You are on page 1of 28

1.

14 Graphene Membranes
Byung Min Yoo and Ho Bum Park, Hanyang University, Seoul, Republic of Korea
r 2017 Elsevier B.V. All rights reserved.

1.14.1 Graphene 358


1.14.1.1 Graphene Synthesis 358
1.14.1.2 Graphene Properties 360
1.14.1.3 Porous Graphene 361
1.14.2 Graphene Oxide 362
1.14.2.1 Graphene Oxide Synthesis 363
1.14.2.2 Graphene Oxide Structure and Properties 363
1.14.2.3 Porous Graphene Oxide 364
1.14.3 Gas Separation 366
1.14.3.1 Gas Separation Properties Using Porous Graphene Membrane 366
1.14.3.2 Gas Separation Properties of Graphene Oxide Membranes 368
1.14.4 Liquid Separation 369
1.14.4.1 Liquid Separation Using Porous Graphene Membrane 369
1.14.4.2 Liquid Separation Using Graphene Oxide Membrane 372
1.14.5 Other Membrane Applications 375
1.14.5.1 Fuel Cell Applications 375
1.14.5.2 Ion-Selective Membranes for Batteries 376
1.14.5.3 Pervaporation 378
1.14.5.4 Membrane Distillation 379
1.14.6 Other 2D Materials for Membrane Applications 380
1.14.7 Conclusions and Perspectives 381
References 383

1.14.1 Graphene

Carbon is the sixth most plentiful element in nature and is a vital element of human life. Carbon exists as different structures in
nature, so-called carbon allotropes. The most general crystalline forms of carbon allotropes are diamond and graphite. Between
them, graphite is a three-dimensional (3D) carbon allotrope with a layered structure and a planar hexagonal carbon network
structure consisting of sp2-hybridized carbons. Each individual layer of these graphite layers is called a graphene layer. These
graphene layers are 0.335 nm apart and the strong carboncarbon covalent s-type bond length within the layer is 0.142 nm. Also,
the p orbitals are delocalized p-type bonds that extend perpendicular to the planes and bring about weak van der Waals attractions
between the layers. Indeed, graphene had long been investigated theoretically and predicted to feature very unusual properties due
to its band structure. Also, the idea to obtain graphene by peeling it off from graphite is very old and dates back to the early 20th
century. The theory behind graphene was rst explored by the physicist Philip Wallace in 1947, who calculated the behavior of
electrons in graphene.1 The name of graphene was not really widely used until professor Geim group in Manchester (UK)
successfully observed a graphene monolayer by repeatedly peeling off few-layer graphite from graphite by using the Scotch-tape
technique in 2004.2 Since that time, graphene has been enormously studied in broad elds because of its unique and unprece-
dented physicalchemical properties such as excellent electrical and thermal conductivity, high charge carrier mobility, a high
Young's modulus, high fracture strength, and large specic surface area. That is, graphene has many outstanding, unique physical
properties such as two-dimensional (2D) atomically thin crystal, transparency (B98%), strongest materials ever measured (200
times of steel), stiffest known materials (two times of diamond), most stretchable crystal (elastic up to 20%), superfast charge
mobility (200 times of silicon), and high thermal conductivity (two times of diamond).3
Since the last decade, graphene has been also pursued for membrane applications mainly because of its atomic thickness,
scalable area, and the ability to engineer selective nanopores in their rigid, but soft crystal lattice. Obviously, for other carbon
materials or other nanomaterials, its availability as a large-area membrane sheet has led to many stimulated ideas to utilize
graphene for membrane applications. Tremendous interests in graphene may be closely associated with current efforts to develop
new membrane materials having much higher permeability and selectivity for more energy-efcient membrane processes.
Although there remain many obstacles for graphene and its derivatives (e.g., graphene oxide, functionalized graphene) for real
membrane applications, graphene will continue to make rapid progress in membrane applications, particularly, such as gas and
liquid separations. This article will cover fundamental knowledge about graphene synthesis, graphene properties, transport
mechanism through porous graphene membrane, and other graphene-based membranes, and their promising membrane
applications.

358 Comprehensive Membrane Science and Engineering II, Volume 1 doi:10.1016/B978-0-12-409547-2.12229-2


Graphene Membranes 359

1.14.1.1 Graphene Synthesis


The methods for obtaining or synthesizing graphene can be categorized into ve major techniques, that is, (1) mechanical
exfoliation, (2) epitaxial growth, (3) chemical vapor deposition (CVD), (4) organic synthesis, and (5) chemical exfoliation
methods listed in Table 1.
1. Mechanical exfoliation or cleavage method is to obtain high-quality graphene sheet by peeling it off from graphite using an
adhesive tape. By repeatedly cleaving high-quality graphite (e.g., highly oriented pyrolytic graphite, HOPG) using Scotch tape,
the graphene monolayer was optically detected by transferring it to a substrate (e.g., SiO2 layer on Si wafer). Because of weak
van der Waals force between graphene layers in graphite (interaction energy B2 eV/nm4), graphene layer can be readily
detached from graphite using an adhesive tape.5 This method can acquire high-quality graphene monolayer suitable for
fundamental studies to explore its physical properties experimentally, but the productivity is extremely so low that is not proper
for commercial mass production.
2. Epitaxial growth method for graphene has to be developed on a substrate (e.g., single-crystal SiC) by vacuum graphitization
at an elevated temperature (B13001C). The silicon atoms will sublime while the carbon-enriched surface undergoes reorga-
nization and then graphitization, resulting in graphene on SiC wafer.6 This epitaxially-grown graphene can be patterned using
lithography methods, suitable for electronic devices.7 However, this method includes some disadvantages: (i) hard to control
graphene thickness, (ii) unusual graphene stacking, and (iii) limited by cost and size of wafers for scalable applications.
3. CVD is up to date one of the most feasible methods of mass production of large-area, monolayer graphene. Using CVD
methods, monolayer or few-layer graphene can be grown straightly on transition-metal substrates such as copper, and nickel via
saturation of carbon upon exposure to hydrocarbon gases at a high temperature.812 Nickel or copper foils are normally used as
substrate with methane. Usually the CVD growth is performed at temperatures B10001C with a hydrocarbon gas. The catalytic
metal substrates can decompose the hydrocarbon gas, as a source of carbon, which is adsorbed into the substrate. When the
substrate is cooled, the solubility of carbon on the substrate decreases, and the carbon precipitates to form mono- to multilayer
graphene lms on the substrate, depending on the type of substrate and the solubility of carbon in the metal substrates. Among
the metal substrates, cold-rolled copper foils have been widely used for CVD graphene growth. Since carbon has very low
solubility in Cu (nearly close to zero), copper has become a potentially promising substrate for the CVD growth of monolayer
graphene.10 Particularly, the growth of high-quality monolayer graphene (498% coverage) can be achieved on polycrystalline
Cu foils. As a whole, the Cu foil is annealed at about 10001C in a H2 atmosphere to remove impurities or to make the substrate

Table 1 Various graphene synthesis methods

Method Illustration Images Advantages Disadvantages

Mechanical High-quality Not scalable


cleavage graphene

Epitaxial High-quality Discontinuous


growth graphene

Chemical High-quality High cost and


vapor and large- complex
deposition area transfer

Bottom-up Potentially Many defects


synthesis suitable for
mass
production

Chemical Low cost and Many defects


exfoliation suitable for
and its mass
reduction production
360 Graphene Membranes

Fig. 1 (A) Continuous roll-to-roll CVD system using selective Joule heating to heat a copper foil suspended between two current-feeding
electrode rollers to B10001C to grow graphene. (B) Reverse gravure coating of a photocurable epoxy resin onto a PET lm and bonding to the
graphene/copper foil, followed by curing of the epoxy resin. (C) Spray etching of the copper foil with a CuCl2 solution. (D) Structure of the
fabricated graphene/epoxy/PET lm.15

surface smooth. A mixture of H2/CH4 is then introduced into the CVD system to initiate graphene growth on the Cu foil
surface. The system is then cooled down to room temperature after a continuous graphene layer is created. Although the CVD
methods have also several hurdles to be overcome for actual applications, the CVD method has emerged as a signicant
technique for mass production with less structural disorder or defects, as compared to any other existing methods.
4. Organic synthesis can be used to prepare graphene-like polycyclic aromatic hydrocarbons (PAHs) which can be a bottom-up
approach to produce graphene. PAHs can be regarded as 2D graphite segments, and they represent one of the most widely
studied classes of carbon compounds in organic synthetic chemistry and materials science. Nanoribbon-like PAHs were
reported by Yang et al.,13 which showed graphene-like properties in the form of graphene nanoribbons (GNRs). If such GNRs
can be extended in size, PAHs can be a new approach for graphene synthesis by total organic synthesis route. Other GNRs was
reported on gold surface using 10,100 -dibromo-9,90 -bianthryl precursor monomers.14 Under ultra-vacuum conditions, the
monomers were thermally deposited on a gold surface to eliminate the halogen substituents from the precursors and provide
the molecular building blocks for the GNRs.

Among the graphene preparation methods, CVD so far provides the best option in terms of graphene quality, scalability, and
relatively low cost of metal substrate (e.g., Cu foil). Samsung already showed the possibility of large-area graphene production on
Cu foil using CVD techniques and demonstrated roll-to-roll transfer of 30-in. sheets of CVD graphene,8 and Sony also displayed
100 m-long graphene-coated lms via roll-to-roll process,15 as shown in Fig. 1. The same strategy can be applied to the roll-to-roll
production of graphene membranes on top of microporous support. However, it must be necessary to study how effective
nanopores can be created on the graphene surface. Also, CVD graphene still contains lots of structural defects, so the technique to
prevent such defects or to heal the unfavorable large pores must be developed.

