Superpermeable Atomic-Thin Graphene Membranes with High

Feb 7, 2017 - Theoretical permeability of membrane is inversely proportional to its thickness, which indicates ultrathin membranes will be extremely p...
1 downloads 12 Views 4MB Size
Download PDF
Superpermeable Atomic-Thin Graphene Membranes with High Selectivity Gaoliang Wei, Xie Quan,* Shuo Chen, and Hongtao Yu Key Laboratory of Industrial Ecology and Environmental Engineering (Ministry of Education, China), School of Environmental Science and Technology, Dalian University of Technology, Dalian 116024, China S Supporting Information *

ABSTRACT: Theoretical permeability of membrane is inversely proportional to its thickness, which indicates ultrathin membranes will be extremely permeable. Inspired by the atomic thickness of graphene, herein we report a four-layered graphene membrane with a thickness of about 2 nm. The ultrathin membrane is facilely fabricated by directly punching a complete graphene sheet through selective removal of some carbon atoms with metal oxide nanoparticles at high temperature. Their perpendicular pore channels spanning the whole thickness could, to a great extent, reduce hydrodynamic resistance for water transport. Experimental tests have revealed a flux of up to 4600 L m−2 h−1 of the membranes with a pore size of 50 nm and pore density of 1.0 × 107 cm−2 at a pressure of 0.2 bar. This flux is 40−400 times higher than those of conventional ceramic membranes and track-etched membranes. The enhancement in water permeance is attributed to their atomic thickness and straight pore channels. High selectivity is also evidenced by selective separation of nanospheres with their narrowly distributed pores. These atomic-thin graphene membranes, in view of their outstanding permeability and selectivity, possess great potential as future advanced membranes and may inspire the design and development of other innovative membranes. KEYWORDS: graphene, membrane, selectivity, atomic thickness, high permeability

M

attracted extensive interest because of its unique physicochemical and mechanical properties.15 These exceptional properties have triggered extensive efforts to explore applications of graphene as advanced membranes through both experimental verifications16−20 and computational simulations.21−23 Benefitting from their extremely thin structure, graphene (and its derivatives) ultrathin membranes were recently prepared by stacking them with their interlayer spaces as pore channels. Performance tests have demonstrated a high permeability of these membranes that is several times higher than that of commercial ones.24,25 In addition to their ultrathin structure, the nearly frictionless surface of graphene also facilitates the rapid flow of water molecules through the membranes.16 These outstanding properties thus endow graphene great advantages as membranes. Nevertheless, in these layered membranes, water molecules follow a quite tortuous path when permeating the interconnected nanochannels, which actually increases their apparent thickness and essentially transmembrane resistance. Further attempts to endow mass permeability to a large-area graphene sheet were consequently made and have been focused on the formation of reactive-ion etched or defect-originated

embrane technology has found wide applications in diverse areas ranging from food industry,1,2 bioengineering,3,4 and water desalination5−7 to wastewater treatment,8,9 and has thus become one of the most important separation technologies. The permeability for water or gas is an essential intrinsic attribute of membranes. Higher permeability usually contributes to higher efficiency and lower cost in a membrane-based separation process. Experimental investigations and theoretical prediction models have confirmed that a thinner membrane can afford a higher permeability with diminished transmembrane resistance.10 Ultrathin membranes are usually hierarchically designed with an asymmetric structure, consisting of two layers: a supporting layer and separation layer. The permeability of the membranes depends, predominantly, on the length of pore channels spanning their thin separation layers. Many associated works have evidentially indicated high fluxes of these heterogeneous membranes (dozens of nanometers in thickness), which are several, even dozens of times higher than that of commercial ones for comparison.10−14 Analogical inference can be that these membranes will be more permeable if their thicknesses are further reduced. However, it may be difficult because of the limitations of the materials used. Graphene, a typical two-dimensional material with a thickness of only one atomic layer of sp2 carbon atoms, has © 2017 American Chemical Society

Received: November 29, 2016 Accepted: February 7, 2017 Published: February 7, 2017 1920

