Recent Development of Novel Membranes For Desalination 2018 Desalination

Desalination 434 (2018) 37–59
Contents lists available at ScienceDirect
Desalination
journal homepage: www.elsevier.com/locate/desal
Recent development of novel membranes for desalination T

a a,b a,⁎
Zhe Yang , Xiao-Hua Ma , Chuyang Y. Tang
a
Department of Civil Engineering, the University of Hong Kong, Pokfulam, Hong Kong
b
School of Chemical Engineering, East China University of Science and Technology, Mei Long Road 130, Shanghai 200237, China
G RA P H I C A L AB S T R A C T
A R T I C L E I N F O A B S T R A C T
Keywords: In the past decades, novel materials (e.g., aquaporin proteins, carbon nanotubes, nanoporous graphene and
Aquaporin membranes graphene oxide) have emerged as promising candidates for synthesizing high performance desalination mem-
Carbon-based membranes branes. These materials can potentially achieve water fluxes of several orders of magnitude higher compared to
Membrane fabrication the state-of-the-art thin-film composite polyamide reverse osmosis membranes. This paper provides a compre-
Desalination
hensive summary of the current progresses and challenges in synthesizing aquaporin-based and carbon-based
Water treatment
membranes. After a detailed review of the material properties of aquaporin proteins, carbon nanotubes, nano-
porous graphene and graphene oxide, a general framework of membrane design and material incorporation is
established. The fabrication methods and separation performance for each type of membrane are summarized.
Future perspectives of aquaporin-based and carbon-based membranes are discussed in lieu with their ultimate
separation performance and commercial scalability.
1. Introduction suggests that ultra-permeable membranes (UPMs) with tripling water

permeability could save up to 15% energy consumption and use 44%
With the increasingly growing world population, water crisis has fewer pressure vessels for seawater desalination. In the context of
become one of the grand challenges in the 21st century [1]. To date, wastewater reclamation, even greater savings (e.g., 45% less energy
reverse osmosis (RO) is the key technology for desalting water to al- input and 63% fewer pressure vessels [4]) can be achieved. Further-
leviate the water scarcity [2]. Membrane is one of the most critical more, increasing membrane selectivity results in improved quality of
components in an RO desalination plant, and it largely determines the the product water. It can also potentially eliminate the second pass that
separation performance of the overall plant [3]. A recent analysis [4] is commonly adopted for boron removal, which significantly reduces
⁎
Corresponding author at: HW 6-19b, Haking Wong Building, Pokfulam Road, Department of Civil Engineering, The University of Hong Kong, Hong Kong.
E-mail address: tangc@hku.hk (C.Y. Tang).
https://doi.org/10.1016/j.desal.2017.11.046
Received 4 May 2017; Received in revised form 29 November 2017; Accepted 29 November 2017
Available online 07 December 2017
0011-9164/ © 2017 Elsevier B.V. All rights reserved.
Z. Yang et al. Desalination 434 (2018) 37–59
the cost of desalination [5]. At the same time, membranes with en- dissolved salts by size exclusion. The water permeability of the TFN
hanced antifouling properties reduce the frequency of membrane membrane was nearly doubled compared to the TFC membrane that has
cleaning and prolong membrane lifespan [6]. no zeolite added. At the same time, the NaCl rejection was not com-
Commercial RO membranes are dominated by thin-film composite promised (93.9 ± 0.3% for TFN vs. 93.4 ± 1.1% for TFC). Other
(TFC) polyamide and its derivatives. These membranes face critical porous nanoparticles, such as mesoporous silica [27,28] and metal or-
challenges of relatively low water permeability, low selectivity and ganic framework (MOF) [29–31], also show similar permeability en-
high fouling tendency [6]. For example, typical water permeability of hancement effect. Deng and co-workers [27] reported that, for meso-
commercial TFC RO membranes range from ~ 1–2 L m− 2 h− 1 bar− 1 porous silica particles with pore size up to 3 nm, the resulting TFN
for seawater reverse osmosis (SWRO) membranes and membranes can maintain their rejection while improving their water
~ 2–8 L m− 2 h− 1 bar− 1 for brackish water reverse osmosis (BWRO) permeability.
[7]. Synthesizing novel RO membranes with improved separation Researchers have also investigated the use of non-porous nano-
properties and better antifouling performance is therefore a key re- particles for the synthesis of TFN membranes. The loading of hydro-
search focus in the field of desalination [6]. philic non-porous nanoparticles such as silica [32], silver [33,34], and
Tremendous advancements have been achieved in making ultra- TiO2 [35–37] often show enhancement of water permeability compared
permeable and antifouling membranes. Much of the exciting progresses to that of the control counterparts. In several studies, researchers found
are fueled by the recent emergence of promising novel materials for reduced salt rejection with the introduction of nanoparticles into the
desalination. Among them, the most notable examples include aqua- polyamide layer, which can often be attributed to the agglomeration of
porin (AQP) proteins [8,9] and some carbon-based materials such as nanoparticles [38,39].
carbon nanotubes (CNTs) [10] and graphene-based materials [11,12]. To date, TFN membranes have successfully reached commerciali-
These novel materials provide new dimensions for designing next- zation (e.g., the zeolite-based LG NanoH2O® membranes), partly be-
generation RO membranes. Fig. 1 shows the increasing number of cause the relative easiness to scale up their production. In this kind of
publications in the recent decade on these topics. membrane structure, polyamide is indeed used as a salt-rejecting matrix
This paper provides a comprehensive review of the recent pro- that helps to minimize defects in the rejection layer and thus maintain
gresses of novel desalination membranes prepared using AQPs, CNTs, high salt rejection. The addition of nanoparticles can also be easily
and/or graphene-based materials. Following a brief presentation of adapted into the existing membrane fabrication lines that are originally
conventional TFC polyamide membranes and its derivative thin film designed for TFC polyamide membrane production. However, the im-
nanocomposite (TFN) membranes (Section 2), we provide a detailed provements in membrane separation performance for the TFN mem-
summary of the material properties of AQPs, CNTs, nanoporous gra- branes are incremental, since their performance is largely limited by the
phene and graphene oxide as well as a general framework for classi- polyamide matrix. The use of the polyamide matrix further makes these
fying novel RO membranes on the basis on how and where the mate- TFN membranes vulnerable to chlorine attack [24]. Therefore, better
rials are loaded into the membranes (Section 3). The fabrication membranes and materials for desalination are needed.
methods using each type of materials and the properties of the corre-
sponding membranes are then reviewed in Section 4–6, and the future 3. Novel materials and methods for synthesizing desalination
perspectives and limitations of these novel membranes are also high- membranes
lighted in Section 7.
3.1. Aquaporin
2. Conventional and novel membrane materials
Nature provides a perfect solution for desalination – water mole-
Although early RO membranes were of asymmetric cellulose acetate cules are transported across biological cells through a group of trans-
membranes, they have been largely replaced by TFC polyamide mem- membrane proteins known as aquaporins (AQPs). AQPs, whose water
branes [13]. Compared to the former, TFC polyamide membranes show transport channels resemble the shape of an hour glass, are highly ef-
better water permeability and salt rejection (e.g., > 99.9% NaCl re- ficient in delivering water molecules with high selectivity [40]. Readers
jection for some SWRO membranes), wider operating temperature interested in the detailed structures and properties of AQPs are referred
range (0–45 °C) and better pH tolerance (1 − 11) [7]. A typical TFC to several existing reviews on this topic [41–43]. A brief description of
membrane consists of three layers (Fig. 2a): a dense polyamide rejec- the structure of AQP1 (the earliest water channel proteins discovered
tion layer, a porous substrate commonly made of polysulfone or poly- by Peter Agre in 1991 [40]) is presented in this section. The hour-glass-
ethersulfone, and a non-woven fabric layer as a mechanical support. shaped channel of AQP1 (Fig. 3) has an internal vestibule of approxi-
The polyamide rejection layer, which determines the water perme- mately 20 Å in length, and the narrowest constriction of the vestibule is
ability and salt rejection of the membrane, can be prepared by the in- approximately 2.8 Å. The asparagine groups near the middle of the
terfacial polymerization (IP) of an amine monomer in aqueous phase vestibule give rise to positively charged sites, and the remaining of the
and acyl chloride monomer in organic phase (Fig. 2b). Researchers have vestibule is considerably hydrophobic. This particular structure forces
applied various attempts to optimize the structure and chemistry of TFC water molecules to line up in the vestibule in a one-dimensional “water
polyamide membranes in order to enhance their separation perfor- wire” with little frictional force from the hydrophobic wall. The tiny
mance and/or antifouling properties [7]. Typical approaches include constriction only allows a single water molecule jump through the
changing monomer types and concentrations [14–18], membrane sur- channel one at a time. The positive charged sites inside the vestibule
face modification [19–22] and post-treatment [23–25]. Despite that ensures the proper orientation of the water molecules, as they undergo
many of these attempts lead to marginal improvements in membrane a “transient dipole reorientation” through the charge interaction and
separation properties, the water permeability of the state-of-the-art TFC hydrogen bonding with the asparagine groups [44]. The combined site
polyamide membranes remain relatively low. restriction and charge repulsion lead to nearly perfect rejection of so-
A major improvement to TFC membranes was introduced by Hoek lutes, including protons [40].
and co-workers in 2007 [26]. These researchers incorporated porous For membrane synthesis, Aquaporin Z (AqpZ, an AQP found in
zeolite nanoparticles (NaA, entrance pore size ~4 Å) into the poly- Escherichia coli cells) is more commonly used due to its relative sim-
amide rejection layer during the interfacial polymerization process. In plicity for harvesting and extraction [45]. The key properties of AqpZ
this membrane structure that is now termed as thin film nanocomposite are summarized in Table 1. Based on stopped-flow measurements [42],
(TFN, see Fig. 2a), the porous zeolite nanoparticles provide preferential AqpZ has a water permeability of approximately 2–10 × 10− 14 cm3/s
water pathways with reduced hydraulic resistance while excluding [46,47]. According to Kumar and co-workers [8], polymeric vesicles
38
Fig. 1. Schematic illustrations of water transport properties of (a) aquaporin, (b) carbon nanotubes and (c) graphene-based membranes for desalination and their related number of
publications in recent years. The publication data were obtained from Scopus® on the basis of the keyword search “aquaporins and membranes”, “carbon nanotubes and membranes” and
“graphene oxide and membrane”, respectively. The data of 2017 is incomplete.
containing AqpZ had water permeability as high as 3.2. Carbon-based materials

~ 600 L m− 2 h− 1 bar− 1 with nearly 100% NaCl rejection. This water
permeability is 2–3 orders of magnitude higher than the existing TFC Carbon based materials (CBMs), e.g., carbon nanotubes (CNTs)
polyamide membranes [41]. These fascinating properties make AQPs [60–64], nanoporous graphene (NPG) [54,65] and graphene oxide (GO)
ideal candidates for preparing high permeability membranes (Section [66–73] have emerged as promising membrane materials [74], thanks
4). to their exceptional water transport properties (Table 1). The rejection
In parallel, researchers have attempted to synthesize artificial su- properties of these materials are highly dependent on their character-
pramolecular water channels through self-assembly [48,49]. Despite istic water channel dimensions as well as chemical modifications (e.g.,
the significant progresses in the recent years, the water permeability of the presence of amine, carboxyl and other groups) [53,75–77]. The
the state-of-the-art artificial water channels is still far below that of the characteristic channel dimensions of CNTs and NPG are governed by
natural AQP water channels [48]. their respective pore sizes, with the pores in NPG (obtained by plasma
etching) more irregularly distributed [54]. Unlike CNTs and NPG,
39
Fig. 2. Schematic diagram of membrane structure of (a) thin-film composite (TFC) membrane and thin-film nanocomposite (TFN) membrane, (b) interfacial polymerization of polyamide
using trimesoyl chloride and m-Phenylene diamine.
schematic diagram of a single CNT, a scanning electron microscopic

(SEM) image of aligned CNTs and an illustration of the transport of
water molecules through these aligned pores. CNTs are generally fab-
ricated by metal-catalyzed chemical vapor deposition method [10].
They can be categorized into single-walled carbon nanotubes (SWCNTs)
and multi-walled carbon nanotubes (MWCNTs), with the latter con-
taining a number of concentric cylindrical tubes. Their exceptional
electrical conductivity, mechanical stability and antimicrobial proper-
ties enable CNTs to be widely applied in electronic devices [82,83], and
biomedicine applications [84–87].
Since the report of vertically-aligned (VA) CNTs by Hinds et al. in
2004 [10], CNTs have drawn great attentions for their potential ap-
plications for desalination and water purification [78]. Holt and co-
workers [53] demonstrated that the experimental water flux obtained
from CNT pores exceeded the theoretical value (calculated using con-
tinuum hydrodynamics assuming non-slippery boundary condition) by
more than three orders of magnitudes. Indeed, the hydrophobic and
atomically smooth interior wall could provide nearly frictionless water
flow inside a CNT channel [10]. Both size exclusion and charge repul-
sion play important roles in water and solute transport in CNTs. Con-
sequently, the pore size and the functionalization of the tips (i.e., the
entrance of CNT channel) have profound impacts on its water perme-
Fig. 3. Schematic structure and mechanism for water transport and proton blocking by an
AQP1 protein channel (cross-sectional view). Three characteristics of the AQP1 channel
ability and salt rejection [10,53]. Reducing pore size or increasing the
determine the selectivity for water: (a) Size restriction. The narrowest site in the AQP1 number of charge groups at the tips tend to greatly increase salt re-
channel is close to 2.8 Å (approximately the diameter of a single water molecule). (b) jection at the expense of reduced water permeability. Furthermore,
Electrostatic repulsion. The amino acid, Arginine (Arg-195), at the narrowest site of the CNTs also have well-known antimicrobial and conductive properties
pore forms a barrier to protons. (c) Water dipole reorientation. Two helices meet at the [57,88]. The use of CNT to fabricate antimicrobial and/or conductive
center of the channel with positively charged dipoles that could reorient the water mo-
nanocomposite membrane for membrane fouling control has also been
lecule. The figure is reprinted from Ref. [44] with copyright permission.
reported [89,90]. Methods for the synthesis of CNT-based desalination
membranes will be further discussed in Section 5.
whose channel sizes are determined by their synthesis conditions, the
characteristic channel size of GO is highly dependent on its degree of
oxidation and solution environment [55]. In this section, detailed ma- 3.2.2. Nanoporous graphene (NPG)
terials properties of CNT, NPG and GO are summarized [74,78–80]. In 2010, the Nobel Prize in Physics was awarded to Geim and
Novoselov for the discovery of graphene [11,91–105]. Graphene con-
sists of a single layer of carbon atoms arranged into a honeycomb lat-
3.2.1. Carbon nanotubes tice; the thickness of this 2D atomic carbon layer is around 0.3 nm
CNTs, first reported by Lijima in 1991 [81], have a hollow cylind- [11,91]. Graphene can be synthesized via mechanical exfoliation,
rical structure composed of a sheet of carbon atoms. Fig. 4a shows a chemical exfoliation, epitaxial growth and chemical vapor deposition
40
[106]. Due to its unique structural, thermal and electrical properties,
channel size depending on the degree of
Water permeability is at least 1010 times

Size-exclusion (and enhanced by charge
Channels formed by adjacent GO layers,
permeability and rejection are sensitive

oxidation or solution environment [55]
from weight-loss measurements by a 1-

μm-thick GO membrane) [12]; water
Antiadhesion (due to hydrophilicity)

graphene has found a wide range of applications [107].
faster than that of helium (obtained

An intact graphene layer is impermeable to water. In order to make
the graphene water permeable, researchers attempt to generate nano-
sized pores on graphene monolayer via laser techniques or chemical
to the interlayer spacing
and antimicrobial [59]

etching methods to produce nanoporous graphene (NPG) (see Fig. 4b
Graphene Oxide (GO)
2-D carbon material
and [54,74]). These nano-sized pores enable ultra-fast convective water

flow across the single-atom-thick NPG layer [95], in direct contrast to
the relatively slow solution-diffusion process in the much thicker re-
repulsion)
jection layers (on the order of 100 nm) of conventional TFC polyamide
membranes. MD simulations [65,108,109] suggest that the water per-
No
meability of NPG can be 2–3 orders of magnitude higher than that of
with non-uniform pore sizes (e.g.,
100% KCl rejection at 40 °C for a
typical commercial TFC membranes (see Table 1). The size of nano-
Size-exclusion (and enhanced by
5-μm-diameter sample (obtained

Nano-sized pores across 1-atom-
~ 3.6 × 106 L m− 2 h− 1; nearly

obtained from plasma etching,
thick graphene layer, possibly
from gravity-driven test in an
pores can have great impact on their separation performance based on
Not reported in literature

the size exclusion mechanism, where a larger pore size results in higher
water permeability and lower solute rejection [109]. The nanopores can
2-D carbon material
also be functionalized with hydrogen and hydroxyl groups to further

charge repulsion)
~ 5–10 Å [54])
enhance the selectivity of NPG by charge repulsion (e.g., a nearly

oven) [54]
complete salt rejection was observed for hydroxylated pores with

0.45 nm in diameter [109]). Methods for fabricating NPG-based desa-
NPG
Yes
lination membranes are presented in Section 6.1.

