Journal of

Materials Chemistry A
View Article Online
PAPER View Journal | View Issue

Sol–gel fabrication of a non-laminated graphene

oxide membrane for oil/water separation†
Published on 14 August 2015. Downloaded by Sabanci University on 12/11/2024 9:46:02 AM.

Cite this: J. Mater. Chem. A, 2015, 3,

19517
Tiefan Huang,a Lin Zhang,*a Huanlin Chena and Congjie Gaoab

A non-laminated graphene oxide membrane crosslinked by polyethyleneimine was prepared via a one-step
sol–gel process. In the as-prepared membrane, the GO nanosheets remain disordered as in the sol state to
form a randomly arranged GO self-assembled structure, which results in a much higher flux compared with
the general laminated GO membrane prepared via vacuum filtration or spin coating because of the lower
flow resistance. Furthermore, this random self-assembly of GO nanosheets gives rise to a hierarchical
micro/nanoscale rough structure on the membrane surface. Along with the crosslinking reaction, PEI
was grafted onto the GO nanosheets to make them hydrophilic. Combining the hydrophilic surface
chemistry with the micro/nanoscale hierarchical surface structure, the non-laminated GO membrane
exhibited the desired superhydrophilic and underwater superoleophobic properties. We have tested the
membrane to separate a series of surfactant-free and surfactant-stabilized oil-in-water emulsions. A high
Received 18th June 2015
Accepted 12th August 2015
separation efficiency (>99%) and flux were achieved using only gravity without any extra power, much
larger than commercial filtration membranes with similar permeation properties. Moreover, the
DOI: 10.1039/c5ta04471e
membrane shows an outstanding antifouling performance for oil droplets and can be recycled easily for
www.rsc.org/MaterialsA long-term use.

investigated its application in nanoltration for nanoparticles

Introduction and dyes.18 Geim et al. constructed a micrometer-thick GO
Graphene oxide (GO), which is an oxidative derivative of membrane via the spin-coating method, which showed unim-
graphene, bears many oxygen functional groups on its basal peded evaporation of water but was completely impermeable to
planes and edges, and the chemical structure endows GO with liquids, vapors, and gases (even helium).19 Gao et al. fabricated
excellent hydrophilicity and dispersity in many solvents.1 This an ultrathin graphene membrane using base-reuxing reduced
improved processing property compared to graphene makes GO GO, which showed high retention rates for organic dyes and
a suitable building block to construct various macro-assem- moderate retention rates for salts.20 Very recently, Jin et al.
blies.2–9 In particular, because of its excellent membrane prepared a GO membrane supported on a ceramic hollow bre
formation characteristics, GO has been widely studied in many membrane for the pervaporation separation of dimethyl
elds.10–15 carbonate/water mixtures.21 In these GO membranes, GO
Previous reports in the literature on membrane-related nanosheets tend to be in horizontally aligned stacks and form
applications of GO have focused on its laminated micro-struc- well-ordered laminated microstructures. This face-to-face
ture. GO membranes with laminated microstructures are arrangement of the GO nanosheets was proposed as a result of
prepared by a ltration-assisted assembly or spin-coating van der Waals attractions, p–p interactions and hydrogen
method.16,17 Using wrinkled hydrazine reduced GO, Li et al. bonding between adjacent nanosheets, as observed in the dried
fabricated a wet graphene membrane via vacuum ltration and graphene paper and theoretical simulation.22,23 To strengthen
the bond between adjacent GO nanosheets to improve GO
a
Key Laboratory of Biomass Chemical Engineering of MOE, Department of Chemical
membrane stability in water, crosslinking between laminated
and Biological Engineering, Zhejiang University, Hangzhou 310027, China. E-mail: microstructure was also studied.24–29
linzhang@zju.edu.cn; Fax: +86-571-87952121; Tel: +86-571-87953802 The nanocapillaries formed between adjacent GO nano-
b
College of Ocean, Zhejiang University of Technology, Hangzhou 310014, China sheets in laminated membranes provide a diffusion route for
† Electronic supplementary information (ESI) available: Porosity calculation, molecule permeation. Because the nanocapillaries are vertical
separation apparatus, high magnication top view SEM images of LGM, sliding
to the apparent permeation orientation, the molecular diffusion
angle measurement of 1,2-dichloroethane (DCE) droplets on the NLGM in
water, separation results for other emulsion systems, SEM images of the NLGM
route is unexpectedly long.19 Therefore, although the laminated
aer the oil/water emulsion separation performance test, photograph of an microstructure has many advantages, such as a molecular-
underwater oil droplet on the NLGM and simulated underwater drag sieving function, it inevitably makes the ow resistance signif-
deformation force tests aer the oil/water emulsion separation test and element icantly large, resulting in a low ux that fails to meet
component percentages of GO and the NLGM. See DOI: 10.1039/c5ta04471e

This journal is © The Royal Society of Chemistry 2015 J. Mater. Chem. A, 2015, 3, 19517–19524 | 19517
 View Article Online

