Forward Osmosis: Where Are We Now?

DES-12318; No of Pages 14
Desalination xxx (2014) xxxxxx
Contents lists available at ScienceDirect
Desalination
journal homepage: www.elsevier.com/locate/desal
Forward osmosis: Where are we now?

Devin L. Shaffer 1, Jay R. Werber 1, Humberto Jaramillo, Shihong Lin, Menachem Elimelech
Department of Chemical and Environmental Engineering, Yale University, New Haven, CT 06520-8286, USA
H I G H L I G H T S
We
We
We
We
We
present a critical review of the current state of forward osmosis (FO).

analyze the energy efciency of FO and emphasize relevant applications.
discuss the key required membrane properties for FO and future implications.
highlight fouling reversibility of FO and relevant benets and applications.
discuss applications where FO outperforms current technologies.
a r t i c l e
i n f o
Article history:
Received 15 September 2014
Received in revised form 17 October 2014
Accepted 21 October 2014
Available online xxxx
Keywords:
Forward osmosis
Desalination
Energy consumption
Fouling
Thin-lm composite
High-salinity brines
a b s t r a c t
Forward osmosis (FO) has been extensively investigated in the past decade. Despite signicant advancements in
our understanding of the FO process, questions and challenges remain regarding the energy efciency and current state of the technology. Here, we critically review several key aspects of the FO process, focusing on energy
efciency, membrane properties, draw solutes, fouling reversibility, and effective applications of this emerging
technology. We analyze the energy efciency of the process, disprove the common misguided notion that FO is
a low energy process, and highlight the potential use of low-cost energy sources. We address the key necessary
membrane properties for FO, stressing the importance of the structural parameter, reverse solute ux selectivity,
and the constraints imposed by the permeabilityselectivity tradeoff. We then dispel the notion that draw solution regeneration can use negligible energy, highlighting the benecial qualities of small inorganic and
thermolytic salts as draw solutes. We further discuss the fouling propensity of FO, emphasizing the fouling reversibility of FO compared to reverse osmosis (RO) and the prospects of FO in treating high fouling potential
feed waters. Lastly, we discuss applications where FO outperforms other desalination technologies and emphasize that the FO process is not intended to replace RO, but rather is to be used to process feed waters that cannot
be treated by RO.
2014 Elsevier B.V. All rights reserved.
Contents
1.
2.
3.
4.
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Is forward osmosis a low energy process? . . . . . . . . . . . . . . . . . . . . .
2.1.
No free lunch for FO . . . . . . . . . . . . . . . . . . . . . . . . . .
2.2.
FORO consumes more electric energy than RO alone . . . . . . . . . . . .
2.3.
Opportunities in leveraging low-cost thermal energy . . . . . . . . . . . .
What are the key required membrane properties for FO? . . . . . . . . . . . . . .
3.1.
Low structural parameter to maximize water ux . . . . . . . . . . . . . .
3.2.
High reverse solute ux selectivity (RSFS) to minimize loss of draw solute . . .
Is nding a magic draw solution the Holy Grail in FO? . . . . . . . . . . . . . . .
4.1.
The choice of draw solute cannot decrease the minimum energy of separation .
4.2.
Simple dissolved salts remain the optimal draw solutes . . . . . . . . . . .
4.3.
Some opportunity remains for draw solute innovation . . . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Corresponding author.
E-mail address: menachem.elimelech@yale.edu (M. Elimelech).
1
These authors contributed equally to this work.
http://dx.doi.org/10.1016/j.desal.2014.10.031
0011-9164/ 2014 Elsevier B.V. All rights reserved.
Please cite this article as: D.L. Shaffer, et al., Forward osmosis: Where are we now?, Desalination (2014), http://dx.doi.org/10.1016/
j.desal.2014.10.031
0
0
0
0
0
0
0
0
0
0
0
0
D.L. Shaffer et al. / Desalination xxx (2014) xxxxxx
5.
Is FO a low fouling process? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

5.1.
FO fouling is highly reversible . . . . . . . . . . . . . . . . . . . . . . . . . .
5.2.
Membrane surface properties are important for fouling control . . . . . . . . . . .
6.
Where does FO outperform other desalination technologies? . . . . . . . . . . . . . . .
6.1.
FO excels with challenging feed waters . . . . . . . . . . . . . . . . . . . . . .
6.2.
Hybrid FO systems can desalinate high-salinity feed waters using less energy. . . . .
6.3.
FO pretreatment can improve the performance of conventional desalination processes
6.4.
FO pretreatment can improve the permeate quality produced by reverse osmosis . . .
6.5.
Osmotic dilution applications of FO are truly low energy . . . . . . . . . . . . . .
7.
Concluding remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1. Introduction
Forward osmosis (FO), an emerging separation/desalination process,
has received increased attention in the past decade in both academic research and industrial development [1,2]. In FO, a semipermeable membrane is placed between two solutions of different concentrations: a
concentrated draw solution and a more dilute feed solution. By using
the osmotic pressure difference to drive the permeation of water across
the membrane, FO can address several shortcomings of hydraulic
pressure-driven membrane processes, such as reverse osmosis (RO).
Early studies focused on various potential applications of FO in the
food, water, and energy sectors [1]. The introduction of the ammonia
carbon dioxide FO process in 2005 as a potential desalination process
that utilizes low-grade thermal energy [3] has stimulated academic
and industrial interest in FO, which resulted in a dramatic increase in
the number of research articles and patents in subsequent years [2,4].
These studies on FO involved membrane development [58], mass
transfer analysis [4,9], membrane characterization [10,11], fouling phenomena [1216], and introduction and characterization of new draw
solutions [1720]. Concurrently, conceptual and bench-scale studies
on various potential applications of FO have been published, including
the use of FO coupled with a draw solution separation/regeneration
stage [21,22], FO in osmotic dilution processes [23,24], and various hybrid systems incorporating FO [2528].
Despite the recent advancements in FO, there remain several challenges to overcome for successful implementation of the technology.
Moreover, confusion exists regarding the energy consumption by the
FO process, triggered by misguided studies dening FO as a low energy
process. Other studies present FO as an alternative to RO, a robust
pressure-driven membrane desalination process. While several review
articles on FO have been published recently [2,4], none has critically addressed the energy efciency of FO, the viability of the technology, and
the applications where FO has clear advantages over conventional desalination processes. A review that analyzes these key points as well as
other enabling aspects of FO is crucially needed.
In this review article, we critically discuss the energy efciency,
membrane performance, optimal draw solutes, and suitable applications of FO. Specically, we address the following key questions: Is FO
a low energy process? What are the key required membrane properties
for FO? Is nding a magic draw solution the Holy Grail in FO? Is FO a low
fouling process? Where does FO outperform other desalination technologies? Addressing these questions and understanding the limits of FO
will provide vital information to further advance the technology and expand the range of its applications.
2. Is forward osmosis a low energy process?
2.1. No free lunch for FO
The fact that water in FO permeates spontaneously through a semipermeable membrane does not mean that FO is more energy efcient as
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
0
0
0
0
0
0
0
0
0
0
0
0
a separation process than other membrane processes. In fact, FO is not