CNTs pore density ≤ 2.5 × 1011 cm− 2 and length of
Antimicrobial [57] (and improved hydrophilicity for

Well-defined cylindrical pores (e.g., ~13–20 Å [53])
flux of an aqueous suspension of gold nanoparticles;

hydrodynamics (obtained from measuring the water
3.2.3. GO-based framework (GOF)

Size-exclusion (and enhanced by charge repulsion)
higher than the calculated results from continuum

Experimental Water permeability is > 1000 times
In addition to generating nanopores in graphene, many recent stu-

Gas permeability is > 10 times higher than the
dies explore the use of derivatives of graphene, particularly graphene

predictions of the Knudsen diffusion model;
oxide (GO) and reduced graphene oxide (rGO), for the preparation
desalination membranes [110]. GO can be prepared by oxidizing the
inexpensive raw material graphite, via Brodie [111], Staudenmaier
[112], Hummers' [113] or modified Hummers' method [114]. The re-
functionalized CNTs [58])
sulting 2D GO nanosheets are composed of elements of C, O and H, with

a great number of hydrophilic functional groups (e.g., epoxy, hydroxyl
1-D carbon material
and carboxyl groups) to make it more water attractive.

Recently, GO has raised great attentions among membrane scientists
~ 3 μm) [53]
to synthesize high performance desalination membranes with ultra-fast

water transport [12,69,110]. GO nanosheets can be stacked into gra-
CNT
phene oxide framework (GOF) for molecular separation, where water/

Yes
solvent molecules transport through the nano-channels formed by ad-

jacent GO interlayers while solutes are retained (see Fig. 4c). The
from stopped-flow measurements
~ 600 L m− 2 h− 1 bar− 1; nearly
channel [52], pore size of ~ 3 Å
of AQP-containing vesicles) [8]

Well-defined hour-glass-shaped
100% NaCl rejection (obtained
specific water transport mechanism of the GOF is depicted as follows:

water molecules are first absorbed at the hydrophilic edges in the GO
Size-exclusion and charge
Not reported in literature
nanosheets, followed by their rapid diffusion along the water channel

between the hydrophobic carbon sheets in the inter-layer [73]. Both the
interlayer spacing and GOF thickness play important roles on the se-
Natural protein
paration performance. An MD simulation study of GOF [115] suggested

repulsion
that the water permeability can be adjusted from ~ 208 to

~16,640 L m− 2 h− 1 bar− 1 by varying the GO membrane interlayer
AqpZ
[40]
No
Material properties of polyamide, AqpZ, CNT, NPG and graphene oxide.
spacing and GOF thickness. However, at an interlayer spacing of 15 Å

(corresponding to the highest water permeability of
pore distribution for some membranes [51]
spectroscopy [50], possibly heterogeneous
characteristic pore diameter of ~ 4–5.8a Å
16,640 L m− 2 h− 1 bar− 1), the NaCl rejection is only ~ 50%. In addi-

~ > 99% NaCl rejection (obtained from
~ 1–2 L m− 2 h− 1 bar− 1 for SWRO and
based on positron annihilation lifetime
Irregular pores in a random network,
tion, GO-based materials have well-known antimicrobial properties,

~ 2–8 L m− 2 h− 1 bar− 1for BWRO;
followed in the order of: graphite oxide < graphite < reduced
GO < GO, thus providing possibilities to apply GO-based membranes
cross-flow filtration tests) [7]
for biofouling control [59]. The hydrophilic nature of GO also allow it

to be used as a hydrophilic modification agent and to impart anti-ad-
Prone to fouling [56]
Cross-linked polymer
hesion properties to membranes. The detailed membrane fabrication

Solution-diffusion
methods of GO-based desalination membranes can be found in Section

6.2.
Polyamide
3.3. A general framework for novel desalination membranes

No
Novel desalination membranes prepared by the aforementioned

conductance
channel size
mechanism
novel materials (AQPs, CNTs, NPG and GO) can be classified into four
properties
properties
Characteristic
Antifouling
general categories (Table 2) on the basis of their location (in the re-
Separation
Transport
Electrical
Material
(Å)
jection layer vs. substrate) and function (materials for forming the
Table 1
primary rejection layer, additives for flux enhancement, or modifiers

for enhancement antifouling performance):
41
Fig. 4. Carbon-based materials for desalination applications. (a) From left to right: a schematic diagram of a carbon nanotube, a scanning electron microscopy micrograph of aligned
CNTs synthesized by an Fe-catalyzed CVD method with the scale bar of 50 μm and a schematic illustration of transport mechanism in a vertically-aligned CNT membranes. (b) From left to
right: a schematic diagram of a nanoporous graphene, a scanning transmission electron microscopy micrograph of a nanoporous graphene (scale bar 2 nm) after 1.5 s exposure to oxygen
plasma with pore size of ~ 1 nm and a schematic illustration of transport mechanism in a nanoporous graphene. (c) From left and right: a schematic diagram of graphene oxide, a scanning
electron microscopy micrograph of graphene oxide-based framework with the scale bar of 1 μm and the schematic illustration of the transport mechanism of the graphene oxide-based
framework. All figures are reprinted with copyright permissions (Fig. 4a from Refs [10] and [78]; Fig. 4b from Refs [54] and [65]; Fig. 4c from Refs [116], [12] and [117]).
1. Membranes using the novel desalting materials to form the primary re- building blocks. The high water permeability nature of AQPs, CNTs,
jection layer (PRL). AQPs, CNTs, NPG, or GO can be used as the NPG, and GOF allow the preparation of ultra-permeable mem-
primary building blocks to form the rejection layer. Notable ex- branes. However, the major challenges of the PRL membranes are
amples include rejection layers prepared from aligned CNTs their scalability, stability and materials cost for industrial applica-
[10,53,77], NPG [54], or GO-based framework (GOF) [12]. Rejec- tions. This design is also more prone to the formation of defects in
tion layers prepared by AQP-containing bilayers without the use of the rejection layer, which could adversely affect solute rejection of
salt-rejecting matrix can also be classified into this category [118]. the resulting membrane.
The separation performance of the resulting membrane is ultimately
governed by the intrinsic properties (Table 1) of the primary 2. Thin-film nanocomposite (TFN) membranes. Materials such as AQPs,
Table 2
Principle, membrane performance and materials of novel desalination membranes.
42
CNTs, and GO can also be incorporated into a desalting matrix (e.g., 4.1. AQP-based PRL membranes
polyamide thin-film layer) to form the membrane rejection layer. In
principle, this design is similar to the zeolite-based TFN membranes The structure of AQP-based PRL membranes resembles that of bio-
introduced in Section 2: the polyamide matrix is used to maintain logical cellular membranes, where the AQPs containing lipid bilayer
the integrity and salt rejection while AQPs, CNTs, and GO can be regulates the transport of water and the retentions of solutes. This de-
used as performance enhancers. For example, AQPs-containing lipid sign allows the potential synthesis of membranes with ultra-fast water
vesicles can be loaded into the polyamide layer during interfacial permeability with relatively high salt rejection.
polymerization [119,120]. These highly permeable vesicles, acting Kaufman and co-workers first developed the vesicles rupture ap-
as faster water transport pathways to shortcut through the poly- proach to form lipid bilayers on a commercial nanofiltration membrane
amide matrix, can enhance water permeability while maintaining [131]. In their approach, liposomes (i.e., lipid vesicles without AQPs)
high salt rejection. Likewise, CNTs or GO can also be loaded in salt- contact the surface of the substrate membrane, and the interaction
rejecting layers for flux enhancement [61,121,122]. In addition, between the liposomes and the substrate causes the vesicles to rupture
these materials are also commonly used for preparing antifouling into planar lipid bilayers. Likewise, Li and co-workers [132] spin-
membranes [58,123]. A key advantage of the TFN approach is its coated a lipid bilayers on a commercial NF membrane. A common issue
easy scale up, e.g., using similar production lines developed for TFC in this approach is that the coating will decrease the overall perme-
membranes. However, it is important to point out that the separa- ability of the coated membrane. This issue is even worsened due to lack
tion properties of TFN membranes are largely limited by the mate- of AQPs in the liposomes. A second major issue is the difficulty to
rial forming the salt-rejecting matrix. Consequently, only partial achieve complete surface coverage, and the uncovered regions will
enhancement in permeability (instead of the orders of magnitude result in severe salt leakages through the membrane.
improvement shown in Table 1) can be achieved. Significant progresses have been achieved in the recent years.
Zhong and co-workers reported several schemes to immobilize the AQP-
3. Surface located nanocomposite (SLN) membranes. Functionalized containing lipid bilayers onto porous substrates [118,133,134]. Fig. 6
CNTs or GO can be attached onto membrane surface via covalent or shows a representative AqpZ-based PRL membrane [118]. The rejection
electrostatic bonding [71,90,124]. This design can be considered as layer is formed by an ABA block copolymer layer that has AqpZ in-
a special case of TFNs, except that the nanomaterials are located at corporated, and this rejection layer is supported on a cellulose acetate
the membrane surface, often for the purpose of enhancing a mem- (CTA) substrate functionalized with methacrylate end groups. The op-
brane's anti(bio)fouling performance. The direct functionalization of timized membrane performance was obtained at 34 L m− 2 h− 1 bar− 1
the membrane surface endows the composite membrane with sig- water permeability with NaCl rejection around 30% at an AqpZ:ABA
nificantly enhanced hydrophilicity and/or biocidal effects [71,124]. ratio of 1:50. Although the membranes reported in this work show
Membrane water permeability based on the SLN approach generally promising water permeability, this method is difficult to scale up (ef-
decreases or maintains at similar level. However, the direct exposure fective membrane area was only 0.071 cm2). Furthermore, this work
of the loaded materials to the feed solution leads to concerns of their reveals the potential issue of defects in the AQP-based PRL membranes,
detachment or leaching from the membrane surface [39]. which can severely limit their salt rejection. Several enhancements have
been proposed in the recent literature, such as further cross-linking by
4. Mixed matrix substrate (MMS) enhanced membranes. Instead of in- disulfide [133] or polydopamine (PDA) modification [134]. Up to date,
corporating nanomaterials in the rejection layer, it is also possible to the highest reported NaCl rejection is 98.9% in the context of FO for a
load them into the porous substrate that acts as a support to the membrane area of 0.096 cm2 [135]. Since the rejection in FO is in-
rejection layer. For example, nanomaterials can be added into the directly measured (calculated on the basis of solute flux and water flux),
polymer dope solution for casting the substrate [58]. The improved this rejection performance is yet to be verified with direct RO testing
pore structure of the substrate can enhance the water permeability using larger membrane areas. Other studies [136] report NaCl rejection
of the overall membrane in an indirect manner [125–129]. In ad- up to 75% (achieved by loading an AQP-containing rejection layer onto
dition, the substrate of a MMS membrane generally exhibits en- a nanofiltration substrate), where the nanofiltration substrate partially
hanced hydrophilicity, porosity and/or antifouling properties, helped to maintain the salt rejection. Minimizing defects and improving
which can be applied in forward osmosis (FO) or pressure retarded rejection are still major challenges for the preparation PRL membranes.
osmosis (PRO) to enhance membrane's water flux and anti(bio) Other major challenges of AQP-based PRL membranes are their
fouling properties [130]. Like SLN membranes, MMS membranes stability and difficulty to scale up for industrial applications. For ex-
also face the risk of leaching of the embedded nanomaterials. amples, the lipids bilayers and AQPs are directly exposed to outside
environment such that they may degrade chemically and biologically
[41].
4. AQP-based desalination membranes
4.2. AQP-based TFN membranes
Kumar and co-workers [8] first tossed the concept of fabricating
AQP-based biomimetic membranes and demonstrated their potentials Compared to the PRL membranes that is difficult to scale up, AQP-
using model AQP-containing vesicles. This pioneering work was quickly based TFN membranes can be fabricated in larger membrane area and
followed by a wave of additional studies by many researchers in the with enhanced membrane stability [26]. The dense salt-rejecting matrix
past decade (see the summary in Table 3). An earlier review [41] (e.g., polyamide) not only ensures adequate salt rejection but also acts
classified these membranes into supported membrane layers (SMLs) as a protection shield to enhance the membrane's chemical and biolo-
and vesicles encapsulated membranes (VEMs). In order to keep a more gical tolerance [41,42,137].
general and consistent classification scheme that is applicable to all The first work on AQP-based TFN membranes was reported by Zhao
types of materials (Table 2), we renamed these biomimetic membranes et al. [119], who incorporated AQPs containing vesicles into a poly-
as: (1) AQP-based PRL membrane (where AQPs assembled in a lipid or amide rejection layer (Fig. 7). These proteolipsomes act as “preferential
polymeric bilayer on a porous substrate to form a rejection layer [118], water pathways” that can enhance the overall water permeability. The
see Fig. 5a) and (2) AQP-based TFN membranes (where AQP containing as-prepared AQP-based TFN membranes had a water permeability of
vesicles are incorporated into membrane thin-film rejection layer ~4 L m− 2 h− 1 bar− 1 and a NaCl rejection of ~ 97% tested under a
[119], see Fig. 5b). pressure of 5 bar. This membrane permeability was ~30% higher
compared to the control membranes, i.e., membranes without or with
43
Table 3
Summary of RO, NF and FO performance of biomimetic membranes.
Type Classification Pw (L m− 2 h− 1 bar− 1) Rejection (%) Testing conditions Results Year and
and membrane area Ref.
(cm2)
PRL
AqpZ-DOPCa NF 3.6 RNaCl = 20% 1 mM NaCl @1 bar DOTAP coated NF270, with both decreased water flux 2012
Area: 28.3 and RNaCl compared to virgin membranes [132]
b
AqpZ-ABA NF 34.2 RNaCl = 32.9% 200 ppm NaCl @5 bar Silanized CA substrate, high Pw with low RNaCl, the 2012
Area: 0.071 amount of AqpZ has huge impact on membrane [118]
performance
AqpZ-ABA NF 16.1 RNaCl = 45.1% 200 ppm NaCl @5 bar Gold coated porous alumina substrate cross-linked 2012
Area: 0.2. with disulfide: high Pw with less defects [133]
AqpZ-DOPC/ NF 5.5 RNaCl = 75% 500 ppm NaCl @4 bar AQP containing lipid bilayers deposited on PSS/PEI/ 2015
DOTAPc RMgCl2 = 97% Area: 19.56 PAN substrate [136]
d −2 −1
AqpZ-ABA FO Jv = 16.4 L m h RNaCl = 98.8% 0.3 M sucrose as DS, Gold and cysteamine coated polycarbonate with UV 2012
200 ppm NaCl as FSe cross-linking [135]
Area: 0.096
AqpZ-DOPC/DOTAP FO/NF Jv = 23.1 L m− 2 h− 1 FO: Js = 3.1 g 2 M MgCl2 as DS, DI AqpZ-DOPC/DOTAP coated on PDA modified porous 2015
NF:6.31 m− 2 h− 1 water as FS polysulfone substrate via amidation reaction to form [134]
NF: RMgCl2 = 90% 2000 ppm MgCl2 @ covalent bonds.
4 bar
Area: 36
TFN
AqpZ-DOPC RO 4 RNaCl = 97% @ 10 mM NaCl @5 bar AqpZ containing vesicles incorporated in PA layer 2012
5 bar Area: > 200 serving as protection layer via IP. Large membrane [119]
area can be obtained.
AqpZ-DOPC RO 8 RNaCl = 97.5% 500 ppm NaCl @5 bar Vesicles embedded in PA rejection layer with superior 2015
Area: 34.2 water flux [120]
AqpZ-DOPC RO 4.1 RNaCl = 97.2% 10 Mm NaCl @10 bar Vesicles embedded in PA rejection layer for long term 2016
Area: 42 stability test [137]
AqpZ-POPC/POPG/ NF ~6 RMgCl2 = 96% 200 ppm MgCl2 @ Vesicles embedded in PSS/PAA LBLf. Membranes with 2012
cholesterolg 4 bar AqpZ showed Pw ↑ 60% with MgCl2 rejection ↑ [140]
Area: 0.785 compared to the control
AqpZ-DOPC NF 36.6 RMgCl2 = 95% 100 ppm MgCl2 @ PDA coated vesicles incorporated in cross-linked PEI 2013
1 bar matrix [138]
Area: 28.3
AqpZ-ABA NF/FO NF: 22.9 RNaCl = 61% 200 ppm salt @5 bar AqpZ-vesicle loaded membrane cross-linked by UV 2013
Jv = 5.6 L m− 2 h− 1 RMgCl2 = 75% 0.3 M sucrose as DS [141]
FO: RNaCl = 50.7% and 200 ppm NaCl as
FS
Area: 0.196
AqpZ-ABA FO Jv = 43.5 L m− 2 h− 1 Js = 8.9 g m− 2 h− 1 0.5 M NaCl as DS, DI Pressure assisted sorption, further coated with 2013
water as FS cysteamine and cross-linked by polydopamine- [142]
Area: 0.196 histidine. The control membrane has FO water flux of
8.6 L m− 2 h− 1 and Js = 6.6 g m− 2 h− 1
AqpZ-POPC/POPG/ FO Jv = 21.8 L m− 2 h− 1 Js = 2.4 g m− 2 h− 1 0.3 M sucrose as DS Magnetic-assisted AQPs embedded membranes 2013
Cholesterol and 200 ppm MgCl2 [143]
as FS
Area: 0.785
a
DOPC: 1,2-dioleoyl-sn-glycero-3-phosphocholine.
b
ABA: methacrylate end functionalized poly(2-methyloxazolineb-dimethylsiloxane-b-2-methyloxazoline) PMOXA(1000)-b-PDMS(4000)-PMOXA(1000) triblock.
c
DOTAP: 1,2-dioleoyl-3-trimethylammonium-propane.
d
Jv: FO water flux; Js: FO solute flux.
e
DS: draw solution; FS: feed solution.
f
LBL: layer by layer deposition of polyacrylic acid (PAA) and polystyrene sulfonate (PSS).
g
POPC: 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine; POPG: 1-palmitoyl-2-oleoyl-sn-glycero-3-phospho-(19-rac-glycerol).
inactiviated AQPs. Similar method has been used to prepare high per- period using real RO feed water obtained from a local wastewater plant.
meability nanofiltration membranes [138,139]. Both flatsheet [119] The salt rejection of both AQP-based TFN and commercial membranes
and hollow fiber [120] membranes have been reported. Nevertheless, did not show major change during the long term test.
the ultimate water permeability in the TFN design is limited by the
dense salt-rejecting matrix, in contrast to the PRL design that can take 5. CNT-based membranes
full advantage of the highly permeable nature of AQPs.
An important consideration for AQP-based membranes is the sta- CNT-based membranes can be classified into: (1) aligned CNT-based
bility of these water channel proteins over prolonged duration. In this PRL membranes, (2) CNT-based TFN, (3) CNT-based SLN membranes
sense, the dense polymeric matrix of the AQP-based TFN membranes and (4) CNT-based MMS membranes. The recently published CNT re-
can provide protection to AQPs against chemical and biological de- lated papers are listed in Table 4.
gradation [41]. For example, Qi et al. [137] reported that the AQP-
based TFN RO membranes were able to maintain their performance 5.1. Aligned CNT-based PRL membranes
after soaking in commonly used chemical cleaning agents (e.g., sodium
hydroxide, ethylenediamine tetraacetic acid, and citric acid, respec- Aligned CNT membranes have drawn growing attentions in the past
tively). These researchers conducted long term test over a three-month decade for their potential application in water purification [10,122].
44
Fig. 5. Schematic diagrams of AQP-based desalination