Journal of Materials Chemistry A Paper

expectations. Under some circumstances, such as ultraltra- physically adsorbed PEI and residual NaOH. The laminated GO
tion, high ux is needed, which laminated GO membranes membrane (LGM) was prepared by a vacuum ltration method
cannot satisfy. Furthermore, on the laminated membrane as a control, using GO (8 mg) and a PEI aqueous solution
surface, GO nanosheets are horizontally aligned, the membrane (16 mg ml1, 0.125 ml).
surface morphology is rather smooth and there is a lack of
obvious structure characteristics, which may be important for Preparation of the oil-in-water emulsion
possessing superwetting properties.30–36 Surfactant-free oil-in-water emulsions were prepared by mixing
In this work, we report a kind of novel hydrogel-like GO water and oil (hexane, octane, toluene and isopar G) at 99 : 1
membrane with a non-laminated microstructure that is (v/v), and the mixtures were sonicated under a power of 2 kW for
completely different from those of previously reported GO 2 h to obtain homogeneous and white emulsions (hexane/water
Published on 14 August 2015. Downloaded by Sabanci University on 12/11/2024 9:46:02 AM.

membranes. Herein, the GO membrane was prepared via a one- (H/W), octane/water (O/W), toluene/water (T/W) and isopar
step sol–gel process using polyethyleneimine (PEI) as the G/water (I/G) emulsions). The droplet sizes of the emulsions
crosslinker to produce randomly assembled GO nanosheets in were in the range of hundreds of nanometers to several
the membrane rather than laminated stacked, forming a non- micrometers, as determined by optical microscopy observation.
laminated microstructure. Therefore, the non-laminated Surfactant-stabilized oil-in-water emulsions (Tween80-stabi-
membrane with more pore space has a much higher ux. lized toluene/water (T80/T/W), SDS-stabilized octane/water
Furthermore, in the non-laminated membrane, the randomly (S/O/W) and SDS-stabilized isopar G/water (S/I/W)) were
arranged graphene oxide assembly gives rise to a micro/nano- prepared by adding surfactant (SDS or Tween80) via a similar
scale hierarchical structure on the membrane surface to endow process. The surfactant concentration in these emulsions was
the membrane with superhydrophilic and underwater super- about 0.02 mg ml1. These emulsions are stable at least for 12 h
oleophobic properties. The non-laminated GO membrane can without obvious demulsication occurred. The droplet sizes are
separate a series of surfactant-free and surfactant-stabilized oil- in the range of several hundreds of nanometer to 2 mm.
in-water emulsions driven solely by gravity without any extra
power with a high separation efficiency (water purity in the Emulsion separation experiment
ltrate aer one-time separation >99 wt%) and a relatively high
ux. Most importantly, due to the ultralow affinity of its surface Oil-in-water emulsion separation experiments were performed
to oil drops, the non-laminated membrane exhibits excellent driven by gravity and the separation apparatus is shown in
antifouling properties, which make it practical for long-term Fig. S1.† 150 ml of the prepared emulsion was poured onto the
recycling. NLGM and spontaneously permeated quickly. The ux was
determined by calculating the time to produce 10 ml of
permeate using the following equation: ux ¼ 10 ml/At, where A
Experimental is the valid area of the membrane (12.56 cm2) and t is the testing
Chemicals and materials time. For each test, three samples were measured to obtain the
average value.
Graphene oxide (GO), prepared using a modied Hummers'
method from ake graphite, was manufactured by XFNANO Instruments and characterization
Material Tech Co. (Nanjing, China). Branched poly-
ethyleneimine (PEI) with a molecular weight of 10 000 was Fourier transform infrared (FTIR) spectra were recorded on a
purchased from Sigma-Aldrich (St. Louis, USA). Non-woven Bruker Tensor27 spectrometer. The samples were prepared as
fabric was purchased from Paper Group (Fremont, USA). Isopar KBr pellets, and the spectra were calculated from a total of
G was purchased from Exxon Mobil Co. (Irving, USA). All other 32 scans. X-ray diffraction (XRD) was carried out using an X'Pert
chemicals were obtained from Sinopharm Chemical Reagent PRO (PANalytical, Netherlands) X-ray diffractometer with Cu Ka
Co. (Shanghai, China). The deionized water (electrical resistivity radiation (l ¼ 1.5406 Å). Scanning electron microscopy (SEM)
> 17 MU cm) used throughout the experimental process was images were obtained on a eld-emission scanning electron
produced in our laboratory. microscope (Hitachi SU8010, Japan). Contact angle measure-
ments were conducted on a drop shape analysis system DIGI-
DROP-DI (GBX Co, France). X-ray photoelectron spectroscopy
Preparation of the non-laminated GO membrane (NLGM) (XPS) was performed on an Axis-Ultra DLD spectrometer (Kratos
The NLGM was prepared using a facile sol–gel process. In brief, Co, USA) using Al Ka radiation at a base pressure of 3  109
GO powder (16 mg) was dispersed in water (2 ml) and ultra- mbar, and the binding energies were referenced to the C 1s line
sonicated for 1.5 h. Then, a NaOH aqueous solution (4 M) was at 284.5 eV from adventitious carbon. Thermogravimetric
added to the dispersion until the pH ¼ 12.8. A PEI aqueous analysis (TGA) was performed on a TA-Q500 (TA Co. USA) with a
solution (16 mg ml1, 0.25 ml) was then added and vigorously temperature increase of 10  C min1 under a nitrogen envi-
stirred, and the resulting sol was slowly drop-cast on non-woven ronment. Optical microscopy images were taken with a
fabric. The nal NLGM was obtained by a sol–gel process, which CFM-330 microscope (Changfang Optic Inc., China) equipped
was performed in a sealed pot at 80  C for 8 h. Before the with a D60 (Nikon, Japan) by dropping the solution on a bio-
characterization or separation processes, the membranes were logical counting board. Total organic carbon (TOC) was deter-
rinsed by ltering copious amounts of water to remove mined using a TOC meter (Liqui TOC, Elementar