only a separation process, but is simultaneously a separation and mixing
process. The water molecules that transport across the membrane from
a salty feed solution mix with the draw solution to reduce its chemical
potential. In order to obtain fresh water as a product, further separation
of the diluted draw solution is required subsequent to the FO process.
Based on thermodynamic principles and practical kinetic requirements,
the theoretical minimal energy for desalination with FO is always higher
than that without FO. In other words, using FO cannot reduce the minimum energy of separation.
This general conclusion regarding the FO desalination energy has
fundamental underlying thermodynamic rationales. In an isothermal
separation process, energy is required to reduce the entropy of the system [29,30]. However, the spontaneity of the FO process implies that entropy is generated [31] and that the system entropy of the intermediate
state (i.e. when draw solution is diluted and feed solution is concentrated) is higher than that of the initial state. Regardless of the process used,
separation of the feed solution to the same degree should result in identical system entropy in the nal state. Therefore, the minimum energy
for the post-FO separation stage, which is required to reduce the system
entropy from the intermediate to the nal state, is obviously higher than
the minimum energy for a standalone separation, which is required to
reduce the system entropy only from the initial to the nal state. To further elucidate this qualitative argument, an analysis is conducted in
Section 2.2 to compare the minimum energy requirement of a
standalone RO process to that of an FORO hybrid process.
2.2. FORO consumes more electric energy than RO alone
For comparison of overall energy consumption between an RO process
and an FORO process, it is a reasonable approximation to consider only
the energy required for the RO separation. This approximation assumes
that the energy requirement in RO for generating the hydraulic pressure
to overcome the osmotic pressure difference between the concentrated
and dilute solutions dominates the overall energy consumption, rendering the energy for ow circulation and other practical considerations in
the RO and FO systems relatively insignicant [29]. Therefore, comparing
the energy consumption of a standalone RO system and a hybrid FORO
system can be approximately reduced to the comparison between the energy consumption of the RO stages in these respective systems.
We now compare the energy consumption between a standalone RO
process (denoted as RO1 in Fig. 1A) and an FORO process (denoted as
FORO2 in Fig. 1A). In the RO1 process, the feed solution is separated
into the brine solution (the red block in Fig. 1A) and the permeate solution (the blue block). In the FO process, water molecules in the feed solution migrate across the semipermeable membrane to mix with the
draw solution (the green block), resulting in the diluted draw solution
(the blue and green composite block) and the same brine solution as
in RO1. The diluted draw solution is then separated by the RO2 process
to produce a permeate solution of the same volume as in the RO1 process, and a concentrated draw solution of the same volume and
j.desal.2014.10.031
Fig. 1. (A) A schematic comparison of a reverse osmosis (RO) process (RO1) with a hybrid forward osmosis and reverse osmosis (FORO) process (FORO2). The blocks with letters B, P, and
D represent the brine, permeate, and draw solutions, respectively. The composite block with both B and P represents the feed solution for both systems, while the block with P and D indicates the diluted draw solution that needs further separation by RO2. The size of the blocks is proportional to the solution volume. Both the RO1 and the FORO2 systems lead to the exact
same separation of the feed solution with identical recovery. (B, C) Qualitative illustration of the osmotic pressures of the feed and draw solutions in FO in a co-current ow module (B) and
in a counter-current ow module (C).
concentration as the initial draw solution. For the RO1 and the FORO2
processes to achieve the same separation, the osmotic pressure of the
brine solution in the RO1 process, BRO1, and the osmotic pressure of the
brine solution in the FO process (i.e. concentrate of the FO feed solution), BFO,feed, have to be equal:
B
RO1 FO;feed :
In an RO process, the minimum specic energy (i.e., energy consumed per unit volume of product water) to achieve a certain recovery
is equal to the brine osmotic pressure, BRO [32]. In an ideal FORO system with perfect salt rejection, the draw solution is conserved, and
the brine of the RO2 process is simply the initial draw solution in the
FO process, 0FO,draw, such that:
B
RO2 FO;draw :
The condition for net driving force in FO dictates that the feed osmotic pressure must always be lower than the draw solution osmotic
pressure in either co-current (Fig. 1B) or counter-current (Fig. 1C)
membrane module conguration. The corollary inequality of the net
driving force condition, BFO,feed b 0FO,draw, when coupled with the conditions given by Eqs. (1) and (2), suggests that:
B
RO2 NRO1 :
The inequality in Eq. (3) indicates that the minimum specic energy for
RO2 is higher than that in RO1 in order for the FORO2 and the RO1 processes to achieve the same recovery. Therefore, a standalone RO process
is more energy efcient than an FORO hybrid process.
Although only the theoretical minimum energy of separation is compared in our analysis, incorporating other practical considerations, such
as concentration polarization and excess pressure, will unlikely change
the relative energetic advantage of RO over FORO. Our analysis, although highly simplied, leads to the same conclusion obtained from
a detailed comparison between an RO and an FORO process when considering other practical factors, such as energy for cross-ow circulation
and the energy requirement for ultraltration pretreatment when RO is
used without FO [33]. The itemized assessment of energy consumption
for a detailed comparison of RO and FORO processes illustrated that
the energy cost of RO dominates the total energy cost for both RO and
FORO processes, which justies the assumption in our simplied analysis. We also note that an intrinsic trade-off exists between the energy
efciency of an FORO process, which is quantied by specic energy,
and the mass transfer kinetics in FO, which can be characterized by average water ux and are controlled by the overall osmotic pressure driving force. The enhanced mass transfer kinetics in FO must be achieved at
the cost of higher energy consumption in the subsequent RO process.
2.3. Opportunities in leveraging low-cost thermal energy
Although incorporation of FO in a separation process cannot reduce
the minimum energy of separation compared to a standalone separation process (RO or other methods), FO hybrid systems may still provide
energy cost savings, if alternative low-cost thermal energy can be used
to power the post-FO separation process. For example, a thermolytic solution of ammoniacarbon dioxide has been proposed as a draw solution in FO, in which low-grade heat energy can be exploited to
regenerate the draw solution by preferentially removing the solutes
that are signicantly more volatile than water [3,34]. Alternatively, FO
can also be coupled with membrane distillation (MD) to desalinate waters that are challenging for a standalone MD process. In a hybrid FO
MD system, FO is applied to mitigate organic fouling and/or mineral
scaling that are detrimental to the MD process, whereas the MD process
separates the water and regenerates the draw solution using low-grade
heat [27]. Further discussion on FO hybrid processes that utilize lowcost thermal energy for separation is provided in Section 6.2.
j.desal.2014.10.031
3. What are the key required membrane properties for FO?

The ideal FO membrane achieves high water ux, completely rejects
feed and draw solutes, and is mechanically robust. Membrane properties for the support and active layers are critical for realizing these
aims. In this section, we highlight two particularly important membrane
design goals: (i) minimizing the structural parameter of the support
layer to mitigate mass transfer limitations and increase water ux, and
(ii) maximizing the reverse solute ux selectivity of the active layer to
limit loss of draw solute.
3.1. Low structural parameter to maximize water ux
The phenomenon of internal concentration polarization (ICP) can
have pronounced performance-limiting effects in FO [3537]. ICP arises
due to the porous support layer serving as an unstirred diffusive boundary layer [6]. As water permeates the membrane active layer from the
feed solution and dilutes the draw solution at the interface between
the active layer and the porous support, the driving force for permeation
decreases. This draw solution dilution is partially counteracted by draw
solute back-diffusion from the bulk draw solution into the support.
However, the result is a concentration prole in which the draw solute
concentration at the active layer surface is lower than in the bulk, reducing the osmotic pressure difference across the active layer and therefore
reducing water ux.
To minimize ICP and maximize water ux, resistance to draw solute
diffusion through the support layer must be minimized. FO membrane
support layer properties are crucial in this aim. The ideal support layer
is thin, highly porous, and minimally tortuous, all of which minimize
the diffusion path and enhance back-diffusion of draw solute. Membrane support chemistry, most notably hydrophilicity, can also play an
important role in the severity of ICP. Hydrophobic support materials,
such as the commonly-used polysulfone and polyester, do not fully
wet when exposed to water, decreasing the effective porosity of the
support layer [38]. These support layer properties, together with the solute diffusivity, comprise the resistance to solute diffusion, K, which can
be expressed as the reciprocal of a thin-lm mass transfer coefcient
[39]:
K
ts
D eff
where D is the solute diffusion coefcient, ts is the support layer thickness, is the support layer tortuosity, and eff is the effective support
layer porosity. The membrane properties can be grouped to dene the
structural parameter S [5,40]:
S KD
ts
:
eff
Since its introduction, the structural parameter has provided a simple

benchmark for the design and comparison of FO membranes.
Until recently, studies of FO predominantly utilized polyamidebased thin lm composite (TFC) RO membranes or the commercial
asymmetric cellulose triacetate (CTA) FO membrane by Hydration
Technology Innovations (HTI, Albany, OR) [38,41]. TFC membranes
are the gold standard for RO because of their high water permeability, high solute selectivity, and chemical and physical stability [5,
42,43]. However, use of TFC-RO membranes in the FO process resulted in very low water uxes due to their dense support layers and
thick fabric backing [5]. These extensive support layers are crucial
for maintaining mechanical integrity under the high hydraulic pressures used in RO, but the same support is unnecessary in FO and leads
to very large structural parameters of nearly 10,000 m [5]. With its
relatively low structural parameter of 500 m, the HTI-CTA FO membrane obtained the best performance in FO, despite having lower
Fig. 2. Forward osmosis (FO) water uxes (active layer facing feed solution) modeled
using Eq. (6) for commercial membranes with deionized water feed and 1 M sodium chloride draw solution. All A, B, and S values used for calculations are values published in peerreviewed journals and obtained at 2025 C using similar methods. TFC-RO, no PET
(Dow) refers to a commercial thin-lm composite reverse osmosis (TFC-RO) sample
(SW30-HR, Dow) with the polyester (PET) fabric removed. A feed-side mass transfer coefcient of 100 L m2 h1 and a diffusion coefcient of 1.48 109 m2 s1 [51] were used
for all calculations. For the blue curve, a B-value of 0.1 L m2 h1 was used.
Data from [5,911,4749].
permeability and selectivity than polyamide TFC-RO membranes [5,

911,41].
Recent research has led to dramatic decreases in the structural
parameter [5,6,8,4448]. A major advance was the formation of robust TFC membranes with phase inversion-formed supports that
were specically tailored to minimize the structural parameter.
Through proper choice of solvents for phase inversion and use of a
thin fabric support, hand-cast TFC membranes were fabricated with
an average S of 492 m, in comparison with 9583 m for a commercial TFC-RO membrane (SW30, Dow Chemical Company, Midland,
MI) [5]. Subsequent studies reported structural parameter values of
less than 100 m for asymmetric membranes formed through
phase inversion [8] and for TFC membranes formed through interfacial polymerization of a polyamide lm directly on an electrospun
polyethersulfone support [46]. However, the properties that contribute to a low structural parameter can also lead to mechanical fragility, and it remains to be seen whether ultra-low S-value membranes
will be robust enough to withstand the hydrodynamic stresses inherent in FO operation.
FO membrane research has rapidly translated into commerciallyavailable, high-ux, TFC-FO membrane modules (Oasys Water Inc., Boston, MA; HTI), most of which have structural parameters less than
500 m [5,911,4749]. FO membrane development and its impact on
water ux performance are demonstrated in Fig. 2, which models ux
based on published water permeability, A, salt permeability, B, and Svalues for commercial membranes considering a 1 M sodium chloride
draw solution and deionized water feed solution. The water ux
model is based on the lm theory [10]:

9
8
Jw S
Jw
>
>
>
>
=
< D exp D F exp k
f

Jw A
>
B
J
J S >
>
>
;
:1
exp w exp w
Jw
D
kf
where Jw is the volumetric water ux, D is the bulk draw solution osmotic pressure, F is the bulk feed solution osmotic pressure, and kf is
the feed-side mass transfer coefcient. For the conditions modeled in
Fig. 2, recently commercialized TFC-FO membranes can achieve FO
j.desal.2014.10.031
water uxes that are 1020 times higher than a TFC-RO membrane
and are twice the water ux of asymmetric CTA FO membranes.
However, despite the achievement of relatively high water uxes,
opportunities for improvement remain. With improved mechanical
integrity, practical FO membranes with ultra-low S-values may be
attained. In addition, all of the asymmetric CTA membranes and
some of the TFC-FO membranes have relatively low selectivity compared to state-of-the-art seawater TFC-RO membranes [42,50]. A
highly selective FO membrane with a low structural parameter has
yet to be developed.
3.2. High reverse solute ux selectivity (RSFS) to minimize loss of draw
solute
Loss of draw solute during FO operation occurs via reverse solute
ux, which refers to the back permeation of draw solutes from the
bulk draw solution through the membrane active layer and into the
feed. The ability of an FO membrane to minimize loss of draw solute is
captured by the reverse solute ux selectivity (RSFS). RSFS represents
the volume of water produced per mass of draw solute lost, and it follows a simple, experimentally-veried relationship based on the van't
Hoff equation [9,10]:
RSFS
Jw
A
nRg T
B
Js
where Js is reverse mass ux of draw solute, n is the number of dissolved

species created by the draw solute (e.g., 2 for NaCl), Rg is the ideal gas
constant, and T is the absolute temperature. The RSFS is independent
of support layer properties (i.e., structural parameter) and bulk draw solution concentration, and it is solely dependent on the permselectivity
of the active layer, namely A and B. To maximize RSFS, the A/B ratio
must be maximized.
Maximizing RSFS is critically important because the loss of draw
solute (i) is economically unfavorable, (ii) can lead to enhanced
fouling and scaling in the feed solution [52,53], and (iii) for some
draw solutes, can have negative impacts upon release to the environment (e.g., ammonia [54]). Using a highly selective membrane
would minimize the detrimental effects of draw solute loss and
would allow for the selection of small, fast-diffusing draw solutes,
such as sodium chloride, magnesium chloride, and ammoniacarbon dioxide [1,55]. As discussed further in Section 4, these solutes
are ideal for minimizing ICP and creating high osmotic pressures.
However, they are prone to relatively high reverse solute uxes
due to their small size [17,55]. Ammoniacarbon dioxide, which
holds great promise as a draw solute due to its ability to be thermally evaporated for facile draw solution regeneration [3,48], suffers
from particularly high membrane solute permeability. Ammonia
carbon dioxide specic reverse salt uxes are nearly an order of
magnitude higher than sodium chloride through asymmetric CTA
membranes and are three times higher than sodium chloride
through TFC-FO membranes [55,56]. Development of highly selective membranes could decrease reverse solute ux to low levels,
even for ammoniacarbon dioxide.
High A/B ratios also increase forward solute ux selectivity, thus decreasing forward permeation of undesired feed solutes. This is crucial in
all FO applications, including hybrid systems such as FORO and FOdistillation. In these systems, solutes that diffuse into the draw solution
can accumulate over time, eventually leading to precipitation in the
draw solution or solute permeation into the nal product water [55,
57]. Maintaining high selectivity in the rst barrier (FO) mitigates
these issues.
While achieving high A/B ratios is critical, the fabrication of highly
selective membranes is constrained by the water permeabilitysolute
selectivity tradeoff [58]. This tradeoff has been postulated to be an intrinsic property of water transport through polymeric membranes,
Fig. 3. Water and salt permeabilities for commercial and hand-cast thin-lm composite
polyamide membranes subjected to chlorine-alkaline modication to alter membrane
permeability. Red dashed line is Eq. (8), with = 0.37 107 cm4 s2, L = 150 nm,
and T = 298 K. Data from Yip and Elimelech [59].
including polyamide-based TFC membranes. Recent analyses led to

the following empirical relationship [58,59]:
2
Rg T
Mw
3
A
where L is the thickness of the active layer, is a tting parameter, and

Mw is the solute molecular weight. This empirical relationship matches
closely with the experimental data shown in Fig. 3 [59], and the cubic relationship of B with A immediately demonstrates that increasing water
permeability comes with rapidly decreasing selectivity. This relationship
provides tremendous insight for the fabrication of selective TFC membranes: to achieve TFC membranes with high RSFS, water permeability
must be sacriced. Ultimately, we anticipate that the best-performing
TFC-FO membrane will have modest water permeability, high selectivity, and an ultra-low structural parameter. Such a membrane would
achieve high water uxes with minimal loss of draw solute.
4. Is nding a magic draw solution the Holy Grail in FO?
Similar to FO membrane properties, the choice of draw solute can
have a large impact on the performance and viability of the FO process.
The ideal draw solute is inexpensive, stable, non-toxic, highly soluble,
and has a molecular size that is large enough to limit reverse draw solute ux through the FO membrane active layer, yet small enough to be
highly mobile and mitigate ICP [17,20]. In addition, draw solutes should
be able to generate high osmotic pressures and be completely regenerated using simple, efcient techniques. Due to these numerous and
sometimes conicting requirements and the large number of potential
solutes that could be used, draw solutes with strikingly different properties have been tested for FO [1720,6064]. It is clear that a focused and
updated discussion of desired draw solute properties is needed. To that
end, this section (i) dispels the notion that draw solutes can be regenerated using negligible energy, (ii) discusses the simple solutes that have
emerged as optimal FO draw solutes, and (iii) assesses where draw solute innovation would be most impactful.
4.1. The choice of draw solute cannot decrease the minimum energy of
separation
Recent research [1720,6064] has led to numerous ideas for draw
solutes, including polyelectrolytes [18], hydrophilic magnetic
j.desal.2014.10.031
Fig. 4. Relationship between osmotic pressure and viscosity for different draw solutions. The solid curves represent draw solutions made of simple small ions. The red open triangles represent a draw solution with polyacrylic acid (MW = 1800 Da) as the draw solute [18]. The black open hexagons indicate a draw solution with a hexavalent phosphazene salt, hexa(4ehtylcarboxylatophenoxyl)phosphazene, as the draw solute [64]. All other symbols are representative of draw solutions using different tertiary amines and dissolved carbon dioxide as
switchable polarity solvents for draw solutes [19,65]. The data for the solid curves were simulated using OLI Stream Analyzer (OLI Systems, Inc. Morris Plains, NJ) at a temperature of
25 C. Due to complexities in modeling the ammoniacarbon dioxide system (i.e., obtaining a continuous curve), aqueous solutions of ammonia carbamate were modeled to represent
the osmotic pressure and viscosity relationship of ammoniacarbon dioxide. All other data were collected from the corresponding literature. The vertical dotted line indicates the viscosity
of pure water at 25 C. The dash-dot lines show the representative osmotic pressures of three feed solutions: shale gas produced water (64.8 bar), seawater (27 bar), and brackish water
(7.7 bar).
nanoparticles [63], hydrogels [61,62], and switchable polarity solvents

[19,65]. A major driver in these studies is nding a draw solute that
can decrease the energy of draw solution regeneration. However, potential energy savings from novel draw solutes are limited. As demonstrated in Section 2.1, the energy requirement of an FO process (with
any draw solution regeneration scheme) cannot be less than the minimum energy of separation. In addition, the energy consumption for
RO in state-of-the-art processes already approaches this minimum energy [32]. Therefore, it is extremely unlikely that any draw solution regeneration technology will prove more energetically favorable than
RO, and when it can be applied, RO will remain the optimal draw solution regeneration technology. Novel draw solution regeneration technologies seeking gains in energy efciency must focus on high osmotic
pressure applications for which RO cannot be used.
A simple example to further illustrate energetic limitations during
draw solution regeneration is the use of polyelectrolytes as a draw solute. Polyelectrolyte draw solution regeneration was performed by ultraltration (UF) [18], which is typically a low-pressure and low-energy
process in comparison with RO. However, if the polyelectrolyte creates
osmotic pressure to drive FO, and is completely regenerated by UF,
then the osmotic pressure of the UF brine, BUF, must be equal to the
inlet osmotic pressure of the FO draw solution:
B
U F FO;draw :
Comparison of Eqs. (9) and (2) results in:

B
U F RO2 :
10
The UF minimum operating hydraulic pressure in an FOUF process is

the same as that for RO in an FORO process, contrary to the usual expectation of UF being low-pressure and low-energy. Thus, the minimum
energy requirements for FOUF and FORO are the same, and using a
large polyelectrolyte does not save any energy compared to FORO or
standalone RO. This conclusion can be generalized for the use of any
pressure-driven membrane process as the draw solution regeneration
step.
4.2. Simple dissolved salts remain the optimal draw solutes
Despite the extensive research efforts into novel draw solutes, simple dissolved inorganic and thermolytic salts (e.g., sodium chloride,
magnesium chloride, and ammoniacarbon dioxide) remain the most
widely-used, and arguably the most effective. These small inorganic solutes have two distinct and important advantages: (i) their ability to
generate high osmotic pressures at low solution viscosities, and (ii)
their high diffusivities that mitigate the ux-limiting effect of ICP.
High osmotic pressures are critical for FO, as the most impactful applications are for high salinity feeds for which RO cannot be used. In
general, the osmotic pressure of a solution, , can be related to the
mass concentration of the solute, c, by the following equation:

cRg T

1
2
A2 c A3 c
M
11
where M is the molecular weight of the solute, and A2 and A3 are the second and third viral coefcients that capture the solutesolvent interactions [66]. For estimating the order of magnitude, the osmotic pressure
is inversely proportional to the molecular weight of the solute. Thus, as
shown in Fig. 4, low molecular weight inorganic solutes easily generate
osmotic pressures exceeding 100 bar at viscosities near that of pure
water (i.e., 1 cP). In comparison, nanoparticles and polymer based
draw solutes, which are extremely large compared to dissolved salts,
are unable to reach an appreciable osmotic pressure without using
very high solute mass concentrations that lead to unacceptably high
j.desal.2014.10.031
Fig. 5. (A) Water ux as a function of diffusion coefcient for membranes with different structural parameters, S. The feed solution has an osmotic pressure, , of 27 bar, representing seawater, and the draw solution has an osmotic pressure of ~87 bar (i.e., 2.0 osmol L1 difference). (B) Water ux as a function of diffusion coefcient with different reference feed solutions,
including deionized water ( = 0 bar), brackish water ( = 7.7 bar), seawater ( = 27 bar), and shale gas produced water ( = 64.8 bar). The draw solution in each case is 2.0 osmol L1
more concentrated than the corresponding feed solution. The structural parameter is 250 m. In both subgures, the following assumptions were made for the calculations (Eq. (6)): the
water permeability of the membrane is 2 L m2 h1 bar1; the solute permeability is 0.1 L m2 h1; the feed-side mass transfer coefcient is 100 L m2 h1; and the temperature is 25 C.
The vertical dashed lines indicate the diffusion coefcients of different draw solutes. For simple ion draw solutes (i.e., NH3CO2, NaCl, and MgCl2), the mutual diffusion coefcients were
calculated using the self-diffusion coefcient of each ion simulated using OLI Stream Analyzer. For poly(acrylic acid) (PAA) and nanoparticles (NPs), the diffusion coefcients were calculated using hydrodynamic radii (RH) based on the StokesEinstein equation. The hydrodynamic radius of PAA (1800 Da) was obtained from Laguecir et al. [67].
viscosity. Low solution viscosity is strongly desired to minimize hydraulic pressure losses across the membrane channel.
The second advantage of small inorganic solutes lies in their relatively
high diffusivities, which decrease ICP and lead to higher water ux. The
importance of limiting ICP was discussed extensively in Section 3.1 as related to membrane properties. As can be seen from Eq. (6) and in Fig. 5,
the draw solute diffusivity is critical, having an exponential effect on the
water ux. Even relatively small variations among the diffusivities of inorganic salts can have pronounced effect on water ux [17]. Compared to
simple inorganic salts, some of the recently proposed nanoparticle and
polyelectrolyte draw solutes are orders of magnitude larger, with characteristic sizes varying from several nanometers to tens of nanometers [18,
63]. Fig. 5A illustrates that these extremely large draw solutes suffer from
signicant ICP and cannot generate high water uxes, even when using
membranes with low structural parameters. Thus, their application to
the treatment of realistic feed solutions, such as brackish water, seawater,
or shale gas produced water, is limited as shown in Fig. 5B.
When choosing draw solutes, there is a tradeoff between small solute size to limit ICP and large solute size to decrease reverse draw solute
ux. Indeed, one of the main advantages of polymer and nanoparticlebased draw solutions is essentially eliminating reverse draw solute
ux. However, with the recent commercialization of TFC-FO membranes, which have much higher RSFS than CTA membranes [10,60], reverse draw solute ux for small inorganic solutes has been greatly
reduced. This trend will continue with the expected commercialization
of more highly selective TFC-FO membranes, further decreasing the
need for exploring large draw solutes, and further promoting the use
of small, highly-mobile solutes.
draw solutes for niche applications. Some FO draw solutes are strategically chosen to serve as the nal product, most notably concentrated fertilizer for agricultural uses [68] and concentrated nutrients in commercial
hydration bags sold by HTI. Large draw solutes may nd application in
low osmotic pressure applications for which even minimal reverse solute
ux is unacceptable (e.g., liquid food processing and pharmaceutical concentration). However, for these low-pressure applications, conventional
pressure-driven membrane processes will likely prove much more efcient and outperform FO.
More signicant opportunity lies in innovative high osmotic pressure draw solutes, such as switchable polarity solutes [19,65], or novel
thermolytic draw solutes, such as trimethylamine (TMA)carbon dioxide [60]. Switchable polarity solutes considered thus far for FO have all
been tertiary amines that phase transition from water-miscible to
water-immiscible depending on the presence of carbon dioxide [19,
65]. Studies of switchable polarity solutes are still preliminary, and energy requirements (i.e. heat input) have yet to be analyzed or disclosed.
However, some of the solutes studied have demonstrated high osmotic
pressures at reasonable viscosities (Fig. 4), and the ability to separate
the draw solute from water in the liquid phase may prove benecial. Reduced reverse solute ux and draw solution regeneration energy are the
main advantages for using TMAcarbon dioxide as a draw solute compared to ammoniacarbon dioxide [60]. While their potential is considerable, switchable polarity solvents and TMA have toxicity concerns
that will likely limit their use to industrial applications [19,60].
5. Is FO a low fouling process?

5.1. FO fouling is highly reversible
4.3. Some opportunity remains for draw solute innovation

While inorganic and thermolytic salts should adequately serve most
FO applications, marginal benets may be obtained using specialized
Fouling involves the deposition and adsorption of feed water constituents, such as organic and inorganic compounds, colloidal particles,
and microbes, to the membrane surface. As water permeates the
j.desal.2014.10.031
Fig. 6. (A) Water ux decline during a forward osmosis (FO) fouling run with a model organic foulant (alginate) (red squares). Fouling solution (pH 5.8) consisted of 200 mg L1 of alginate, 50 mM sodium chloride, and 0.5 mM calcium chloride. Fouling run was conducted with a cellulose triacetate FO membrane (Hydration Technology Innovations) at an initial
water ux of 27 L m2 h1, with a cross-ow velocity of 8.5 cm s1, and at a temperature of 25.5 C. Cleaning was performed by rinsing with a 50 mM sodium chloride solution at
high cross-ow velocity (21 cm s1) for 15 min. Data from Mi and Elimelech [13]. (B) Flux recovery with different types of organic and inorganic foulants: alginate and bovine serum
albumin (BSA) as model organic foulants, and silica and gypsum as model scalants. Data sources: alginate fouling [13], silica scaling [73], and gypsum scaling [72]. BSA results are
based on unpublished data of Mi and Elimelech.
semipermeable membrane, foulants in the feed solution accumulate on

the membrane surface, forming a cake layer, which creates hydraulic resistance and cake-enhanced concentration polarization that reduce the
net driving force for permeation [52,69]. Fouling deteriorates membrane performance and decreases membrane lifespan, both of which increase operating costs for membrane desalination [50,70].
Recent studies indicate that FO is a low fouling process because the
foulant layer formed in FO is structurally different from fouling in
pressure-driven membrane processes [12,26,52,71]. The structure of
the cake layer formed during RO is densely compacted, whereas the
cake layer in FO is thicker but much less compact [52]. Because of the
loosely-packed fouling layer, FO can recover as much as 80100% of
the initial water ux through periodic rinses to clean the membrane surface [71]. In comparison, RO fouling is mostly irreversible without chemical cleaning [52].
Fig. 6A demonstrates how simple rinsing at a high cross-ow velocity with a low ionic strength solution accomplishes complete removal of
an alginate fouling layer from an FO membrane [13]. Reversibility is illustrated by the recovery of the water ux after rinsing (blue circles)
to the initial water ux prior to fouling (red squares). Such high reversibility is similarly observed with scaling by gypsum [72] and silica [73],
and fouling by proteins (Fig. 6B).
Membrane fouling in FO can also be mitigated by introducing hydrodynamic shear forces near the membrane surface to prevent foulant accumulation. Spacers placed alongside the membrane induce mixing and
have been shown to prevent the attachment of foulants [74,75]. Other
hydrodynamic strategies that mitigate fouling in FO include operating
at a higher cross-ow velocity [52], use of ow pulsation [74], and introducing air bubbles [13]. However, these hydrodynamic approaches will
increase the energy consumption of the system, which is a consideration for any fouling mitigation strategy.
High reversibility of fouling in FO enables the treatment of raw feed
waters with high fouling potential, such as those heavily laden with organic material (e.g., natural and efuent organic matter) [76]. Several
studies have demonstrated that a steady water ux is maintained during wastewater treatment [74,76] and sludge dewatering [77] by FO.
Nonetheless, fouling is undesirable as it decreases process performance
and increases operational costs. Therefore, understanding fouling
mechanisms and developing fouling mitigation methods continue to
be active areas of research [74].
5.2. Membrane surface properties are important for fouling control

Fouling propensity in FO is dictated by hydrodynamic operation conditions, as discussed in Section 5.1, and also by the afnity between
foulants and the membrane surface [32]. As such, a fouling-resistant
FO membrane is characterized by an inert surface chemistry to prevent
attracting foulants and a smooth surface topography that impedes
foulant entrapment.
Membranes utilized thus far in FO have predominantly been either
asymmetric cellulose triacetate (CTA) or polyamide thin lm composites
(TFC). Although organic fouling of FO has been a prevalent topic in recent
publications, the vast majority of the fouling studies were conducted with
CTA membranes [1214,26,52,7182] because no TFC-FO membranes
were commercially available until recently. As discussed in Section 3,
TFC membranes are the current state-of-the-art for FO because they exhibit a higher water permeability and salt rejection than CTA membranes
[5]. Thus, there is a critical need for systematic research on the fouling resistance of TFC-FO membranes, which are inherently prone to fouling because of their high surface roughness, relative hydrophobicity, and the
presence of carboxyl groups on the membrane surface [32].
In the few studies on FO fouling that have used TFC membranes, efforts have focused on surface modication of the membrane active layer
to mitigate their fouling propensity [15,8385]. A common method to
impart antifouling properties is by grafting hydrophilic polymers, such
as poly(ethylene glycol) (PEG), to the membrane active layer to create
a polymer brush, which acts as a steric barrier to the adsorption of
foulants [15,83,86]. These modied membranes exhibit a signicant decrease in water contact angle and reduced foulantmembrane adhesion
forces, indicating an increase in fouling resistance. However, this improvement in fouling resistance is achieved at the cost of reduced
water permeability that is attributed to the hydraulic resistance of the
grafted polymer layer [15,83].
In organic fouling tests, FO membranes modied with PEG demonstrated reduced ux decline due to organic fouling and were capable
of recovering 100% of the initial water ux after rinsing, as illustrated
in Fig. 7 [83]. As such, surface modication reduces the propensity for
membrane fouling and improves the reversibility of fouling, thereby enabling the treatment of high fouling potential feed waters by FO. A review on surface modications for antifouling membranes can be
found elsewhere [87].
j.desal.2014.10.031
Fig. 7. (A) Representative fouling curves for a control forward osmosis (FO) polyamide thin-lm composite (TFC) membrane and membrane modied in situ with poly(ethylene glycol).
Fouling conditions were as follows: synthetic secondary wastewater efuent solution (pH 7.4) supplemented with 250 mg L1 alginate as model organic foulant, initial permeate water
ux of ~20 L m2 h1, and cross-ow velocities of 8.5 cm s1 for the feed and draw solutions. After the fouling run, the system was cleaned by circulating 15 mM sodium chloride solution
for 15 min at a cross-ow velocity of 21.4 cm s1 through the feed and draw solution channels. (B) Water ux decline and water ux recovery results for triplicate FO organic fouling
experiments with control polyamide (black) and poly(ethylene glycol) in situ modied (blue) membranes. The percentage of water ux recovered after the physical cleaning step is
shown as blank columns.
Data from Lu et al. [83].
6. Where does FO outperform other desalination technologies?