membranes: (a) AQP-based PRL membranes where
AQPs are assembled in a lipid or polymeric bilayer on
a porous substrate (gray) to form a rejection layer; (b)
AQP-based TFN membranes where AQP containing
vesicles (proteoliposomes or proteo-polymersomes)
are encapsulated in a dense polymer layer (green) on a
porous substrate (gray). (For interpretation of the re-
ferences to colour in this figure legend, the reader is
referred to the web version of this article.)
Hinds et al. [10] pioneered the fabrication of vertically-aligned (VA) substrate using vacuum filtration (referred by the authors as “semi-
CNTs, whose water permeability was 2–3 orders of magnitude higher aligned” CNTs) to fabricate high performance RO membranes (Fig. 9).
compared to results predicted by the conventional fluid flow theory. The resultant CNT-incorporated membranes showed quadrupled water
This result was attributed to the “frictionless” interior wall of CNTs. The permeability (1.3 L m− 2 h− 1 bar− 1 for TFN vs. 0.3 L m− 2 h− 1 bar− 1
selectivity of CNTs can be fine-tuned by introducing different functional for TFC) with near identical NaCl rejection (98.6% for TFN vs. 97.6%
groups on the tips of the CNTs [80]. In many existing studies on aligned for TFC). The authors attributed the enhanced membrane performance
CNT-based PRL membranes, their performance were predicted from to the ultra-fast transport properties of CNTs' that were uniformly de-
molecular dynamics simulations [76,77,144,145]. Future studies need posited in thin-film rejection layer.
to experimentally demonstrate NaCl rejection in addition to water Recently, Xue and co-workers [61] functionalized MWCNTs with
permeability in order to verify their feasibility for desalination. The three different functional groups, namely carboxyl (MWCNT-COOH),
issues related to scaling up shall also be adequately addressed. hydroxyl (MWCNT-OH) and amino groups (MWCNT-NH) (see Fig. 10),
In a recent work, Baek and co-workers [89] reported ultrafiltration followed by the incorporation the functionalized CNT into the piper-
membranes with vertically aligned CNTs (VA-CNTs) in an epoxy azine (PIP) aqueous solution to fabricate TFN membranes. The impacts
(Fig. 8). The resulting membranes, with CNT pore diameter of 4.8 nm of the functional groups on membrane's separation performance were
and pore density of 6 × 1012 pores/cm2, showed a 3 times higher water systematically investigated. At the optimized MWCNT concentration of
permeability compared to a commercial UF membrane with similar 0.01% (w/v), all the MWCNTs incorporated membranes exhibited im-
rejection of polyethylene oxide (PEO, molecular weight ~ 100 kDa). proved pure water permeability and salt rejection. Among the three
Furthermore, the VA-CNT membranes exhibited strong biocidal effect types of MWCNTs, MWCNT-OH TFN membrane shows the highest
with almost 2 log scale bacteria reduction. water flux of 41.4 L m− 2 h− 1 and a Na2SO4 rejection of 97.6% at a
given pressure of 6 bar. The authors attributed this performance to the
synergistic effect of –OH groups in MWCNTs and –NH2 groups in PIP. In
5.2. CNT-based TFN membranes addition, MWCNT-NH2 incorporated membranes showed enhanced salt
rejection and stability than those of MWCNT-COOH incorporated
Due to the difficulty to prepare membranes with aligned CNTs while membranes due to the adhesive strength between –NH2 and –COOH in
maintaining adequate salt rejection, many researchers adopted the the PA matrix.
CNT-based TFN design where CNTs are introduced into rejection layers Several studies have also reported the use of CNT-based TFN
(e.g., during the interfacial polymerization to form polyamide) membranes for improved antifouling performance [74]. CNTs have
[61,122,146,147]. Because of the inherently hydrophobic and non-re- strong antimicrobial effects, which prompts researchers to use them for
active nature of CNTs (which often results in incompatibility with the preparation of antibiofouling membranes (particularly by loading CNTs
polymer matrix), various chemical or physical modifications have been onto membrane surfaces, see the discussion on SLN membranes in 5.3.).
developed to improve the dispersion of CNTs in the matrix [148]. At the same time, the loading of CNTs in the TFN membranes allows
Among them, acid treatment is regarded as the most efficient method researchers to prepare electrically conductive rejection layers and to
that forms hydroxyl (− OH) and carboxyl groups (−COOH) at the ends control fouling electrochemically [150–152]. For example, Lannoy and
of CNTs and thus makes them more hydrophilic and reactive [149]. co-workers [153] employed CNT-incorporated conductive TFN mem-
Functionalized CNTs then can be incorporated into the thin-film poly- branes. A CNTs layer was first deposited onto the polyethersulfone ul-
amide rejection layer [61,147], which can have profound impact on trafiltration membrane, followed by interfacial polymerization process
membrane physicochemical properties (e.g., hydrophilicity, porosity, to form the polyamide thin-film rejection layer (Fig. 11a). The CNT
charge density, and additional water channels) [39]. interlayer provided excellent conductivity (~ 400 S/m) for the
Chan and co-workers [122] loaded zwitterionic CNTs onto a
Fig. 6. Schematic diagrams of AQP embedded membranes

(a): ABA layer on a silanized CA substrate. (b) AqpZ in-
corporated ABA CA membrane. All figures are reproduced
from Ref. [118] with copyright permission.
45
Fig. 7. AQPs TFN membranes. (a) Schematic diagram of the synthesis routines of the AQP-based TFN membranes, (b) conceptual model of water transport mechanism of the TFN
membranes. All figures are reproduced from Ref. [119] with copyright permission.
nanocomposite membrane. In a membrane biofouling test, the appli- membrane is covalently bonded with functionalized CNTs in the pre-
cation of a 1.5 V, 16 mHz voltage to the membrane significantly slowed sence of EDC/NHS system. The separation performance of the modified
down the flux reduction (Fig. 11b). According to these researchers, the TFC membranes did not change significantly, implying that this method
alternating electrical potential caused the instabilities in the local pH, does not affect the integrity of the polyamide. In the antimicrobial test,
which creates a non-ideal environmental for bacteria to produce extra- the modified TFC membranes showed a 44% inactivation of cultural
polymeric substances. bacteria on the membrane surface compared to the control TFC mem-
brane.
5.3. CNT-based SLN membranes

5.4. CNT-based MMS membranes
Immobilization of functionalized CNTs onto membrane surface can
be an effective method to mitigate membrane (bio)fouling due to the Functionalized CNTs can also be incorporated into the substrate
enhanced hydrophilicity and antimicrobial properties of the functio- matrix to enhance its water permeability [154], rejection [155], hy-
nalized CNTs [57]. Tiraferri and co-workers [90] presented a novel drophilicity [156], conductivity [152], mechanical stability [157] and
strategy to immobilize carboxylated CNTs onto polyamide rejection anti(bio)fouling properties [58]. An interfacial polymerization reaction
layer by strong covalent bonds (Fig. 12). Prior to surface modification, can subsequently perform to prepare the CNT-based MMS membrane.
the CNTs were first functionalized with carboxylic groups. Then, a so- Wang and co-workers [157] reported a slightly enhanced water per-
lution of N-hydroxysuccinimide/1-ethyl-3-[3-(dimethylamino)propyl] meability (2.7 ± 0.2 L m− 2 h− 1 bar− 1) of the CNT-based MMS
carbodiimide (EDC/NHS) was used to transform the native carboxylic membrane compared to that of the control membrane
groups of the polyamide thin-film layer into semistable amine-reactive (2.4 ± 0.3 L m− 2 h− 1 bar− 1) upon the initial increase in MWCNTs
esters, followed by reaction with ethylenediamine (ED). The resultant loading (up to 0.5 wt%), and further increase in loading (up to 2.5 wt%)
Table 4
Recent developments of CNTs-based desalination membranes.
Type Polymer Classification Observations and performance Year and

Ref.
PRL Polystyrene Gas separation Great potential in separation and sensing; Functionalization of CNT tips would gate molecular transport 2004 [10]
through CNT pores
PRL – Microfiltration High removal efficiency for heavy metals and hydrocarbons; Filtered poliovirus of sizes around 25 to 2004 [158]
30 nm with a molecular mass of 8.5 × 106 Da
PRL SiNx Gas separation Nitrogen flow permeability of 4.7 × 10− 4 mol m− 2 Pa− 1 at a pore density of 4.0 × 1010 cm− 2 with 2004 [159]
estimated pore size of 66 nm.
PRL Si3N4 Nano-scale filtration Ultra-fast water flow with pores smaller than 2 nm to enhance selectivity 2006 [53]
PRL – Gravity-driven Superhydrophobic and superoleophilic VA CNT filters for rapid water/diesel separation 2010 [160]
PRL Epoxy Solvent filtration CNTs pores packing density of 2.4 × 1010 cm− 2; water flux of 2 cm × 2 cm CNT membrane was 2011 [161]
6.75 × 10− 2 mL cm− 2 min− 1
PRL Urethane/ethanol Ultrafiltration CNTs pores density 3.0 × 1012 with average pore size around 4.1 nm. The VA CNT membrane showed 2016 [162]
938 times higher Pw than the UF membrane.
TFN Polysulfone FO FO water flux ↑ 60% 2013 [163]
TFN Polysulfone/polyamide NF Conductive NF membrane with antibiofouling properties 2013 [153]
TFN Polyamide NF Water flux ↑ 62% with Na2SO4 rejection over 99% 2013 [164]
TFN Polyamide NF -COOH,-OH and –NH MWCNTs embedded into PA layer with all water flux ↑ 2016 [61]
SLN Polyamide RO Slightly reduced NaCl rejection; Enhanced inactivation of culturable bacteria by 44% 2011 [90]
SLN Polyethyleneimine NF Vacuum assisted deposition of carboxylated MWCNTs; Water flux ↑ 44% with constant salt rejection 2013 [165]
MMS Polyethersulfone FO Substrate hydrophilicity ↑ and FO water flux ↑ 2013 [157]
MMS Polysulfone Ultrafiltration Membrane pore size and hydrophilicity increased upon the addition of MWCNTs; the PSF membrane 2006 [154]
with 4.0 wt% MNCNTs showed higher flux and rejection
MMS Polysulfone Ultrafiltration Surface roughness and hydrophilicity of the modified membranes improved; however, the tensile 2008 [166]
strength didn't improve
MMS Poly (vinylidene fluoride) Microfiltration antimicrobial properties ↑ 2008 [167]
MMS Polyethersulfone Ultrafiltration hydrophilicity ↑ and antifouling properties against protein ↑ 2013 [156]
MMS Polysulfone Ultrafiltration membrane separation performance ↑, hydrophilicity, and antifouling properties ↑ 2013 [58]
MMS Polysulfone Ultrafiltration Thermal stability ↑ and rejection of heavy metal ↑ 2013 [155]
46
Fig. 8. Schematic illustration of the fabrication of