1.14.1.2 Graphene Properties


As mentioned above, graphene is a single atomic layer (of 0.345 nm) of graphite, which is an allotrope of carbon consisting of
tightly bonded carbon atoms organized into a hexagonal lattice. Actually, it was commonly believed that two-dimensional
materials could not exist owing to thermal instability when separated before monolayer graphene was experimentally isolated by
using a Scotch tape method.2 In the case of graphene, the carbon to carbon bonds in graphene lattice are so small (0.142 nm long)
and strong to prevent thermal uctuations from destabilizing it. One of the most promising properties of graphene is denitely
ultrahigh electrical conductivity. In graphene, each carbon atom is connected to three other carbon atoms on the 2D plane, leaving
one electron freely available in third dimension for electronic conduction. That is, p-electrons as highly mobile electrons are placed
Graphene Membranes 361

Fig. 2 Relations between different types of topological defect in graphene. (A) Disclinations, (B) dislocations, and (C) grain boundaries (linear
defects).21

above and below graphene sheet. Basically, the electron properties of graphene can be elucidated by the bonding (the valance
band) and antibonding (the conduction band) of these p orbitals. Really, electrons in graphene behave very much like photons in
their mobility due to their lack of mass. Another outstanding property of graphene is its excellent mechanical strength,
which makes graphene suitable membrane matrix. Graphene is known as the strongest material ever discovered with a tensile
strength of 130 GPa (compared to A36 structural steel (0.4 GPa) and Kevlar (polyaramid, 0.38 GPa). Also, graphene is very light
(0.77 mg m2), so 1 g of graphene can roughly cover a whole football eld (57,600 ft2) with single layer. Basically graphite,
consisting of a number of graphene, is known to have a negative thermal expansion coefcient (TEC) in the temperature range of
0700 K.16 Likewise, the TEC of single-layer graphene is strongly dependent on temperature and also measured negative TEC in
the whole temperature range with a room temperature value of (  8.0(70.7)  106 K1).17 Graphene also exhibits superior heat
conduction properties compared with other materials.18,19 Single-layer graphene prepared by either CVD or mechanical cleavage
starts to show defects at B5001C in air.20 The defects are initially sp3-hybridized type and become vacancy-like at higher
temperature. It is known that bilayer graphene shows better thermal stability. Although perfect graphene without any defects is one
of the most promising materials in nanotechnology, the outstanding physical properties of graphene are deteriorated with intrinsic
or extrinsic defects.21 The representative defects in graphene include point defects (e.g., StoneWales defect (generated by a pure
reconstruction of a graphene lattice (switching between pentagons, hexagons, and heptagons)), single vacancies, and multiple
vacancies) and one-dimensional defects (e.g., dislocation-like defects, defects at the edges of a graphene layer) (see Fig. 2).
Particularly, the linear defect corresponding to grain boundaries in graphene should be very important because the properties of
polycrystalline graphene are often dominated by the grain size and by the atomic structure of the grain boundaries.

1.14.1.3 Porous Graphene


In simulations, precise subnanometer pores or nanopores can be ideally created into a single-layer graphene membrane with a
number of pore density. Narrow pore size engineering with high pore density on a large-area graphene sheet without defects would
be technically challenging. So far, several top-down methods have been developed to create nanopores on monolayer graphene
sheet. Focused electron beam irradiation can create well-dened nanopores (pore size B3.5 nm) in graphene sheet, but this
method can be used in a scanning electron microscope (SEM) or scanning transmission electron microscope (STEM), so this is less
efcient and not scalable for large-area graphene membrane with high pore density. That is, it is limited to small-area graphene
sheet (Fig. 3A).22 Likewise, focused ion beam irradiation has been also employed to obtain tunable and well-dened pore size in
graphene.28 However, the pore size is relatively larger (5100 nm) than focused electron beam, and also limited to small area.
Ultraviolet-induced oxidative etching was used to create pores (o10 nm), and Bunch et al.23 showed the selective gas transport of
362 Graphene Membranes

Fig. 3 (A) TEM image of multiple nanopores made by electron beam ablation.22 (B) AFM image of UV-etched suspended graphene. Red areas are
pits created by the UV etching.23 (C) STEM image of pores created in graphene membrane by ion bombardment.24 (D) Aberration-corrected STEM
image of graphene after exposure to oxygen plasma.25 (E) TEM image of graphene nanomesh.26 (F) STM image of the polymer network composed
of CHP backbone.27

nanoporous graphene membrane prepared using this method (Fig. 3B). This method seems to be scalable but the pore size
distribution is too broad to perform molecular sieving separation. OHern et al.24 employed ion bombardment to create isolated
and reactive defects on monolayer graphene and further enlarged these defects to nanopores by chemical oxidative etching
(Fig. 3C). As compared to focused electron beam method, this method can treat large-area graphene with tunable pore size (o1
nm) and moderate pore density. Oxygen plasma etching can give rise to tunable pore size engineering, treating large-area graphene
samples with moderate pore density (Fig. 3D).25 In 2010, Bai et al.26 demonstrated isoporous graphene membranes, so-called
graphene nanomesh, using block copolymer (e.g., poly(styrene-block-methyl methacrylate)) lithography method (Fig. 3E). As a
result, they obtained well-dened isoporous graphene membrane with the pore sizes of o20 nm. Actually, this work was not for
membrane applications and also the pore size obtained from this method is still too large to be used for small gas separations, but
the results opened up the feasibility to prepare isoporous graphene membrane.
Although there were a few reports on bottom-up synthesis of porous graphene, Bieri and coworkers27 successfully created a
form of porous graphene, with a regular distribution of the pores. Remarkably, they produced a 2D polyphenylene networks with
single-atom wide pores and subnanometer periodicity by the arylaryl coupling on a metal surface of the cyclohexa-m-phenylene
(CHP), as shown in Fig. 3F. Actually, this work showed a great possibility of porous graphene with precise pore dimension using
bottom-up synthesis. However, there exist many challenging issues to use such porous graphene membranes, such as reducing
defects, scale up, and safe transfer.

1.14.2 Graphene Oxide

Graphene oxide (GO) is a monolayer of graphite oxide. Graphite oxide is highly oxidized graphite. GO can be obtained from
exfoliation of graphite oxide. GO is basically a wrinkled 2D carbon sheet with various oxygen-containing functional groups on its
basal planes and at edges, with the thickness around 1 nm and lateral dimensions between a few nanometers to several micro-
meters. GO becomes more and more popular in the scientic community because it is believed to be a signicant precursor to
graphene, and GO itself has also unique properties applicable to many promising applications like graphene. In addition, GO can
be cheaply synthesized in a mass scale by strong oxidation and exfoliation of graphite.29,30 GO sheets can be well dispersed in
water due to the electrostatic repulsion of their ionized functional groups (e.g., hydroxyl and carboxyl groups).31,32 In common
sense, the preparation of multi-layered GO membrane will be much easier and more scalable, as compared with the preparation of
nanoporous graphene membrane. Also, various coating methods (e.g., vacuum ltration, drop-casting, layer-by-layer deposition,
spray coating, and spin casting) can be available to create thin GO layers onto existing porous supports.3335 In general, the 2D
Graphene Membranes 363

Table 2 Summary of synthetic methods used to synthesize GO43

Method Oxidant Reaction media C/O ratio Raman spectral ID/IG ratio Notes

Staudenmaier KClO3 HNO3 H2SO4 1.17 0.89


Brodie KClO3 Fuming HNO3 KClO3 added stepwise than in a single bolus
Hofmann KClO3 Non-fuming HNO3 1.15 0.87
Hummers KMnO4 NaNO3 Conc. H2SO4 0.84 0.87 Modications can eliminate the need for NaNO3
Tour KMnO4 H3PO4 H2SO4 0.74 0.85

nanochannels between neighboring GO sheets can be considered for passages for molecules and ions smaller than the interlayer
spacing of GO sheets while hindering large species.

1.14.2.1 Graphene Oxide Synthesis


Basically, GO can be synthesized via various methods, such as Brodie,36 Staudenmaier,37,38 Hummers,39 Modied Hummers,40,41
and, more recently, Tour method,42 listed in Table 2. Each method produces graphene oxide with different degree of oxidation,
chemical structure, and structural characteristics (e.g., defects).

(1) Brodie method. Graphite oxide was rst synthesized about 150 years ago by B.C. Brodie, a British chemist in the 19th century.
Brodie treated graphite repeatedly with potassium chlorate (KClO3) and fuming nitric acid (HNO3).36
(2) Staudenmaier method. Staudenmaier37,38 tried to improve previous Brodie method by changing two things: (1) use of con-
centrated sulfuric acid for increasing the acidity of the mixture, and (2) addition of more potassium chlorate solution into the
reaction mixture during the reaction. As a result, he obtained a highly oxidized GO product in a single batch reaction, thus
reducing the experimental procedure. However, this method is known to be hazardous and time-consuming, so now this
method is seldom used for GO production.
(3) Hummers and modied Hummers method. Hummers and Offeman developed a different chemical recipe for producing GO39
using a mixture of sodium nitrate (NaNO3), potassium permanganate (KMnO4), and concentrated sulfuric acid (H2SO4).
The mixture was maintained below 451C for the oxidation of graphite. The oxidation time is within about 2 h, resulting in a
highly oxidized product. However, original Hummers method led to incomplete oxidization, that is, not fully oxidized
product. Thus, a preexpansion process was introduced by Kovtyukhova in 199940 to obtain higher degree of oxidation. In this
method, graphite was pretreated with a mixture of concentrated H2SO4, K2S2O8, and P2O5 at 801C for several hours. The
mixture was ltered, rinsed, and dried before the Hummers method. These days, the modied Hummers method, including
more addition of potassium permanganate, is more popularly used for GO synthesis. Typically, the modied Hummers
method produces monolayer GO platelets with 1 nm in thickness and from a several nanometers to a few micrometer in
lateral dimensions. Usually, the chemical composition is nearly close to be C:O:H 4:2.95:2.540 (i.e., C/O ratio1.36). The
degree of oxidation and oxidation time of GO is much improved as compared to Brodie method, but the impurities should
be completely removed via several separation and purication steps to obtain high-quality GO.
(4) Tour method. More recently, in 2010, Tour group at Rice University developed a new recipe for GO synthesis.42 They did not
use sodium nitride (NaNO3), and rather introduced a new acid, that is, phosphoric acid into the reaction step. Highly
oxidized GO was prepared with six equivalents of KMnO4 in a 9:1 mixture of H2SO4/H3PO4. This method does not produce
toxic gases (e.g., NO2, N2O4, or ClO2) during the oxidation reaction due to the absence of NaNO3, and produces more intact
graphitic basal planes.