6.1. FO excels with challenging feed waters
FO must be paired with a draw solution regeneration process for desalination applications, and the most studied of these hybrid FO desalination systems is FORO. As discussed in Section 2.2, a hybrid FORO
system will not consume less energy for separation than an RO system,
and FO will likely not replace RO as a standalone process for brackish
water or seawater desalination.
While FO is not more energetically favorable than RO for separation,
in applications with feed waters that are challenging to treat because of
high salinity, high fouling potential, or the presence of specic contaminants, FO can outperform or enhance RO and other desalination technologies. Hybrid FO-distillation systems may be used to treat highsalinity feed waters with osmotic pressures that exceed the working
pressure of RO. By employing thermolytic draw solutes, these hybrid
systems may use less energy for desalination than standalone thermal
technologies. FO can also improve the performance and water quality
of conventional membrane and thermal desalination processes when
used as pretreatment because of its reduced fouling propensity and ability to remove dissolved constituents. Osmotic dilution may be
employed in the treatment of impaired waters to reduce energy consumption and mitigate environmental impacts of treatment. For all FO
hybrid systems, the potential benets and expanded applications that
result from incorporating FO are achieved at the cost of the required
membrane area and associated materials and equipment for the FO
system.
6.2. Hybrid FO systems can desalinate high-salinity feed waters using less
energy
Hybrid FO systems will not consume less energy than the comparable
standalone desalination technology to achieve the same separation, and
this difference in energy consumption is especially evident when operating near the minimum specic energy of separation. However, hybrid FO
systems have the potential to consume less total energy than other technologies when desalinating high-salinity feed waters using draw solutions composed of thermolytic salts. To the extent that high-salinity
waters also have high organic fouling potential, FO hybrid systems are
expected to effectively treat these difcult feed waters because of the
fouling resistance of the FO process, as discussed in Section 5.
High-salinity feed waters include brines, such as concentrate
from seawater RO [88] or oil and gas produced water [57]. Other
high-salinity sources are feed waters that must be treated to very
high recovery and corresponding feed water salinity, such as desalination of landll leachate [89], industrial wastewater [90], and ue
gas desulfurization wastewater [91], for which the concentrate is
treated as hazardous waste and must be minimized. Brackish water
desalination in inland locations, where brine disposal options are
limited and costly, is also a source of high-salinity concentrate
streams [92]. Zero-liquid discharge treatment schemes are pursued
in situations where economic and regulatory limitations make the
disposal or discharge of high-salinity water infeasible.
Thermal desalination is required for separation of high-salinity feed
waters because the osmotic pressure of the feed water exceeds the operating pressure of RO, which is approximately 70 bar [42,50]. Membrane separation by electrodialysis is also not feasible because the
required energy for separation is proportional to the feed water salinity
and the required permeate water quality, making desalination of highsalinity solutions uneconomical [93]. Multi-effect distillation (MED),
multi-stage ash (MSF), and mechanical vapor compression (MVC)
are the technologies commonly used for desalinating high-salinity
feed waters. These distillation technologies heat the high-salinity feed
solution to vaporize the water, leaving behind salts and other nonvolatile feed water contaminants, and they operate far from the minimum specic energy referenced in Section 2.2.
Hybrid FO systems that utilize thermolytic draw solutions may be
energetically favorable to distillation technologies for treating highsalinity feed solutions because only relatively small volumes of
thermolytic draw solute must be vaporized and recovered from the
draw solution, as opposed to vaporizing and recovering the solvent
(water) in conventional distillation [57]. The energy required for draw
solute recovery is related to the draw solute vapor pressure, and draw
solutes with high vapor pressures require less total energy for recovery.
By using low-temperature distillation processes for thermolytic draw
solute recovery, hybrid FO systems can also leverage low-cost thermal
energy sources, such as solar thermal energy, geothermal energy, and
j.desal.2014.10.031
10
Fig. 8. Schematic of a zero-liquid discharge treatment system incorporating forward osmosis with a thermolytic draw solution. The forward osmosis process serves to further concentrate
the brine from the reverse osmosis process before the concentrated brine is treated in a crystallizer.
industrial waste heat, to reduce the energy cost of separation. FO combined with membrane distillation (FOMD) is one example [27,94].
In FO applications, thermolytic salts with vapor pressures much
higher than the vapor pressure of water are desirable as draw solutes
because they can be vaporized from the draw solution by a distillation
process operating at relatively low temperatures. The most studied
thermolytic draw solution is the ammoniacarbon dioxide system [3,
95]. In this draw solution, ammonium bicarbonate and ammonium hydroxide salts, which have low molecular weights and high solubility, are
dissolved in water to create a concentrated draw solution with high osmotic pressure [41]. Upon moderate heating, the ammonium salts are
decomposed into ammonia and carbon dioxide gases, which may be recovered and reconstituted into the draw solution [3,96].
Oasys Water Inc. recently demonstrated the use of a hybrid FO system with an ammoniacarbon dioxide draw solution and a distillation
column for draw solute recovery for desalination of high-salinity shale
gas produced water [48,97]. The specic energy consumption of the
Oasys hybrid FO system was reported to be signicantly lower than
other thermal distillation methods [48]. Oasys Water also recently announced plans to install a similar system for treatment and concentration of RO brine from treatment of ue gas desulfurization wastewater
from a coal-red power plant. The Oasys hybrid FO-distillation system
will be integrated into a zero-liquid discharge facility design as a brine
concentrator upstream of a crystallizer [98]. An example schematic of
a zero-liquid discharge treatment system incorporating a hybrid FO process with a thermolytic draw solution is shown in Fig. 8.
6.3. FO pretreatment can improve the performance of conventional desalination processes
Conventional desalination technologies include both membranebased separation processes, such as RO, nanoltration (NF), and
electrodialysis, and thermal desalination technologies like MED, MSF,

and MVC. Feed water pretreatment is critical to protect the physical
equipment of these conventional processes from damage by feed
water constituents and to facilitate their performance by maintaining
consistent quality of the pretreated feed water. Current pretreatment
technologies for desalination are designed to reduce the fouling potential of the feed water by removing natural organic matter and
suspended solids. However, pretreatment technologies typically are
not designed to remove dissolved solids [50].
Inorganic scaling in membrane and thermal desalination processes
caused by sparingly-soluble dissolved salts in the feed water limits the
operating conditions and the overall system recovery of these processes.
In membrane desalination processes, scaling limits the degree to which
the feed water may be concentrated (i.e. limits system recovery) [99]
and establishes corresponding limits on the system operating pressures
and uxes. In MED and MSF, scaling reduces heat transfer efciency, restricts operating temperatures, and lowers the system recovery
[100102].
In order to prevent the deleterious effects of scaling, FO can function as an advanced pretreatment technology that removes both
dissolved organic material and dissolved inorganic scalants from
the feed water, in addition to suspended constituents. When FO
pretreatment is used, the coupled conventional desalination process for draw solution recovery is only exposed to an engineered
draw solution that has negligible fouling and scaling potential,
such as a sodium chloride solution or an ammoniacarbon dioxide
solution. As discussed in Section 5.1, the reversibility of fouling in
FO indicates that under proper hydrodynamic conditions, FO membranes can maintain their ux and performance when in contact with
raw feed waters that have a high fouling potential. A schematic of FO
pretreatment for a conventional membrane or thermal desalination
process is shown in Fig. 9.
j.desal.2014.10.031
11
Fig. 9. Schematic of forward osmosis (FO) pretreatment for a conventional membrane (reverse osmosis (RO) and nanoltration (NF)) or thermal (multi-effect distillation (MED) and
multi-stage ash (MSF)) desalination process. Feed water foulants and scalants are excluded from the draw solution, enabling the conventional desalination process to operate at high
recovery for draw solution reconcentration.
Using FO as a pretreatment can enhance the performance of conventional desalination processes by removing sparingly-soluble scalants
from the feed water. Coupled desalination processes may achieve
higher system recoveries by operating at higher pressure or temperature without the risk of scaling. Process modeling of an FORO system
[103] and bench-scale testing of both FORO [104] and FONF [21] systems have demonstrated the practicability of implementing FO as a pretreatment process. The potential benets of FO pretreatment for
achieving higher system recoveries in membrane desalination were
demonstrated by pilot-scale studies of brackish water desalination
with ion exchange pretreatment [105] and seawater desalination with
NF pretreatment [106]. The advantage of FO pretreatment over ion exchange or NF, in addition to the lower fouling propensity, is the ability
to remove all dissolved species from the feed water, not just specic cations or anions.
Simulation of MSF and MED thermal processes for seawater desalination showed that FO pretreatment signicantly reduced the concentrations of multivalent ions in the pretreated feed water, thus
reducing the scaling tendencies in these processes and enabling them
to operate at higher temperatures and recovery rates [28,107]. Process
modeling of FO pretreatment for thermal desalination technologies
builds upon extensive modeling and pilot-scale studies of NF pretreatment for these technologies. Modeling results [108,109] and data from
pilot-scale testing [110,111] of NF pretreatment for MSF showed that
the scaling potential was reduced and higher operating temperatures
were possible. FO pretreatment rejects all dissolved constituents from
the feed water, which is an advantage over NF pretreatment. In addition,
NF increases operating costs [107,112] and fouling propensity because it
is a pressurized system.
6.4. FO pretreatment can improve the permeate quality produced by reverse
osmosis
used in FO and RO, is affected by characteristics of the solute, characteristics of the membrane, and complex interactions between the feed
water and membrane surface [113]. Two types of dissolved contaminants for which incomplete rejection by RO is a concern for potable
water production and water reuse are trace organic compounds
(TrOCs) and boron. TrOCs in the aquatic environment are derived
from human sources, and they have been the subject of regulatory interest and public concern [114,115]. Boron exists naturally in seawater as
uncharged boric acid, and its removal is important to protect human
health and to enable eventual reuse of the desalinated seawater for agricultural irrigation [42]. FO pretreatment provides an additional barrier
for dissolved constituents, reducing the concentrations of TrOCs and
boron in the product water compared to conventional RO desalination.
FO demonstrated higher rejection of representative TrOCs than RO
in bench-scale studies [116], and hybrid FORO systems have demonstrated excellent rejection of TrOCs (N99%) because of the dual
membrane barrier to these contaminants [117]. FO has also been demonstrated to more effectively reject boron than RO because of the inuence of reverse solute ux in FO [118]. Current practice is to employ
additional RO passes in series to achieve high boron rejection in seawater desalination [42]. However, process modeling of FO pretreatment for
seawater RO desalination indicates that boron removal could be
achieved more energy-efciently with an FORO system than a conventional two-pass RO process [119]. While the FORO system shows
promise for TrOC and boron removal, different rejection rates for
TrOCs or boron in the FO and RO processes may result in their accumulation in the draw solution of a closed-loop FORO system and may potentially degrade product water quality [119,120]. Adsorption by
activated carbon, ultraviolet light oxidation [27], and draw solution
bleeding [119] have been suggested as mechanisms to mitigate contaminant accumulation in the closed-loop draw solution.
6.5. Osmotic dilution applications of FO are truly low energy
The ability of FO pretreatment to remove dissolved constituents