19518 | J. Mater. Chem. A, 2015, 3, 19517–19524 This journal is © The Royal Society of Chemistry 2015
 View Article Online

Paper Journal of Materials Chemistry A

Analysensysteme GmbH). The tests were repeated three times

and the results given were the average values.

Results and discussion

Preparation and characterization of the NLGM
As shown in Fig. 1a, the NLGMs were fabricated by a one-step
sol–gel process, wherein PEI served as the chemical crosslinker
and a non-woven fabric was used as the substrate. In the sol
Published on 14 August 2015. Downloaded by Sabanci University on 12/11/2024 9:46:02 AM.

state, the reaction between PEI and GO is insufficient, so almost

all the GO nanosheets in solution were free. During the heating
treatment, sol–gel inversion occurred: PEI crosslinked free GO
nanosheets in the sol, constructing a 3D network structure of
the membrane, in which GO nanosheets maintained their
Fig. 2 FTIR spectra of the GO(I), NLGM (II) and PEI(III).
random orientation as in the sol.
The successful crosslinking reaction of GO by PEI was
conrmed by FTIR spectroscopy and XPS. Compared with the
FTIR spectrum of GO (curve I, Fig. 2), an increase in the peak component intensities were reduced, indicating that GO was
intensity at 1114 cm1, which corresponds to the stretching partially reduced by PEI, which is consistent with FTIR results.
vibration of the new C–N bonds and the residual group of PEI, In addition, there is a predominant peak component at
was observed in the NLGM (curve II, Fig. 2). The peak intensity approximately 286 eV, corresponding to the C–N component,
of the NLGM at 2930 cm1 and 2845 cm1 (inset in Fig. 2), suggesting that PEI successfully reacted with GO.40 For conve-
which corresponds to the C–H stretching vibrations of PEI nient comparison, the percentages of each carbon component
(curve III, Fig. 2), substantially increased. The peak in the region (excluding C–N) calculated from the peak area integral for the
corresponding to the –COOH stretching vibration at 1726 cm1 GO membrane and NLGM are listed in Table 1. Aer reacting
decreased, which is attributed to the possible electrostatic with PEI, the change in the C–O–C and C–OH peak intensities of
interaction between GO and PEI.37 In addition, a peak appeared the NLGM showed an opposite trend: the peak intensity of the
at 1578 cm1, and this peak corresponds to the unoxidized C–O–C showed a substantial decline, whereas that of the C–OH
aromatic domain of GO. This result demonstrates that PEI peak increased. These observations suggest a nucleophilic
reduced GO to some extent along with the crosslinking substitution reaction, in which an amine of the PEI attacks the
reaction.38 carbon atom of the epoxy group of GO; therefore, the three-
Fig. 3a presents the survey XPS spectra of GO and the NLGM; membered epoxide ring was opened, resulting in C–N and
the element component percentages of GO and the NLGM are C–OH groups.28 In this reaction, we infer that PEI not only
displayed in Table S1.† For the NLGM spectra, due to the crosslinked individual GO nanosheets together, integrating the
reaction with PEI, an obvious peak at 396.4 eV was observed, membrane, but also graed on the GO nanosheets, making it
which can be attributed to the N1s binding energy, whereas only more hydrophilic.
a tiny peak in this position was observed for GO, which was In order to determine the amount of PEI in the NLGM, TGA
probably introduced during the preparation of GO. To further measurements were carried out, as shown in Fig. 4. GO was
discuss the elemental analyses of the composite membranes, thermally unstable and starts to lose mass upon heating even
the C 1s region of the XPS spectra was deconvoluted. As shown below 100  C, with a moderate weight loss that resulted from
in Fig. 3b, the unmodied composite membrane shows curve the removal of adsorbed water. Starting from 140  C, a rapid
ttings with ve peaks representing different component weight loss was observed, attributable to the removal of labile
species: C–C/C]C, C–OH, C–O–C, C]O and O–C]O, when the oxygen-containing functional groups. The PEI showed a rapid
binding energy was 284.5, 285.1, 286.4, 287.3, and 288.5 eV, weight loss between 325  C and 410  C. As to the curve of the
respectively.39 Aer treatment of PEI, some of the peak NLGM, compared with that of GO and PEI, its decomposition
curve showed the combination of both of them. The weight loss
region from 290  C to 410  C could be attributed to the
decomposition of PEI with a weight loss of 19.8%, indicating
that almost all PEI was reacted for GO nanosheet crosslinking in
the sol–gel process. Since the GO was partially reduced by PEI as
well, the removal of thermally labile oxygen functional groups
resulted in a much increased thermal stability for the reduced
GO, which results in a lower weight loss of NLGM than GO.41,42
Due to PEI cross-linking, the mechanical stability of the
membrane was signicantly improved. In order to prove this,
Fig. 1 Schematic illustration of the formation of an NLGM via the we prepared NLGM without PEI crosslinking, via a similar
sol–gel process.