1.14.2.2 Graphene Oxide Structure and Properties


In 1939, Hofmann and Holst proposed the chemical structure of GO. Their proposed structure consists of epoxy groups spreading
across the basal planes of graphene with C/O ratio 244 (Fig. 4A). In 1946, Ruess modied the Hofmann model by introducing
hydroxyl groups into corrugating basal plane45 (Fig. 4B). His model includes 1,3-ether on a cyclohexane ring with the four-
position hydroxylated and also stoichiometrically ordered. Later, Scholz and Boehm proposed a new model in which epoxy and
ether groups were completely removed from Ruess model and put hydroxyl groups at the four-position of 1,2-oxidized cyclo-
hexane rings (Fig. 4C).46 In 1988, Nakajima and Matsuo proposed a stage 2-type model (C2F)n in uorinated graphite and tried to
make the oxide analog for GO chemical structure (Fig. 4D).47 The most recent models are Lerf and Klinowski (Fig. 4E)4851 and
Szabo and Dekany models (Fig. 4F).52 Between them, these days, Lerf and Klinowski model has been widely adopted in scientic
community, even though the precise chemical structure of GO remains controversial. In the Lerf and Klinowski model, GO was
considered as a nonstoichiometric, amorphous structure. Previous models were based on elemental analysis, chemical reactivity,
and X-ray diffraction data, while Lerf and Klinowski model used solid-state 13C-nuclear magnetic resonance (SSNMR) technique
for GO characterization. Throughout SSNMR analysis and GO reactivity experiments with various compounds, they concluded
that GO has epoxy and hydroxyl groups on the basal plane, and keto groups are more preferred at the periphery of GO than
carboxyl acids. Also, the acidity of GO could be due to the keto groups with a ketoenol tautomerization and the proton exchange
on the enol site. Also they recognized strong hydrogen bonding between GO themselves and water molecules. SzaboDekany
364 Graphene Membranes

Fig. 4 Proposed GO models. (A) Hoffman, (B) Ruess, (C) ScholzBoehm, (D) Nakajima-Matsuo, (E) Lerf-Klinowski, and (F) Dekany model.43

Fig. 5 DFT energy diagram for GO reduction.53

model somewhat is similar to Lerf and Klinowski model, but this model is based on the logic in Ruess and ScholzBoehm models
expressing corrugating nature of the carbon network. In spite of many studies on GO structure, the exact chemical structure of GO
remained unsolved owing to large sample-to-sample variations mainly due to different synthesis methods, diverse degree of
oxidation, various graphite sources, metastability of GO, and amorphous, nonstoichiometric nature of GO, etc.
Although GO potentially has a wide variety of applications, the stability of GO toward external stimuli (e.g., temperature,
humidity, and light) has been debated. According to Kim et al.,53 multilayer GO produced by the Hummers method is a
metastable material whose structure and chemistry evolve at room temperature (Fig. 5). They proposed that GO reaches a nearly
reduced C/O ratio at quasi-equilibrium, and shows a structure deprived of epoxide groups and enriched in hydroxyl groups. That
is, the hydrogen content in GO is a signicant factor to control the structure, chemistry, and properties of GO.
The role of intercalated water molecules in multilayer GO sheets also plays an important role in changing the whole interlayer
chemistry as compared with single-layer GO. Acik et al.54 conrmed that both defects formation and carbonyl formation are
driven by intercalated water (Fig. 6). As such, the stability of GO membranes in various environments should be further studied
with more fundamental understanding of mass transport through GO membranes.

1.14.2.3 Porous Graphene Oxide


Differing from porous graphene membranes, mass transport channels through GO membranes are void spacing between GO
sheets by considering the laminated microstructure. Actually, the lateral size of GO sheets ranges from 100 nm to 5 mm, so the
mass transport pathway through even ultrathin multilayer GO membranes will be pretty long due to its high aspect ratio, leading
to low mass ux. To minimize the resistance for mass transport due to its laminar structure, the size of the GO sheets should be
reduced, and proper interlayer spacing between neighboring GO sheets should be controlled. Another way is to create pores on GO
basal plane. Some experiments have already suggested that decreasing GO size and increasing porosity of GO sheets can enhance
the permeability of the GO membranes.55,56 As compared with the pore formation in graphene, the formation of pores within GO
sheets during the synthesis is hardly controlled, and inadequate effort has been dedicated to the synthesis of porous GO sheets.
Fan et al.57 developed porous graphene nanosheets by the etching of carbon atoms on the graphene sheets by MnO2 (Fig. 7).
Graphene Membranes 365

Fig. 6 Comparison of infrared differential spectra of (i) single layer and (ii) three layers of GO in the 602501C range. Absorbance changes in
functional groups are shown within temperatures of (a) 60751C, (b) 751001C, (c) 1001251C, (d) 1251501C, (e) 1501751C, (f) 1752001C,
(g) 2002251C, and (h) 2252501C. Loss of epoxides and carboxyls is shown within red and yellow solid lines while formation/loss of carbonyls
is in pink solid lines and formation of ethers in green and CQC in gray. Absorbance unit is abbreviated as a.u. Frequency regions such as a-, b-,
and g- are highlighted with blue, orange, and green dashed lines, respectively. Infrared enhancement is shown with aqua regions.54

The aqueous GO solution was mixed with KMnO4, and the suspension was heated using a microwave oven. The reaction
mechanism is 4MnO4  3C H2 O-4MnO2 CO3 2 2HCO3  .
Koinuma et al.58 reported a simple method to create nanopores in GO sheets using photoreaction in O2 under UV irradiation
(Fig. 8). The oxygen content decreased from 34% to 24% after the UV irradiation in O2, leading to reduced GO with some
nanopores. The proposed reaction mechanisms are as follows:

2C O H2 Oad-COH COOHin O2 1

C O COOH-CO2 COH in O2 2

COC H2 O ad-2COH in O2 3

4COH O2 -4C O 2H2 O in O2 4

The authors proposed that the reaction (1) is the trigger reaction for creating nanopores in GO. Some of the COOH groups
might react with CQO to evolve CO2 (reaction (2)), which might lead to very small-size nanopores. In O2, many CQO will be
formed by reaction (4), promoting the reaction (2) increasing the size and/or the number of the nanopores. The water generated
by reaction (4) will promote the production of COOH (reaction (1)) and COH (reactions (1) and (3)) in O2. However, this
method cannot create the graphene sheets because graphene sheets do not contain the sp3-oxygenated groups. Other methods to
create pores in GO sheets can be found in the literature.59,60 So far, most of the methods to create nanopores in GO often lead to
the chemical or thermal reduction of GO sheets, thus resulting in the reduction of interlayer spacing between the GO sheets, which
may increase the energy barrier for mass transport. Therefore, the method to create nanopores in GO without any reduction should
be developed.
366 Graphene Membranes

Fig. 7 TEM image of porous graphene nanosheets by etching.57

Fig. 8 AFM image of GO nanosheets after the photoreaction in oxygen for 1 h.58

1.14.3 Gas Separation

1.14.3.1 Gas Separation Properties Using Porous Graphene Membrane


In general, many gas separation membranes, such as dense polymeric membranes, can be performed by solutiondiffusion
model.6164 Here, the selectivity comes from a product of diffusivity selectivity and solubility selectivity in the membrane
materials. Basically, diffusivity is governed by thermal motion of polymer chains giving rise to size-dependent diffusion of smaller
molecules while selectivity can be governed by the free volume (i.e., porosity), chemical afnity, and molecular structure, etc.
Differing from conventional gas separation membranes, separation mechanism through porous graphene membrane will be
followed by size-sieving mechanism. Also, a big difference in typical size-sieving membrane materials is the fact that the thickness
is one-atom-thick, that is, there is no long transport channel. Inherently, graphene lattice is a rigid body in which do not allow any
molecules or ions to penetrate because of the hexagonal carbon aromatic rings with high electron density. Bunch et al.65
experimentally showed that graphene is impermeable to helium and other gases, which rather triggered the membrane researches
Graphene Membranes 367

using graphene with a potential pore engineering. That is, to make use of graphene as a gas-selective membrane material, it is
certainly necessary to create proper nanopores in graphene plane. Until the porous graphene membrane was experimentally
achieved, a number of theoretical studies6669 have been performed to predict the gas transport behaviors through imaginary
porous graphene or similar materials such as graphdiyne,70,71 graphyne,72 silicene,73 C2N-h2D crystal,74 and 2D polyphenylene
network.7578 In these simulations, density functional theory or other quantum mechanical methods were used to calculate the
energy barrier for gas molecules of interest to penetrate a pore. Actually, due to its one-atom-thick nature, gas permeability is not
meaningful. Instead, gas permeance was calculated using a transition state method with kinetic energy distribution of gas
molecules. The selectivity can be obtained from the ratio of Arrhenius factors.69 Another approach to estimate the gas permeance
through porous graphene membrane is to use molecular dynamics (MD). The gas permeance and selectivity through porous
graphene membrane, as expected, was strongly dependent on molecular size, surface adsorption, attractive or repulsive interac-
tions with functional groups decorated on the pore rim, and molecular mass, etc.79,80 The simulation results to predict gas
permeance and selectivity were varied, strongly depending on the simulation methods. Generally, when the pore size in graphene
membrane is larger than the gas molecules, the gas transport becomes effusion. Effusion is a molecular ow through a thin
aperture smaller than the mean free path of gas. If the pore size becomes the same size with the gas molecule or slightly smaller
than the size of the gas molecule, the pore can act as an energy barrier against gas transport (Fig. 9). Using these simulation
techniques, porous graphene membrane (without any consideration about support membrane) showed extremely high gas
permeance if proper pore size and pore density can be achieved. The steric exclusion model and the activated transport model can
be applied to explain the gas transport across the porous graphene membrane, by considering the relative ratio of pore size and
molecular size of gas. Probably, pore size is not only a dominating factor to determine the gas selectivity in porous graphene
membrane. Surface adsorption and diffusion of gas molecules on graphene membrane should be further considered with
interaction with heteroatoms or functional groups on the pore rim.
In 2012, Koenig et al.23 for the rst time reported selective gas transport through porous graphene membrane. They used
bilayer graphene membranes to make sure that there were initially no defects in their graphene membranes, rather than single-
layer graphene probably containing intrinsic defects. By UV/ozone etching, they wanted to create pores on the impermeable
graphene membranes. By monitoring gas transport during the UV/ozone etching, they successfully created subnanometer pores on
their bilayer or single graphene membranes. In single-layer, porous graphene membrane, they reported gas permeance order as
He4Ne4H24Ar with increasing kinetic diameter of gases. However, they obtained anomalously high gas permeance for CO2 and
N2O, which might be due to strong polar interaction with some polar groups in the pore rims. Although there was no chemical
information on pore rim, one could imagine that UV/ozone etching at ambient atmosphere might lead to oxidation of graphene.
The oxidized sites can be interacted with such polar gases (e.g., CO2), nally enhancing gas permeance (Fig. 10).

Fig. 9 Effects of pore functionalization and comparison with experimental results. (A) Comparison of MD results for pores without
functionalization with existing experimental results. (B) Comparison of MD results for four selected pores (P-10, Fun-1, Fun-2, and Fun-3) with
existing experimental results. (CE) Sketches of the functionalized pores Fun-1, Fun-2, and Fun-3, respectively. Blue spheres denote C atoms in
graphene, red spheres denote H atoms, and pink spheres denote C atoms in the functional groups.79
368 Graphene Membranes

Fig. 10 Compilation of measured leak rates. Leak rates out of the microcavity for the Bi-3.4 A membrane before etching and after etching, the Bi-
4.9 A membrane after etching, and the average before etching of 24 membranes (12 for N2) on the same graphene ake as Bi-3.4 A membrane
(note that these last symbols are hidden by black squares for several gases).23

More recently, Celebi et al.28 demonstrated more upscaled porous graphene membranes created by using a focused ion beam
for gas separation. However, the pore size was still too large (B7.6 nm) to separate effectively small gas molecules. H2/CO2
selectivity was consistent with effusion, but H2 permeance was B102 mol m2 s Pa, which was more than three orders of
magnitude higher than existing gas separation membranes with similar selectivity due to ultrathin membrane thickness.
Unfortunately, experimental approaches to prepare large-area porous graphene membranes for gas separation are very few and
limited because of lack of precise pore engineering techniques and difculties of high-quality large-area graphene membrane
production. However, this situation will be getting better in the future because of so rapid advances in graphene fabrication
methods.