from the feed water can also help conventional RO desalination meet
stringent product water quality requirements. The degree to which a
dissolved solute is rejected by desalination membranes, like those
Osmotic dilution is the concept of utilizing the salinity difference between two solutions to drive water permeation in FO, without a draw
solution recovery process [23]. In osmotic dilution, a dilute stream
j.desal.2014.10.031
12
becomes concentrated and a concentrated stream is diluted as permeation occurs across a semipermeable FO membrane. Both the diluted
and concentrated solutions are the products of osmotic dilution, and
no draw solution recovery is necessary. By eliminating the energyintensive draw solution recovery, osmotic dilution is truly a lowenergy FO process and can achieve the free lunch referenced in
Section 2.1.
Osmotic dilution is often conceived as the coupling of an impaired
feed water (e.g., wastewater efuent) with saline sources (e.g., brackish
or seawater) to concentrate the impaired feed water and dilute the saline
feed. Osmotic dilution can be potentially adapted at two stages in a conventional seawater desalination facility to (i) dilute the feed water to
the RO process to a lower salinity, resulting in reduced energy requirement, and (ii) dilute the concentrated brine prior to discharge to mitigate
detrimental impacts to the environment [23]. Osmotic dilution offers the
potential for energy and cost savings in an RO facility by lowering the operating hydraulic pressure. Environmental impacts may be diminished by
reducing electricity requirements, and also by discharging brines with
lower salinity to the aquatic ecosystem. Moreover, reducing the volume
of the impaired water offers an additional benet [24].
Although osmotic dilution has been demonstrated to be an effective
process at the bench-scale [23,74] and pilot-scale [23,76], osmotic dilution continues to face challenges before it is commercially implemented.
An osmotic dilution system requires a large membrane area due to the
relatively low osmotic pressure difference and resulting low water
ux, which corresponds to large footprint and capital costs [121]. Additionally, the use of an impaired water for osmotic dilution of the
feed water to a drinking-water treatment facility, such as a seawater
reverse osmosis facility, continues to be perceived negatively by consumers [24], limiting deployment of this practice.
7. Concluding remarks
Through simple thermodynamic arguments, we have shown that FO
cannot reduce the minimum energy required for desalination, regardless of the type of draw solution used. An FORO desalination system
cannot consume less energy than a standalone RO system to achieve a
certain recovery. However, FO hybrid systems can outperform other desalination technologies when they are applied to the desalination of
high-salinity feed waters using thermolytic draw solutions to consume
less total energy for separation.
Research into FO membrane fabrication has yielded commerciallyavailable, high water ux FO membranes with low structural parameters. While extensive research on FO continues to be conducted with
commercial asymmetric CTA membranes, successful applications of FO
will utilize TFC polyamide membranes specically engineered for FO
that demonstrate relatively high permeability and selectivity. High selectivity is crucially important for FO membranes for operational, economic, and environmental reasons. In particular, FO treatment of
high-salinity feed waters requires high osmotic pressure draw solutions
generated by low molecular weight draw solutes, further underscoring
the need for FO membranes with very high selectivity. Simple dissolved
inorganic and thermolytic salts remain the most effective draw solutes
because they generate high osmotic pressures to drive permeation
and maintain high diffusivities to mitigate ICP. The design of highly selective TFC membranes is constrained by the inherent permeabilityselectivity tradeoff, and we expect that the best-performing TFC-FO
membrane will have modest water permeability, high selectivity, inert
surface chemistry, and an ultra-low structural parameter.
One apparent advantage of the FO process is the relatively lower
fouling propensity and higher fouling reversibility compared to RO. A
thermodynamic energy analysis of the FORO process does not capture
this important benet of FO, which can potentially outweigh the energetic costs of using FO in hybrid systems and may enable these systems
to treat feed waters with high fouling potential. Discerning fouling
mechanisms in FO and developing fouling mitigation strategies will
allow FO to reach its full commercial potential in a variety of applications. Development of FO membranes should continue to be focused
on increasing fouling resistance. FO excels with challenging feed waters,
and its future applications will be in treating high-salinity or high fouling potential waters, or waters that contain specic contaminants like
boron or TrOCs. The unique advantages of FO that will drive these future
applications are its low fouling propensity, capacity for high osmotic
pressure driving forces that exceed the operating limits of RO, and
high rejection of feed water contaminants. Hybrid FO systems
employing thermolytic draw solutions, such as the ammoniacarbon dioxide system, can be used to desalinate high-salinity feed waters while
consuming less total energy than applicable thermal desalination technologies. FO may also be applied as advanced pretreatment to improve
water quality and enable higher system recovery for conventional desalination technologies by rejecting feed water foulants, scalants, and contaminants of concern.
Acknowledgments
Financial support from the Department of Defense through the Strategic Environmental Research and Development Program (SERDP, Project ER-2217) is gratefully acknowledged. We also acknowledge a
STAR Fellowship awarded by the US Environmental Protection Agency
to D.L.S. and Graduate Research Fellowships awarded by the US National
Science Foundation to J.R.W. and H.J.
References
[1] T.Y. Cath, A.E. Childress, M. Elimelech, Forward osmosis: principles, applications,
and recent developments, J. Membr. Sci. 281 (2006) 7087.
[2] S. Zhao, L. Zou, C.Y. Tang, D. Mulcahy, Recent developments in forward osmosis:
opportunities and challenges, J. Membr. Sci. 396 (2012) 121.
[3] J.R. McCutcheon, R.L. McGinnis, M. Elimelech, A novel ammoniacarbon dioxide
forward (direct) osmosis desalination process, Desalination 174 (2005) 111.
[4] C. Klaysom, T.Y. Cath, T. Depuydt, I.F. Vankelecom, Forward and pressure retarded
osmosis: potential solutions for global challenges in energy and water supply,
Chem. Soc. Rev. 42 (2013) 69596989.
[5] N.Y. Yip, A. Tiraferri, W.A. Phillip, J.D. Schiffman, M. Elimelech, High performance
thin-lm composite forward osmosis membrane, Environ. Sci. Technol. 44
(2010) 38123818.
[6] A. Tiraferri, N.Y. Yip, W.A. Phillip, J.D. Schiffman, M. Elimelech, Relating performance of thin-lm composite forward osmosis membranes to support layer formation and structure, J. Membr. Sci. 367 (2011) 340352.
[7] J. Wei, C. Qiu, C.Y. Tang, R. Wang, A.G. Fane, Synthesis and characterization of atsheet thin lm composite forward osmosis membranes, J. Membr. Sci. 372 (2011)
292302.
[8] S. Zhang, K.Y. Wang, T.-S. Chung, H. Chen, Y.C. Jean, G. Amy, Well-constructed cellulose acetate membranes for forward osmosis: minimized internal concentration
polarization with an ultra-thin selective layer, J. Membr. Sci. 360 (2010) 522535.
[9] W.A. Phillip, J.S. Yong, M. Elimelech, Reverse draw solute permeation in forward
osmosis: modeling and experiments, Environ. Sci. Technol. 44 (2010) 51705176.
[10] A. Tiraferri, N.Y. Yip, A.P. Straub, S. Romero-Vargas Castrillon, M. Elimelech, A
method for the simultaneous determination of transport and structural parameters of forward osmosis membranes, J. Membr. Sci. 444 (2013) 523538.
[11] T.Y. Cath, M. Elimelech, J.R. McCutcheon, R.L. McGinnis, A. Achilli, D. Anastasio, A.R.
Brady, A.E. Childress, I.V. Farr, N.T. Hancock, J. Lampi, L.D. Nghiem, M. Xie, N.Y. Yip,
Standard methodology for evaluating membrane performance in osmotically driven membrane processes, Desalination 312 (2013) 3138.
[12] B. Mi, M. Elimelech, Chemical and physical aspects of organic fouling of forward osmosis membranes, J. Membr. Sci. 320 (2008) 292302.
[13] B. Mi, M. Elimelech, Organic fouling of forward osmosis membranes: fouling reversibility and cleaning without chemical reagents, J. Membr. Sci. 348 (2010)
337345.
[14] Y. Liu, B. Mi, Combined fouling of forward osmosis membranes: synergistic foulant
interaction and direct observation of fouling layer formation, J. Membr. Sci.
407408 (2012) 136144.
[15] S. Romero-Vargas Castrilln, X. Lu, D.L. Shaffer, M. Elimelech, Amine enrichment
and poly(ethylene glycol) (PEG) surface modication of thin-lm composite forward osmosis membranes for organic fouling control, J. Membr. Sci. 450 (2014)
331339.
[16] Y. Wang, F. Wicaksana, C.Y. Tang, A.G. Fane, Direct microscopic observation of forward osmosis membrane fouling, Environ. Sci. Technol. 44 (2010) 71027109.
[17] A. Achilli, T.Y. Cath, A.E. Childress, Selection of inorganic-based draw solutions for
forward osmosis applications, J. Membr. Sci. 364 (2010) 233241.
[18] Q. Ge, J. Su, G.L. Amy, T.-S. Chung, Exploration of polyelectrolytes as draw solutes in
forward osmosis processes, Water Res. 46 (2012) 13181326.
j.desal.2014.10.031