This journal is © The Royal Society of Chemistry 2015 J. Mater. Chem. A, 2015, 3, 19517–19524 | 19519
 View Article Online

Journal of Materials Chemistry A Paper

Published on 14 August 2015. Downloaded by Sabanci University on 12/11/2024 9:46:02 AM.

Fig. 4 TGA of GO, PEI and NLGM.

intensive peak was observed at 2q ¼ 9.62 . This indicates an

ordered arrangement of GO nanosheets in the membrane,
similar to previous related reports.28,29 However, no obvious
peaks appear in the NLGM, which demonstrates that, as we
anticipated, the GO nanosheets in the NLGM were completely
separated as in the sol state and remained in a rather disor-
dered arrangement aer the crosslinking reaction during the
sol–gel process.43–45
Scanning electron microscopy (SEM) images of the NLGM
and LGM at different magnications are shown in Fig. 7. From
the cross-sectional view, it is obvious that the NLGM exhibits a
homogeneous sponge-like structure with a thickness of 250 mm
(Fig. 7a), which is much thicker than that of the LGM prepared
via the vacuum ltration method (Fig. 7b) because most GO
Fig. 3 (a) Survey XPS spectra and (b and c) C 1s XPS spectra of the GO nanosheets in the NLGM were not stacked horizontally but in
and NLGM, respectively. random arrangements. The structure of the NLGM is charac-
terized by interconnected pores, and these pores may form
Table 1 C 1s component peak areas as the percentages of the total
numerous tortuous and continuous passageways for liquid to
area of the GO and NLGM ow. From the top view, the surface of the NLGM is composed
of numerous GO nanosheet clusters at the micrometer scale
C–C/C]C [%] C–OH [%] C–O–C [%] C]O [%] O–C]O [%] (circled in Fig. 7d for instance). These clusters are connected
with each other but are disorderly arranged, which makes the
GO 38.1 2.08 45.8 8.30 5.71
NLGM 61.1 10.7 12.0 10.3 5.8 membrane surface rather rough. On the surfaces of these
clusters, ner protrusion structures at the nanometer scale
(circled in Fig. 7e for instance) are clearly recognized. These
preparation method, as a control. Fig. 5 shows the NLGMs with
and without PEI crosslinking aer oscillation (60 rpm) for 1 h
and 1 day in water, respectively. During the oscillation process,
the NLGM without PEI crosslinking disintegrated rapidly, which
resulted in the water becoming yellow brown. In contrast, the
integrity of the NLGM with PEI crosslinking was maintained aer
an oscillation period as long as 1 day, and the water still
remained clear and clean, which indicated that the NLGM
stability signicantly improved due to PEI crosslinking. This
ensured the potential of the membrane for practical applications.

Effect of the sol–gel process on the membrane microstructure

Fig. 5 Photographic images of the NLGMs with (right) and without
The XRD spectra of the NLGM are given in Fig. 6, and the LGM (left) PEI crosslinking after oscillation (60 rpm) for 1 h and 1 day in
is also displayed. For the LGM prepared by vacuum ltration, an water, respectively.

19520 | J. Mater. Chem. A, 2015, 3, 19517–19524 This journal is © The Royal Society of Chemistry 2015
 View Article Online

Paper Journal of Materials Chemistry A

Published on 14 August 2015. Downloaded by Sabanci University on 12/11/2024 9:46:02 AM.

Fig. 6 XRD patterns of the NLGM and LGM.

