which

Fig. 1. The schematic diagram for thefabrication of PS-b-P2VP compositemembranes on water surface. (a)Dosages of PS-b-P2VP solution addeddropwise to the water filled in aweighing bottle with PVDF substratepreset at the bottom; (b) slower eva-poration of the solvent in the container;(c) solvent evaporation in open air; (d)the as-formed PS-b-P2VP films col-lected directly by the preset substrates;(e) the composite membranes with PS-b-P2VP layer atop PVDF substrate afterthermal treatment and (f) the compo-sites with mesopores cavitated in thePS-b-P2VP layer by selective swelling.

Fig. 2. The measured and calculated thicknesses of the BCP films prepared onwater surface using various dosages of BCP solutions.

N. Yan et al. Journal of Membrane Science 568 (2018) 40–46

42

are very similar to those of PS homopolymers [27]. In stark contrast,the bottom surface of the film, which contacted the water in the processof film preparation, shows a contact angle of ca. 55°. This value is alsoclose to that of the pure P2VP films [28]. Therefore, we speculate thatthere is always an enrichment of PS blocks on the top surface due to thelower surface tension of PS (γPS = 39–40.7 mNm-1 [27,29]) than P2VP(γP2VP = 47mNm-1 [29]) and the hydrophilic P2VP blocks are enrichedon the bottom surface of the films so as to lower the interfacial energywith water in the film-forming course. We note that the janus surfaceproperties of the films can be obtained even if the thickness is only ~17 nm, which may provide a new pathway for fabricating thin film withjanus surfaces [30,31].

We further had an observation on the microphase-separatedmorphologies of these as-prepared PS-b-P2VP films with TEM analysis.The samples were prepared on water surface with a thickness of ~60 nm as predicted by Fig. 2 and collected directly by using the coppergrids as the substrates. As shown in Fig. 4, the P2VP phase appeareddarker as they were selectively stained with iodine. We note that the as-prepared PS-b-P2VP film has a dense and nonporous morphology withthe P2VP cylinders randomly distributed in the PS matrix. Meanwhile,although the P2VP blocks are enriched at the bottom surface of thefilms as revealed by the WCA results (Fig. 3), the PS microdomains stillexist at the bottom of the film. The P2VP phase can be roughly dividedinto two forms: circular form and elongated form, which corresponds tothe perpendicular and in-plane orientations of the P2VP blocks in the

PS matrix. The diameter of the P2VP microdomains can be roughlydetermined as ~12 nm according to this image.

3.2. Preparation of PS-b-P2VP composite membranes

As the thickness of the BCP layer prepared on water surface can beeasily tuned by changing the BCP dosages, we have the flexibility toprepare composite membranes having BCP selective layers with dif-ferent thicknesses. However, when small BCP dosages are applied, theintegrity of the membranes cannot be guaranteed according to both theSEM observations and the water permeabilities of the composites as themacropores of the PVDF substrate cannot be fully covered by the ul-trathin BCP layers. Therefore, 100 μL PS-b-P2VP solutions were used forcomposites preparation in order to insure the membrane integrity. TheBCP layer was first prepared with macroporous PVDF membrane as thesubstrate, following the process described in Fig. 1a-d. With the eva-poration of the solvent, a thin layer of PS-b-P2VP film was graduallyformed on water surface. Then the designated PVDF membrane presetat the bottom of the bottle was moved upward for the collection of theas-formed layer directly (Fig. 1d). After natural drying for at least 1 h,the PS-b-P2VP layer was composited with the supports. However, theadhesion between the PS-b-P2VP layer and the PVDF substrate is quitelow as the top layer can be easily peeled off from the substrate uponcontacting water. To solve this problem, the composite membraneswere thermally treated at 110 °C, which is slightly higher than the glass-transition temperature of both blocks (TgPS is ~ 104 °C and TgP2VP is ~95 °C, respectively, according to the supplier), for 0.5 h, to enhance theadhesion between the two layers. After thermal treatment, the PS-b-P2VP layer shows a strong adhesion with the macroporous membranes(Fig. 1e). Meanwhile, thermal treatment can lead to complete eva-poration of the residual solvent in the membrane, and also eliminatesome potential defects. Then the composite membranes were soaked inethanol bath at 55 °C for different durations to cavitate the BCP filmsbased on the mechanism of selective swelling induced pore generation[15,16]. Briefly, as the composites were immersed in ethanol which isselective to the hydrophilic blocks, ethanol diffuses into the BCP layerand is preferentially enriched in the P2VP domains, leading to theswollen and expansion of the domains distributed continuously in thePS matrix. Upon withdrawing the membranes from the ethanol bath,the P2VP chains shrink and collapse, resulting in the formation of in-terconnected nanopores at the position previously occupied by theswollen chains while the framework of the layer were maintained withthe frozen PS matrix. After swelling treatment, the BCP layer remainedstrongly adhered to the PVDF substrates and no defects or exfoliationcan be observed (Fig. 1f).