DOI: 10.1021/acsnano.6b08000 ACS Nano 2017, 11, 1920−1926

Article

www.acsnano.org

Article

ACS Nano pores.26−28 These graphene membranes have demonstrated excellent selectivity or permeability for both gas and water; however, their future wide applications are still challenging because this task demands easy and low-cost methods for the preparation of these porous graphene membranes with many controlled pores. Herein we report a facile preparation method of atomically thin porous graphene membranes. These membranes are fabricated by directly punching complete graphene sheets through selective removal of some carbon atoms with metal oxide nanoparticles (NPs) derived from the decomposition of metal nitrate at high temperature. By controlling the size and distribution density of metal oxide NPs on graphene, the corresponding pore size and porosity of the resultant graphene membrane can be tunable for broader applications. Unlike the membranes composed of graphene fragments reported previously,24,25 these membranes with regular pores are derived from entire pieces of graphene. The almost infinitesimal thickness and perpendicular pore channels spanning the whole thickness will be expected to, to a great extent, reduce the transport resistance for water molecules across the membrane.

To produce pores on graphene, a selective etching approach is performed with metal oxide NPs such as, CuO NPs, at high temperature, the mechanism of which is based on eq 1. To obtain well-distributed CuO NPs on graphene, 0.1−20 g L−1 Cu(NO 3 ) 2 −acetone solution containing 1 or 8 g L −1 poly(methyl methacrylate) (PMMA) is spin-coated on graphene/Cu foil with a size of 1 cm × 1 cm. This solution will form a thin liquid film if dropped on the rotating graphene at high speed and leave a uniform Cu(NO3)2/PMMA layer after the evaporation of acetone (Figure S2a,b). Mapping analysis of element N on an energy dispersive spectrometer (EDS, Figure S2c) reveals a homogeneous distribution of Cu(NO3)2 on graphene. In contrast, in the absence of PMMA, an acetone solution of Cu(NO 3 ) 2 tends to exist in discontinuous drops when spin-coated on graphene and finally forms large Cu(NO3)2 particles, as shown in Figure S2d. PMMA can be almost completely decomposed during consecutive thermal calcination processes with few carbons remaining even in nitrogen gas. Structurally perfect monolayer graphene has an intrinsic strength of 130 GPa predicted to exceed that of any other material. 30 However, the inevitable presence of grain boundaries and defects in graphene derived from CVD can weaken its mechanical strength.30 To guarantee that the prepared graphene membranes have no nonignorable defects (for example, holes and cracks) that must be avoided in a separation process, a layer-by-layer stack of graphene is used to form a four-layered graphene film for further puncture (schematically shown in Figure S3). Mutually staggered grain boundaries and defects of four monolayer graphenes can strengthen the resultant porous graphene membranes to withstand applied pressure.28 In addition to the important role of maintaining good distribution of Cu(NO3)2, the PMMA layer also allows the crack-free transfer of graphene, which has been demonstrated in many other works.

RESULTS AND DISCUSSION Preparation of Porous Graphene Membranes. The graphene membranes are fabricated with a method schematically shown in Figure 1. This method essentially contains two

800 ° C

(1)

C + CuO ⎯⎯⎯⎯⎯→ Cu + CO 170 ° C

2Cu(NO3)2 ⎯⎯⎯⎯⎯→ 2CuO + 4NO2 + O2

(2)

Over 170 °C, Cu(NO3)2 on graphene can be converted into CuO NPs following a thermal decomposition reaction (eq 2). The resultant CuO NPs can react with the graphene which interfacially contacts with them at 800 °C in Ar flow. As a result, site-specific carbon atoms are removed to form pores. After removal of the Cu substrate by floating the sample on FeCl3/HCl solution and several washes, the graphene left on water surface is transferred to a TEM grid for SEM observation. As shown in Figure 2a, the graphene is transparent and can be distinguishable only by its grain boundaries. Closer observation

Figure 1. Scheme for preparation of porous graphene membranes. Note: this scheme does not demonstrate the stack process of four monolayer graphenes in detail.