[19] M.L. Stone, C. Rae, F.F. Stewart, A.D. Wilson, Switchable polarity solvents as draw
solutes for forward osmosis, Desalination 312 (2013) 124129.
[20] Q. Ge, M. Ling, T.-S. Chung, Draw solutions for forward osmosis processes: developments, challenges, and prospects for the future, J. Membr. Sci. 442 (2013)
225237.
[21] S. Zhao, L. Zou, D. Mulcahy, Brackish water desalination by a hybrid forward osmosisnanoltration system using divalent draw solute, Desalination 284 (2012)
175181.
[22] O.A. Bamaga, A. Yokochi, B. Zabara, A.S. Babaqi, Hybrid FO/RO desalination system:
preliminary assessment of osmotic energy recovery and designs of new FO membrane module congurations, Desalination 268 (2011) 163169.
[23] T.Y. Cath, N.T. Hancock, C.D. Lundin, C. Hoppe-Jones, J.E. Drewes, A multi-barrier
osmotic dilution process for simultaneous desalination and purication of impaired water, J. Membr. Sci. 362 (2010) 417426.
[24] L.A. Hoover, W.A. Phillip, A. Tiraferri, N.Y. Yip, M. Elimelech, Forward with osmosis:
emerging applications for greater sustainability, Environ. Sci. Technol. 45 (2011)
98249830.
[25] K.Y. Wang, M.M. Teoh, A. Nugroho, T.-S. Chung, Integrated forward osmosismembrane distillation (FOMD) hybrid system for the concentration of protein solutions, Chem. Eng. Sci. 66 (2011) 24212430.
[26] A. Achilli, T.Y. Cath, E.A. Marchand, A.E. Childress, The forward osmosis membrane
bioreactor: a low fouling alternative to MBR processes, Desalination 239 (2009)
1021.
[27] M. Xie, L.D. Nghiem, W.E. Price, M. Elimelech, A forward osmosismembrane distillation hybrid process for direct sewer mining: system performance and limitations,
Environ. Sci. Technol. 47 (2013) 1348613493.
[28] A. Altaee, A. Mabrouk, K. Bourouni, P. Palenzuela, Forward osmosis pretreatment of
seawater to thermal desalination: high temperature FO-MSF/MED hybrid system,
Desalination 339 (2014) 1825.
[29] R. Semiat, Energy issues in desalination processes, Environ. Sci. Technol. 42 (2008)
81938201.
[30] K.H. Mistry, R.K. McGovern, G.P. Thiel, E.K. Summers, S.M. Zubair, J.H. Lienhard, Entropy generation analysis of desalination technologies, Entropy 13 (2011)
18291864.
[31] S.I. Sandler, Chemical, Biochemical, and Engineering Thermodynamics, John Wiley
& Sons, Hoboken, NJ, 2006.
[32] M. Elimelech, W.A. Phillip, The future of seawater desalination: energy, technology,
and the environment, Science 333 (2011) 712717.
[33] R.K. McGovern, J.H. Lienhard V, On the potential of forward osmosis to energetically outperform reverse osmosis desalination, J. Membr. Sci. 469 (2014) 245250.
[34] R.L. McGinnis, J.R. McCutcheon, M. Elimelech, A novel ammoniacarbon dioxide
osmotic heat engine for power generation, J. Membr. Sci. 305 (2007) 1319.
[35] J.R. McCutcheon, M. Elimelech, Inuence of concentrative and dilutive internal
concentration polarization on ux behavior in forward osmosis, J. Membr. Sci.
284 (2006) 237247.
[36] J.R. McCutcheon, M. Elimelech, Modeling water ux in forward osmosis: implications for improved membrane design, AICHE J. 53 (2007) 17361744.
[37] S. Loeb, L. Titelman, E. Korngold, J. Freiman, Effect of porous support fabric on osmosis through a LoebSourirajan type asymmetric membrane, J. Membr. Sci. 129
(1997) 243249.
[38] J.R. McCutcheon, M. Elimelech, Inuence of membrane support layer hydrophobicity on water ux in osmotically driven membrane processes, J. Membr. Sci. 318
(2008) 458466.
[39] K.L. Lee, R.W. Baker, H.K. Lonsdale, Membranes for power generation by pressureretarded osmosis, J. Membr. Sci. 8 (1981) 141171.
[40] K. Gerstandt, K.V. Peinemann, S.E. Skilhagen, T. Thorsen, T. Holt, Membrane processes
in energy supply for an osmotic power plant, Desalination 224 (2008) 6470.
[41] J.R. McCutcheon, R.L. McGinnis, M. Elimelech, Desalination by ammoniacarbon dioxide forward osmosis: inuence of draw and feed solution concentrations on process performance, J. Membr. Sci. 278 (2006) 114123.
[42] C. Fritzmann, J. Lwenberg, T. Wintgens, T. Melin, State-of-the-art of reverse osmosis desalination, Desalination 216 (2007) 176.
[43] R.J. Petersen, Composite reverse osmosis and nanoltration membranes, J. Membr.
Sci. 83 (1993) 81150.
[44] N.-N. Bui, M.L. Lind, E.M.V. Hoek, J.R. McCutcheon, Electrospun nanober supported thin lm composite membranes for engineered osmosis, J. Membr. Sci. 385386
(2011) 1019.
[45] N.N. Bui, J.R. McCutcheon, Hydrophilic nanobers as new supports for thin lm
composite membranes for engineered osmosis, Environ. Sci. Technol. 47 (2013)
17611769.
[46] X. Song, Z. Liu, D.D. Sun, Nano gives the answer: breaking the bottleneck of internal
concentration polarization with a nanober composite forward osmosis membrane for a high water production rate, Adv. Mater. 23 (2011) 32563260.
[47] J. Ren, J.R. McCutcheon, A new commercial thin lm composite membrane for forward osmosis, Desalination 343 (2014) 187193.
[48] R.L. McGinnis, N.T. Hancock, M.S. Nowosielski-Slepowron, G.D. McGurgan, Pilot
demonstration of the NH3/CO2 forward osmosis desalination process on high salinity brines, Desalination 312 (2013) 6774.
[49] A.P. Straub, N.Y. Yip, M. Elimelech, Raising the bar: increased hydraulic pressure allows unprecedented high power densities in pressure-retarded osmosis, Environ.
Sci. Technol. Lett. 1 (2014) 5559.
[50] L.F. Greenlee, D.F. Lawler, B.D. Freeman, B. Marrot, P. Moulin, Reverse osmosis desalination: water sources, technology, and today's challenges, Water Res. 43
(2009) 23172348.
[51] V. Lobo, Mutual diffusion coefcients in aqueous electrolyte solutions (technical
report), Pure Appl. Chem. 65 (1993) 26132640.
13
[52] S. Lee, C. Boo, M. Elimelech, S. Hong, Comparison of fouling behavior in forward osmosis (FO) and reverse osmosis (RO), J. Membr. Sci. 365 (2010) 3439.
[53] S. Phuntsho, F. Lot, S. Hong, D.L. Shaffer, M. Elimelech, H.K. Shon, Membrane scaling and ux decline during fertiliser-drawn forward osmosis desalination of brackish groundwater, Water Res. 57 (2014) 172182.
[54] USEPA, Aquatic Life Ambient Water Quality Criteria for AmmoniaFreshwater,
United States Environmental Protection Agency, 2013. 242.
[55] N.T. Hancock, T.Y. Cath, Solute coupled diffusion in osmotically driven membrane
processes, Environ. Sci. Technol. 43 (2009) 67696775.
[56] J.S. Yong, Reverse Draw Solute Transport in Forward Osmosis Systems(Doctoral
dissertation) Chemical and Environmental Engineering, Yale University, 2012. 148.
[57] D.L. Shaffer, L.H. Arias Chavez, M. Ben-Sasson, S. Romero-Vargas Castrilln, N.Y. Yip,
M. Elimelech, Desalination and reuse of high-salinity shale gas produced water:
drivers, technologies, and future directions, Environ. Sci. Technol. 47 (2013)
95699583.
[58] G.M. Geise, H.B. Park, A.C. Sagle, B.D. Freeman, J.E. McGrath, Water permeability
and water/salt selectivity tradeoff in polymers for desalination, J. Membr. Sci.
369 (2011) 130138.
[59] N.Y. Yip, M. Elimelech, Performance limiting effects in power generation from salinity gradients by pressure retarded osmosis, Environ. Sci. Technol. 45 (2011)
1027310282.
[60] C. Boo, Y.F. Khalil, M. Elimelech, Performance evaluation of trimethylaminecarbon
dioxide thermolytic draw solution for engineered osmosis, J. Membr. Sci. 473
(2015) 302309.
[61] D. Li, X. Zhang, J. Yao, G.P. Simon, H. Wang, Stimuli-responsive polymer hydrogels
as a new class of draw agent for forward osmosis desalination, Chem. Commun. 47
(2011) 17101712.
[62] D. Li, X. Zhang, J. Yao, Y. Zeng, G.P. Simon, H. Wang, Composite polymer hydrogels
as draw agents in forward osmosis and solar dewatering, Soft Matter 7 (2011)
1004810056.
[63] M.M. Ling, K.Y. Wang, T.-S. Chung, Highly water-soluble magnetic nanoparticles as
novel draw solutes in forward osmosis for water reuse, Ind. Eng. Chem. Res. 49