1.14.3.2 Gas Separation Properties of Graphene Oxide Membranes


As an alternative approach to such porous graphene membranes, gas-selective graphene-based membranes can be achieved using
multilayer graphene oxide (GO) membranes. The concept using graphene oxide for gas separation is to use the interlayer spacing
for gas transport between GO sheets. Basically, the interlayer distance between graphic layers is as narrow as 0.334 nm, so even
small gas molecules cannot diffuse through the interlayer spacing. On the other hand, the interlayer spacing between GO sheets is
known to be in the range from 0.6 to 1.2 nm.81 Also this interlayer distance can be further controlled by inserting molecular spacers
or decorating functional groups on the GO basal plane. However, in sharp contrast to above expectation, Nair et al.82 showed that
dry GO membrane (B0.5 mm thick) acts as a barrier material for even small gas molecules (e.g., helium). As compared to PET lm,
they did not detect He permeability through the dry GO membrane. When dry GO membrane was exposed to humid atmosphere,
He started to permeate through the hydrated GO membranes, and also the water transport rate through the hydrated GO
membrane was unexpectedly fast (Fig. 11). This was mainly caused by the hydrophilic nature of GO sheets because GO contains
many oxygen-containing functional groups (e.g., hydroxyl, epoxy, and carboxyl) on its basal plane and at edges.83 With increasing
the relative humidity, the interlayer spacing between GO sheets becomes enlarged due to more intercalation of water molecules,
resulting in fast water permeability. At the same time, gas molecules also can permeate through hydrated GO sheets in the presence
of water molecules because the interlayer spacing between hydrated GO sheets is sufciently large for the gas transport.
In 2013, Kim et al.84 and Li et al.85 almost at the same time reported the gas permeation properties of ultrathin GO membranes
(o10 nm) on different porous supports (e.g., microporous polymer membrane or porous inorganic membrane) using spin
coating or vacuum ltration. Li et al.85 prepared ultrathin GO (1.8 nm thick) membrane on the anodized aluminum lter disc and
measured various gas permeabilities. They found that ultrathin GO membranes have much higher gas permeability than relatively
thick GO membranes, indicating that the gas permeability of such laminated GO membranes strongly depends on the membrane
thickness (Fig. 12). Owing to long diffusional channel caused by laminated structure, the energy barrier for diffusion would be
high as compared with other membrane materials and also the pressure gradient through the membrane will be different from
polymeric dense membranes. As mentioned above, they used anodized aluminum disc for coating thin GO membranes, so their
GO membrane would be ionically crosslinked with multivalent aluminum ions, so leading to high H2/CO2 or H2/N2 separation
properties.
Kim et al. also investigated the gas permeabilities of ultrathin GO coated on microporous polymer (polyethersulfone, PES)
membranes as well as those of freestanding thick, dry membranes by applying a high transmembrane pressure. In sharp contrast to
Graphene Membranes 369

Fig. 11 Permeation through GO. Weight loss for a container sealed with a GO lm (hE1 mm; aperture's areaE1 cm2). No loss was detected for
ethanol, hexane, etc., but water evaporated from the container as freely as through an open aperture (blue curves). The measurements were
carried out at room temperature in zero humidity.82

Geim group's previous result,82 small gas molecules can permeate through the freestanding thick GO membranes with different
permeation rates, although the gas permeabilities are quite low. The order of gas permeability in dry GO membranes follows the
kinetic diameter of the gases tested in this study (i.e., He4H24CO24O24N24CH4), indicating that the separation can be
dominated by molecular sieving mechanism as often observed in pyrolytic carbon molecular sieve membranes.86,87 They also
found that average GO size signicantly affected the gas permeability of the resultant GO membranes. They prepared different GO
membranes using different GO sizes, prepared using different sonication time in GO suspension. As expected, the gas permeability
was signicantly increased with decreasing the average GO size, caused by the fact that the diffusional pathway can be reduced
with decreasing the GO size (Fig. 13A). Furthermore, they studied the effect of stacking manner of GO sheets on the gas
permeability of GO membranes and also investigated the effect of relative humidity on the gas permeability of GO membranes.
They coated ultrathin GO layer (o10 nm) on top of the PES membranes using two different methods: (1) Method 1: contacting
the surface of a support membrane to the air/liquid interface of a GO solution and then spin casting to form a relatively
heterogeneous GO coating, (2) Method 2: direct drop spin casting of GO solution on the top surface of a PES membrane to
prepare a more highly interlocked GO thin layer. The gas permeabilities of the GO membranes obtained from different two
coating methods were quite different, strongly depending on different GO stacking structures. The thin Go membranes prepared
by method 1 exhibited typical gas permeation behavior following Knudsen transport in nanoporous membranes. Even after
several coating times, the selectivity just follows Knudsen selectivity, meaning that nanopores created by the edges of less
interlocked GO sheets are seldom covered with the same coating method (Fig. 13B). On the other hand, the GO membranes
prepared by method 2 showed high gas selectivity following molecular sieving mechanism. The GO membranes were dried at
ambient atmospheres, so some of intercalated water molecules still exist inside GO layers. The presence of water molecules inside
GO layers contributed to higher CO2 gas permeance than any other gases tested in this study, leading to high CO2/N2 selectivity
because the solubility of CO2 in water is much higher than that of N2 in water (Fig. 13C). In general, the presence of water vapor in
gas mixtures strongly deteriorates the separation performance of conventional membrane materials (e.g., hydrophobic glassy
polymers, zeolite membranes), leading to signicant reduction of both permeability and selectivity by water condensation on
membrane surfaces or inside pores. In this regard, water-enhanced gas separation in GO membranes can be used in post-
combustion CO2 capture process because dehydration process in the permeate side is much easier than water removal in the high-
pressure feed side. Also they tested the gas permeabilities of reduced GO membranes by thermal treatment below 2001C. During
thermal treatment, H2O and CO2 were evolved, leading to some defects and decrease in interlayer spacing. As a result, such
thermally reduced GO membranes showed high H2/CO2 selectivity (Fig. 13D).

1.14.4 Liquid Separation

1.14.4.1 Liquid Separation Using Porous Graphene Membrane


Transport of water molecules (B2.8 A diameter) through porous graphene membrane is mainly determined by pore size. Also, for
water/ion selectivity, the pore size should be carefully designed. Similar to gas separations, the experimental studies on liquid
separation, particularly for desalination, are at infant stage. Surwade et al.25 showed high water/ion selectivity using a single-layer
graphene suspended on a micron-scale aperture. To create pores in a single-layer graphene, they used oxygen plasma, and then
successfully introduced B1 nm pores at a density B1012 cm2 in graphene. With a proper plasma treatment condition, they
reported 100% NaCl rejection while water permeated fast through such porous graphene membrane (Fig. 14). This might be the
rst experimental report on the desalination properties of porous graphene membrane with subnanometre pores, although a
370 Graphene Membranes

Fig. 12 50:50 H2/CO2 and H2/N2 gas mixture separations and comparison with literature data. (A) and (B) show separation results for a 1.8-nm-
thick GO membrane, (C) and (D) for 9-nm membrane, and (E) and (F) for an 18-nm membrane. (G) Comparison of ultrathin GO membranes with
polymeric membranes and inorganic microporous membranes for H2/CO2 mixture separation: selectivity versus H2 permeance.85
Graphene Membranes 371

Fig. 13 Gas transport behavior through ultrathin GO membranes. (A) H2 permeance of thin GO membranes prepared by method two as a
function of applied feed pressure. (B) Gas permeance of GO membranes as a function of molecular weight (method one; dashed line represents
the ideal Knudsen selectivity) under dry and humidied conditions. (C) Gas permeance of GO membranes as a function of kinetic diameter
(method two) under dry and humidied conditions. (D) Gas transport behavior through thermally reduced GO membranes. H2 and CO2
permeability of thermally treated GO membranes as a function of temperature.84

number of simulation work have predicted graphene with subnanometre pores could act as a highly selective and permeable
membrane with better membrane performance than current state-of-the-art polymer membranes.
Although extensive efforts have been made to synthesize large-area graphene membranes, monolayer graphene without defects
over large areas still remains highly challenging as mentioned earlier. Currently available method to produce large-area graphene is
CVD. CVD graphene also includes multilayer spots, StoneWales (57 rings) defects, tears, and other intrinsic defects. Therefore,
the use of multilayer graphene can be more plausible for a potentially cheap, scalable, and high-yield approach to defect-free
graphene membrane fabrication without further healing process for unfavorable large pores responsible for nonselective ow.
Supposed that such multiplayer graphene membrane (note that each graphene sheet contains proper pores and pore density)
would be much closer to practical applications, Cohen-Tanugi et al.88 showed that multilayer NPG would be much better than
monolayer nanoporous graphene (NPG) by considering various congurational variables such as the number of graphene layer,
interlayer distance, pore size and pore alignment (Fig. 15). Although interlayer distance and pore alignment between graphene
layers, including appropriate pore diameter (69 A ), seem to be unrealistic, their simulation results give experimental scientist
useful material guidelines about multiplayer nanoporous graphene membrane.
Not for liquid separation, Kim et al.84 studied the effect of the number of graphene layers on gas separation properties.
Actually, CVD graphene has lots of structural defects (e.g., grain boundaries, point defects), so even multilayer graphene sheets still
can allow gas transport through interlayer spacing enlarged by folding or wrinkles and defects (Fig. 16). That is, if inserting
372 Graphene Membranes

Fig. 14 Water transport measurements and desalination experiments. (A) Porous graphene membrane assembly for water ux measurements. A
graphene membrane on a silicon chip with a 5-mm hole in a 300-nm-thick SiN membrane is sealed on a glass vial lled with deionized water. The
vial is turned upside down and placed in an oven at 401C. Water loss is measured by monitoring the mass of the vial. (B) Water loss after 24 h
and ionic conductivity through the same porous graphene membranes etched at various exposure times. C1 and C2 are controls with large tears
or completely broken graphene membranes, respectively. (C) Water/salt selectivity as a function of ID/IG ratio showing exceptionally high selectivity
for a short etching time. Selectivity was calculated as the ratio of water ux to ionic conductivity from (B), normalized to the water ux and
conductivity of the pore in SiN without graphene.25

appropriate molecular spacers and creating pores in graphene sheets can be achieved, such multilayer graphene membranes can be
used for molecular separation.