procedures: growth of graphene and its perforation. The lowpressure chemical vapor deposition (CVD) approach is performed to grow high-quality monolayer graphene on Cu foil (100 μm in thickness) with CH4 as a carbon source. The scanning electron microscopy (SEM) image in Supporting Information, Figure S1a of Cu foil after 5 min CVD shows the typical grain boundaries. After removal of Cu foil with 2.5 M FeCl3/0.5 M HCl solution, a floating transparent film is visually observed. Further observation (Figure S1b) on the transmission electron microscopy (TEM) shows a typical graphene-like thin film structure with some wrinkles. Raman spectra (Figure S1c) of the samples reveals two peaks centered at 1580 cm−1 (G band) and 2672 cm −1 (2D band), which are typical characteristics of graphene. The calculated intensity ratio of 2D band to G band (I2D/IG) from the Raman spectra is about 2.0, indicating that a monolayer graphene forms on the Cu foil.29

Figure 2. (a) SEM image of a porous graphene on a TEM grid. (b) Enlarged view of the marked area in (a). The inset shows a TEM image of the sample. 1921

DOI: 10.1021/acsnano.6b08000 ACS Nano 2017, 11, 1920−1926

Article

ACS Nano

resultant nanoparticles increase in their number with the increase of Cu(NO3)2 concentrations, evidencing reasonability of the inference. However, a further increase in Cu(NO3)2 concentration will induce wide distribution of resultant pores (Figure S10). In summary, by controlling the concentrations of Cu(NO3)2 and PMMA, and calcination durations, the pore size and pore density can be thus tunable. High Permeability and Selectivity of Graphene Membranes. To obtain crack-free transfer of this porous graphene onto other substrates for separation tests, 10 μL of 2.0 g L−1 ethanol solution of polyvinyl butyral (PVB) is spincoated on porous graphene/Cu foil. After cutting off its 0.5 cmwidth four sides as schematically shown in the inset of Figure 4a, the quadrilateral sample is floated on 2.5 M FeCl3/0.5 M

reveals some pores that appear as dark dots in graphene (Figure 2b). TEM image in inset of Figure 2b of the sample displays a circular graphene-free area, further confirming the formation of pores in graphene. Obviously, every pore corresponds to a Cu/ CuO (or Cu) NP located near it (Figure 2b), which evidently indicates these pores are derived from the etching of CuO NPs. Additionally, the spherical structure of CuO NPs creates circular pores spanning the whole thickness, which are an ideal pore channel architecture for membranes. This method allows the possibility of fabricating low-cost, high-quality, and largearea graphene membranes. Tunable Pore Size and Porosity of Graphene Membranes. Since CuO NPs can react with graphene around them, it is reasonable to infer that their sizes can determine the pore sizes of graphene membranes. It is also inferred that a higher concentration of Cu(NO3)2 will result in larger CuO NPs. To verify the inferences, 10 μL of Cu(NO3)2−PMMA− acetone solutions with 1.0, 3.0, 9.0, and 20.0 g L−1 Cu(NO3)2 and 8.0 g L−1 PMMA are spin-coated on graphene/Cu foil, respectively. As shown in Figure 3a and Figure S4, the average

Figure 3. (a) Average pore size as a function of concentrations of Cu(NO3)2. (b) Average pore size as a function of calcination times. The inset schematically shows that CuO NPs can remove more carbon atoms around them for a longer duration.

pore size of resultant graphene membranes increases with an increase in the concentration of Cu(NO3)2. Specifically, at a concentration of 1.0 g L−1, the obtained membranes feature a 10 nm pores, which is increased to about 100 nm at 20.0 g L−1. It is also found that higher concentrations, for example, 30.0 g L−1, will induce the discontinuity of the PMMA layer and nonhomogeneous distribution of Cu(NO3)2 (Figure S5). Alternatively, the pore size can be also tunable by varying calcination durations. As shown in Figure 3b and Figure S6, the pore size increases with prolonging calcination durations, which can be explained by that CuO NPs can remove more carbon atoms around them for a longer duration to produce larger pores. Interestingly, if the concentration of PMMA in Cu(NO3)2− PMMA−acetone solution is decreased to 1.0 g L−1, resultant pore size does not increase with the increase of Cu(NO3)2 concentrations (from 0.1, 0.3, 0.9, to 2.0 g L−1). Instead, it demonstrates an increase in the number of pores (Figure S7). The reason for different results is inferred that, if the PMMA layer is relatively thick, the size of Cu(NO3)2 particles embedded in the PMMA increases with an increase in its concentration. However, a thin PMMA layer, as a result of its low concentration, will confine the size of Cu(NO3)2 to the thickness of the PMMA layer. Thus, more CuO NPs, instead of larger CuO NPs, will form to produce more pores (schematically shown in Figure S8). SEM observations in Figure S9 of the samples before and after calcination at 800 °C indicate almost all Cu(NO3)2 are embedded in the PMMA layer, and