The surface morphology of the as-prepared composite membraneswas shown in Fig. 5a. Before swelling, the membranes were nonporouswith a smooth surface. After soaking the membranes in ethanol at 55 °Cfor various durations, pores were generated in the BCP skin layer, asshown in Fig. 5b-f. Swelling at this mild condition for only 1 h, somesmall round pores appeared on the surface and only few elongatedpores were observed at raised positions (Fig. 5b). Further increasing theswelling time to 3 h, a minority of these small round pores contactedand merged with each other, leading to larger elongated pores (Fig. 5c).As the duration comes to 5, 8 and 12 h, a much rougher topography wasproduced with the elongated pores becoming much clearer at the raisedpositions (Fig. 5d-f). Moreover, due to the porous structure and en-richment of P2VP chains [20], the membrane surface was highly hy-drophilic after swelling, much different to the as-prepared BCP filmsbefore swelling (Fig. 3).

We also examined the cross-sectional morphologies of the compo-site membranes prepared with different swelling durations. As shown inFig. 6a-f, the BCP layers were all jointed closely with the supportingmembranes before and after swelling, which verified the strong adhe-sion between the BCP and the substrates after thermal treatment.Meanwhile, before swelling the membranes exhibit an extremely dense

Fig. 3. The WCAs of the upper and bottom surfaces of the films prepared onwater surface with various dosages.

Fig. 4. The TEM image of the stained PS-b-P2VP films with a thickness of ~60 nm.

N. Yan et al. Journal of Membrane Science 568 (2018) 40–46

43

state across the whole thickness of the BCP layer (Fig. 6a). However,different from the porous morphology observed from the surface SEMcharacterizations, few pores were observed in the membranes swelledfor 1 h (Fig. 6b). The pores were also unclear as the duration prolongs to

3 h (Fig. 6c) while they can be observed distinctly until swelling for 5 h,which is coupled with an obvious increase of the membrane thickness(Fig. 6d). There is no further change except a much more rugged to-pography was observed with the following swelling process (Fig. 6e-f).

Fig. 5. The surface SEM images of the composite membranes (a) without swelling and exposed to ethanol at 55 °C for (b) 1 h; (c) 3 h; (d) 5 h; (e) 8 h; (f) 12 h. All theimages have the same magnification and the scale bar is shown in (f).

Fig. 6. The cross-sectional SEM images of the PS-b-P2VP composite membranes (a) before swelling and exposed to ethanol at 55 °C for (b) 1 h; (c) 3 h; (d) 5 h; (e) 8 h;(f) 12 h. All the images have the same magnification and the scale bar is shown in (f).

N. Yan et al. Journal of Membrane Science 568 (2018) 40–46

44

Although it is difficult to measure the thickness of the porous BCPlayers atop the PVDF substrate, it is still obvious that the BCP layersexhibited increasing thicknesses with rising swelling durations. To de-termine the porosities of these membranes prepared with differentswelling durations, ellipsometry was employed. The BCP layers col-lected by silicon wafers were used to simulate those attached on mac-roporous PVDF substrates as the composite membrane itself is notsuitable for ellipsometrical characterization. Then the as-collected filmson silicon wafers were treated with the same procedures includingthermal treatment and selective swelling as those to the PVDF com-posite membranes. The porosities of these membranes with differentswelling durations were calculated through the thickness changes afterswelling and the results were shown in Fig. 7. We note that the porosityis nearly linearly increased in the first 5 h of swelling. For instance, theporosity is estimated to be 6.1% in the first hour, and it is increased to~16.3% and ~25.5% with the duration extending to 3 h and 5 h, re-spectively. However, the increase rate slows down from 5 h to 12 h, andthe porosity is estimated to be ~38.3% for the membrane prepared witha swelling duration of 12 h.