1.14.4.2 Liquid Separation Using Graphene Oxide Membrane


Since GO has hydrophilic nature and GO sheets can be well dispersed in water, GO membranes have been extensively studied for
water purication and desalination. Essentially, facile membrane fabrication from aqueous GO dispersion has also led to many
experimental studies using GO or modied GO membranes. Particularly, the studies on liquid separation, using GO membranes,
have been tremendously performed since Geim group reported the unimpeded evaporation of water through thin GO mem-
branes.82 Actually, in 1980, Boehm, Clauss, and Hofmann already published membrane properties of graphite oxide (not fully
exfoliated) and found that the membranes are permeable to water.89 Indeed, such tremendous interest on superfast water
transport through GO membranes is similar to the case of frictionless ow of water through carbon nanotube. However,
hydrophilic domains (decorated by oxygen-functional groups) in GO sheets cover above 80%, and thus the small amount of
graphitic domains would not be interconnected with each other. That is, it is still questionable how water molecules can diffuse
Graphene Membranes 373

Fig. 15 Effects of membrane congurational parameters on the water ow rate per pore (AC) and salt rejection (DF) of bilayer membranes: (A,
D) effects of pressure, (B,E) effects of layer separation at different pore offsets (i.e., O 0 A and O19.6 A ), and (C,F) effects of pore offset at a
layer separation of 8 A .88

through laminated GO layers so fast. After that, Joshi et al.90 demonstrated selective water/ion behaviors of freestanding GO
membranes. Initially, they measured the permeation rate of different liquids (e.g., water, glycerol, toluene, benzene, dimethyl
sulfoxide, and ethanol) using a two-chamber diffusion cell. One chamber was lled with water and other chamber was lled with
a 1 M sucrose solution. Only water molecules can permeate fast through the GO membrane by osmotic pressure (Fig. 17A). Using
374 Graphene Membranes

Fig. 16 Few-layered graphene-coated polymer membranes. (A) SEM images of graphene/PTMSP membrane surfaces. (B) Changes in gas
permeability and sheet resistance of graphene/PTMSP membranes as a function of the number of graphene layers. (C) AFM images of a graphene/
PTMSP membrane surface. (D) Relation between O2 permeability and O2/N2 selectivity of graphene/PTMSP membranes.84

the same experiment, salt permeabilities through GO membranes were measured by monitoring the increase of ion conductivity in
a receiving chamber as a function of time. Several salt ions were used to evaluate the salt permeability through GO membranes in
terms of hydrated ion size. Small ions (e.g., NaCl) can permeate through GO membranes until hydrated ion sizes exceed 4.5 A
(Fig. 17B).
Raidongia et al.91 showed the possibility of selective ion transport through nanochannels between GO sheets under a bias
voltage. Also, Sun et al. observed the selective water/ion transport properties of GO membranes using a two-chamber diffusion
cell.92 Small ions (e.g., NaCl) penetrated through GO membranes rapidly while the permeabilities of heavy metal ions were
relatively low, that is, selective ion separation can be performed by GO membranes (Fig. 18).
Actually, so far most of studies on graphene oxide or its derivatives for liquid separation membranes have used vacuum or high
pressure-driven ltration methods to obtain thin GO membranes on mechanically stable, porous supports. These methods can be
easily available in the laboratory, but they would not be suitable for industrial scale, such as roll-to-roll process. More recently,
Akbari et al.93 demonstrated the possibility of scale up of GO-based membranes, particularly for liquid separation. The GO
membrane structure in this work was somewhat different from ones reported in the literature because they prepared thin GO layer
on a porous polymeric support (e.g., Nylon66, pore size 0.2 mm) using highly viscous GO solutions (B60 mg mL1). High shear
coating rates led to highly ordered, continuous thin lms of multilayer GO composite membranes. These GO thin-lm composite
membranes showed high rejection (490%) for charged and uncharged organic molecules (the hydrated radius Z5 A ) (Fig. 19B
and C), but low rejection (30%40%) of monovalent and divalent ions (Fig. 19D). Also, they reported that such a highly ordered
GO sheets in the plane of the membrane could make more well-dened transport channels and then can enhance the water
permeability (71 m2 h1bar1 for 150 nm thick GO membranes) (Fig. 19A).
Apart from the liquid separation properties of GO membranes reported in the literature, the stability of GO membranes should
be further considered for practical applications. Depending on pH, ion impurities, and preparation conditions, multilayer GO
Graphene Membranes 375

Fig. 17 (A) Sieving through the atomic scale mesh. The shown permeation rates are normalized per 1 M feed solution and measured by using 5
mm thick membranes. No permeation could be detected for the solutes shown within the gray area during measurements lasting for at least 10
days. The thick arrows indicate our detection limit, which depends on a solute. Several other large moleculesincluding benzoic acid, DMSO, and
toluenewere also tested and exhibited no detectable permeation. The dashed curve is a guide to the eye, showing an exponentially sharp cutoff
at 4.5 A , with a width of E0.1 A . (B) Simulations of molecular sieving. Permeation rates for NaCl, CuCl2, MgCl2, propanol, toluene, and octanol
for capillaries. For octanol poorly dissolved in water, the hydrated radius is not known, and we use its molecular radius. Blue marks indicate
permeation cutoff for an atomic cluster (inset) for graphene capillaries accommodating two and three layers of water (widths of 9 and 13 A ,
respectively).90

Fig. 18 The penetration processes of different ionic compounds through GO membranes.92

membranes can be disintegrated in the aqueous solution. In many cases, actual liquid separation processes use crossow ltration
system to prevent concentration polarization and membrane fouling. There are few reports on the structural stability issue of
pristine GO membranes in such practical condition. According to more recent work on the stability,94 neat GO membranes easily
disintegrate in water because GO sheets become negatively charged on hydration and the electrostatic repulsion repels GO sheets
with the same negative charges. They also found that GO preparation condition should be carefully checked. Many researchers
used anodized aluminum oxide lter disk during the preparation of GO membranes. During the ltration, aluminum ions can be
released, which can crosslink GO sheets via electrostatic attraction. As a result, the GO membranes from anodized aluminum lter
did not show severe disintegration in water (Fig. 20A). The pH of solution will be also a signicant factor to determine the
structural stability of the resultant GO membranes (Fig. 20B).

1.14.5 Other Membrane Applications

1.14.5.1 Fuel Cell Applications


In 2013, Tateishi et al.95 reported that GO shows relatively high proton conductivity even at low relatively humidity (RH) and
temperatures. The main oxygen-containing functional group in GO is epoxide, which acts as the site for proton transfer after water
376 Graphene Membranes

Fig. 19 (A) Water ux versus applied pressure for three different membranes: SAM (red) with a thickness of 150715 nm, vacuum ltration
(blue) with a thickness of 170720 nm, and NF270, a commercial nanoltration membrane (green). SAM showed a retention of 9072% for
methyl red, while the vacuum ltration membrane and NF270 showed 5075% and 9071.5% retention, respectively. (B) Retention performance of
the 150715 nm thick shear-aligned membrane, as a function of hydrated radius, for probe molecules with different charges and sizes. (MV is
methyl viologen, MR is methyl red, MnB is methylene blue, MO is methyl orange, OG is orange G, Ru is Tris (bipyridine) ruthenium (II) chloride,
RB is Rhodamine B, RosB is Rose Bengal, MB is methylene blue, BB is brilliant blue. The green, red, and blue symbols represent electroneutral,
negatively, and positively charged probe molecules, respectively. (C) Retention details of the membrane for the probe molecules. (d) Salt retention
by the 150715 nm thick SAM, for four different salt solutions. Error bars in these gures are from ve measurements showing the maximum and
minimum values.93

molecules bind to epoxide, even at low RH and room temperature. The proton conductivities are 105104 S cm1 at 10%20%
RH at 251C. This feature of GO can be used in various electrochemical cells and batteries. Although the proton conductivity of
thick GO paper is lower than that of fully hydrated Naon membrane at high RH (B100%), commercially used for fuel cell
applications, the proton conductivity of GO lms is strongly dependent on the thickness, so with thinner GO lm can compete
with Naon membrane. Fig. 21 shows the performance of the GO fuel cell with Pt/C electrodes. The GOFC with Pt/C electrodes
shows better fuel cell performance than a membrane electrode assembly (MEA) consisting of Pt/C electrodes and Naon mem-
brane with B150 mm thickness.
Hu et al.96 reported an interesting feature of graphene and other 2D materials. That is, monolayers of graphene and hexagonal
boron nitride (hBN) are highly permeable to thermal protons under ambient conditions while no proton transport was monitored
for thicker crystals such as monolayer molybdenum disulde (MoS2), bilayer graphene, or multilayer hBNs. The authors proposed
that the proton conductivity could be further improved by introducing the graphene and hBN membranes with catalytic metal
nanoparticles, which will lead to promising hydrogen-based technologies (Fig. 22).

1.14.5.2 Ion-Selective Membranes for Batteries


Since GO membranes have ion selectivity, researchers have investigated the use of GO membranes for ion-selective applications
such as lithium-sulfur batteries97 and vanadium redox ow batteries.98 Huang et al.97 used ion-selective ultrathin GO membrane
to develop a novel, ion selective but highly permeable battery separator for signicantly improving both energy density and power
density of lithium-sulfur batteries. Lithium-sulfur battery is one of the promising next-generation batteries. Here polysuldes are
materials generated at the cathode side, diffuse from through the membrane, react with lithium anode, and shuttle back. During
the process, polysuldes dissolve and irreversibly react with metal lithium and organic components, including the destruction of
Graphene Membranes 377

Fig. 20 (A) GO membranes obtained from aaO and Teon lters have similar microstructures but drastically different mechanical properties and
stability in water. Photos of GO lm prepared on Teon and AAO. GO (Teon) readily disintegrates in water, whereas GO (AAO) remains intact. (B)
Effect of pH on the corrosion of AAO and the stability of the corresponding GO membranes in water. Concentration of released Al as a function of
pH after 1 day. Photos of GO (AAO) prepared from GO solutions at pH 3, pH 5.5 and pH 8.5 before and after being soaked in water for 30 min
and 2 h, respectively.94

the cathode structure, depletion of the lithium anode, and loss of active sulfur materials. So far, porous polymer membranes such
as porous polypropylene or polyethylene have been used for the separator to separate physically two electrodes but no ion-
selective nature. On the other hand, GO membranes can block polysulde by electrostatic repulsion and steric exclusion, which are
efcient in long-term inhabitation of the shuttle effect (Fig. 23).
For needs for large-capacity energy storage system, the vanadium redox ow battery (VRB) is an ideal candidate due to its high
efciency, long cycle times, and comparatively low cost. However, the VRB commercial membrane Naon exhibits several
drawbacks, such as high cost ($500700 m2) and low ion selectivity. Thus, considerable efforts have been made to develop a
hydrocarbon polymer membrane with high ion selectivity and adjustable ion conductivity for high-performance VRBs. In general,
sulfonated hydrocarbon membranes have been developed. However, high degree of sulfonation of hydrocarbon polymers to
obtain high ion conductivity often lead to reduced mechanical strength as well as high swelling degree that affects vanadium ion
crossover through the membrane, which limits the VRB performance. To prevent such negative effects, GO was incorporated into
such highly sulfonated polymer electrolytes because GO itself could provide a proton transport pathway via surface diffusion.95
Aziz et al.98 reported a highly ion-selective membrane for vanadium redox ow batteries consisting of sulfonated poly(arylene
378 Graphene Membranes

Fig. 21 Polarization and power density curves for the membrane electrode assembly (MEA) consisting of GO electrolyte and Pt/C electrodes (red)
and that composed of Naon electrolyte and Pt/C electrodes (blue) at 251C.95

Fig. 22 Proton transport through two-dimensional crystals. (A) Examples of IV characteristics for monolayers of hBN, graphite, and MoS2. (B)
Histograms for two-dimensional crystals that are found to exhibit measurable proton conductivity.96

ether ketone) coupled with GO. With incorporation of GO, the GO composite membrane exhibited an effectively low self-
discharge rate and excellent Coulombic efciency in VRBs (Fig. 24).