VA CNT-based PRL UF membrane (a) transfer the
VA CNTs to a tape (b) pouring epoxy into the
vacant areas of VA CNTs using the cast. All fig-
ures are reproduced from Ref. [89] with copy-
right permission.
Fig. 9. Cross-sectional schematics of the fabrication techniques for the semialigned CNT polyamide membranes. (A) PES ultrafiltration membrane was soaked in a surfactant solution to
increase the pore sizes and its hydrophilicity, then fixed by two PTFE holders, (B) zwitterion functionalized CNTs were vacuum-filtered onto the pretreated PES substrate, and (C)
interfacial polymerization technique was performed between the semi-VA CNTs at which MPD aqueous solution came in contact with organic TMC solution with (D) a photograph of the
final membrane product. All figures are reproduced from Ref. [122] with copyright permission.
led to a reduction in both water permeability and NaCl rejection. Due to flux (Fig. 13b). The flux enhancement was attributed to the mitigated
its ability for improving the hydrophilicity of the substrate by loading internal concentration polarization (ICP) as a result of more porous and
functionalized CNTs, this method has also been commonly used for hydrophilic carboxyl functionalized MWCNT containing substrate.
preparing FO and PRO membranes [157]. Wang and co-workers [157] The incorporation of functionalized CNTs could enhance substrate's
embedded carboxyl functionalized MWCNTs into polyethersulfone antifouling properties. Daraei and co-workers [156] functionalized
(PES) dope solution to generate the MWCNTs/PES substrate, followed MWCNTs with citric acid, acrylic acid and acrylamide to achieve large
by interfacial polymerization to form the CNT-based MMS membrane. number of functional groups on MWCNTs. The ultrafiltration mem-
The CNT embedded substrate showed a significantly improved tensile branes loaded with the functionalized MWCNTs showed a smooth and
strength and hydrophilicity with the increasing concentration of hydrophilic membrane surface with the enhanced fouling resistance
MWCNTs (Fig. 13a), which is accompanied with an improved FO water against whey proteins. Since the CNTs-based mixed matrix
47
Fig. 10. Schematic preparation procedures of different nanocomposite NF membranes via interfacial polymerization by incorporating MWCNTs with different functional groups. All
figures are reproduced from Ref. [61] with copyright permission.
ultrafiltration membrane shows enhanced membrane's anti(bio)fouling Cohen-Tanugi and co-workers [109] employed MD simulations to
properties [156], similar effect is expected for CNTs-based MMS analyze the water permeability and NaCl rejection of NPG (see Fig. 14).
membranes. Future study may further explore the combined hydro- The separation performance of NPG membrane is largely dependent on
philicity enhancement and antimicrobial effects in the context of FO its pose size and pore chemistry [109]. The authors further evaluated
and PRO application. the mechanical stability of the NPG membrane, which exhibited a ro-
bust stability to withstand a high pressure of approximately 570 bar
6. Graphene-based membranes [168].
Even though extensive computational results have showed that NPG
6.1. Nanoporous graphene membranes membranes have the potential to achieve extremely high water per-
meability and high NaCl retentions, only few experimental studies have
A pristine defect-free graphene monolayer is non-porous and hy- been reported so far. For example, Surwade and co-workers [54] ex-
drophobic and is impermeable to helium gas [92]. Through the creation perimentally investigated the transport properties of water and ions
of nano-sized pores (e.g., with plasma etching [54]), the resulting na- across a monolayer NPG based PRL membrane, where the nanopores
noporous graphene has outstanding molecular sieving properties [65]. were generated by plasma etching. The NPG membrane exhibited a
Water can easily transport through these sub-nanometer pores while water flux of 3.6 × 106 L m− 2 h− 1 at 40 °C with near 100% rejection
other larger molecular species are rejected on the basis of size exclusion of KCl using gravity and vapor pressure as driving force [109]. Despite
(and possibly charge interaction). Since it is extremely challenging to the outstanding separation performance, the effective membrane dia-
experimentally characterize NPG membranes, the performance of NPG meter of the fabricated NPG is only 5 μm.
membranes are often analyzed using MD simulations.
Fig. 11. (a) TEM cross-sectional structure of a TFN membrane loaded with CNTs and (b) Normalized water flux reduction with time in a befouling test: with or without voltage on TFN
membrane and control membrane (with no CNT loading). Red circle represents membrane cleaning. All figures are reproduced from Ref. [153] with copyright permission. (For
interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
48
Fig. 12. Schematic preparation procedures

of CNTs immobilized PA TFC membranes.
The figure is reproduced from Ref. [90] with
copyright permission.
Fig. 13. (a) tensile strength value of MWCNTs incorporated PES substrates with different MWCNTs loadings. The corresponding curves are labeled with the weight concentration of
MWCNTs in the PES substrates; (b) FO water flux and salt rejection of control commercial HTI membrane, control FO membrane and MWCNTs incorporated FO membranes. All figures
are reproduced from Ref. [157] with copyright permission.
6.2. Graphene oxide-based membranes membranes have been prepared by using vacuum filtration methods
[66,67]. Wang and co-workers [67] vacuum filtered a GO dispersion to
Similar to CNT-based desalination membranes, GO-based mem- form a GOF-based PRL membrane on a PAN nanofibrous substrate (see
brane can also be classified into: (1) GO-based PRL membranes, (2) GO Fig.16). The resulting GOF membranes showed high rejection for dyes
incorporated TFN membranes, (3) GO-based SLN membranes and (4) and moderate rejection for divalent ions (~ 100% for Congo red and
GO-based MMS membranes. The recently published papers related to ~56.7% for Na2SO4) with water permeability of approximately
GO-based membranes are shown in Table 5. 2 L m− 2 h− 1 bar− 1. The relatively low salt rejection typically found in
GO-based PRL membranes may be attributed to the relatively wide
channel dimension, as the typical inter-layer spacing (d) of is around
6.2.1. GO-based PRL membranes 13.5 Å for GO nanoplates when swelled in water [170,171].
The most common methods to fabricate the GOF-based PRL mem- In a more recent study, Abraham and co-workers [55] fabricated GO
branes include layer-by-layer (LBL) assembly and vacuum filtration PRL membranes with superior molecular permeation properties via
[67,69,169]. The first study of GOF-based PRL membrane was reported vacuum filtration. They successfully controlled the interlayer spacing
by Nair and co-workers in 2012 [12], who synthesized a layer-by-layer between the GO laminates from ~ 9.8 Å to 6.4 Å, providing opportu-
GOF PRL membrane through LBL spray- or spin- coating method (see nities to sieve ions at the molecular level. In an FO testing using 3 M
Fig. 15). The resultant GO PRL membrane exhibited an unimpeded sugar as DS and 0.1 M NaCl as FS, the NaCl rejection of the resultant GO
water vapor permeation with almost completely impermeable of he- membrane was around 97% (although the FO water flux was only
lium, in which water permeates through the GO membrane at least 1010 around 0.5 L m− 2 h− 1, in comparison to the typical water flux of
times faster than helium. These authors further attributed the extremely 5 L m− 2 h− 1 for a commercial FO membrane). Future studies may
high water permeability and high selectivity of the GO membrane to the explore whether the water flux can be significantly improved while still
low-friction water flow in the interlayer channels. maintaining adequate salt rejection by reducing membrane thickness.
Inspired by Nair's pioneering work, various GO-based PRL In a recent publication in Nature, Chen et al. [172] fabricated a
49
Fig. 14. Monolayer graphene with hydrogenated (a)

and hydroxylated (b) pore, and (c) schematic illus-
tration of water transport through NPG in a compu-
tational system. All figures are reproduced from Ref.
[109] with copyright permission.
freestanding GO membrane by drop-casting a 5 mg mL− 1 GO suspen- coated Al2O3 substrate and then post-treated the fabricated GO mem-
sion on a homemade ceramic membrane (average pore size of 200 nm). brane in an oven. It was found that GO membrane without thermal
Through hydrated cation-π interactions between the hydrated cations treatment was not stable and can be easily damaged by a gentle finger
(e.g., K+, Na+, Ca2 +, Li+, or Mg2 + ions) and the GOF, they could touch or rinsing with water. However, after the thermal treatment, the
successfully controlled the interlayer spacing of the GOF down to only GO nanosheets can be strongly bonded onto the PDA coated substrate
1 Å to achieve ionic and molecular sieving. Such fine control of inter- surface, leading to a significantly enhanced stability. In desalting of
layer spacing, once verified by more studies, has the potential to enable 3.5 wt% NaCl seawater feed solution, a high water flux of
a wide range of applications in addition to water treatment and gas 48.4 L m− 2 h− 1 with NaCl rejection of 99.7% can be obtained at 90 °C
separation. through pervaporation (PV).
In addition to vacuum filtration, some works improved the stability
of GO PRL membranes using cross-linkers [69], electrostatic interlayer
force [169], or thermal based post-treatment [173]. Hu and Mi [69] 6.2.2. GO-based TFN membranes
demonstrated the use of TMC as a cross-linker to covalently bond GO GO can be incorporated into a polyamide rejection layer through
nanosheets on a polydopamine (PDA) coated PSF substrate. The re- interfacial polymerization to form GO-based TFN membranes. The in-
sultant GOF membrane had water permeability ranging from corporated GO nanoplates can improve membrane hydrophilicity and
~ 8–27.6 L m− 2 h− 1 bar− 1, together with relatively low rejection of the interlayer channels between these nanoplates may provide water
monovalent ions and divalent ions (6–46%) and high rejection of dyes channels to further enhance the membrane water permeability. Several
(93–95%). Interestingly, membrane's water permeability and salt re- studies have also reported improved chemical resistance of GO loaded
jection had no obvious variations with the increasing GO layers. These TFN membranes (e.g., improved chlorine resistance attributed to the
authors attributed this unique phenomenon to the frictionless water hydrogen bonding between GO and polyamide that impedes the first
transport between the GO interlayers. Apart from fabricating GOF by step of chlorination of the replacement of amidic hydrogen with
covalent bonds, Hu and Mi [169] presented another work on the basis chlorine [174]; enhanced resistance to strong oxidants and chemical
of relying on electrostatic interaction to synthesis GO membrane (see agents due to the formation of substantial amounts of polyester bonds
Fig. 17). The water permeability of this GO membrane was around one [175]).
order of magnitude higher than a commercial FO membrane. However, GO can be either dispersed in aqueous MPD solution [176] or or-
the electrostatic interaction between the GO layers would be weakened ganic TMC solution [121] to fabricate the GO-based TFN membranes.
after immersing in high ionic strength solution, resulting in a deterio- Chae and co-workers [176] dispersed the as-prepared GO nanoplates in
rated membrane. a MPD aqueous solution, followed by interfacial polymerization with
Xu and co-workers [173] also successfully fabricated PRL GO TMC. The resultant GO-TFN membrane exhibited an 80% enhancement
membrane by vacuum filtering a GO suspension onto a polydopamine in water permeability and significantly enhanced antibiofouling prop-
erties compared to those of the pristine TFC membrane. These authors
50
Z. Yang et al.
Table 5
Summary of recent developments of GO-based membranes.
Type Method Polymer Classification Membrane performance Year and Ref.
−2 −1 −1
Test conditions Flux (L m h bar ) Rejection (%)
PRL LBL PAA/hPANa NF @ 5 bar ~ 0.84 RNa + = 43.2%; RMg2 + = 92.6% 2012 [188]
PRL Vacuum filtration TiO2/Polycarbonate NF 10 ppm dyes @100 kPa 7 RMethyl_orange = 100% 2013 [66]
PRL LBL TMC/PDA/PSF NF 10 mM Na2SO4; 27.6 RNa2SO4 = 46% 2013 [69]
7.5 mg/L R-WT @3.4 bar RRhodamine-WT = 93–95%
PRL LBL PAHb/hPAN NF/FO NF: @3.45 bar 6 FO: Rsucrose = 99% 2014 [169]
FO: DI as FS; 1 M sucrose as DS
PRL Vacuum filtration PAN nanofiber NF @ 1 bar 2 RNa2SO4 = 56.7% 2016 [67]
RCongo-red = 100%
−2 −1
PRL Vacuum filtration PDA/Al2O3 PV 3.5 wt% NaCl@90 °C 48.4 L m h RNaCl = 99.7% 2016 [173]
Pressure: N/A
PRL Vacuum filtration Epoxy FO 3 M sugar as DS; 0.1 M NaCl as FS Jv = 0.5 L m− 2 h− 1 RNaCl = 97% 2017 [55]
PRL Drop-casting KCl-controlled GOF (280 nm) FO 0.25 M NaCl and 0.25 M KCl as DS; DI water as FS Jv = 0.36 L m− 2 h− 1 Na+ permeation rate: 0.48 × 10− 2 mol m− 2 h− 1 2017 [172]
TFN IP PA RO 2000 ppm NaCl@ 15.5 bar ~ 1.06 RNaCl = 99% 2015 [176]
TFN IP PA RO 2000 ppm NaCl@ 20.5 bar ~ 2.9 RNaCl = 93.8% 2015 [121]
51
SLN LBL PA RO 2000 ppm NaCl@ 15.5 bar ~ 0.9 RNaCl = 96.4% 2013 [71]
SLN Covalent bonding NHS/EDC/ED/PA RO 50 mM NaCl@ 27.6 bar ~ 1.45 RNaCl = 97.8% 2013 [124]
SLN Coordination bonds TiO2/PSF UF 50 mg/L MB@ 0.69 bar ~ 65.2 N/A 2014 [72]
MMS Phase inversion PSF UF @ 1 bar 450 RBSA = 99% 2013 [123]
MMS Phase inversion PVDFc UF @ 1 bar 26.49 RBSA = 79% 2013 [182]
MMS Phase inversion Isocyanate/PSF UF @ 1 bar 135 RBSA = 95% 2013 [187]
MMS Phase inversion HPEId/PSF UF @ 1 bar 153.5 RPEG20,000 = 85%; 2013 [185]
RPVA = 90%
MMS Phase inversion CNT/PVDF UF @ 1 bar 410 N/A 2013 [184]
MMS Phase inversion PES UF @ 4 bar 6 Rdyes = 96% 2013 [181]
MMS Phase inversion APTS/PVDF UF @ 1 bar 401.39 RBSA = 57% 2014 [186]
MMS Phase inversion PVDF MF @0.25 bar 1298 RYeast = 80% 2014 [183]
MMS Phase inversion/IP C3N4/PES/PA FO 2 M NaCl as DS; DI water as FS; Jv = 41.4 L m− 2 h− 1 Jv = 9.35 g m− 2 h− 1 2014 [189]
MMS Phase inversion/IP PSF/PA FO 0.5 M NaCl as DS; DI water as FS; Jv = 19.77 L m− 2 h− 1 Jv = 3.44 g m− 2 h− 1 2015 [179]
a
PAA: poly(acrylic acid); hPAN: hydrolyzed polyacrylonitrile.
b
PAH: poly(allylamine hydrochloride).
c
PVDF: Polyvinylidene difluoride.
d
HPEI: hyperbranched polyethylenimine.
Desalination 434 (2018) 37–59
Fig. 15. Fabrication of a freestanding GO mem-