Figure 4. (a) Digital photo of the phenomenon after removal of Cu foil. The inset schematically shows the way to cut off 0.5 cm-width four sides of the quadrilateral sample. (b) Digital photo of polycarbonate membrane with a PVB/graphene film on it. (c) SEM image of polycarbonate membrane with an average pore size of 1 μm. Because of the quite weak conductivity of polycarbonate membrane, its SEM image reveals some deformations and distortions. (d) SEM image of porous graphene on polycarbonate membrane. Benefitting from the good electroconductibility and completeness of porous graphene, the polycarbonate membrane can be clearly observed. (e) Magnified SEM image of the marked area in panel d. (f) SEM image of the porous graphene on a TEM grid. The inset shows the pore size distribution of the membrane.

HCl solution. Graphene on the back of the Cu foil can thus automatically sink to the bottom of the glass dish (Figure 4a). Owing to the porous structure of graphene, the Cu/CuO NPs (or simply Cu NPs) generated after the etching reaction can be also removed. The PVB-on-graphene can be transferred to various substrates, for example, polycarbonate membranes (Figure 4b). As shown in Figure 4c,d, the removal of PVB in ethanol and the subsequent wash with (NH4)2S2O8/HCl solution do not induce structural cracks of porous graphene 1922

DOI: 10.1021/acsnano.6b08000 ACS Nano 2017, 11, 1920−1926

Article

ACS Nano

Figure 5. (a) Water flux of graphene membrane as a function of pressure. (b) UV−vis spectra of Au NP dispersion before and after filtration. Left inset: scheme for gluing graphene edges. Right inset: the permeation paths of water across the graphene membranes. (c) SEM image of graphene membrane after filtration of Au NPs. (d) Water fluxes of five ultrafiltration membranes (from left to right): α-alumina membrane with 50 nm pore size, polycarbonate membrane with 50 nm pore size, polyvinylidene fluoride (PVDF) membrane hollow fiber membrane with 100 nm pore size, PVDF flat-sheet membrane with 100 nm pore size and graphene membrane with 50 nm pore size. The fluxes of αalumina membrane and polycarbonate membrane are obtained from the proportional conversion of the data reported in ref 32 and ref 33, respectively. PVDF1 and PVDF2 membranes are commercially available from Hangzhou Kaijie Corporation and Millipore Corporation, respectively.

permeation paths of water across the membranes (right inset of Figure 5a). The flux of the porous graphene is almost 40−400 times higher than that of an α-alumina membrane and polycarbonate membrane with the same pore size of 50 nm (Figure 5d). Despite the larger pore size (100 nm) of other two ultrafiltration membranes for comparison, their fluxes are still significantly lower than that featured by this graphene membrane with 50 nm pore size (Figure 5d). It has been theoretically predicted that, based on the Hagen− Poiseuille eq (eq 3: ε, surface porosity; rp, pore radius; Δp, applied pressure; τ, tortuosity), the flux (J) of a membrane is inversely proportional to its thickness (δ) and viscosity (μ) of pore channels. A conventional membrane usually possesses a thickness ranging from several to hundreds of micrometers. Contrastingly, the thickness of graphene membranes is only about 2 nm,28 which will greatly reduce hydrodynamic resistance. Unlike other graphene-based ultrathin membranes with tortuous pore channels,24,25 the porous graphene membranes possess perpendicular pore channels spanning their whole thickness. Water molecules can thus penetrate the membranes along the shortest routes, also diminishing the permeance resistance. It should be admitted that permeability of the graphene membranes is still restricted by their relatively low porosity. Future efforts should be focused on improving the pore numbers to further enhance their permeability.