1.14.5.3 Pervaporation
Owing to rapid water transport rate through hydrophilic GO membranes and high rejection properties of large organic solvents,
GO membranes have been studied for dehydration of aqueous organic solution using pervaporation process. Huang et al.99
prepared GO-coated ceramic hollow bers by using a vacuum suction method. They tried to separate dimethyl carbonate(DMC)/
water mixtures using GO-coated ceramic membranes, and reported high water permeation ux up to 2100 m2 h1 at 401C with a
high separation factor of 740. Such a high separation factor cannot be explained only by size-exclusion mechanism, based on the
kinetic diameters of water (0.265 nm) and DMC (0.47 nmodDMCo0.63 nm) molecules, and interlayer distance of GO sheets
(0.61.0 nm). Usually, pervaporation membrane follows solutiondiffusion mechanism. Thus, total selectivity can be affected by
both diffusion-selectivity and solubility-selectivity. They checked the sorption amount of water and DMC on GO surfaces and
revealed that the sorption amount of water on GO surfaces is larger than that of DMC, leading to high solubility selectivity
(Fig. 25).
Graphene Membranes 379

Fig. 23 Electrochemical performance of the GO membrane-incorporated lithium-sulfur batteries. (A) Cycling performance at a rate of 0.1 C with/
without a GO membrane; (B) galvanostatic chargedischarge proles at a rate of 0.1 C; (C) CV proles with a GO membrane at a scan rate of 0.1
mV s1; (D) open circuit voltage proles showing the self-discharge behavior; (E) electrochemical impedance spectra (inset is the enlarged EIS at
the high-frequency region); and (F) rate performance of lithium-sulfur batteries with/without a GO membrane.97

1.14.5.4 Membrane Distillation


Global demand for safe drinking water is more and more increasing in the upcoming decades and energy-saving desalination
processes over thermal distillation and reverse osmosis are of great interest. Membrane distillation (MD) is a membrane-based
thermal evaporation process that may offer energy-efcient water production at relatively low temperatures (50901C). Thus, MD
can be performed effectively using low-grade heat sources and solar power.100102 In MD, a hydrophobic porous membrane acts as
a barrier between a hot feed and a cool permeate. As the heated brine passes on the feed side of the membrane, it is partially
transformed to water vapor and passes through the pores to be condensed on the permeate side. Bhadra et al.103 showed that
immobilization of GO on very hydrophobic polytetrauoroethylene (PTFE) membrane surface signicantly could enhance the
overall permeate ux with complete salt rejection, and the ux reached as high as 97 kg m2 h at 801C (Fig. 26).
380 Graphene Membranes

Fig. 24 (A) Discharge capacity retention of the VRB and (B) self-discharge curves of the VRBs constructed with Naon-212, SPAEK, SPAEK/PW-
mGO (1%), and SPAEK/PW-mGO (2%) membranes. (C) Cycle efciencies of the VRB constructed with the SPAEK/PW-mGO (1%) membrane as a
function of cycling numbers at a current density of 40 mA cm2 and (D) different current density values.98

Fig. 25 (A) Total permeation ux and (B) water content in permeate for the separation of DMC/water mixture. (C) QCM responses of the GO
sorption capacity for water (black), methanol (red), and DMC (blue).99

The presence of oxygen-containing functional groups on the GO sheets is known to create nanoscale wrinkles and structural
defects that provide passage for water transport while the hydrophobic graphitic domains form a near-frictionless surface for rapid
water ow across the membrane. This may help enhance the overall water ux through GO-assisted membranes. In general,
graphene has high thermal conductivity while the thermal transport properties of GO remain unknown. Simulation technique
predicted that the thermal conductivity of GO is signicantly reduced as the oxygen content increases. Since the GO is only present
on the surface, this helps in reducing the temperature polarization without contributing toward conductive heat losses, which are
key factors for increasing ux in MD.104

1.14.6 Other 2D Materials for Membrane Applications

Actually, graphene has triggered a great deal of attention toward 2D materials in membrane science and technology. Owing to their
atomic thickness and scalable lateral dimension, many 2D materials have been developed extensively as an amazing membrane
Graphene Membranes 381

Fig. 26 Effect of (A) temperature and (B) ow rate on water vapor ux.103

platform for energy-efcient technologies. So far, we discussed the rapid development of graphene or graphene-based membranes
for many membrane applications such as gas separation, water purication and desalination, pervaporation, fuel cell applications,
membrane distillation, and batteries. Although all 2D materials do not always have atomic thickness like graphene, studies to
prepare 2D nanosheets from bulk crystal materials with well-dened pores have been extensively performed. For examples,
zeolites and MOFs will be such candidate materials. Many researchers highly expect the use of 2D zeolite nanosheets for molecular
separation. Recently, the Tsapatsis group successfully obtained high-purity exfoliated zeolite (MFI) nanosheets (1.5-unit-cell thick,
300-nm lateral size) by using a polymer-melt-compounding exfoliation technique combined with a density gradient centrifuga-
tion method.105 They showed thin zeolite lms consisting of zeolite nanosheets with high aspect ratio can be fabricated on porous
supports as shown in Fig. 27. Because of high aspect ratio of nanosheets, any nanosheets did not penetrate into large pores of
support membranes, indicating that high-ux zeolite membranes can be fabricated. However, defects between adjacent nanosheets
still exist, so these defects should be healed to obtain high selective membranes. More recently, the nanosheets of metal organic
frameworks (MOFs) have been also extensively studied for membrane applications.106 The addition of MOF nanosheets into
polymer matrix leads to many positive properties, for example, antiplasticization107 or increasing both gas permeability and
selectivity.108 Similar studies will be continued because of their versatility.
These days, there are many methods to obtain 2D nanosheets. Among them, liquid exfoliation of layered materials will be one
of the promising methods.109 Generally, layered materials consist of 2D platelets weakly stacked to form 3D structures. The
representative materials are graphite, hBN, transition-metal dichalcogenides, transition metal trichalcogenides, metal halides,
oxides, IIIVI layered semiconductors, layered zirconium phosphates and phosphonates, clay (layered silicates), layered double
hydroxides (LDHs), ternary transition meal carbides, and nitrides. Some of these layered materials have been already studied to
develop new membrane materials in the form of mixed matrix membranes. In any sense, 2D nanosheets can be good candidate
llers to be able to affect the polymeric membrane performance by incorporating a small amount of them, by considering the
actual selective membrane layer is as thin as below 100 nm. Fig. 28 shows a schematic description of the main liquid exfoliation
mechanisms.

1.14.7 Conclusions and Perspectives

Graphene and other 2D materials triggered by graphene provide a new avenue to explore high-performance membrane materials
with high permeability and selectivity. While other nanomaterials or nanoporous materials have been studied mainly as rein-
forcing or transport-enhanced llers in conventional membrane materials (e.g., polymers), 2D materials like graphene have
382 Graphene Membranes

Fig. 27 Images of the MFI nanosheet coating on porous supports. (A) SEM image (top view) of the coating of MFI nanosheets on an Anopore
disk. The top half of the image shows the bare Anopore support, whereas the bottom half shows a uniform coating of nanosheet on the 200-nm
pores of the support. (B) SEM image (top view) of the coating of an MFI nanosheet on a homemade porous a-alumina support. (C) FIB image of
the cross section of the coating in (B). The image was taken by a Ga ion source (30 kV) at a tilt angle of 52 degree. The nanosheet coating is
sandwiched between the FIB-deposited platinum (to protect the coating from milling) and the alumina support. (D) TEM image of the cross section
of the coating in (B). The dark layer on top of the coating is FIB-deposited platinum. (E) HRTEM image of the coating cross section. Scale bars in
(A) to (D), 200 nm; in (E), 20 nm.105

already membrane form. As such, by controlling nanopores in these materials at the nanoscale, these materials can be used for
membrane applications such as gas and liquid separations. However, for actual membrane applications, it will be a long way for
them to be realized in the real membrane platform (e.g., thin-lm composite membrane). Precise pore engineering has been
already achieved, but subnanometer pore size is still greatly challenging. We can gure out how difcult it is to make isoporous
membranes by considering that commercially available isoporous membranes are currently polycarbonate track-etched mem-
branes and anodic aluminum oxide membranes at best. They have been used only for research and analytic applications. Scale up
is not as a signicant issue as other membrane materials because graphene can be already synthesized in any large size using CVD
technique. However, such CVD graphene still suffers from many structural defects (e.g., line defects or point defects), which
hinders early applications in the eld of electronics. Such defects, rather, can be used for membrane applications if these defects
can be properly controlled or healed with functionalization. Another difcult thing will be in a queue. Graphene growth can occur
on catalytic metal substrate (e.g., Cu foil). That is, either graphene or porous graphene on metal substrate should be carefully
transferred to appropriate porous supports for real applications. Many graphene transfer methods have been actively developed,
but most of them look like unsuitable for existing membrane fabrications. Nondestructive graphene transfer method should be
developed, too. Owing to one-atomic nature of graphene, it is hard to handle such an ultrathin material for membrane fabrication.
To avoid the formation of defects during the synthesis or transfer will be very difcult. Therefore, a few layer stacking method may
be in a sense the better choice for real membrane congurations. Finally, support membranes suitable for such 2D materials
should be developed. In many cases, the importance of support membranes has been often underscored. At present, the surface
porosity of existing support membranes (in the ultraltration level) reaches B5% at best. That is, even though super-permeable,
ultrathin 2D-selective layers could be developed, the combination with existing support membranes consequently leads to low
permeance and selectivity.
Graphene Membranes 383

Fig. 28 Schematic description of the main liquid exfoliation mechanisms (ref). (A) Ion intercalation, (B) ion exchange, (C) sonication-assisted
exfoliation.109

Despite many persistent challenging issues, graphene and graphene-like 2D materials will offer a great opportunity to open a
new area in the membrane eld or beyond separation eld. Such ultra-thin, atom-thick-membranes can be used for other
applications such as batteries, fuel cells, sensors, energy harvesting, and biomedical applications, which is not only for large-scale
separation processes. The understanding of selective mass transport through single nanopore, as occurred in biological membranes
(e.g., aquaporins), is a persistent longing issue in the scientic community so far. In this regard, 2D membrane materials with one
atom thickness would be a great platform to study the selective mass transport through single pore or functionalized single pore,
which will be a rst step for much ideal separation using membranes.