brane and the characterization of Helium leakage
rate. (A) a photograph of a 1-μm-thick free-
standing GO membrane, (B) SEM micrograph of
the film's cross-sectional features, (C) a schematic
illustration of the water transport through GO
laminates with a typical L d− 1–1000, and (D)
data of helium leakage rate versus its applied
pressure. All figures are reproduced from Ref.
Fig. 16. Schematic diagram of assembly freestanding LBL GO as a rejection layer via vacuum suction on a nanofibrous substrate. The figure is reproduced from Ref. [67] with copyright
permission.
suggested that both the size and the concentration of GO had great 59.4 ± 0.4 L m− 2 h− 1 with slightly decreased salt rejection of both
impact on the performance of the resultant TFN membrane. On the NaCl and Na2SO4. The decreased salt rejection can be attributed to the
other hand, Yin and co-workers [121] prepared GO-based TFN mem- relatively large lateral size of the GO nanoplates (from 500 nm to sev-
brane by dispersing the GO nanoplates in the TMC/hexane organic eral micrometers), which could potentially result in defects in the
solution. With the increasing GO concentration from 0 to 0.015 wt%, polyamide thin-film layer (~ 200 nm).
the water flux of the GO-TFN increased from 39.0 ± 1.6 to
52
Fig. 17. Schematic diagram of assembly LBL GO membrane via electrostatic interaction by alternatively depositing GO and PAH on a HPAN support substrate. The figure is reproduced
from Ref. [169] with copyright permission.
6.2.3. GO-based SLN membranes by either TiO2 or GO alone. In contrast, conventional mixed matrix
GO-based SLN membranes are prepared by loading GO onto mem- membrane could only provide limited enhancement due to the majority
brane surface via covalent bonds [124], electrostatic interaction [71] or of the TiO2 incorporated inside membrane matrix without direct surface
coordination bonds [72] to enhance the membrane's antifouling and interactions [178].
antimicrobial performance. Perreault and co-workers [124] im-
mobilized GO onto the surface of a polyamide rejection layer via 6.2.4. GO-based MMS membranes
covalent bonds using EDC/NHS/ED as bridging agents. This method is Similar to CNT-based MMS membranes, GO can be incorporated
similar to the fabrication routine in Section 5.3 related to the surface into the membrane substrate to enhance its water permeability, rejec-
loaded carboxylated CNTs. The immobilization of GO on the PA re- tion, hydrophilicity, mechanical stability and antimicrobial properties
jection layer did not detrimentally affect its intrinsic transport proper- [79]. Park and co-workers [179] prepared an FO membrane by per-
ties (e.g., water permeability and NaCl rejection). Furthermore, the forming interfacial polymerization on a GO incorporated substrate. At
direct contact of bacteria on the GO modified TFC membranes resulted the optimum GO concentration of 0.25 wt% in PSF substrate, these
in 65% bacterial activation within 1-h of contact time. authors reported a significantly enhanced FO water flux of
In addition to covalent bonds, GO can be bonded onto membrane ~20 L m− 2 h− 1 of the GO modified TFC membrane in the active layer
surface via electrostatic interaction. Choi and co-workers [71] modified facing feed solution orientation (Fig. 20) together with improved
PA-TFC membrane via layer-by-layer deposition of positively charged membrane selectivity. According to the existing literature, GO has been
aminated GO and pristine negatively charge GO (see Fig. 18). Thanks to loaded into various polymeric matrix to prepare GO-loaded MMS (e.g.,
the ultrafast water permeation of GO nanosheets, the surface located GO/PSF [123,180], GO/PES [181], GO/PVDF UF membrane [182],
TFC membranes did not show decreased water permeability compared GO/PVDF MF membrane [183] and GO/CNT/PVDF [184]). Several
to the control TFC membranes, whereas conventional coating techni- recent studies also reported the further functionalization (e.g., HPEI/
ques would generally cause a flux decline [19,177]. Furthermore, the GO [185], APTS/GO [186], isocyanate/GO [187]) of GO before its in-
GO-loaded membrane showed reduced fouling tendency and improved corporation in MMS).
chlorine resistance. These authors attributed the enhanced antifouling
properties to the smoother and more hydrophilic membrane surface 7. Future perspectives
after GO modification.
The GO-based SLN technique has also been adopted for modifying Fig. 21 compares the various membranes (novel AQP and carbon-
UF membrane. For example, Gao and co-workers [72] modified the based membranes vs. conventional TFC and TFN membranes) in a
surface of PSF membrane via the hydrogen bonds between GO and TiO2 qualitative manner. The horizontal axis represents the commercial
(Fig. 19). The GO/TiO2 surface modification rendered the membrane viability on the basis of scalability of technology and cost of fabrication,
with strong photocatalytic properties, showing 3–4 times faster pho- and the vertical axis reveals the (potential) membrane performance
todegradation of methylene blue compared to PSF membranes modified enhancement with respect to water permeability, selectivity and
Fig. 18. Schematic diagram of layer-by-layer coating of

positively charged aminated GO (AGO) and pristine GO
on a polyamide TFC membrane surface by electrostatic
interaction. The figure is reprinted from Ref. [71] with
53
Fig. 19. (a) Photographs and (b) schematic

diagram of experimental procedures for
GO/TiO2 deposition on a polysulfone mem-
brane. All figures are reproduced from Ref.
incorporate novel materials such as AQPs, CNTs, and GO. Due to their
extremely permeable nature (Table 1), AQPs, CNTs, and GO-based TFNs
can potentially outperform conventional TFN membranes. At the
meantime, the salt-rejecting matrix (e.g., polyamide) maintains the
high salt rejection required for desalination applications. Nevertheless,
the ultimate permeability of a TFN membrane is largely determined by
the matrix, with the loaded nanomaterials only acting as “enhancers”.
Therefore, this membrane design is difficult to fully unleash the re-
volutionary performance of AQPs, CNTs, and GO. In general, TFN
membranes can be easily scaled up for commercial production; notable
commercial products include NanoH2O® by LG (zeolite-based TFNs
[26,190]) and AIMs® by Aquaporin D/S (aquaporin-based TFNs
[119,120]). Despite the vast number of studies on TFN membranes, the
rejection mechanisms of these membranes need to be further in-
vestigated in viewing of the often different trends (e.g., increased
[191,192] vs. decreased salt rejection [193,194]) reported in the lit-
erature. In addition to the intrinsic rejection properties of AQPs, CNTs,
and GO, the incorporation of these materials into the rejection layer
may decrease its cross-linking degree [195] or even creates defects in
the rejection layer [146]. Such issues need to be more adequately ad-
dressed in the future studies. In addition, long term leachability of the
nanofillers from TFN membranes needs to be systematically studied.
The issue of leachability is even more critical for the case of SLNs,
where the loaded nanomaterials are directly exposed to the feed water.
Due to the potential nanotoxicity of some nanomaterials (e.g., CNTs and
graphene) [196], their leaching into the product water or to the en-
vironment can be an important concern that needs to be further ad-
dressed.
Unlike the TFN approach where nanomaterials are directly loaded
to the rejection layer, the MMS approach loads the nanomaterials into
the substrate. Interestingly, the improved substrate properties (e.g.,
hydrophilicity and pore structure) can still significantly enhance the
overall membrane separation performance (e.g., with doubled mem-
brane permeability [127]). This approach has also been widely adopted
in the FO and PRO literature as an effective method of substrate en-
hancement to achieve reduced internal concentration polarization and
Fig. 20. Effect of GO loading amounts in PSF substrate on the TFC-FO performance (DI improved antifouling performance [130]. Similar to the TFN mem-
water as feed solution and 0.5 M NaCl as draw solution). (a) FO water flux, (b) reverse salt branes, the MMS membranes are relatively straightforward to prepare,
flux and (c) reverse flux selectivity. All figures are reproduced from Ref. [179] with
and their production can be readily integrated with existing commercial
RO production lines with the exception of some minor modifications.
The scalability of the technology means that MMS and TFN membranes
antifouling properties. are more readily to be commercialized compared to PRL membranes
Compared to the state-of-the-art TFC membranes, conventional TFN made of AQPs or carbon-based materials.
membranes prepared by incorporating various nanoparticles into the Despite the great promises of novel desalting materials such as
polyamide rejection layer tend show enhanced separation performance AQPs, CNTs, NPG, and GO, processing these materials into PRL mem-
[26,39]. In recent years, this membrane structure has been extended to branes are challenging. Most of the existing studies report membrane
54
Fig. 21. Comparison of membrane performance enhancement