on a polycarbonate membrane with an average pore size of 1 μm (Figure 4c,d), and can eliminate the possibility that Cu or CuO residues are still left on the graphene surface or within the pore channels (Figure S11 and Table S1). Closer observation clearly reveals some pores (Figure 4e) with a density of about 1.0 × 107 cm−2 (Figure 4f) and a narrow distribution centering at 50 nm (inset of Figure 4f). To prevent structural breakage of the atomic-thin graphene membranes when assembled into a membrane module or during water flux measurements, the edge of its effective membrane area is glued to the underlying polycarbonate support with Sealant-704 (schematically shown in left inset of Figure 5a). Before measurements, about 0.5 mL of ethanol is dripped on the graphene membrane to completely wet its pores. The flux of graphene/polycarbonate membranes is measured to be about 700 L m−2 h−1 at a pressure difference of 0.2 bar. In view of the intrinsic surface porosity (15%) of polycarbonate membranes, the porous graphene membranes should have a flux of 4600 L m−2 h−1, which is 6.7 times higher than that of graphene/polycarbonate membranes. Additionally, their flux demonstrates a linear dependence on pressure ranging from 0 to 0.2 bar (Figure 5a), suggesting the structural stability of their pore channels. Filtration of 65 nm Au NPs has demonstrated a >99% rejection (Figure 5b). Combined with the SEM image of the graphene membrane after filtration (Figure 5c), they evidence the structural intactness of the ultrathin graphene membranes during filtration and validity of obtained permeance values. It has been reported that the average interlayer spacing of randomly stacked graphene layers is 0.355 nm,31 which is significantly smaller than pore size (50 nm). Therefore, it is impossible for water molecules to transport through such narrow gaps, and pores are the only

J = επrp2Δp/8μδτ

(3)

In addition to their circular configuration, the pores of graphene membranes also feature a narrow distribution in size, which is expected for precise and selective separations. To separate 10 nm polystyrene (PS) nanospheres and 35 nm PS nanospheres from their mixture, as a proof-of-concept, the 1923

DOI: 10.1021/acsnano.6b08000 ACS Nano 2017, 11, 1920−1926

Article

ACS Nano

Figure 6. UV−vis spectra of 10 nm PS nanosphere dispersion (a), 35 nm PS nanosphere dispersion (b), and their mixture (c) before and after filtration. Insets in panels a and b show the TEM images of 10 and 35 nm PS nanospheres, respectively. These PS nanosphere dispersions are commercially available from Suzhou Nanomicro Technology Company, Ltd. (d) The percentage of 35 nm PS nanospheres in a mixture as a function of separation times.

work demonstrates the feasibility of the ultrathin graphenebased membranes and provides fundamental guidance toward developing highly permeable membranes, which is important for designing other advanced membranes with both high permeability and selectivity.

graphene membranes with an average size of 20 nm (shown in Figure S12) are fabricated and first applied for the filtration of a 0.001 wt % 10 nm PS nanosphere dispersion. As shown in Figure 6a, most of nanospheres can penetrate the membrane because they are smaller than the pores ranging from 11 to 25 nm. On the contrary, the membrane can remove >90% of the 35 nm PS nanospheres, as certified in Figure 6b. The results suggest that these pores in graphene membranes allow permeance of 10 nm PS nanospheres and enable retention of 35 nm PS nanospheres, demonstrating the excellent selectivity of the graphene membranes. Filtration of their mixture also evinces the single pass because only 10 nm PS nanospheres are detected in the filtrate (Figure 6c and Figure S13). In view of their selectivity, they can thus separate 10 and 35 nm PS nanospheres from the mixture. As shown in Figure 6d and Figure S14, the percentage of 35 nm PS nanospheres in mixture is increased from an initial 25% to 67%. After recovery of the nanospheres retained on membranes with careful washing and ultrasonic assistance (see the Supporting Information), they are filtered once again. After the third filtration, the percentage is increased to >95%. Benefitting from the narrow distribution of their pore sizes, the graphene membranes can be thus used for selective and precise separations.