3.3. Water permeabilities and separation performances

We then investigated the water permeance and separation perfor-mances of the prepared composite membranes. Fig. 8 shows the waterpermeabilities and BSA retentions of the composite membranes as afunction of swelling durations. Prior to swelling, the membranes were

demonstrated as no water permeation under the pressure of 0.2 bar,indicating the dense nature of the as-prepared membranes after thermaltreatment. By contrast, the membranes showed a permeability of319.2 Lm-2 h-1 bar-1 after a mild swelling in ethanol at 55 °C for only1 h. This confirms the generation of pores throughout the entiremembranes with selective swelling although the pores cannot be clearlyobserved under cross-sectional SEM characterizations. After soaked inhot ethanol for 3 h, the membrane showed a dramatic increase of per-meation to 832.4 Lm-2 h-1 bar-1. The permeation was further increasedto 955.7 Lm-2 h-1 bar-1 with a swelling duration of 5 h. This should beowing to the increased porosities with longer swelling durations, asdiscussed above. However, the thickness increase of BCP layer wouldalso produce a greater mass-transfer resistance so as to lead a nonlinearaugment of permeability though the porosity seems to be increasedlinearly [22]. The permeation showed slight increase if further in-creasing the swelling duration to 8 h and 12 h, which also coincideswith the decreased rate of porosity change. In contrast, the BSA re-tention was decreased with swelling durations. The membrane pre-pared with a swelling duration of 1 h exhibited a BSA retention as highas 93.3%. As the duration prolonged to 3 h and 5 h, the BSA rejectionsdecreased to 81.0% and 77.5%, respectively. As the swelling durationwas further increased to 8 h and 12 h, the retention of BSA dropped to66.0% and 49.7%, correspondingly. The decreased retentions shouldalso be account for increased porosities and also the elongated nano-pores. Therefore, the permselectivity of the membranes can be regu-lated over a wide range simply by changing the swelling durations.Meanwhile, we note that, the membranes exposed to ethanol for 3 hwith over 80% rejection rate of BSA proteins exhibited a permeabilityover 800 Lm-2 h-1 bar-1. Such a high flux is several times higher thanthe commercial ultrafiltration membranes with comparable retentions[32,33]. This is also superior to the performances of many BCP ultra-filtration membranes reported in other works [20,34–36], which shouldbe ascribed to the ultrathin nature of the porous BCP selective layers.Moreover, the selective swelling induced PS-b-P2VP membranes areexpected to find applications in selective separation of varied macro-molecules [20–24] such as lysozyme (LYZ) and BSA.

4. Conclusion

We have demonstrated a facile strategy for the fabrication of com-posite membranes having ultrathin BCP selective layers which are en-abled by evaporating dilute BCP solutions on water surface. In thisstrategy, dilute BCP solutions spread on water surface are allowed toevaporate to form continuous BCP films, which are then collected onPVDF macroporous substrates. Subsequent soaking in hot ethanol ca-vitated the BCP layer supported on the PVDF substrates, producingcomposite membranes with nanoporous structures as the selectivelayers. The thickness of the BCP layers formed on the water surface canbe easily tuned by changing the volume of the used BCP solutions, andlarge-area continuous films with thickness down to ~ 17 nm can beproduced. However, the ~ 260 nm-thick layer was utilized for com-posite membrane preparation to guarantee the membrane integrity.Thus-produced composite membranes exhibit improved permeabilitythan many ultrafiltration membranes with comparable separation per-formance due to their ultrathin nature of the BCP layer. The permes-electivity of the membranes can be flexibly tuned by regulating theswelling durations. This simple and flexible method provides a cost-effective and up-scalable possibility to produce highly permeable BCPcomposite membranes. Also, this method is expected to find importantapplications in producing ultrathin polymer coatings and films for adiversity of usages.

Acknowledgements

Financial support from the National Basic Research Program ofChina (2015CB655301), National Natural Science Foundation of China

Fig. 7. The porosities of PS-b-P2VP films as a function of swelling durations inethanol at 55 °C.

Fig. 8. The water permeance and BSA retentions of the composite membranesprepared by swelling in ethanol at 55 °C for various durations.

N. Yan et al. Journal of Membrane Science 568 (2018) 40–46

45

(21776126), and the Natural Science Foundation of Jiangsu Province(BK20150063) is gratefully acknowledged. We also thank the Programof Excellent Innovation Teams of Jiangsu Higher Education Institutionsand the Project of Priority Academic Program Development of JiangsuHigher Education Institutions (PAPD).