and commercial viability of the conventional TFC membrane,
conventional TFN membrane and the emerging novel membrane
materials. The horizontal axis represents the commercial po-
tential on the basis of scalability of technology and cost of fab-
rication (“easy” for membranes that are already commercialized
with large scale production; “possible” for membranes that can
be potentially scaled up for commercialization in the near term;
and “difficult” for membranes that are difficult to scale up or too
costly to produce in the near term). The vertical axis represents
the (potential) membrane performance improvement in terms of
water permeability, selectivity and antifouling properties com-
pared to the state-of-the-art TFC membranes, with “revolu-
tionary” indicating the possibility of orders of magnitude im-
provement and “marginal” indicating more moderate
improvement. The figure adopted an evaluation framework si-
milar to the one published by Pendergast and Hoek [197].
areas on the order of mm (or even μm) and often involves highly spe- PRL membrane structure has the greatest potential in achieving re-
cialized fabrication techniques, with the important exception of GO volutionary breakthroughs (e.g., orders of magnitude improvements in
membranes that can be processed at large scales. Meanwhile, many water permeability), PRL membranes on the basis of AQPs, CNTs, NPG
improvements are still required to demonstrate the technical feasibility and GO are still far from commercialization. In addition to the com-
of these membranes for desalination. A critical gap is the salt rejection. monly reported water permeability, key issues to be addressed for PRLs
Although molecular level simulations often reveal extremely high se- include their scalability, stability, and salt rejection in the context of RO
lectivity (e.g., nearly 100% NaCl rejection for AQPs [41]), experimental desalination.
works have generally show relatively low salt rejection (≪90% for
NaCl in most cases, See Tables 3–5). Such low rejection performance is Acknowledgement
likely caused by the presence of large amount of defects in the rejection
layer and/or the instability of the rejection layer (e.g., GO layer swells The study receives financial support from the General Research
in water). Among the few studies reporting NaCl rejection of > 90%, Fund of the Research Grants Council (Project # 17207514). The partial
the rejection was often obtained by indirect measurement using very funding support from Strategic Research Theme on Clean Energy at the
small membrane areas (e.g., using a 5 μm nanoporous graphene [54]). University of Hong Kong and the Seed Grant for Basic Research
Therefore, future studies need to further validate the rejection perfor- (104003453) are also appreciated. The authors also gratefully ac-
mance, particularly for NaCl, using pressurized RO tests with larger knowledge the Hong Kong Scholars Program (No. XJ2015015).
membrane areas. At the same time, many PRL membranes show ade-
quate rejections against small organic molecules (e.g., dyes) and diva- References
lent ions, which enable their potential applications for nanofiltration/
ultrafiltration processes. [1] D. Seckler, R. Barker, U. Amarasinghe, Water scarcity in the twenty-first century,
Int. J. Water Resour. Dev. 15 (1999) 29–42.
[2] M. Elimelech, W.A. Phillip, The future of seawater desalination: energy, tech-
8. Conclusions nology, and the environment, Science 333 (2011) 712–717.
[3] K.P. Lee, T.C. Arnot, D. Mattia, A review of reverse osmosis membrane materials
for desalination—development to date and future potential, J. Membr. Sci. 370
The material properties, major fabrication methods, and membrane (2011) 1–22.
performance of novel desalination membranes using AQPs, CNTs, NPG [4] D. Cohen-Tanugi, R.K. McGovern, S.H. Dave, J.H. Lienhard, J.C. Grossman,
Quantifying the potential of ultra-permeable membranes for water desalination,
and GO are comprehensively reviewed. These membranes can be gen- Energy Environ. Sci. 7 (2014) 1134–1141.
erally classified into PRL (in which the nanomaterials form the primary [5] J.R. Werber, A. Deshmukh, M. Elimelech, The critical need for increased se-
rejection layer), TFN (by loading them into a salt-rejecting matrix), SLN lectivity, not increased water permeability, for desalination membranes, Environ.
Sci. Technol. Lett. 3 (2016) 112–120.
(by immobilizing them onto an existing membrane surface), and MMS [6] J.R. Werber, C.O. Osuji, M. Elimelech, Materials for next-generation desalination
(by incorporating them into the substrate). The TFN and MMS ap- and water purification membranes, Nat. Rev. Mater. 1 (2016) 16018.
proaches often result in moderate improvements in water permeability [7] A. Fane, C. Tang, R. Wang, Membrane technology for water: microfiltration, ul-
trafiltration, nanofiltration, and reverse osmosis, Treatise Water Sci. 2011.
while maintaining relatively high salt rejection, whereas the SLN ap-
[8] M. Kumar, M. Grzelakowski, J. Zilles, M. Clark, W. Meier, Highly permeable
proach can be highly effective against membrane fouling. Although the
55
polymeric membranes based on the incorporation of the functional water channel covalently bonded with modified mesoporous silica nanoparticles, J. Membr. Sci.
protein Aquaporin Z, Proc. Natl. Acad. Sci. 104 (2007) 20719–20724. 428 (2013) 341–348.
[9] Y.-x. Shen, P.O. Saboe, I.T. Sines, M. Erbakan, M. Kumar, Biomimetic membranes: [39] J. Yin, B. Deng, Polymer-matrix nanocomposite membranes for water treatment, J.
a review, J. Membr. Sci. 454 (2014) 359–381. Membr. Sci. 479 (2015) 256–275.
[10] B.J. Hinds, N. Chopra, T. Rantell, R. Andrews, V. Gavalas, L.G. Bachas, Aligned [40] P. Agre, Aquaporin water channels, Biosci. Rep. 24 (2004) 127–163.
multiwalled carbon nanotube membranes, Science 303 (2004) 62–65. [41] C. Tang, Z. Wang, I. Petrinić, A.G. Fane, C. Hélix-Nielsen, Biomimetic aquaporin
[11] A.K. Geim, K.S. Novoselov, The rise of graphene, Nat. Mater. 6 (2007) 183–191. membranes coming of age, Desalination 368 (2015) 89–105.
[12] R. Nair, H. Wu, P. Jayaram, I. Grigorieva, A. Geim, Unimpeded permeation of [42] C. Tang, Y. Zhao, R. Wang, C. Hélix-Nielsen, A. Fane, Desalination by biomimetic
water through helium-leak–tight graphene-based membranes, Science 335 (2012) aquaporin membranes: review of status and prospects, Desalination 308 (2013)
442–444. 34–40.
[13] R.J. Petersen, Composite reverse osmosis and nanofiltration membranes, J. [43] P. Agre, L.S. King, M. Yasui, W.B. Guggino, O.P. Ottersen, Y. Fujiyoshi, A. Engel,
Membr. Sci. 83 (1993) 81–150. S. Nielsen, Aquaporin water channels–from atomic structure to clinical medicine,
[14] C. Feng, J. Xu, M. Li, Y. Tang, C. Gao, Studies on a novel nanofiltration membrane J. Physiol. 542 (2002) 3–16.
prepared by cross-linking of polyethyleneimine on polyacrylonitrile substrate, J. [44] D. Kozono, M. Yasui, L.S. King, P. Agre, Aquaporin water channels: atomic
Membr. Sci. 451 (2014) 103–110. structure molecular dynamics meet clinical medicine, J. Clin. Invest. 109 (2002)
[15] R. Han, Formation and characterization of (melamine–TMC) based thin film 1395–1399.
composite NF membranes for improved thermal and chlorine resistances, J. [45] D.F. Savage, J.D. O'Connell, L.J. Miercke, J. Finer-Moore, R.M. Stroud, Structural
Membr. Sci. 425 (2013) 176–181. context shapes the aquaporin selectivity filter, Proc. Natl. Acad. Sci. 107 (2010)
[16] L.-F. Liu, Z.-B. Cai, J.-N. Shen, L.-X. Wu, E.M. Hoek, C.-J. Gao, Fabrication and 17164–17169.
characterization of a novel poly (amide-urethane@ imide) TFC reverse osmosis [46] J.C. Mathai, S. Tristram-Nagle, J.F. Nagle, M.L. Zeidel, Structural determinants of
membrane with chlorine-tolerant property, J. Membr. Sci. 469 (2014) 397–409. water permeability through the lipid membrane, J. Gen. Physiol. 131 (2008)
[17] H. Wang, Q. Zhang, S. Zhang, Positively charged nanofiltration membrane formed 69–76.
by interfacial polymerization of 3, 3′, 5, 5′-biphenyl tetraacyl chloride and pi- [47] N.T. Hovijitra, J.J. Wuu, B. Peaker, J.R. Swartz, Cell-free synthesis of functional
perazine on a poly (acrylonitrile)(PAN) support, J. Membr. Sci. 378 (2011) aquaporin Z in synthetic liposomes, Biotechnol. Bioeng. 104 (2009) 40–49.
243–249. [48] X.-B. Hu, Z. Chen, G. Tang, J.-L. Hou, Z.-T. Li, Single-molecular artificial trans-
[18] J. Wei, X. Liu, C. Qiu, R. Wang, C.Y. Tang, Influence of monomer concentrations membrane water channels, J. Am. Chem. Soc. 134 (2012) 8384–8387.
on the performance of polyamide-based thin film composite forward osmosis [49] M. Barboiu, Y. Le Duc, A. Gilles, P.-A. Cazade, M. Michau, Y.M. Legrand, A. Van
membranes, J. Membr. Sci. 381 (2011) 110–117. Der Lee, B. Coasne, P. Parvizi, J. Post, An artificial primitive mimic of the
[19] Z. Yang, Y. Wu, J. Wang, B. Cao, C.Y. Tang, In situ reduction of silver by poly- Gramicidin-A channel, Nat. Commun. 5 (2014) 4142.
dopamine: a novel antimicrobial modification of a thin-film composite polyamide [50] T. Fujioka, N. Oshima, R. Suzuki, W.E. Price, L.D. Nghiem, Probing the internal
membrane, Environ. Sci. Technol. 50 (2016) 9543–9550. structure of reverse osmosis membranes by positron annihilation spectroscopy:
[20] S. Yu, Y. Zheng, Q. Zhou, S. Shuai, Z. Lü, C. Gao, Facile modification of poly- gaining more insight into the transport of water and small solutes, J. Membr. Sci.
propylene hollow fiber microfiltration membranes for nanofiltration, Desalination 486 (2015) 106–118.
298 (2012) 49–58. [51] V. Freger, Nanoscale heterogeneity of polyamide membranes formed by interfacial
[21] Y. Zhou, S. Yu, C. Gao, X. Feng, Surface modification of thin film composite polymerization, Langmuir 19 (2003) 4791–4797.
polyamide membranes by electrostatic self deposition of polycations for improved [52] J.S. Jung, G.M. Preston, B.L. Smith, W.B. Guggino, P. Agre, Molecular structure of
fouling resistance, Sep. Purif. Technol. 66 (2009) 287–294. the water channel through aquaporin CHIP. The hourglass model, J. Biol. Chem.
[22] J.T. Arena, B. McCloskey, B.D. Freeman, J.R. McCutcheon, Surface modification of 269 (1994) 14648–14654.
thin film composite membrane support layers with polydopamine: enabling use of [53] J.K. Holt, H.G. Park, Y. Wang, M. Stadermann, A.B. Artyukhin, C.P. Grigoropoulos,
reverse osmosis membranes in pressure retarded osmosis, J. Membr. Sci. 375 A. Noy, O. Bakajin, Fast mass transport through sub-2-nanometer carbon nano-
(2011) 55–62. tubes, Science 312 (2006) 1034–1037.
[23] G.-D. Kang, C.-J. Gao, W.-D. Chen, X.-M. Jie, Y.-M. Cao, Q. Yuan, Study on hy- [54] S.P. Surwade, S.N. Smirnov, I.V. Vlassiouk, R.R. Unocic, G.M. Veith, S. Dai,
pochlorite degradation of aromatic polyamide reverse osmosis membrane, J. S.M. Mahurin, Water desalination using nanoporous single-layer graphene, Nat,
Membr. Sci. 300 (2007) 165–171. Nanotechnology 10 (2015) 459–464.
[24] V.T. Do, C.Y. Tang, M. Reinhard, J.O. Leckie, Effects of chlorine exposure condi- [55] J. Abraham, K.S. Vasu, C.D. Williams, K. Gopinadhan, Y. Su, C.T. Cherian, J. Dix,
tions on physiochemical properties and performance of a polyamide membrane– E. Prestat, S.J. Haigh, I.V. Grigorieva, P. Carbone, A.K. Geim, R.R. Nair, Tunable
mechanisms and implications, Environ. Sci. Technol. 46 (2012) 13184–13192. sieving of ions using graphene oxide membranes, Nat. Nanotechnol. 12 (2017)
[25] N.Y. Yip, M. Elimelech, Thermodynamic and energy efficiency analysis of power (546–545).
generation from natural salinity gradients by pressure retarded osmosis, Environ. [56] L.F. Greenlee, D.F. Lawler, B.D. Freeman, B. Marrot, P. Moulin, Reverse osmosis
Sci. Technol. 46 (2012) 5230–5239. desalination: water sources, technology, and today's challenges, Water Res. 43
[26] B.-H. Jeong, E.M. Hoek, Y. Yan, A. Subramani, X. Huang, G. Hurwitz, A.K. Ghosh, (2009) 2317–2348.
A. Jawor, Interfacial polymerization of thin film nanocomposites: a new concept [57] S. Kang, M. Pinault, L.D. Pfefferle, M. Elimelech, Single-walled carbon nanotubes
for reverse osmosis membranes, J. Membr. Sci. 294 (2007) 1–7. exhibit strong antimicrobial activity, Langmuir 23 (2007) 8670–8673.
[27] J. Yin, E.-S. Kim, J. Yang, B. Deng, Fabrication of a novel thin-film nanocomposite [58] J. Yin, G. Zhu, B. Deng, Multi-walled carbon nanotubes (MWNTs)/polysulfone
(TFN) membrane containing MCM-41 silica nanoparticles (NPs) for water pur- (PSU) mixed matrix hollow fiber membranes for enhanced water treatment, J.
ification, J. Membr. Sci. 423 (2012) 238–246. Membr. Sci. 437 (2013) 237–248.
[28] Z. Yang, J. Yin, B. Deng, Enhancing water flux of thin-film nanocomposite (TFN) [59] S. Liu, T.H. Zeng, M. Hofmann, E. Burcombe, J. Wei, R. Jiang, J. Kong, Y. Chen,
membrane by incorporation of bimodal silica nanoparticles, AIMS Envrion. Sci. 3 Antibacterial activity of graphite, graphite oxide, graphene oxide, and reduced
(2016) 185–198. graphene oxide: membrane and oxidative stress, ACS Nano 5 (2011) 6971–6980.
[29] S. Sorribas, P. Gorgojo, C. Téllez, J. Coronas, A.G. Livingston, High flux thin film [60] X. Song, L. Wang, C.Y. Tang, Z. Wang, C. Gao, Fabrication of carbon nanotubes
nanocomposite membranes based on metal–organic frameworks for organic sol- incorporated double-skinned thin film nanocomposite membranes for enhanced
vent nanofiltration, J. Am. Chem. Soc. 135 (2013) 15201–15208. separation performance and antifouling capability in forward osmosis process,
[30] X.-H. Ma, Z. Yang, Z.-K. Yao, Z.-L. Xu, C.Y. Tang, A facile preparation of novel Desalination 369 (2015) 1–9.
positively charged MOF/chitosan nanofiltration membranes, J. Membr. Sci. 525 [61] S.-M. Xue, Z.-L. Xu, Y.-J. Tang, C.-H. Ji, Polypiperazine-amide nanofiltration
(2017) 269–276. membrane modified by different functionalized multiwalled carbon nanotubes
[31] X. Liu, N.K. Demir, Z. Wu, K. Li, Highly water-stable zirconium metal-organic (MWCNTs), ACS Appl. Mater. Interfaces 8 (2016) 19135–19144.
framework UiO-66 membranes supported on alumina hollow fibers for desalina- [62] M. Tian, R. Wang, K. Goh, Y. Liao, A.G. Fane, Synthesis and characterization of
tion, J. Am. Chem. Soc. 137 (2015) 6999–7002. high-performance novel thin film nanocomposite PRO membranes with tiered
[32] G.L. Jadav, P.S. Singh, Synthesis of novel silica-polyamide nanocomposite mem- nanofiber support reinforced by functionalized carbon nanotubes, J. Membr. Sci.
brane with enhanced properties, J. Membr. Sci. 328 (2009) 257–267. 486 (2015) 151–160.
[33] X. Liu, S. Qi, Y. Li, L. Yang, B. Cao, C.Y. Tang, Synthesis and characterization of [63] A. Khalid, A.A. Al-Juhani, O.C. Al-Hamouz, T. Laoui, Z. Khan, M.A. Atieh,
novel antibacterial silver nanocomposite nanofiltration and forward osmosis Preparation and properties of nanocomposite polysulfone/multi-walled carbon
membranes based on layer-by-layer assembly, Water Res. 47 (2013) 3081–3092. nanotubes membranes for desalination, Desalination 367 (2015) 134–144.
[34] J. Wu, C. Yu, Q. Li, Regenerable antimicrobial activity in polyamide thin film [64] M. Tian, Y.-N. Wang, R. Wang, Synthesis and characterization of novel high-per-
nanocomposite membranes, J. Membr. Sci. 476 (2015) 119–127. formance thin film nanocomposite (TFN) FO membranes with nanofibrous sub-
[35] H.S. Lee, S.J. Im, J.H. Kim, H.J. Kim, J.P. Kim, B.R. Min, Polyamide thin-film strate reinforced by functionalized carbon nanotubes, Desalination 370 (2015)
nanofiltration membranes containing TiO2 nanoparticles, Desalination 219 (2008) 79–86.
48–56. [65] D. Cohen-Tanugi, J.C. Grossman, Nanoporous graphene as a reverse osmosis
[36] C. Kong, T. Shintani, T. Tsuru, “Pre-seeding”-assisted synthesis of a high perfor- membrane: recent insights from theory and simulation, Desalination 366 (2015)
mance polyamide-zeolite nanocomposite membrane for water purification, New J. 59–70.
Chem. 34 (2010) 2101–2104. [66] C. Xu, A. Cui, Y. Xu, X. Fu, Graphene oxide–TiO2 composite filtration membranes
[37] D. Emadzadeh, W. Lau, M. Rahbari-Sisakht, H. Ilbeygi, D. Rana, T. Matsuura, and their potential application for water purification, Carbon 62 (2013) 465–471.
A. Ismail, Synthesis, modification and optimization of titanate nanotubes-poly- [67] J. Wang, P. Zhang, B. Liang, Y. Liu, T. Xu, L. Wang, B. Cao, K. Pan, Graphene oxide
amide thin film nanocomposite (TFN) membrane for forward osmosis (FO) ap- as an effective barrier on a porous nanofibrous membrane for water treatment,
plication, Chem. Eng. J. 281 (2015) 243–251. ACS Appl. Mater. Interfaces 8 (2016) 6211–6218.
[38] H. Wu, B. Tang, P. Wu, Optimizing polyamide thin film composite membrane [68] P. Sun, M. Zhu, K. Wang, M. Zhong, J. Wei, D. Wu, Z. Xu, H. Zhu, Selective ion
56
penetration of graphene oxide membranes, ACS Nano 7 (2012) 428–437. characterization of large-area, semiconducting nanoperforated graphene mate-
[69] M. Hu, B. Mi, Enabling graphene oxide nanosheets as water separation mem- rials, Nano Lett. 10 (2010) 1125–1131.
branes, Environ. Sci. Technol. 47 (2013) 3715–3723. [102] C.A. Merchant, K. Healy, M. Wanunu, V. Ray, N. Peterman, J. Bartel,
[70] Y. Han, Z. Xu, C. Gao, Ultrathin graphene nanofiltration membrane for water M.D. Fischbein, K. Venta, Z. Luo, A.T. Johnson, M. Drndic, DNA translocation
purification, Adv. Funct. Mater. 23 (2013) 3693–3700. through graphene nanopores, Nano Lett. 10 (2010) 2915–2921.
[71] W. Choi, J. Choi, J. Bang, J.-H. Lee, Layer-by-layer assembly of graphene oxide [103] G.F. Schneider, S.W. Kowalczyk, V.E. Calado, G. Pandraud, H.W. Zandbergen,
nanosheets on polyamide membranes for durable reverse-osmosis applications, L.M. Vandersypen, C. Dekker, DNA translocation through graphene nanopores,
ACS Appl. Mater. Interfaces 5 (2013) 12510–12519. Nano Lett. 10 (2010) 3163–3167.
[72] Y. Gao, M. Hu, B. Mi, Membrane surface modification with TiO2–graphene oxide [104] J. Schrier, Helium separation using porous graphene membranes, J. Phys. Chem.
for enhanced photocatalytic performance, J. Membr. Sci. 455 (2014) 349–356. Lett. 1 (2010) 2284–2287.
[73] W.-S. Hung, Q.-F. An, M. De Guzman, H.-Y. Lin, S.-H. Huang, W.-R. Liu, C.-C. Hu, [105] M.E. Suk, N.R. Aluru, Water transport through ultrathin graphene, J. Phys. Chem.
K.-R. Lee, J.-Y. Lai, Pressure-assisted self-assembly technique for fabricating Lett. 1 (2010) 1590–1594.
composite membranes consisting of highly ordered selective laminate layers of [106] Y. Zhang, L. Zhang, C. Zhou, Review of chemical vapor deposition of graphene and
amphiphilic graphene oxide, Carbon 68 (2014) 670–677. related applications, Acc. Chem. Res. 46 (2013) 2329–2339.
[74] Y. Manawi, V. Kochkodan, M.A. Hussein, M.A. Khaleel, M. Khraisheh, N. Hilal, [107] B.M. Yoo, H.J. Shin, H.W. Yoon, H.B. Park, Graphene and graphene oxide and
Can carbon-based nanomaterials revolutionize membrane fabrication for water their uses in barrier polymers, J. Appl. Polym. Sci. 131 (2014) 39628.
treatment and desalination? Desalination 391 (2016) 69–88. [108] D. Cohen-Tanugi, J.C. Grossman, Water permeability of nanoporous graphene at
[75] M. Majumder, N. Chopra, R. Andrews, B.J. Hinds, Nanoscale hydrodynamics: realistic pressures for reverse osmosis desalination, J. Chem. Phys. 141 (2014)
enhanced flow in carbon nanotubes, Nature 438 (2005) 44. 074704.
[76] G. Hummer, J.C. Rasaiah, J.P. Noworyta, Water conduction through the hydro- [109] D. Cohen-Tanugi, J.C. Grossman, Water desalination across nanoporous graphene,
phobic channel of a carbon nanotube, Nature 414 (2001) 188–190. Nano Lett. 12 (2012) 3602–3608.
[77] B. Corry, Designing carbon nanotube membranes for efficient water desalination, [110] B. Mi, Graphene oxide membranes for ionic and molecular sieving, Science 343
J. Phys. Chem. B 112 (2008) 1427–1434. (2014) 740–742.
[78] R. Das, M.E. Ali, S.B.A. Hamid, S. Ramakrishna, Z.Z. Chowdhury, Carbon nanotube [111] B. Brodie, Sur le poids atomique du graphite, Ann. Chim. Phys. 59 (1860) e472.
membranes for water purification: a bright future in water desalination, [112] L. Staudenmaier, Verfahren zur darstellung der graphitsäure, Eur. J. Inorg. Chem.
Desalination 336 (2014) 97–109. 32 (1899) 1394–1399.
[79] H.M. Hegab, L.D. Zou, Graphene oxide-assisted membranes: fabrication and po- [113] W.S. Hummers Jr., R.E. Offeman, Preparation of graphitic oxide, J. Am. Chem.
tential applications in desalination and water purification, J. Membr. Sci. 484 Soc. 80 (1958) 1339.
(2015) 95–106. [114] D.C. Marcano, D.V. Kosynkin, J.M. Berlin, A. Sinitskii, Z. Sun, A. Slesarev,
[80] K. Sears, L. Dumée, J. Schütz, M. She, C. Huynh, S. Hawkins, M. Duke, S. Gray, L.B. Alemany, W. Lu, J.M. Tour, Improved synthesis of graphene oxide, ACS Nano
Recent developments in carbon nanotube membranes for water purification and 4 (2010) 4806–4814.