EXPERIMENTAL SECTION Growth of Graphene. Monolayer graphene was grown on Cu foil with a low-pressure CVD method with CH4 as carbon source. In a typical procedure, polished Cu foil (detailed process was shown in Supporting Information) was placed in a tube furnace and hearted to 1000 °C for 30 min in a mixture of 10 sccm H2 and 400 sccm N2. Then, 1 sccm CH4 was introduced for 2 min, while a total pressure of 10 Torr was maintained. The furnace was subsequently cooled down to ambient temperature. Preparation of Grapheme Membrane. Graphene membranes were prepared by punching a graphene sheet with metal oxide nanoparticles at high temperature, as schematically shown in Figure 1. Typically, 10 μL of 1−20 g L−1 Cu(NO3)2−acetone solution containing 1 or 8 g L−1 PMMA (detailed preparation was shown in Supporting Information) was spin-coated on graphene/Cu foil with a size of 1 cm × 1 cm at a speed of 1500 r/min. Layer-by-layer stack of graphenes was subsequently used to form a four-layer graphene film, as schematically shown in Figure S3. The sample was then calcined at 800 °C for various durations in 200-sccm Ar flow, followed by cooling down to ambient temperature. Subsequently, 10 μL of 2.0 g L−1 ethanol solution of PVB was spin-coated on it. After removal of the Cu foil by floating it on the surface of 2.5 M FeCl3/0.5 M HCl solution for 12 h and being washed for several times with ultrapure water, the PVB/graphene was transferred onto porous polycarbonate membrane. After its dryness, the PVB/graphene/polycarbonate was immersed into ethanol solution for 2 h. This was repeated three times to completely remove PVB. The porous graphene/polycarbonate was also further rinsed with 0.5 wt % (NH4)2S2O8/0.5 M HCl solution and finally with clean water for three times. Characterizations. The morphologies of samples were observed by field-emission scanning electron microscopy (FESEM, Hitachi S4800, Japan) and transmission electron microscopy (TEM, FEI-Tecnai

CONCLUSIONS Atomic-thin porous graphene membranes have been fabricated with a facile method. Such membranes have controlled pore sizes and enable ultrafast permeation of water molecules. Orders-of-magnitude enhancements are experimentally observed for water permeance compared with that by conventional membranes, which is attributed to infinitesimal hydrodynamic resistance of the graphene membranes as a result of their atomic-thin architecture and perpendicular pore channels. High selectivity is also evidenced by selective separation of nanospheres with their narrowly distributed pores. Our present 1924

DOI: 10.1021/acsnano.6b08000 ACS Nano 2017, 11, 1920−1926

Article

ACS Nano G2 F30). For samples revealed in Figure 2 and Figures S4 and S6, they were slightly swept with a soft brush to migrate some Cu/CuO (or Cu) NPs before removal of Cu foil. The pore size and density of graphene membranes were obtained from SEM images. The concentrations of Au NPs and PS nanospheres were analyzed on a Shimadzu V-550 UV−vis spectrophotometer. Raman spectroscopy was recorded on a DXR microscope (Thermo Fisher, USA) at an excitation wavelength of 532 nm.