References

[1] S.L. Loo, A.G. Fane, W.B. Krantz, T.T. Lim, Emergency water supply: a review ofpotential technologies and selection criteria, Water Res. 46 (2012) 3125–3151.

[2] X.Y. Qiu, H.Z. Yu, M. Karunakaran, N. Pradeep, S.P. Nunes, K.V. Peinemann,Selective separation of similarly sized proteins with tunable nanoporous block co-polymer membranes, ACS Nano 7 (2013) 768–776.

[3] J. Feng, S. Xiong, Z. Wang, Z. Cui, S. Sun, Y. Wang, Atomic layer deposition of metaloxides on carbon nanotube fabrics for robust, hydrophilic ultrafiltration mem-branes, J. Membr. Sci. 550 (2018) 246–253.

[4] Z. Shi, W.B. Zhang, F. Zhang, X. Liu, D. Wang, J. Jin, L. Jiang, Ultrafast separationof emulsified oil/water mixtures by ultrathin free-standing single-walled carbonnanotube network films, Adv. Mater. 25 (2013) 2422–2427.

[5] B.J. Hinds, N. Chopra, T. Rantell, R. Andrews, V. Gavalas, L.G. Bachas, Alignedmultiwalled carbon nanotube membranes, Science 303 (2004) 62–65.

[6] H.Z. Yu, X.Y. Qiu, N. Moreno, Z.W. Ma, V.M. Calo, S.P. Nunes, K.V. Peinemann,Self-assembled asymmetric block copolymer membranes: bridging the gap fromultra- to nanofiltration, Angew. Chem.-Int. Ed. 54 (2015) 13937–13941.

[7] J. Hahn, J.I. Clodt, V. Filiz, V. Abetz, Protein separation performance of self-as-sembled block copolymer membranes, Rsc Adv. 4 (2014) 10252–10260.

[8] H. Fan, J. Gu, H. Meng, A. Knebel, J. Caro, High-flux imine-linked covalent organicframework COF-LZU1 membranes on tubular alumina supports for highly selectivedye separation by nanofiltration, Angew. Chem. (Int. Ed. Engl.) (2018).

[9] G. Li, K. Zhang, T. Tsuru, Two-dimensional covalent organic framework (COF)membranes fabricated via the assembly of exfoliated COF nanosheets, Acs Appl.Mater. Interfaces 9 (2017) 8433–8436.

[10] E.A. Jackson, M.A. Hillmyer, Nanoporous membranes derived from block copoly-mers: from drug delivery to water filtration, ACS Nano 4 (2010) 3548–3553.

[11] W. van Zoelen, G. ten Brinke, Thin films of complexed block copolymers, SoftMatter 5 (2009) 1568–1582.

[12] R.M. Ho, W.H. Tseng, H.W. Fan, Y.W. Chiang, C.C. Lin, B.T. Ko, B.H. Huang,Solvent-induced microdomain orientation in polystyrene-b-poly (L-lactide) diblockcopolymer thin films for nanopatterning, Polymer 46 (2005) 9362–9377.

[13] K.V. Peinemann, V. Abetz, P.F.W. Simon, Asymmetric superstructure formed in ablock copolymer via phase separation, Nat. Mater. 6 (2007) 992–996.

[14] V. Abetz, Isoporous block copolymer membranes, Macromol. Rapid Commun. 36(2015) 10–22.

[15] Y. Wang, F. Li, An emerging pore-making strategy: confined swelling-induced poregeneration in block copolymer materials, Adv. Mater. 23 (2011) 2134–2148.

[16] Y. Wang, Nondestructive creation of ordered nanopores by selective swelling ofblock copolymers: toward homoporous membranes, Acc. Chem. Res. 49 (2016)1401–1408.

[17] Z. Xu, Y. Liu, H. Chen, M. Yang, H. Li, Bamboo-like, oxygen-doped carbon tubeswith hierarchical pore structure derived from polymer tubes for supercapacitorapplications, J. Mater. Sci. 52 (2017) 7781–7793.

[18] A. Jung, S. Rangou, C. Abetz, V. Filiz, V. Abetz, Structure formation of integral

asymmetric composite membranes of polystyrene-block-Poly(2-vinylpyridine) on anonwoven, Macromol. Mater. Eng. 297 (2012) 790–798.

[19] W. Chen, M. Wei, Y. Wang, Advanced ultrafiltration membranes by leveragingmicrophase separation in macrophase separation of amphiphilic polysulfone blockcopolymers, J. Membr. Sci. 525 (2017) 342–348.