gas separation, Materials 3 (2010) 127–149. [115] A. Nicolaï, B.G. Sumpter, V. Meunier, Tunable water desalination across graphene
[81] S. Iijima, Helical microtubules of graphitic carbon, Nature 354 (1991) 56–58. oxide framework membranes, Phys. Chem. Chem. Phys. 16 (2014) 8646–8654.
[82] S. Ravindran, S. Chaudhary, B. Colburn, M. Ozkan, C.S. Ozkan, Covalent coupling [116] C.K. Chua, M. Pumera, The reduction of graphene oxide with hydrazine: eluci-
of quantum dots to multiwalled carbon nanotubes for electronic device applica- dating its reductive capability based on a reaction-model approach, Chem.
tions, Nano Lett. 3 (2003) 447–453. Commun. 52 (2016) 72–75.
[83] M.S. Strano, C.A. Dyke, M.L. Usrey, P.W. Barone, M.J. Allen, H. Shan, C. Kittrell, [117] Y. You, V. Sahajwalla, M. Yoshimura, R.K. Joshi, Graphene and graphene oxide for
R.H. Hauge, J.M. Tour, R.E. Smalley, Electronic structure control of single-walled desalination, Nano 8 (2016) 117–119.
carbon nanotube functionalization, Science 301 (2003) 1519–1522. [118] P.S. Zhong, T.-S. Chung, K. Jeyaseelan, A. Armugam, Aquaporin-embedded bio-
[84] W. Yang, P. Thordarson, J.J. Gooding, S.P. Ringer, F. Braet, Carbon nanotubes for mimetic membranes for nanofiltration, J. Membr. Sci. 407 (2012) 27–33.
biological and biomedical applications, Nanotechnology 18 (2007) 412001. [119] Y. Zhao, C. Qiu, X. Li, A. Vararattanavech, W. Shen, J. Torres, C. Helix-Nielsen,
[85] A. Bianco, K. Kostarelos, C.D. Partidos, M. Prato, Biomedical applications of R. Wang, X. Hu, A.G. Fane, Synthesis of robust and high-performance aquaporin-
functionalised carbon nanotubes, Chem. Commun. (2005) 571–577. based biomimetic membranes by interfacial polymerization-membrane prepara-
[86] Z. Liu, S. Tabakman, K. Welsher, H. Dai, Carbon nanotubes in biology and med- tion and RO performance characterization, J. Membr. Sci. 423 (2012) 422–428.
icine: in vitro and in vivo detection, imaging and drug delivery, Nano Res. 2 [120] X. Li, S. Chou, R. Wang, L. Shi, W. Fang, G. Chaitra, C.Y. Tang, J. Torres, X. Hu,
(2009) 85–120. A.G. Fane, Nature gives the best solution for desalination: aquaporin-based hollow
[87] Z. Liu, S.M. Tabakman, Z. Chen, H. Dai, Preparation of carbon nanotube bio- fiber composite membrane with superior performance, J. Membr. Sci. 494 (2015)
conjugates for biomedical applications, Nat. Protoc. 4 (2009) 1372–1381. 68–77.
[88] S. Kang, M. Herzberg, D.F. Rodrigues, M. Elimelech, Antibacterial effects of carbon [121] J. Yin, G. Zhu, B. Deng, Graphene oxide (GO) enhanced polyamide (PA) thin-film
nanotubes: size does matter, Langmuir 24 (2008) 6409–6413. nanocomposite (TFN) membrane for water purification, Desalination 379 (2016)
[89] Y. Baek, C. Kim, D.K. Seo, T. Kim, J.S. Lee, Y.H. Kim, K.H. Ahn, S.S. Bae, S.C. Lee, 93–101.
J. Lim, High performance and antifouling vertically aligned carbon nanotube [122] W.-F. Chan, H.-y. Chen, A. Surapathi, M.G. Taylor, X. Shao, E. Marand,
membrane for water purification, J. Membr. Sci. 460 (2014) 171–177. J.K. Johnson, Zwitterion functionalized carbon nanotube/polyamide nano-
[90] A. Tiraferri, C.D. Vecitis, M. Elimelech, Covalent binding of single-walled carbon composite membranes for water desalination, ACS Nano 7 (2013) 5308–5319.
nanotubes to polyamide membranes for antimicrobial surface properties, ACS [123] J. Lee, H.-R. Chae, Y.J. Won, K. Lee, C.-H. Lee, H.H. Lee, I.-C. Kim, J.-M. Lee,
Appl. Mater. Interfaces 3 (2011) 2869–2877. Graphene oxide nanoplatelets composite membrane with hydrophilic and anti-
[91] Y. Zhang, Y.-W. Tan, H.L. Stormer, P. Kim, Experimental observation of the fouling properties for wastewater treatment, J. Membr. Sci. 448 (2013) 223–230.
quantum Hall effect and Berry's phase in graphene, Nature 438 (2005) 201–204. [124] F. Perreault, M.E. Tousley, M. Elimelech, Thin-film composite polyamide mem-
[92] J.S. Bunch, S.S. Verbridge, J.S. Alden, A.M. Van Der Zande, J.M. Parpia, branes functionalized with biocidal graphene oxide nanosheets, Environ. Sci.
H.G. Craighead, P.L. McEuen, Impermeable atomic membranes from graphene Technol. Lett. 1 (2013) 71–76.
sheets, Nano Lett. 8 (2008) 2458–2462. [125] A.K. Ghosh, E.M. Hoek, Impacts of support membrane structure and chemistry on
[93] K. Sint, B. Wang, P. Král, Selective ion passage through functionalized graphene polyamide–polysulfone interfacial composite membranes, J. Membr. Sci. 336
nanopores, J. Am. Chem. Soc. 130 (2008) 16448–16449. (2009) 140–148.
[94] U. Bangert, M.H. Gass, A.L. Bleloch, R.R. Nair, A.K. Geim, Manifestation of ripples [126] L. Huang, J.R. McCutcheon, Impact of support layer pore size on performance of
in free-standing graphene in lattice images obtained in an aberration-corrected thin film composite membranes for forward osmosis, J. Membr. Sci. 483 (2015)
scanning transmission electron microscope, Phys. Status Solidi (a) 206 (2009) 25–33.
1117–1122. [127] N. Ma, J. Wei, S. Qi, Y. Zhao, Y. Gao, C.Y. Tang, Nanocomposite substrates for
[95] M. Bieri, M. Treier, J. Cai, K. Ait-Mansour, P. Ruffieux, O. Groning, P. Groning, controlling internal concentration polarization in forward osmosis membranes, J.
M. Kastler, R. Rieger, X. Feng, K. Mullen, R. Fasel, Porous graphenes: two-di- Membr. Sci. 441 (2013) 54–62.
mensional polymer synthesis with atomic precision, Chem. Commun. (Camb.) [128] D. Emadzadeh, W.J. Lau, A.F. Ismail, Synthesis of thin film nanocomposite for-
(2009) 6919–6921. ward osmosis membrane with enhancement in water flux without sacrificing salt
[96] K.V. Emtsev, A. Bostwick, K. Horn, J. Jobst, G.L. Kellogg, L. Ley, J.L. McChesney, rejection, Desalination 330 (2013) 90–99.
T. Ohta, S.A. Reshanov, J. Rohrl, E. Rotenberg, A.K. Schmid, D. Waldmann, [129] D. Emadzadeh, W.J. Lau, T. Matsuura, A.F. Ismail, M. Rahbari-Sisakht, Synthesis
H.B. Weber, T. Seyller, Towards wafer-size graphene layers by atmospheric pres- and characterization of thin film nanocomposite forward osmosis membrane with
sure graphitization of silicon carbide, Nat. Mater. 8 (2009) 203–207. hydrophilic nanocomposite support to reduce internal concentration polarization,
[97] Ç.Ö. Girit, J.C. Meyer, R. Erni, M.D. Rossell, C. Kisielowski, L. Yang, C.-H. Park, J. Membr. Sci. 449 (2014) 74–85.
M. Crommie, M.L. Cohen, S.G. Louie, Graphene at the edge: stability and dy- [130] X. Liu, L.-X. Foo, Y. Li, J.-Y. Lee, B. Cao, C.Y. Tang, Fabrication and character-
namics, Science 323 (2009) 1705–1708. ization of nanocomposite pressure retarded osmosis (PRO) membranes with ex-
[98] D.-e. Jiang, V.R. Cooper, S. Dai, Porous graphene as the ultimate membrane for gas cellent anti-biofouling property and enhanced water permeability, Desalination
separation, Nano Lett. 9 (2009) 4019–4024. 389 (2016) 137–148.
[99] K.S. Kim, Y. Zhao, H. Jang, S.Y. Lee, J.M. Kim, K.S. Kim, J.H. Ahn, P. Kim, [131] Y. Kaufman, A. Berman, V. Freger, Supported lipid bilayer membranes for water
J.Y. Choi, B.H. Hong, Large-scale pattern growth of graphene films for stretchable purification by reverse osmosis, Langmuir 26 (2010) 7388–7395.
transparent electrodes, Nature 457 (2009) 706–710. [132] X. Li, R. Wang, C. Tang, A. Vararattanavech, Y. Zhao, J. Torres, T. Fane,
[100] S. Blankenburg, M. Bieri, R. Fasel, K. Mullen, C.A. Pignedoli, D. Passerone, Porous Preparation of supported lipid membranes for aquaporin Z incorporation, Colloids
graphene as an atmospheric nanofilter, Small 6 (2010) 2266–2271. Surf. B: Biointerfaces 94 (2012) 333–340.
[101] M. Kim, N.S. Safron, E. Han, M.S. Arnold, P. Gopalan, Fabrication and [133] P.H. Duong, T.-S. Chung, K. Jeyaseelan, A. Armugam, Z. Chen, J. Yang, M. Hong,
57
Planar biomimetic aquaporin-incorporated triblock copolymer membranes on carbon nanotubes, J. Membr. Sci. 435 (2013) 233–241.
porous alumina supports for nanofiltration, J. Membr. Sci. 409 (2012) 34–43. [164] J. nan Shen, C. chao Yu, H. min Ruan, C. jie Gao, B. Van der Bruggen, Preparation
[134] W. Ding, J. Cai, Z. Yu, Q. Wang, Z. Xu, Z. Wang, C. Gao, Fabrication of an and characterization of thin-film nanocomposite membranes embedded with poly
aquaporin-based forward osmosis membrane through covalent bonding of a lipid (methyl methacrylate) hydrophobic modified multiwalled carbon nanotubes by
bilayer to a microporous support, J. Mater. Chem. A 3 (2015) 20118–20126. interfacial polymerization, J. Membr. Sci. 442 (2013) 18–26.
[135] H. Wang, T.S. Chung, Y.W. Tong, K. Jeyaseelan, A. Armugam, Z. Chen, M. Hong, [165] K. Goh, L. Setiawan, L. Wei, W. Jiang, R. Wang, Y. Chen, Fabrication of novel
W. Meier, Highly permeable and selective pore-spanning biomimetic membrane functionalized multi-walled carbon nanotube immobilized hollow fiber mem-
embedded with aquaporin Z, Small 8 (2012) 1185–1190. branes for enhanced performance in forward osmosis process, J. Membr. Sci. 446
[136] M. Wang, Z. Wang, X. Wang, S. Wang, W. Ding, C. Gao, Layer-by-layer assembly of (2013) 244–254.
aquaporin Z-incorporated biomimetic membranes for water purification, Environ. [166] L. Brunet, D.Y. Lyon, K. Zodrow, J.-C. Rouch, B. Caussat, P. Serp, J.-C. Remigy,
Sci. Technol. 49 (2015) 3761–3768. M.R. Wiesner, P.J. Alvarez, Properties of membranes containing semi-dispersed
[137] S. Qi, R. Wang, G.K.M. Chaitra, J. Torres, X. Hu, A.G. Fane, Aquaporin-based carbon nanotubes, Environ. Eng. Sci. 25 (2008) 565–576.
biomimetic reverse osmosis membranes: stability and long term performance, J. [167] A.S. Brady-Estévez, S. Kang, M. Elimelech, A single-walled-carbon-nanotube filter
Membr. Sci. 508 (2016) 94–103. for removal of viral and bacterial pathogens, Small 4 (2008) 481–484.
[138] X. Li, R. Wang, F. Wicaksana, C. Tang, J. Torres, A.G. Fane, Preparation of high [168] D. Cohen-Tanugi, J.C. Grossman, Mechanical strength of nanoporous graphene as
performance nanofiltration (NF) membranes incorporated with aquaporin Z, J. a desalination membrane, Nano Lett. 14 (2014) 6171–6178.
Membr. Sci. 450 (2013) 181–188. [169] M. Hu, B. Mi, Layer-by-layer assembly of graphene oxide membranes via elec-
[139] G. Sun, T.-S. Chung, K. Jeyaseelan, A. Armugam, Stabilization and immobilization trostatic interaction, J. Membr. Sci. 469 (2014) 80–87.
of aquaporin reconstituted lipid vesicles for water purification, Colloids Surf. B: [170] P. Sun, K. Wang, H. Zhu, Recent developments in graphene-based membranes:
Biointerfaces 102 (2013) 466–471. structure, mass-transport mechanism and potential applications, Adv. Mater. 12
[140] G. Sun, T.-S. Chung, K. Jeyaseelan, A. Armugam, A layer-by-layer self-assembly (2016) 2287–2310.
approach to developing an aquaporin-embedded mixed matrix membrane, RSC [171] R. Joshi, P. Carbone, F.-C. Wang, V.G. Kravets, Y. Su, I.V. Grigorieva, H. Wu,
Adv. 3 (2013) 473–481. A.K. Geim, R.R. Nair, Precise and ultrafast molecular sieving through graphene
[141] W. Xie, F. He, B. Wang, T.-S. Chung, K. Jeyaseelan, A. Armugam, Y.W. Tong, An oxide membranes, Science 343 (2014) 752–754.
aquaporin-based vesicle-embedded polymeric membrane for low energy water [172] L. Chen, G. Shi, J. Shen, B. Peng, B. Zhang, Y. Wang, F. Bian, J. Wang, D. Li,
filtration, J. Mater. Chem. A 1 (2013) 7592–7600. Z. Qian, G. Xu, G. Liu, J. Zeng, L. Zhang, Y. Yang, G. Zhou, M. Wu, W. Jin, J. Li,
[142] H.L. Wang, T.-S. Chung, Y.W. Tong, K. Jeyaseelan, A. Armugam, H.H.P. Duong, H. Fang, Ion sieving in graphene oxide membranes via cationic control of inter-
F. Fu, H. Seah, J. Yang, M. Hong, Mechanically robust and highly permeable layer spacing, Nature 550 (2017) 380.
AquaporinZ biomimetic membranes, J. Membr. Sci. 434 (2013) 130–136. [173] K. Xu, B. Feng, C. Zhou, A. Huang, Synthesis of highly stable graphene oxide
[143] G. Sun, T.-S. Chung, N. Chen, X. Lu, Q. Zhao, Highly permeable aquaporin-em- membranes on polydopamine functionalized supports for seawater desalination,
bedded biomimetic membranes featuring a magnetic-aided approach, RSC Adv. 3 Chem. Eng. Sci. 146 (2016) 159–165.
(2013) 9178–9184. [174] S.G. Kim, D.H. Hyeon, J.H. Chun, B.-H. Chun, S.H. Kim, Novel thin nanocomposite
[144] A. Alexiadis, S. Kassinos, Molecular simulation of water in carbon nanotubes, RO membranes for chlorine resistance, Desalin. Water Treat. 51 (2013)
Chem. Rev. 108 (2008) 5014–5034. 6338–6345.
[145] J.A. Thomas, A.J. McGaughey, O. Kuter-Arnebeck, Pressure-driven water flow [175] Y. Zhang, Y. Su, J. Peng, X. Zhao, J. Liu, J. Zhao, Z. Jiang, Composite nanofil-
through carbon nanotubes: insights from molecular dynamics simulation, Int. J. tration membranes prepared by interfacial polymerization with natural material
Therm. Sci. 49 (2010) 281–289. tannic acid and trimesoyl chloride, J. Membr. Sci. 429 (2013) 235–242.
[146] W.-F. Chan, E. Marand, S.M. Martin, Novel zwitterion functionalized carbon na- [176] H.-R. Chae, J. Lee, C.-H. Lee, I.-C. Kim, P.-K. Park, Graphene oxide-embedded thin-
notube nanocomposite membranes for improved RO performance and surface film composite reverse osmosis membrane with high flux, anti-biofouling, and
anti-biofouling resistance, J. Membr. Sci. 509 (2016) 125–137. chlorine resistance, J. Membr. Sci. 483 (2015) 128–135.
[147] H. Wu, B. Tang, P. Wu, Optimization, characterization and nanofiltration prop- [177] H. Guo, Y. Deng, Z. Tao, Z. Yao, J. Wang, C. Lin, T. Zhang, B. Zhu, C.Y. Tang, Does
erties test of MWNTs/polyester thin film nanocomposite membrane, J. Membr. hydrophilic Polydopamine coating enhance membrane rejection of hydrophobic
Sci. 428 (2013) 425–433. endocrine-disrupting compounds? Environ. Sci. Technol. Lett. 3 (2016) 332–338.
[148] O.-k. Park, N.H. Kim, K.-t. Lau, J.H. Lee, Effect of surface treatment with po- [178] J.Y.X.F. Wang, B.L. Deng, Ultrafiltration of Surface Water by Poly (vinylidene
tassium persulfate on dispersion stability of multi-walled carbon nanotubes, fluoride)(PVDF)/TiO2 Mixed Matrix Hollow Fiber Membranes (HFMs) With
Mater. Lett. 64 (2010) 718–721. Advanced Antifouling Properties Under Visible Light Irradiation, University of
[149] K. Balasubramanian, M. Burghard, Chemically functionalized carbon nanotubes, Missouri–Columbia, 2014.
Small 1 (2005) 180–192. [179] M.J. Park, S. Phuntsho, T. He, G.M. Nisola, L.D. Tijing, X.-M. Li, G. Chen, W.-
[150] Z. Wang, J. Ma, C.Y. Tang, K. Kimura, Q. Wang, X. Han, Membrane cleaning in J. Chung, H.K. Shon, Graphene oxide incorporated polysulfone substrate for the
membrane bioreactors: a review, J. Membr. Sci. 468 (2014) 276–307. fabrication of flat-sheet thin-film composite forward osmosis membranes, J.
[151] W. Duan, A. Ronen, S. Walker, D. Jassby, Polyaniline-coated carbon nanotube Membr. Sci. 493 (2015) 496–507.
ultrafiltration membranes: enhanced anodic stability for in situ cleaning and [180] B. Ganesh, A.M. Isloor, A.F. Ismail, Enhanced hydrophilicity and salt rejection
electro-oxidation processes, ACS Appl. Mater. Interfaces 8 (2016) 22574–22584. study of graphene oxide-polysulfone mixed matrix membrane, Desalination 313
[152] C.-F. de Lannoy, D. Jassby, D. Davis, M. Wiesner, A highly electrically conductive (2013) 199–207.
polymer–multiwalled carbon nanotube nanocomposite membrane, J. Membr. Sci. [181] S. Zinadini, A.A. Zinatizadeh, M. Rahimi, V. Vatanpour, H. Zangeneh, Preparation
415 (2012) 718–724. of a novel antifouling mixed matrix PES membrane by embedding graphene oxide
[153] C.-F.O. de Lannoy, D. Jassby, K. Gloe, A.D. Gordon, M.R. Wiesner, Aquatic bio- nanoplates, J. Membr. Sci. 453 (2014) 292–301.
fouling prevention by electrically charged nanocomposite polymer thin film [182] C. Zhao, X. Xu, J. Chen, F. Yang, Effect of graphene oxide concentration on the
membranes, Environ. Sci. Technol. 47 (2013) 2760–2768. morphologies and antifouling properties of PVDF ultrafiltration membranes, J.
[154] J.-H. Choi, J. Jegal, W.-N. Kim, Fabrication and characterization of multi-walled Environ. Chem. Eng. 1 (2013) 349–354.
carbon nanotubes/polymer blend membranes, J. Membr. Sci. 284 (2006) [183] C. Zhao, X. Xu, J. Chen, F. Yang, Optimization of preparation conditions of poly
406–415. (vinylidene fluoride)/graphene oxide microfiltration membranes by the Taguchi
[155] P. Shah, C. Murthy, Studies on the porosity control of MWCNT/polysulfone experimental design, Desalination 334 (2014) 17–22.
composite membrane and its effect on metal removal, J. Membr. Sci. 437 (2013) [184] J. Zhang, Z. Xu, M. Shan, B. Zhou, Y. Li, B. Li, J. Niu, X. Qian, Synergetic effects of
90–98. oxidized carbon nanotubes and graphene oxide on fouling control and anti-fouling
[156] P. Daraei, S.S. Madaeni, N. Ghaemi, M.A. Khadivi, B. Astinchap, R. Moradian, mechanism of polyvinylidene fluoride ultrafiltration membranes, J. Membr. Sci.
Enhancing antifouling capability of PES membrane via mixing with various types 448 (2013) 81–92.
of polymer modified multi-walled carbon nanotube, J. Membr. Sci. 444 (2013) [185] L. Yu, Y. Zhang, B. Zhang, J. Liu, H. Zhang, C. Song, Preparation and character-
184–191. ization of HPEI-GO/PES ultrafiltration membrane with antifouling and anti-
[157] Y. Wang, R. Ou, Q. Ge, H. Wang, T. Xu, Preparation of polyethersulfone/carbon bacterial properties, J. Membr. Sci. 447 (2013) 452–462.
nanotube substrate for high-performance forward osmosis membrane, [186] Z. Xu, J. Zhang, M. Shan, Y. Li, B. Li, J. Niu, B. Zhou, X. Qian, Organosilane-
Desalination 330 (2013) 70–78. functionalized graphene oxide for enhanced antifouling and mechanical properties
[158] A. Srivastava, O. Srivastava, S. Talapatra, R. Vajtai, P. Ajayan, Carbon nanotube of polyvinylidene fluoride ultrafiltration membranes, J. Membr. Sci. 458 (2014)
filters, Nat. Mater. 3 (2004) 610–614. 1–13.
[159] J.K. Holt, A. Noy, T. Huser, D. Eaglesham, O. Bakajin, Fabrication of a carbon [187] H. Zhao, L. Wu, Z. Zhou, L. Zhang, H. Chen, Improving the antifouling property of
nanotube-embedded silicon nitride membrane for studies of nanometer-scale mass polysulfone ultrafiltration membrane by incorporation of isocyanate-treated gra-
transport, Nano Lett. 4 (2004) 2245–2250. phene oxide, Phys. Chem. Chem. Phys. 15 (2013) 9084–9092.
[160] C. Lee, S. Baik, Vertically-aligned carbon nano-tube membrane filters with su- [188] N. Wang, S. Ji, G. Zhang, J. Li, L. Wang, Self-assembly of graphene oxide and
perhydrophobicity and superoleophilicity, Carbon 48 (2010) 2192–2197. polyelectrolyte complex nanohybrid membranes for nanofiltration and perva-
[161] F. Du, L. Qu, Z. Xia, L. Feng, L. Dai, Membranes of vertically aligned superlong poration, Chem. Eng. J. 213 (2012) 318–329.
carbon nanotubes, Langmuir 27 (2011) 8437–8443. [189] Y. Wang, R. Ou, H. Wang, T. Xu, Graphene oxide modified graphitic carbon nitride
[162] K.-J. Lee, H.-D. Park, The most densified vertically-aligned carbon nanotube as a modifier for thin film composite forward osmosis membrane, J. Membr. Sci.
membranes and their normalized water permeability and high pressure durability, 475 (2015) 281–289.
J. Membr. Sci. 501 (2016) 144–151. [190] M.L. Lind, A.K. Ghosh, A. Jawor, X. Huang, W. Hou, Y. Yang, E.M. Hoek, Influence
[163] M. Amini, M. Jahanshahi, A. Rahimpour, Synthesis of novel thin film nano- of zeolite crystal size on zeolite-polyamide thin film nanocomposite membranes,
composite (TFN) forward osmosis membranes using functionalized multi-walled Langmuir 25 (2009) 10139–10145.
58
[191] H. Wu, B. Tang, P. Wu, MWNTs/polyester thin film nanocomposite membrane: an B. Chu, Thin-film nanofibrous composite ultrafiltration membranes based on
approach to overcome the trade-off effect between permeability and selectivity, J. polyvinyl alcohol barrier layer containing directional water channels, Ind. Eng.
Phys. Chem. C 114 (2010) 16395–16400. Chem. Res. 49 (2010) 11978–11984.
[192] H. Huang, X. Qu, H. Dong, L. Zhang, H. Chen, Role of NaA zeolites in the inter- [195] M.L. Lind, D. Eumine Suk, T.-V. Nguyen, E.M. Hoek, Tailoring the structure of thin
facial polymerization process towards a polyamide nanocomposite reverse osmosis film nanocomposite membranes to achieve seawater RO membrane performance,
membrane, RSC Adv. 3 (2013) 8203–8207. Environ. Sci. Technol. 44 (2010) 8230–8235.
[193] Z. Yang, Y. Wu, H. Guo, X.-H. Ma, C.-E. Lin, Y. Zhou, B. Cao, B.-K. Zhu, K. Shih, [196] M. Chen, X. Qin, G. Zeng, Biodegradation of carbon nanotubes, graphene, and
C.Y. Tang, A novel thin-film nano-templated composite membrane with in situ their derivatives, Trends Biotechnol. 35 (2017) 836–846.
silver nanoparticles loading: separation performance enhancement and implica- [197] M.M. Pendergast, E.M. Hoek, A review of water treatment membrane nano-
tions, J. Membr. Sci. 544 (2017) 351–358. technologies, Energy Environ. Sci. 4 (2011) 1946–1971.
[194] H. Ma, K. Yoon, L. Rong, M. Shokralla, A. Kopot, X. Wang, D. Fang, B.S. Hsiao,
59