clusters and ner protrusion structures originated from the

twisted GO nanosheets, which is caused by the high reaction
temperature and high pH value conditions during the sol–gel
process.46,47 The crosslinking effect of PEI also can make the GO
nanosheets corrugated, because PEI molecules would be unre-
stricted in the reaction at either termini with different sites of
Fig. 7 Cross-sectional view of the NLGM (a) and the LGM (b); top view
the graphene sheet, potentially leading to intrasheet bonding,
SEM images of the NLGM (c–e) and the LGM (f).
induced curvature and folding.48 In contrast, this micro/nano-
scale hierarchical structure is not observed on the surface of the
LGM prepared by the ltration method; its surface is fairly trapped water molecules greatly decrease the contact area
smooth with only slight waves, as shown in Fig. 7f. (Fig. S2† for between the oil droplet and the solid surface. Hence, these
higher magnication). The porosity of the NLGM can be results clearly demonstrated that the micro/nanoscale hierar-
calculated from the material density49 and membrane thickness chical structure is essential for membranes to have super-
perspectives, and the two calculated porosities, nearly the same, wetting properties. The oil-droplet sliding angle of the NLGM in
are 97.4% and 98.4%, respectively (see ESI†), which are similar water was found to be about 2.2 , as shown in Fig. S3.†
to other reports of GO-based hydrogels.50–53 To further characterize the dynamic wetting behavior of oil
on the NLGM surface, simulated underwater drag deformation
force tests of oil-droplets were performed, as demonstrated in
Effect of the membrane surface microstructure on the
Fig. 8c. A droplet was squeezed to sufficiently contact the
underwater superoleophobic property
membrane surface with an obvious deformation, and it was
We examined the wetting behavior of oil and water on the top then lied up. There is no deformation observed during the
surface of the NLGM, as shown in Fig. 8. A water contact angle
(CA) of approximately 0 in air and an underwater oil
(1,2-dichloroethane is taken as an example) contact angle (OCA)
of approximately 160 are observed, suggesting super-
hydrophilic and underwater superoleophobic properties. As a
control, the OCA of the LGM was 143 , less than the common
standard superwetting property (Fig. 8b).54 Because the two
types of membranes have the same surface chemistry, this
wetting difference can be attributed to the surface microstruc-
ture distinction. According to the Cassie–Baxter theory, the
micro/nanoscale hierarchical structure is important for mate-
rials possessing superwetting properties.55 Combined with the
hydrophilic PEI-graing GO nanosheets, the NLGM exhibited
superhydrophilic and underwater superoleophobic properties.
In water, the NLGM contained water in its balanced state. When
an oil droplet comes into contact with the membrane surface,
Fig. 8 (a) Photographs of a water droplet and an underwater oil
water can be trapped in the rough micro/nanoscale structures droplet on the NLGM, respectively. (b) Photograph of an underwater
and forms an oil/water/solid composite interface, showing oil droplet on the LGM. (c) Photographs of the simulated underwater
excellent underwater superoleophobic properties. These drag deformation force tests of an oil-droplet on the NLGM.

This journal is © The Royal Society of Chemistry 2015 J. Mater. Chem. A, 2015, 3, 19517–19524 | 19521
 View Article Online

Journal of Materials Chemistry A Paper

leaving process, thus indicating extremely weak oil affinity to

the membrane surface.

Separation performance of the NLGM for oil/water emulsions

Based on the superhydrophilic and underwater super-
oleophobic properties, we have applied the NLGM for oil/water
emulsion separation. Usually, in practical industry separation
processes, water has a higher density than oils; hence, the
superhydrophilic and underwater superoleophobic properties
Published on 14 August 2015. Downloaded by Sabanci University on 12/11/2024 9:46:02 AM.

of the NLGM would form the water-layer on the membrane

surface and avoid direct contact between oil-droplets and
membrane surfaces. It could greatly alleviate the membrane
fouling phenomenon and maintain a high separation efficiency
with time. To test the oil/water separation performance of the
NLGM, both surfactant-free oil-in-water emulsions (T/W, H/W,
O/W and I/G) and surfactant-stabilized oil-in-water emulsions
(T80/T/W, S/O/W and S/I/W) were prepared. The as-prepared
emulsions were then poured onto the NLGMs to perform
ltration separation. The whole process is driven solely by
gravity. The collected ltrate (right) is totally transparent
compared to the original milky white feed emulsion (le), as
shown in Fig. 9a and b, which takes the toluene/water emulsion
and Tween80-stabilized toluene/water as an example. Under an Fig. 9 (a and b) Separation results of the toluene/water emulsions and
optical microscope, it can be seen that in the feed solution, Tween80-stabilized toluene/water as examples; scale bar ¼ 10 mm. (c)
numerous oil droplets with sizes from hundreds of nanometers Separation performance for a series of oil/water emulsions.
to several micrometers ood the entire view, whereas no oil
droplets were observed in the collected ltrate, indicating the
for oil droplets, we infer that it has excellent antifouling
effectiveness of the NLGMs in separating various toluene/water
properties. Here, we took the SDS-stabilized isopar G/water
emulsions. Similar effective separation is also achieved for
emulsion as an example to test the NLGM antifouling properties
other oil/water systems (Fig. S4†). The precise oil content in the
because isopar G is an oen-used solvent in industry and is
ltrate was measured by using a total organic carbon (TOC)
composed of various types of isoparaffin. The separation of the
analyzer. As shown in Fig. 9c, for various oil/water emulsions,
SDS-stabilized isopar G/water emulsion is similar to real
including hexane, octane and isopar G, the oil contents aer
conditions in practical industry applications. Fig. 9 depicts the
separation were almost less than 10 ppm, which suggested a
recyclability using the SDS-stabilized isopar G-in-water emul-
high separation efficiency (>99.5%) of the membrane, which
sion. In every cycle, the membrane was rinsed with water aer a
can be attributed to its superhydrophilic and underwater super
certain amount of ltrate was collected. As shown in Fig. 10,
oil-repelling properties. Unlike the three other aliphatic
over 10 cycles, the membrane still maintained stable
hydrocarbons that are nearly insoluble in water, the oil content
in the ltrate aer separation of the toluene/water emulsion is
relatively high due to the fact that toluene has a slight solubility
in water,56 which resulted in relatively low separation efficiency
(>91.5%). The permeability of the membrane was calculated by
measuring the time needed to collect a certain volume of the
ltrate during the process of separating various oil/water
emulsions. The uxes for T/W, H/W, O/W, I/G, T80/T/W, S/O/W
and S/I/W emulsions were 634, 688, 672, 644, 592, 603,
597 l m2 h1, respectively. Compared to the permeation
performances of common commercial membranes (regen-
erated cellulose membrane C100F, Microdyne-Nadir GmbH)57
(the ux for emulsion with an oil concentration of 0.5–1.5% is
less than 200 l m2 h1 at 0.5–1.5 MPa under pH 5–7), the ux
results suggest the remarkable permeability of the membrane,
considering that no external pushing force was used in the
separation process.
In the treatment of oily wastewater, antifouling properties Fig. 10 Cycling performance of the membrane using the SDS-stabi-
are very important. Because the NLGM shows ultralow affinity lized isopar G/water emulsion as the model system.