[20] Z. Wang, X. Yao, Y. Wang, Swelling-induced mesoporous block copolymer mem-branes with intrinsically active surfaces for size-selective separation, J. Mater.Chem. 22 (2012) 20542–20548.

[21] X. Shi, Z. Wang, Y. Wang, Highly permeable nanoporous block copolymer mem-branes by machine-casting on nonwoven supports: an upscalable route, J. Membr.Sci. 533 (2017) 201–209.

[22] Z. Wang, L. Guo, Y. Wang, Isoporous membranes with gradient porosity by selectiveswelling of UV-crosslinked block copolymers, J. Membr. Sci. 476 (2015) 449–456.

[23] W. Sun, Z. Wang, X. Yao, L. Guo, X. Chen, Y. Wang, Surface-active isoporousmembranes nondestructively derived from perpendicularly aligned block copoly-mers for size-selective separation, J. Membr. Sci. 466 (2014) 229–237.

[24] L. Guo, Y. Wang, Nanoslitting of phase-separated block copolymers by solventswelling for membranes with ultrahigh flux and sharp selectivity, Chem. Commun.50 (2014) 12022–12025.

[25] W.A. Phillip, B. O'Neill, M. Rodwogin, M.A. Hillmyer, E.L. Cussler, Self-assembledblock copolymer thin films as water filtration membranes, Acs Appl. Mater.Interfaces 2 (2010) 847–853.

[26] W.A. Hamilton, G.S. Smith, N.A. Alcantar, J. Majewski, R.G. Toomey, T.L. Kuhl,Determining the density profile of confined polymer brushes with neutron re-flectivity, J. Polym. Sci. Part B-Polym. Phys. 42 (2004) 3290–3301.

[27] M.L. Wadey, I.F. Hsieh, K.A. Cavicchi, S.Z.D. Cheng, Solvent dependence of themorphology of spin-coated thin films of polydimethylsiloxane-rich polystyrene-block-polydimethylsiloxane copolymers, Macromolecules 45 (2012) 5538–5545.

[28] K.S. Iyer, B. Zdyrko, S. Malynych, G. Chumanov, I. Luzinov, Reversible sub-mergence of nanoparticles into ultrathin block copolymer films, Soft Matter 7(2011) 2538–2542.

[29] Y. Lin, A. Boker, J.B. He, K. Sill, H.Q. Xiang, C. Abetz, X.F. Li, J. Wang, T. Emrick,S. Long, Q. Wang, A. Balazs, T.P. Russell, Self-directed self-assembly of nano-particle/copolymer mixtures, Nature 434 (2005) 55–59.

[30] H.C. Yang, Y. Xie, J. Hou, A.K. Cheetham, V. Chen, S.B. Darling, Janus membranes:creating asymmetry for energy efficiency, Adv. Mater. (2018), https://doi.org/10.1002/adma.201801495.

[31] H.C. Yang, R.Z. Waldman, M.B. Wu, J. Hou, L. Chen, S.B. Darling, Z.K. Xu,Dopamine: just the right medicine for membranes, Adv. Funct. Mater. 28 (2018)1705327.

[32] Dow Ultrafiltration Operating Specifications, ⟨https://www.dow.com/en-us/water-and-process-solutions/products/ultrafiltration⟩, (accessed 5 August 2018).

[33] Submerged HSU Series Brochure, ⟨http://www.toraywater.com/products/mf/mf_002.html⟩, (accessed 5 August 2018).

[34] H. Yang, Z. Wang, Q. Lan, Y. Wang, Antifouling ultrafiltration membranes by se-lective swelling of polystyrene/poly(ethylene oxide) block copolymers, J. Membr.Sci. 542 (2017) 226–232.

[35] Y.C. Tang, X. Lin, K. Ito, L. Hong, T. Ishizone, H. Yokoyama, M. Ulbricht, Tunablemagneto-responsive mesoporous block copolymer membranes, J. Membr. Sci. 544(2017) 406–415.

[36] H. Ahn, S. Park, S.-W. Kim, P.J. Yoo, D.Y. Ryu, T.P. Russell, Nanoporous blockcopolymer membranes for ultrafiltration: a simple approach to size tunability, ACSNano 8 (2014) 11745–11752.

N. Yan et al. Journal of Membrane Science 568 (2018) 40–46

46