(10) Karan, S.; Samitsu, S.; Peng, X.; Kurashima, K.; Ichinose, I. Ultrafast Viscous Permeation of Organic Solvents Through DiamondLike Carbon Nanosheets. Science 2012, 335, 444−447. (11) Ling, S.; Jin, K.; Kaplan, D. L.; Buehler, M. J. Ultrathin FreeStanding Bombyx mori Silk Nanofibril Membranes. Nano Lett. 2016, 16, 3795−3800. (12) Zhang, Q. G.; Deng, C.; Soyekwo, F.; Liu, Q. L.; Zhu, A. M. Sub-10 nm Wide Cellulose Nanofibers for Ultrathin Nanoporous Membranes with High Organic Permeation. Adv. Funct. Mater. 2016, 26, 792−800. (13) Zhang, X.; Zhang, T.; Ng, J.; Sun, D. D. High-Performance Multifunctional TiO2 Nanowire Ultrafiltration Membrane with a Hierarchical Layer Structure for Water Treatment. Adv. Funct. Mater. 2009, 19, 3731−3736. (14) Ke, X. B.; Zhu, H. Y.; Gao, X. P.; Liu, J. W.; Zheng, Z. F. HighPerformance Ceramic Membranes with a Separation Layer of Metal Oxide Nanofibers. Adv. Mater. 2007, 19, 785−790. (15) Perreault, F.; Fonseca de Faria, A.; Elimelech, M. Environmental Applications of Graphene-Based Nanomaterials. Chem. Soc. Rev. 2015, 44, 5861−5896. (16) Nair, R. R.; Wu, H. A.; Jayaram, P. N.; Grigorieva, I. V.; Geim, A. K. Unimpeded Permeation of Water Through Helium-Leak-Tight Graphene-Based Membranes. Science 2012, 335, 442−444. (17) Joshi, R. K.; Carbone, P.; Wang, F. C.; Kravets, V. G.; Su, Y.; Grigorieva, I. V.; Wu, H. A.; Geim, A. K.; Nair, R. R. Precise and Ultrafast Molecular Sieving Through Graphene Oxide Membranes. Science 2014, 343, 752−754. (18) Aba, N. F. D.; Chong, J. Y.; Wang, B.; Mattevi, C.; Li, K. Graphene Oxide Membranes on Ceramic Hollow FibersMicrostructural Stability and Nanofiltration Performance. J. Membr. Sci. 2015, 484, 87−94. (19) Hu, M.; Mi, B. Enabling Graphene Oxide Nanosheets as Water Separation Membranes. Environ. Sci. Technol. 2013, 47, 3715−3723. (20) Cheng, C.; Jiang, G.; Garvey, C. J.; Wang, Y.; Simon, G. P.; Liu, J. Z.; Li, D. Ion Transport in Complex Layered Graphene-Based Membranes with Tunable Interlayer Spacing. Sci. Adv. 2016, 2, e1501272. (21) Boukhvalov, D. W.; Katsnelson, M. I.; Son, Y.-W. Origin of Anomalous Water Permeation Through Graphene Oxide Membrane. Nano Lett. 2013, 13, 3930−3935. (22) Lin, L.-C.; Grossman, J. C. Atomistic Understandings of Reduced Graphene Oxide as an Ultrathin-Film Nanoporous Membrane for Separations. Nat. Commun. 2015, 6, 8335. (23) Cohen-Tanugi, D.; Grossman, J. C. Water Desalination Across Nanoporous Graphene. Nano Lett. 2012, 12, 3602−3608. (24) Liu, H.; Wang, H.; Zhang, X. Facile Fabrication of Freestanding Ultrathin Reduced Graphene Oxide Membranes for Water Purification. Adv. Mater. 2015, 27, 249−254. (25) Han, Y.; Xu, Z.; Gao, C. Ultrathin Graphene Nanofiltration Membrane for Water Purification. Adv. Funct. Mater. 2013, 23, 3693− 3700. (26) O’Hern, S. C.; Boutilier, M. S. H.; Idrobo, J.-C.; Song, Y.; Kong, J.; Laoui, T.; Atieh, M.; Karnik, R. Selective Ionic Transport Through Tunable Subnanometer Pores in Single-Layer Graphene Membranes. Nano Lett. 2014, 14, 1234−1241. (27) O’Hern, S. C.; Stewart, C. A.; Boutilier, M. S. H.; Idrobo, J.-C.; Bhaviripudi, S.; Das, S. K.; Kong, J.; Laoui, T.; Atieh, M.; Karnik, R. Selective Molecular Transport Through Intrinsic Defects in a Single Layer of CVD Graphene. ACS Nano 2012, 6, 10130−10138. (28) Celebi, K.; Buchheim, J.; Wyss, R. M.; Droudian, A.; Gasser, P.; Shorubalko, I.; Kye, J.-I.; Lee, C.; Park, H. G. Ultimate Permeation Across Atomically Thin Porous Graphene. Science 2014, 344, 289− 292. (29) Li, X.; Cai, W.; An, J.; Kim, S.; Nah, J.; Yang, D.; Piner, R.; Velamakanni, A.; Jung, I.; Tutuc, E.; et al. Large-Area Synthesis of High-Quality and Uniform Graphene Films on Copper Foils. Science 2009, 324, 1312−1314.