19522 | J. Mater. Chem. A, 2015, 3, 19517–19524 This journal is © The Royal Society of Chemistry 2015
 View Article Online

Paper Journal of Materials Chemistry A

performance. The ux was nearly unchanged with increasing 2 H. Bai, C. Li, X. L. Wang and G. Q. Shi, J. Phys. Chem. C, 2011,
cycle time, indicating that the membrane has excellent anti- 115, 5545–5551.
fouling properties. This result may be attributed to the ultralow 3 X. J. Zhu, Y. W. Zhu, S. Murali, M. D. Stollers and R. S. Ruoff,
adhesion of oil to the membrane surface, which was demon- ACS Nano, 2011, 5, 3333–3338.
strated by the SEM images of the membrane surface aer 4 Y. T. Chen, F. Guo, A. Jachak, S. P. Kim, D. Datta, J. Y. Liu,
emulsion separation, as shown in Fig. S5.† We can see that the I. Kulaots, C. Vaslet, H. D. Jang, J. X. Huang, A. Kane,
membrane surface remained almost unchanged and the V. B. Shenoy and R. H. Hurt, Nano Lett., 2012, 12, 1996–2002.
integrity of the membrane structure was maintained aer 5 X. C. Dong, H. Xu, X. W. Wang, Y. X. Huang, M. B. Chan-Park,
repeated separation processes. Hence as we expected, the H. Zhang, L. H. Wang, W. Huang and P. Chen, ACS Nano,
wetting properties of the membrane were maintained as well. 2012, 6, 3206–3213.
Published on 14 August 2015. Downloaded by Sabanci University on 12/11/2024 9:46:02 AM.