See also: 2.6 Polymeric Membranes for Gas Separation. 2.7 Design and Preparation of Pervaporation Membranes.
2.12 Electromembrane Processes: Basic Aspects and Applications

References

1. Wallace, P. R. The Band Theory of Graphite. Phys. Rev. 1947, 71, 622.
2. Novoselov, K. S.; et al. Electric Field Effect in Atomically Thin Carbon Films. Science 2004, 306, 666669.
3. Geim, A. K.; Novoselov, K. S. The Rise of Graphene. Nat. Mater. 2007, 6, 183191.
4. Fernandez-Moran, H. Single Crystals of Graphite and Mica as Specimen Support for Electron Microscopy. J. Appl. Phys. 1960, 31, 1840.
5. Zhang, Y. B.; Small, J. P.; Pontius, W. V.; Kim, P. Fabrication and Electric-Field-Dependent Transport Measurements of Mesoscopic Graphite Devices. Appl. Phys. Lett.
2005, 86, 073104.
6. Soldano, C.; Mahmood, A.; Durardin, E. Production, Properties and Potential of Graphene. Carbon 2010, 48, 21272150.
7. Berger, C.; Song, Z. M.; Li, X. B.; Wu, X. S.; Brown, N.; Naud, C.; Mayou, D.; Li, T. B.; Hass, J.; Marchenkov, A. N.; Conrad, E. H.; First, P. N.; de Heer, W. A.
Electronic Connement and Coherence in Patterned Epitaxial Graphene. Science 2006, 312, 11911196.
8. Bae, S.; et al. Roll-to-Roll Production of 30-inch Graphene Films for Transparent Electrodes. Nat. Nanotechnol. 2010, 5, 574578.
9. Kim, K. S.; et al. Large-Scale Pattern Growth of Graphene Films for Stretchable Transparent Electrodes. Nature 2009, 457, 706710.
10. Li, X.; et al. Large-Area Synthesis of High-Quality and Uniform Graphene Films on Copper Foils. Science 2009, 324, 13121314.
11. Cai, W. W.; Moore, A. L.; Zhu, Y. W.; Li, X. S.; Chen, S. S.; Shi, L.; Ruoff, R. S. Thermal Transport in Suspended and Supported Monolayer Graphene Grown by
Chemical Vapor Deposition. Nano Lett. 2010, 10, 16451651.
12. Li, X. S.; et al. Large-Area Graphene Single Crystals Grown by Low-Pressure Chemical Vapor Deposition of Methane on Copper. J. Am. Chem. Soc. 2011, 133,
28162819.
13. Yang, X. Y.; Dou, X.; Rouhanipour, A.; Zhi, L. J.; Rader, H. J.; Mullen, K. Two-Dimensional Graphene Nanoribbons. J. Am. Chem. Soc. 2008, 130, 42164217.
14. Cai, J. M.; Rufeux, P.; Jaafar, R.; Bieri, M.; Braun, T.; Blankenburg, S.; Muoth, M.; Seitsonen, A. P.; Saleh, M.; Feng, X. L.; Mullen, K.; Fasel, R. Atomically Precise
Bottom-Up Fabrication of Graphene Nanoribbons. Nature 2010, 466, 470473.
15. Kobayahsi, T.; et al. Production of a 100-m-Long High-Quality Graphene Transparent Conductive Film by Roll-to-Roll Chemical Vapor Deposition and Transfer Process.
Appl. Phys. Lett. 2013, 102, 023112.
16. Steward, E. G.; Cook, B. P.; Kellett, E. A. Dependence on Temperature of the Interlayer Spacing in Carbons of Different Graphitic Perfection. Nature 1960, 187,
10151016.
384 Graphene Membranes

17. Yoon, D.; Son, Y.-W.; Cheong, H. Negative Thermal Expansion Coefcient of Graphene Measured by Raman Spectroscopy. Nano Lett. 2011, 11, 32273231.
18. Balandin, A. A. Thermal Properties of Graphene and Nanostructured Carbon Materials. Nat. Mater. 2011, 10, 569.
19. Ghosh, S.; Nika, D. L.; Pokatilov, E. P.; Balandin, A. A. Heat Conduction in Graphene: Experimental Study and Theoretical Interpretation. New J. Phys. 2009, 11, 095012.
20. Nan, H. Y.; Ni, Z. H.; Wang, J.; Zafar, Z.; Shi, Z. X.; Wang, Y. Y. The Thermal Stability of Graphene in Air Investigated by Raman Spectroscopy. J. Raman Spectrosc.
2013, 44, 10181023.
21. Banhart, F.; Kotakoski, J.; Krasheninnikov, A. V. Structural Defects in Graphene. ACS Nano 2011, 5, 2541.
22. Fischbein, M. D.; Drndic, M. Electron Beam Nanosculpting of Suspended Graphene Sheets. Appl. Phys. Lett. 2008, 93, 113107.
23. Koenig, S. P.; Wang, L.; Pellegrino, J.; Bunch, J. S. Selective Molecular Sieving Through Porous Graphene. Nat. Nanotechnol. 2012, 7, 728732.
24. OHern, 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, 12341241.
25. Surwade, S. P.; Smirnov, S. N.; Vlassiouk, I. V.; Unocic, R. R.; Veith, G. M.; Dai, S.; Mahurin, S. M. Water Desalination Using Nanoporous Single-Layer Graphene. Nat.
Nanotechnol. 2015, 10, 459464.
26. Bai, J.; Zhong, X.; Jiang, S.; Huang, Y.; Duan, X. Graphene Nanomesh. Nat. Nanotechnol. 2010, 5, 190194.
27. Bieri, M.; Treier, J.; Cai, J.; Ait-Mansour, K.; Rufeux, P.; Groning, O.; Groning, P.; Kastler, M.; Rieger, R.; Feng, X.; Mullen, K.; Fasel, R. Porous Graphenes: Two-
Dimensional Polymer Synthesis With Atomic Precision. Chem. Commun. 2009, 45, 69196921.
28. Celebi, K.; et al. Ultimate Permeation Across Atomically Thin Porous Graphene. Science 2014, 344, 289292.
29. Segal, M. Selling Graphene by the Ton. Nat. Nanotechnol. 2009, 4, 611613.
30. Park, S.; Ruoff, R. S. Chemical Methods for the Production of Graphenes. Nat. Nanotechnol. 2009, 4, 217224.
31. Bai, H.; Li, C.; Wang, X.; Shi, G. Q. On the Gelation of Graphene Oxide. J. Phys. Chem. C 2011, 115, 55455551.
32. Huang, L.; Li, C.; Yuan, W. J.; Shi, G. Q. Strong Composite Films With Layered Structures Prepared by Casting Silk Fibroin-Graphene Oxide Hydrogels. Nanoscale
2013, 5, 37803786.
33. Putz, K. W.; Compton, Q. C.; Segar, C.; An, Z.; Nguyen, S. T.; Brinston, L. C. Evolution of Order During Vacuum-Assisted Self-Assembly of Graphene Oxide Paper and
Associated Polymer Nanocomposites. ACS Nano 2011, 5, 66016609.
34. Choi, W.; Choi, J.; Bang, J.; Lee, J.-H. Layer-by-Layer Assembly of Graphene Oxide Nanosheets on Polyamide Membranes for Durable Reverse-Osmosis Applications.
ACS Appl. Mater. Interfaces 2013, 5, 1251012519.
35. Hu, M.; Mi, B. Enabling Graphene Oxide Nanosheets as Water Separation Membranes. Environ. Sci. Technol. 2013, 47, 37153723.
36. Brodie, B. C. On the Atomic Weight of Graphite. Philos Trans. R. Soc. Lond. B: Biol. Sci. 1859, 149, 249259.
37. Staudenmaier, L. Verfahren Zur Darstellung der Graphitsure. Ber. Dtsch. Chem. Ges. 1898, 31, 14811487.
38. Staudenmaier, L. Verfahren Zur Darstellung der Graphitsure. Ber. Dtsch. Chem. Ges. 1899, 32, 13941399.
39. Hummers, W. S.; Offeman, R. E. Preparation of Graphitic Oxide. J. Am. Chem. Soc. 1958, 80, 1399.
40. Kovtyukhova, N. I.; Ollivier, P. J.; Martin, B. R.; Mallouk, T. E.; Chizhik, S. A.; Buzaneva, E. V.; Gorchinskiy, A. D. Layer-by-Layer Assembly of Ultrathin Composite Films
From Micron-Sized Graphite Oxide Sheets and Polycations. Chem. Mater. 1999, 11, 771778.
41. Chen, J.; Yao, B.; Li, C.; Shi, G. An Improved Hummers Method for Eco-Friendly Synthesis of Graphene Oxide. Carbon 2013, 64, 225229.
42. Marcano, D. C.; Kosynkin, D. V.; Berlin, J. M.; Sinitskii, A.; Sun, Z.; Slessarev, A.; Alemany, L. B.; Lu, W.; Tour, J. M. Improved Synthesis of Graphene Oxide. ACS
Nano 2010, 4, 48064814.
43. Dreyer, D. R.; Todd, A. D.; Bielawski, C. W. Harnessing the Chemistry of Graphene Oxide. Chem. Soc. Rev. 2014, 43, 52885301.
44. Hofmann, U.; Holst, R. The Acidic Nature and the Methylation of Graphitoxide. Ber. Dtsch. Chem. Ges. 1939, 72, 754771.
45. Ruess, G. ber das Graphitoxyhydroxyd (Graphitoxyd). Monatsch. Chem. 1947, 76, 381417.
46. Scholz, W.; Boehm, H. P. Graphite Oxide. 6. Structure of Graphite Oxide. Z. Anorg. Allg. Chem. 1969, 369, 327340.
47. Nakajima, T.; Mabuchi, A.; Hagiwara, R. A New Structure Model of Graphite Oxide. Carbon 1988, 26, 357361.
48. Lerf, A.; He, H. Y.; Forster, M.; Klinowski, J. Structure of Graphite Oxide Revisited. J. Phys. Chem. B 1998, 102, 44774482.
49. He, H. Y.; Riedl, T.; Lerf, A.; Klinowski, J. Solid-State NMR Studies of the Structure of Graphite Oxide. J. Phys. Chem. 1996, 100, 1995419958.
50. Lerf, A.; He, H. Y.; Riedl, T.; Forster, M.; Klinowski, J. C-13 and H-1 MAS NMR Studies of Graphite Oxide and Its Chemically Modied Derivatives. Solid State Ion
1997, 101, 857862.
51. He, H.; Klinowski, J.; Forster, M.; Lerf, A. A New Structural Model for Graphite Oxide. Chem. Phys. Lett. 1998, 287, 5356.
52. Szabo, T.; Berkesi, O.; Forgo, P.; Josepovits, K.; Sanakis, Y.; Petridis, D.; Dekany, I. Evolution of Surface Functional Groups in a Series of Progressively Oxidized
Graphite Oxides. Chem. Mater. 2006, 18, 27402749.
53. Kim, S.; Zhou, S.; Hu, Y.; Acik, M.; Chabal, Y. J.; Berger, C.; de Heer, W.; Bongiorno, A.; Riedo, E. Room-Temperature Metastability of Multilayer Graphene Oxide Films.
Nat. Mater. 2012, 11, 544549.
54. Acik, M.; Mattevi, C.; Gong, C.; Lee, G.; Cho, K.; Chhowalla, M.; Chabal, Y. J. The Role of Intercalated Water in Multilayered Graphene Oxide. ACS Nano 2010, 4,
58615868.
55. Paneri, A.; Moghaddam, S. Impact of Synthesis Conditions on Physicochemical and Transport Characteristics of Graphene Oxide Laminates. Carbon 2015, 86, 245255.
56. Han, Y.; Xu, Z.; Cao, C. Ultrathin Graphene Nanoltration Membrane for Water Purication. Adv. Funct. Mater. 2013, 23 (29), 36933700.
57. Fan, Z.; Zhao, Q.; Li, T.; Yan, J.; Ren, Y.; Feng, J.; Wei, T. Easy Synthesis of Porous Graphene Nanosheets and Their Use in Supercapacitors. Carbon 2012, 50,
16991712.
58. Koinuma, M.; Ogata, C.; Kamei, Y.; Hatakeyama, K.; Tateishi, H.; Watanabe, Y.; Taniguchi, T.; Gezuhara, K.; Hayami, S.; Funatsu, A.; Sakata, M.; Kuwahara, Y.; Kurihara,
S.; Matsumoto, Y. Photochemical Engineering of Graphene Oxide Nanosheets. J. Phys. Chem. C 2012, 116 (37), 1982219827.
59. Russo, P.; Hu, A.; Compagnini, G. Synthesis, Properties and Potential Applications of Porous Graphene: A Review. NanoMicro Lett. 2013, 5 (4), 260273.
60. Yu, C.; Zhang, B.; Yan, F.; Zhao, J.; Li, L.; Li, J. Engineering Nano-Porous Graphene Oxide by Hydroxyl Radicals. Carbon 2016, 105, 291296.
61. Baker, R. W.; Low, B. T. Gas Separation Membrane Materials: A Perspective. Macromolecules 2014, 47, 69997013.
62. Baker, R. W. Membrane Technology and Applications; Wiley: Chichester, 2004.
63. Yampolskii, Y. Polymeric Gas Separation Membranes. Macromolecules 2012, 45, 32983311.
64. Baker, R. W. Future Directions of Membrane Gas Separation Technology. Ind. Eng. Chem. Res. 2002, 41, 13931411.
65. Bunch, J. S.; et al. Impermeable atomic membranes from graphene sheets. Nano Lett. 2008, 8, 24582462.
66. Jiang, D.; Cooper, V. R.; Dai, S. Porous Graphene as the Ultimate Membrane for Gas Separation. Nano Lett. 2009, 9, 40194024.
67. Du, H.; et al. Separation of Hydrogen and Nitrogen Gases With Porous Graphene Membrane. J. Phys. Chem. C 2011, 115, 2326123266.
68. Schrier, J. Helium Separation Using Porous Graphene Membranes. J. Phys. Chem. Lett. 2010, 1, 22842287.
69. Blankenburg, S.; et al. Porous Graphene as an Atmospheric Nanolter. Small 2010, 6, 22662271.
70. Cranford, S. W.; Buehler, M. J. Selective Hydrogen Purication Through Graphdiyne Under Ambient Temperature and Pressure. Nanoscale 2012, 4, 4587.
71. Jiao, Y.; et al. Graphdiyne: A Versatile Nanomaterial for Electronics and Hydrogen Purication. Chem. Commun. 2011, 47, 11843.
72. Zhang, H.; et al. Tunable Hydrogen Separation in spsp2 Hybridized Carbon Membranes: A First-Principles Prediction. J. Phys. Chem. C 2012, 116, 1663416638.
73. Hu, W.; Wu, X.; Li, Z.; Yang, J. Porous Silicone as a Hydrogen Purication Membrane. Phys. Chem. Chem. Phys. 2013, 15, 57535757.
74. Mahood, J.; et al. Nitrogenated Holey Two-Dimensional Structures. Nat. Commun. 2015, 6, 6486.
Graphene Membranes 385