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsnano.6b08000. Polish of Cu foil, preparation of Cu(NO3)2-PMAAacetone solution, fabrication of Au NPs, recovery of polystyrene nanospheres, SEM, TEM, and Raman spectrum of graphene, layer-by-layer stack of monolayer graphenes, SEM image of graphene/Cu foil, SEM images of porous graphene membranes obtained at various conditions, SEM image and pore size distribution of the porous graphene membrane for selective separation of PS nanospheres, TEM image of the filtrate (PDF)

AUTHOR INFORMATION Corresponding Author

*E-mail: [email protected]. ORCID

Xie Quan: 0000-0003-3085-0789 Notes

The authors declare no competing financial interest.

ACKNOWLEDGMENTS This work was supported by the National Natural Science Foundation of China (21437001) and the Programme of Introducing Talents of Discipline to Universities (B13012). REFERENCES (1) Cushen, M.; Kerry, J.; Morris, M.; Cruz-Romero, M.; Cummins, E. Nanotechnologies in the Food IndustryRecent Developments, Risks and Regulation. Trends Food Sci. Technol. 2012, 24, 30−46. (2) Charcosset, C. Preparation of Emulsions and Particles by Membrane Emulsification for the Food Processing Industry. J. Food Eng. 2009, 92, 241−249. (3) Bueno, S. M. A.; Legallais, C.; Haupt, K.; Vijayalakshmi, M. A. Experimental Kinetic Aspects of Hollow Fiber Membrane-Based Pseudobioaffinity Filtration: Process for IgG Separation from Human Plasma. J. Membr. Sci. 1996, 117, 45−56. (4) Saxena, A.; Tripathi, B. P.; Kumar, M.; Shahi, V. K. MembraneBased Techniques for the Separation and Purification of Proteins: An overview. Adv. Colloid Interface Sci. 2009, 145, 1−22. (5) Sun, P.; Wang, K.; Zhu, H. Recent Developments in GrapheneBased Membranes: Structure, Mass-Transport Mechanism and Potential Applications. Adv. Mater. 2016, 28, 2287−2310. (6) Yang, H. Y.; Han, Z. J.; Yu, S. F.; Pey, K. L.; Ostrikov, K.; Karnik, R. Carbon Nanotube Membranes with Ultrahigh Specific Adsorption Capacity for Water Desalination and Purification. Nat. Commun. 2013, 4, 2220. (7) Gethard, K.; Sae-Khow, O.; Mitra, S. Water Desalination Using Carbon-Nanotube-Enhanced Membrane Distillation. ACS Appl. Mater. Interfaces 2011, 3, 110−114. (8) Pendergast, M. T. M.; Hoek, E. M. V. A Review of Water Treatment Membrane Nanotechnologies. Energy Environ. Sci. 2011, 4, 1946−1971. (9) Yin, J.; Deng, B. Polymer-Matrix Nanocomposite Membranes for Water Treatment. J. Membr. Sci. 2015, 479, 256−275. 1925

DOI: 10.1021/acsnano.6b08000 ACS Nano 2017, 11, 1920−1926

Article

ACS Nano (30) Lee, C.; Wei, X.; Kysar, J. W.; Hone, J. Measurement of the Elastic Properties and Intrinsic Strength of Monolayer Graphene. Science 2008, 321, 385−388. (31) Kim, H. W.; Yoon, H. W.; Yoon, S.-M.; Yoo, B. M.; Ahn, B. K.; Cho, Y. H.; Shin, H. J.; Yang, H.; Paik, U.; Kwon, S.; et al. Selective Gas Transport Through Few-Layered Graphene and Graphene Oxide Membranes. Science 2013, 342, 91−95. (32) Hua, F. L.; Tsang, Y. F.; Wang, Y. J.; Chan, S. Y.; Chua, H.; Sin, S. N. Performance Study of Ceramic Microfiltration Membrane for Oily Wastewater Treatment. Chem. Eng. J. 2007, 128, 169−175. (33) Liang, H.-W.; Wang, L.; Chen, P.-Y.; Lin, H.-T.; Chen, L.-F.; He, D.; Yu, S.-H. Carbonaceous Nanofiber Membranes for Selective Filtration and Separation of Nanoparticles. Adv. Mater. 2010, 22, 4691−4695.

1926

DOI: 10.1021/acsnano.6b08000 ACS Nano 2017, 11, 1920−1926