For example, the underwater oil droplet angle of the membrane 6 P. K. Santra and P. V. Kamat, J. Am. Chem. Soc., 2012, 134,
is 157 aer the SDS/isorpar G/water emulsion and the 2508–2511.
membrane surface still has low affinity to oil droplets (Fig. S6†). 7 J. N. Tiwari, K. Nath, S. Kumar, R. N. Tiwari, K. C. Kemp,
These results suggest the capability of the membrane for long- N. H. Le, D. H. Youn, J. S. Lee and K. S. Kim, Nat.
term use in the treatment of oil/water mixtures. Commun., 2013, 4, 7.
8 M. Z. Seyedin, J. M. Razal, P. C. Innis, R. Jalili and
Conclusions G. G. Wallace, Adv. Funct. Mater., 2015, 25, 94–104.
9 W. Li, J. S. Wang, J. S. Ren and X. G. Qu, Adv. Mater., 2013, 25,
In summary, a novel hydrogel-like non-laminated GO 6737–6743.
membrane crosslinked by polyethyleneimine (PEI) was 10 D. A. Dikin, S. Stankovich, E. J. Zimney, R. D. Piner,
prepared via a one-step sol–gel process. The merit of this G. H. B. Dommett, G. Evmenenko, S. T. Nguyen and
method lies in its easy applicability, which is especially favor- R. S. Ruoff, Nature, 2007, 448, 457–460.
able for scaling-up in industry. Using various characterization 11 X. W. Yang, L. Qiu, C. Cheng, Y. Z. Wu, Z. F. Ma and D. Li,
techniques, it was proved that GO successfully reacted with PEI Angew. Chem., Int. Ed., 2011, 50, 7325–7328.
and that GO nanosheets in the membrane were randomly 12 H. B. Huang, Z. G. Song, N. Wei, L. Shi, Y. Y. Mao, Y. L. Ying,
arranged, forming a sponge-like non-laminated microstructure. L. W. Sun, Z. P. Xu and X. S. Peng, Nat. Commun., 2013, 4.
This special structure endows the membrane with a lower ow 13 R. K. Joshi, P. Carbone, F. C. Wang, V. G. Kravets, Y. Su,
resistance compared with usual laminated GO membranes; I. V. Grigorieva, H. A. Wu, A. K. Geim and R. R. Nair,
therefore, a much higher ux was achieved. Furthermore, Science, 2014, 343, 752–754.
combining the hydrophilic surface chemistry with the 14 B. X. Mi, Science, 2014, 343, 740–742.
micro/nanoscale hierarchical surface structure, the membrane 15 H. Y. Liu, H. T. Wang and X. W. Zhang, Adv. Mater., 2015, 27,
shows superhydrophilic and underwater superoleophobic prop- 249–254.
erties, which can effectively separate a wide range of surfactant- 16 H. W. Kim, H. W. Yoon, S. M. Yoon, B. M. Yoo, B. K. Ahn,
free and surfactant-stabilized undissolved oil-in-water emulsions Y. H. Cho, H. J. Shin, H. Yang, U. Paik, S. Kwon, J. Y. Choi
with droplet sizes from the micrometer to the nanometer range. and H. B. Park, Science, 2013, 342, 91–95.
Using only gravity, a high separation efficiency (>99%) and ux 17 H. Li, Z. N. Song, X. J. Zhang, Y. Huang, S. G. Li, Y. T. Mao,
were achieved. More signicantly, due to the low affinity of this H. J. Ploehn, Y. Bao and M. Yu, Science, 2013, 342, 95–98.
membrane to oil droplets in water, it shows excellent antifouling 18 L. Qiu, X. H. Zhang, W. R. Yang, Y. F. Wang, G. P. Simon and
properties and is easy to recycle for long-term use. This work is D. Li, Chem. Commun., 2011, 47, 5810–5812.
the rst study on non-laminated GO membranes, which may 19 R. R. Nair, H. A. Wu, P. N. Jayaram, I. V. Grigorieva and
inspire future studies on designing novel GO membranes for A. K. Geim, Science, 2012, 335, 442–444.
separation elds as well as many other applications. 20 Y. Han, Z. Xu and C. Gao, Adv. Funct. Mater., 2013, 23, 3693–
3700.
Acknowledgements 21 K. Huang, G. P. Liu, Y. Y. Lou, Z. Y. Dong, J. Shen and
W. Q. Jin, Angew. Chem., Int. Ed., 2014, 53, 6929–6932.
The authors acknowledge nancial support for this work from 22 N. V. Medhekar, A. Ramasubramaniam, R. S. Ruoff and
the Zhejiang Provincial National Science Foundation of China V. B. Shenoy, ACS Nano, 2010, 4, 2300–2306.
(No. LR12B06001); the National Basic Research Program of 23 C. N. Yeh, K. Raidongia, J. J. Shao, Q. H. Yang and
China (No. 2015CB655303); the National Natural Science J. X. Huang, Nat. Chem., 2015, 7, 166–170.
Foundation of China (No. 51578485, 51238006); and the 24 S. Park, K. S. Lee, G. Bozoklu, W. Cai, S. T. Nguyen and
Research Fund for the Doctoral Program of Higher Education of R. S. Ruoff, ACS Nano, 2008, 2, 572–578.
China (No. 20130101110064). 25 S. Park, D. A. Dikin, S. T. Nguyen and R. S. Ruoff, J. Phys.
Chem. C, 2009, 113, 15801–15804.
Notes and references 26 Z. An, O. C. Compton, K. W. Putz, L. C. Brinson and
S. T. Nguyen, Adv. Mater., 2011, 23, 3842–3846.
1 D. R. Dreyer, S. Park, C. W. Bielawski and R. S. Ruoff, Chem. 27 M. Hu and B. X. Mi, Environ. Sci. Technol., 2013, 47, 3715–
Soc. Rev., 2010, 39, 228–240. 3723.

This journal is © The Royal Society of Chemistry 2015 J. Mater. Chem. A, 2015, 3, 19517–19524 | 19523
 View Article Online

Journal of Materials Chemistry A Paper

28 W. S. Hung, C. H. Tsou, M. De Guzman, Q. F. An, Y. L. Liu, 43 Z. Tang, S. Shen, J. Zhuang and X. Wang, Angew. Chem., Int.
Y. M. Zhang, C. C. Hu, K. R. Lee and J. Y. Lai, Chem. Mater., Ed., 2010, 49, 4603–4607.
2014, 26, 2983–2990. 44 Y. X. Xu, K. X. Sheng, C. Li and G. Q. Shi, ACS Nano, 2010, 4,
29 T. F. Huang, L. Zhang, H. L. Chen and C. J. Gao, J. Mater. 4324–4330.
Chem. B, 2015, 3, 1605–1611. 45 X. Zhang, Z. Sui, B. Xu, S. Yue, Y. Luo, W. Zhan and B. Liu, J.
30 Z. X. Xue, S. T. Wang, L. Lin, L. Chen, M. J. Liu, L. Feng and Mater. Chem., 2011, 21, 6494–6497.
L. Jiang, Adv. Mater., 2011, 23, 4270–4273. 46 R. L. D. Whitby, A. Korobeinyk, V. M. Gun'ko, R. Busquets,
31 M. L. Huang, Y. Si, X. M. Tang, Z. G. Zhu, B. Ding, L. F. Liu, A. B. Cundy, K. Laszlo, J. Skubiszewska-Zieba, R. Leboda,
G. Zheng, W. J. Luo and J. Y. Yu, J. Mater. Chem. A, 2013, 1, E. Tombacz, I. Y. Toth, K. Kovacs and S. V. Mikhalovsky,
14071–14074. Chem. Commun., 2011, 47, 9645–9647.
Published on 14 August 2015. Downloaded by Sabanci University on 12/11/2024 9:46:02 AM.