75. Schrier, J. Fluorinated and Nanoporous Graphene Materials as Sorbents for Gas Separations. ACS Appl. Mater. Interfaces 2011, 3, 44514458.
76. Li, Y.; Zhou, Z.; Shen, P.; Chen, Z. Two-Dimensional Polyphenylene: Experimentally Available Porous Graphene as a Hydrogen Purication Membrane. Chem. Commun.
2010, 46, 3672.
77. Schrier, J.; McClain, J. Thermally-Driven Isotope Separation Across Nanoporous Graphene. Chem. Phys. Lett. 2012, 521, 118124.
78. Tian, Z.; Dai, S.; Jiang, D. Expanded Porphyrins as Two-Dimensional Porous Membranes for CO2 Separation. ACS Appl. Mater. Interfaces 2015, 7, 1307313079.
79. Sun, C.; et al. Mechanisms of Molecular Permeation Through Nanoporous Graphene Membranes. Langmuir 2014, 30, 675682.
80. Hauser, A. W.; Schwerdtfeger, P. Nanoporous Graphene Membranes for Efcient 3He/4He Separation. J. Phys. Chem. Lett. 2012, 3, 209213.
81. Dikin, D. A.; et al. Preparation and Characterization of Graphene Oxide Paper. Science 2007, 448, 457460.
82. 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, 442444.
83. Dreyer, D. R.; Park, S.; Bielawski, C. W.; Ruoff, R. S. The Chemistry of Graphene Oxide. Chem. Soc. Rev. 2010, 39 (1), 228240.
84. 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.; Choi, J.; Park, H. B. Selective Gas Transport
Through Few-Layered Graphene and Graphene Oxide Membranes. Science 2013, 342, 9195.
85. Li, H.; Song, Z.; Zhang, X.; Huang, Y.; Li, S.; Mao, Y.; Ploehn, H. J.; Bao, Y.; Yu, M. Ultrathin, Molecular-Sieving Graphene Oxide Membranes for Selective Hydrogen
Separation. Science 2013, 342, 9598.
86. Park, H. B.; Suh, I. Y.; Lee, Y. M. Novel Pyrolytic Carbon Membranes Containing Silica: Preparation and Characterization. Chem. Mater. 2002, 14 (7), 30343046.
87. Suda, H.; Haraya, K. Gas Permeation Through Micropores of Carbon Molecular Sieve Membranes Derived From Kapton Polyimide. J. Phys. Chem. B 1997, 101 (20),
39883994.
88. Cohen-Tanugi, D.; Lin, L.-C.; Grossman, J. C. Multilayer Nanoporous Graphene Membranes for Water Desalination. Nano Lett. 2016, 16, 10271033.
89. Boehm, H. P.; Clauss, A.; Hofmann, U. Graphite Oxide and Its Membrane Properties. J. Chim. Phys. 1960, 58 (12), 110117.
90. 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, 752754.
91. Raidongia, K.; Huang, J. Nanouidic Ion Transport Through Reconstructed Layered Materials. J. Am. Chem. Soc. 2012, 134, 1652816531.
92. Sun, P.; Zhu, M.; Wang, K.; Zhong, M.; Wei, J.; Wu, D.; Xu, Z.; Zhu, H. Selective Ion Penetration of Graphene Oxide Membranes. ACS Nano 2013, 7, 428437.
93. Akbari, A.; Sheath, P.; Martin, S. T.; Shinde, D. B.; Shaibani, M.; Banerjee, P. C.; Tkacz, R.; Bhattacharyya, D.; Majumder, M. Large-Area Graphene-Based Nanoltration
Membranes by Shear Alignment of Discotic Nematic Liquid Crystals of Graphene Oxide. Nat. Commun. 2016, 7, 10891.
94. Yeh, C.-N.; Raidongia, K.; Shao, J.; Yang, Q.-H.; Huang, J. On the Origin of the Stability of Graphene Oxide Membranes in Water. Nat. Chem. 2015, 7, 166170.
95. Tateishi, H.; et al. Graphene Oxide Fuel Cell. J. Electrochem. Soc. 2013, 160 (11), F1175F1178.
96. Hu, S.; et al. Proton Transport Through One-Atom-Thick Crystals. Nature 2014, 516, 227230.
97. Huang, J.-Q.; Zhuang, T.-Z.; Zhang, Q.; Peng, H.-J.; Chen, C.-M.; Wei, F. Permselective Graphene Oxide Membrane for Highly Stable and Anti-Self-Discharge
LithiumSulfur Batteries. ACS Nano 2015, 9 (3), 30023011.
98. Aziz, M. A.; Oh, K.; Shanmugam, S. A Sulfonated Poly(Arylene Ether Ketone)/Polyoxometalate-Graphene Oxide Composite: A Highly Ion Selective Membrane for All
Vanadium Redox Flow Batteries. Chem. Commun. 2017, 53, 917920.
99. Huang, K.; Liu, G.; Lou, Y.; Dong, Z.; Shen, J.; Jin, W. A Graphene Oxide Membrane With Highly Selective Molecular Separation of Aqueous Organic Solution. Angew.
Chem. Int. Ed. 2014, 53, 69296932.
100. Qtaishat, M. R.; Banat, F. Desalination by Solar Powered Membrane Distillation Systems. Desalination 2013, 308, 186197.
101. Khayet, M. Solar Desalination by Membrane Distillation: Dispersion in Energy Consumption Analysis and Water Production Costs. (A Review). Desalination 2013, 308,
89101.
102. Gude, V. G. Energy Storage for Desalination Processes Powered by Renewable Energy and Waste Heat Sources. Appl. Energy 2015, 137, 877898.
103. Bhadra, B.; Roy, S.; Mitra, S. Desalination Across a Graphene Oxide Membrane Via Direct Contact Membrane Distillation. Desalination 2016, 378, 3743.
104. Qtaishat, M.; Matsuura, T.; Kruczek, B.; Khayet, M. Heat and Mass Transfer Analysis in Direct Contact Membrane Distillation. Desalination 2008, 219, 272292.
105. Varron, K.; et al. Dispersible Exfoliated Zeolite Nanosheets and Their Application as a Selective Membrane. Science 2011, 334, 7275.
106. Peng, Y.; Li, Y.; Ban, Y.; Jin, H.; Jiao, W.; Liu, X.; Weishen, Y. MetalOrganic Framework Nanosheets as Building Blocks for Molecular Sieving Membranes. Science
2014, 346, 13561359.
107. Rodenas, T.; Luz, I.; Prieto, G.; Seoane, B.; Miro, H.; Corma, A.; Kapteijn, F.; Llabres, I.; Xamena, F. X.; Gason, J. MetalOrganic Framework Nanosheets in Polymer
Composite Materials for Gas Separation. Nat. Mater. 2015, 14, 4855.
108. Bachman, J. E.; Smith, Z. P.; Li, T.; Xu, T.; Long, J. R. Enhanced Ethylene Separation and Plasticization Resistance in Polymer Membranes Incorporating MetalOrganic
Framework Nanocrystals. Nat. Mater. 2016, 15, 845849.
109. Nicolosi, V.; Chhowalla, M.; Kanatzidis, M. G.; Strano, M. S.; Coleman, J. N. Liquid Exfoliation of Layered Materials. Science 2013, 340, 14201438.

You might also like