Recent Development of Novel Membranes For Desalination 2018 Desalination

Uploaded by

Copyright:

Available Formats

Recent Development of Novel Membranes For Desalination 2018 Desalination

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Recent Development of Novel Membranes For Desalination 2018 Desalination

Uploaded by

Copyright:

Available Formats

What are some of the challenges facing current reverse osmosis membranes?

What are some of the challenges facing current reverse osmosis membranes?

What are some promising novel materials that have been explored for developing new desalination membranes?

What are some promising novel materials that have been explored for developing new desalination membranes?

Desalination 434 (2018) 37–59

Contents lists available at ScienceDirect

Recent development of novel membranes for desalination T

1. Introduction suggests that ultra-permeable membranes (UPMs) with tripling water

containing AqpZ had water permeability as high as 3.2. Carbon-based materials

schematic diagram of a single CNT, a scanning electron microscopic

[106]. Due to its unique structural, thermal and electrical properties,

channel size depending on the degree of

Water permeability is at least 1010 times

Channels formed by adjacent GO layers,

permeability and rejection are sensitive

from weight-loss measurements by a 1-

Antiadhesion (due to hydrophilicity)

faster than that of helium (obtained

to the interlayer spacing

and antimicrobial [59]

2-D carbon material

and [54,74]). These nano-sized pores enable ultra-fast convective water

100% KCl rejection at 40 °C for a

5-μm-diameter sample (obtained

~ 3.6 × 106 L m− 2 h− 1; nearly

from gravity-driven test in an

pores can have great impact on their separation performance based on

Not reported in literature

also be functionalized with hydrogen and hydroxyl groups to further

enhance the selectivity of NPG by charge repulsion (e.g., a nearly

complete salt rejection was observed for hydroxylated pores with

lination membranes are presented in Section 6.1.

Antimicrobial [57] (and improved hydrophilicity for

ﬂux of an aqueous suspension of gold nanoparticles;

3.2.3. GO-based framework (GOF)

higher than the calculated results from continuum

In addition to generating nanopores in graphene, many recent stu-

dies explore the use of derivatives of graphene, particularly graphene

sulting 2D GO nanosheets are composed of elements of C, O and H, with

and carboxyl groups) to make it more water attractive.

to synthesize high performance desalination membranes with ultra-fast

phene oxide framework (GOF) for molecular separation, where water/

solvent molecules transport through the nano-channels formed by ad-

of AQP-containing vesicles) [8]

100% NaCl rejection (obtained

speciﬁc water transport mechanism of the GOF is depicted as follows:

Not reported in literature

nanosheets, followed by their rapid diﬀusion along the water channel

paration performance. An MD simulation study of GOF [115] suggested

that the water permeability can be adjusted from ~ 208 to

spacing and GOF thickness. However, at an interlayer spacing of 15 Å

16,640 L m− 2 h− 1 bar− 1), the NaCl rejection is only ~ 50%. In addi-

tion, GO-based materials have well-known antimicrobial properties,

for biofouling control [59]. The hydrophilic nature of GO also allow it

hesion properties to membranes. The detailed membrane fabrication

methods of GO-based desalination membranes can be found in Section

3.3. A general framework for novel desalination membranes

Novel desalination membranes prepared by the aforementioned

primary rejection layer, additives for ﬂux enhancement, or modiﬁers