32 F. Zhang, W. B. Zhang, Z. Shi, D. Wang, J. Jin and L. Jiang, 47 S. Ansari and E. P. Giannelis, J. Polym. Sci., Part B: Polym.
Adv. Mater., 2013, 25, 4192–4198. Phys., 2009, 47, 888–897.
33 W. B. Zhang, Z. Shi, F. Zhang, X. Liu, J. Jin and L. Jiang, Adv. 48 R. L. D. Whitby, A. Korobeinyk, S. V. Mikhalovsky, T. Fukuda
Mater., 2013, 25, 2071–2076. and T. Maekawa, J. Nanopart. Res., 2011, 13, 4829–4837.
34 X. Gao, L. P. Xu, Z. Xue, L. Feng, J. Peng, Y. Wen, S. Wang and 49 S. Xie, X. Liu, B. Zhang, H. Ma, C. Ling, M. Yu, L. Li and J. Li,
X. Zhang, Adv. Mater., 2014, 26, 1771–1775. J. Mater. Chem. A, 2015, 3, 2552–2558.
35 M. M. Tao, L. X. Xue, F. Liu and L. Jiang, Adv. Mater., 2014, 50 H. Hu, Z. Zhao, W. Wan, Y. Gogotsi and J. Qiu, Adv. Mater.,
26, 2943–2948. 2013, 25, 2219–2223.
36 W. B. Zhang, Y. Z. Zhu, X. Liu, D. Wang, J. Y. Li, L. Jiang and 51 N. Zhang, H. Qiu, Y. Si, W. Wang and J. Gao, Carbon, 2011,
J. Jin, Angew. Chem., Int. Ed., 2014, 53, 856–860. 49, 827–837.
37 S. Park, J. H. An, R. D. Piner, I. Jung, D. X. Yang, 52 M. Chen, C. Zhang, X. Li, L. Zhang, Y. Ma, L. Zhang, X. Xu,
A. Velamakanni, S. T. Nguyen and R. S. Ruoff, Chem. F. Xia, W. Wang and J. Gao, J. Mater. Chem. A, 2013, 1,
Mater., 2008, 20, 6592–6594. 2869–2877.
38 T. Szabo, O. Berkesi, P. Forgo, K. Josepovits, Y. Sanakis, 53 N. Li, Q. Zhang, S. Gao, Q. Song, R. Huang, L. Wang, L. Liu,
D. Petridis and I. Dekany, Chem. Mater., 2006, 18, 2740–2749. J. Dai, M. Tang and G. Cheng, Sci. Rep., 2013, 3, 1604.
39 D. Yang, A. Velamakanni, G. Bozoklu, S. Park, M. Stoller, 54 A. Tuteja, W. Choi, M. L. Ma, J. M. Mabry, S. A. Mazzella,
R. D. Piner, S. Stankovich, I. Jung, D. A. Field, G. C. Rutledge, G. H. McKinley and R. E. Cohen, Science,
C. A. Ventrice and R. S. Ruoff, Carbon, 2009, 47, 145–152. 2007, 318, 1618–1622.
40 H. L. Ma, H. B. Zhang, Q. H. Hu, W. J. Li, Z. G. Jiang, Z. Z. Yu 55 A. B. D. Cassie, Discuss. Faraday Soc., 1948, 3, 11–16.
and A. Dasari, ACS Appl. Mater. Interfaces, 2012, 4, 1948– 56 H. Stephen and T. Stephen, Solubilities of Inorganic and
1953. Organic Compounds, Binary systems, Part 1, Pergamon
41 S. Stankovich, D. A. Dikin, R. D. Piner, K. A. Kohlhaas, Press, Oxford, 1979, vol. 1, p. 67.
A. Kleinhammes, Y. Jia, Y. Wu, S. T. Nguyen and 57 M. Hesampour, A. Krzyzaniak and M. Nystrom, J. Membr.
R. S. Ruoff, Carbon, 2007, 45, 1558–1565. Sci., 2008, 325, 199–208.
42 M. X. Chen, C. C. Zhang, X. C. Li, L. Zhang, Y. L. Ma,
L. Zhang, X. Y. Xu, F. L. Xia, W. Wang and J. P. Gao, J.
Mater. Chem. A, 2013, 1, 2869–2877.

19524 | J. Mater. Chem. A, 2015, 3, 19517–19524 This journal is © The Royal Society of Chemistry 2015