(2010) 58695876.
[64] M.L. Stone, A.D. Wilson, M.K. Harrup, F.F. Stewart, An initial study of hexavalent
phosphazene salts as draw solutes in forward osmosis, Desalination 312 (2013)
130136.
[65] A.D. Wilson, F.F. Stewart, Structurefunction study of tertiary amines as switchable
polarity solvents, RSC Adv. 4 (2014) 1103911049.
[66] M. Rubinstein, R. Colby, Polymer Physics, University Press, Oxford, 2003.
[67] A. Laguecir, S. Ulrich, J. Labille, N. Fatin-Rouge, S. Stoll, J. Bufe, Size and pH effect
on electrical and conformational behavior of poly(acrylic acid): simulation and experiment, Eur. Polym. J. 42 (2006) 11351144.
[68] S. Phuntsho, H.K. Shon, S. Hong, S. Lee, S. Vigneswaran, A novel low energy fertilizer driven forward osmosis desalination for direct fertigation: evaluating the performance of fertilizer draw solutions, J. Membr. Sci. 375 (2011) 172181.
[69] E.M.V. Hoek, M. Elimelech, Cake-enhanced concentration polarization: a new fouling mechanism for salt-rejecting membranes, Environ. Sci. Technol. 37 (2003)
55815588.
[70] N. Misdan, W.J. Lau, A.F. Ismail, Seawater Reverse Osmosis (SWRO) desalination by
thin-lm composite membranecurrent development, challenges and future
prospects, Desalination 287 (2012) 228237.
[71] Y. Kim, M. Elimelech, H.K. Shon, S. Hong, Combined organic and colloidal fouling in
forward osmosis: fouling reversibility and the role of applied pressure, J. Membr.
Sci. 460 (2014) 206212.
[72] B. Mi, M. Elimelech, Gypsum scaling and cleaning in forward osmosis: measurements and mechanisms, Environ. Sci. Technol. 44 (2010) 20222028.
[73] B. Mi, M. Elimelech, Silica scaling and scaling reversibility in forward osmosis, Desalination 312 (2013) 7581.
[74] C. Boo, M. Elimelech, S. Hong, Fouling control in a forward osmosis process integrating seawater desalination and wastewater reclamation, J. Membr. Sci. 444
(2013) 148156.
[75] R. Valladares Linares, S.S. Bucs, Z. Li, M. AbuGhdeeb, G. Amy, J.S. Vrouwenvelder,
Impact of spacer thickness on biofouling in forward osmosis, Water Res. 57
(2014) 223233.
[76] N.T. Hancock, P. Xu, M.J. Roby, J.D. Gomez, T.Y. Cath, Towards direct potable reuse
with forward osmosis: technical assessment of long-term process performance at
the pilot scale, J. Membr. Sci. 445 (2013) 3446.
[77] N.C. Nguyen, S.-S. Chen, H.-Y. Yang, N.T. Hau, Application of forward osmosis on
dewatering of high nutrient sludge, Bioresour. Technol. 132 (2013) 224229.
[78] M.M. Motsa, B.B. Mamba, A. D'Haese, E.M.V. Hoek, A.R.D. Verliefde, Organic fouling
in forward osmosis membranes: the role of feed solution chemistry and membrane structural properties, J. Membr. Sci. 460 (2014) 99109.
[79] H. Yoon, Y. Baek, J. Yu, J. Yoon, Biofouling occurrence process and its control in the
forward osmosis, Desalination 325 (2013) 3036.
[80] Z.-Y. Li, V. Yangali-Quintanilla, R. Valladares-Linares, Q. Li, T. Zhan, G. Amy, Flux
patterns and membrane fouling propensity during desalination of seawater by forward osmosis, Water Res. 46 (2012) 195204.
[81] Q. She, X. Jin, Q. Li, C.Y. Tang, Relating reverse and forward solute diffusion to membrane fouling in osmotically driven membrane processes, Water Res. 46 (2012)
24782486.
[82] V. Parida, H.Y. Ng, Forward osmosis organic fouling: effects of organic loading, calcium and membrane orientation, Desalination 312 (2013) 8898.
[83] X. Lu, S. Romero-Vargas Castrilln, D.L. Shaffer, J. Ma, M. Elimelech, In situ surface
chemical modication of thin-lm composite forward osmosis membranes
for enhanced organic fouling resistance, Environ. Sci. Technol. 47 (2013)
1221912228.
j.desal.2014.10.031
14
[84] M. Ben-Sasson, X. Lu, E. Bar-Zeev, K.R. Zodrow, S. Nejati, G. Qi, E.P. Giannelis, M.
Elimelech, In situ formation of silver nanoparticles on thin-lm composite reverse
osmosis membranes for biofouling mitigation, Water Res. 62 (2014) 260270.
[85] K.R. Zodrow, M.E. Tousley, M. Elimelech, Mitigating biofouling on thin-lm composite polyamide membranes using a controlled-release platform, J. Membr. Sci.
453 (2014) 8491.
[86] M. Morra, On the molecular basis of fouling resistance, J. Biomater. Sci. Polym. Ed.
11 (2000) 547569.
[87] D. Rana, T. Matsuura, Surface modications for antifouling membranes, Chem. Rev.
110 (2010) 24482471.
[88] M. Kurihara, H. Yamamura, T. Nakanishi, High recovery/high pressure membranes
for brine conversion SWRO process development and its performance data, Desalination 125 (1999) 915.
[89] R. Rautenbach, T. Linn, L. Eilers, Treatment of severely contaminated waste water
by a combination of RO, high-pressure RO and NF potential and limits of the process, J. Membr. Sci. 174 (2000) 231241.
[90] B. Van der Bruggen, L. Lejon, C. Vandecasteele, Reuse, treatment, and discharge of
the concentrate of pressure-driven membrane processes, Environ. Sci. Technol.
37 (2003) 37333738.
[91] R.K. Srivastava, W. Jozewicz, Flue gas desulfurization: the state of the art, J. Air
Waste Manag. Assoc. 51 (2001) 16761688.
[92] D. Hasson, R. Segev, D. Lisitsin, B. Liberman, R. Semiat, High recovery brackish
water desalination process devoid of precipitation chemicals, Desalination 283
(2011) 8088.
[93] H. Strathmann, Electrodialysis, a mature technology with a multitude of new applications, Desalination 264 (2010) 268288.
[94] S. Lin, N.Y. Yip, T.Y. Cath, C.O. Osuji, M. Elimelech, Hybrid pressure retarded osmosismembrane distillation system for power generation from low-grade heat:
thermodynamic analysis and energy efciency, Environ. Sci. Technol. 48 (2014)
53065313.
[95] R.L. McGinnis, M. Elimelech, Energy requirements of ammoniacarbon dioxide forward osmosis desalination, Desalination 207 (2007) 370382.
[96] R.L. McGinnis, M. Elimelech, Global challenges in energy and water supply: the
promise of engineered osmosis, Environ. Sci. Technol. 42 (2008) 86258629.
[97] B.D. Coday, P. Xu, E.G. Beaudry, J. Herron, K. Lampi, N.T. Hancock, T.Y. Cath, The
sweet spot of forward osmosis: treatment of produced water, drilling wastewater,
and other complex and difcult liquid streams, Desalination 333 (2014) 2335.
[98] WDR, FO: RO's new best friend, Water Desalination Report, 50, 2014, p. 1.
[99] A. Antony, J.H. Low, S. Gray, A.E. Childress, P. Le-Clech, G. Leslie, Scale formation
and control in high pressure membrane water treatment systems: a review, J.
Membr. Sci. 383 (2011) 116.
[100] H.T. El-Dessouky, H.M. Ettouney, Fundamentals of Salt Water Desalination, Elsevier
Science Ltd, 2002.
[101] B. Van der Bruggen, C. Vandecasteele, Distillation vs. membrane ltration: overview of process evolutions in seawater desalination, Desalination 143 (2002)
207218.
[102] A.N.A. Mabrouk, H.E.-b.S. Fath, Techno-economic analysis of hybrid high performance MSF desalination plant with NF membrane, Desalin. Water Treat. 51
(2012) 844856.
[103] F. Zaviska, L. Zou, Using modelling approach to validate a bench scale forward osmosis pre-treatment process for desalination, Desalination 350 (2014) 113.
[104] O.A. Bamaga, A. Yokochi, E.G. Beaudry, Application of forward osmosis in pretreatment of seawater for small reverse osmosis desalination units, Desalin. Water
Treat. 5 (2009) 183191.
[105] S.G.J. Heijman, H. Guo, S. Li, J.C. van Dijk, L.P. Wessels, Zero liquid discharge: heading for 99% recovery in nanoltration and reverse osmosis, Desalination 236
(2009) 357362.
[106] W.L. Ang, A.W. Mohammad, N. Hilal, C.P. Leo, A review on the applicability of integrated/hybrid membrane processes in water treatment and desalination plants,
Desalination (2014) (In press).
[107] A. Altaee, A. Mabrouk, K. Bourouni, A novel forward osmosis membrane pretreatment of seawater for thermal desalination processes, Desalination 326 (2013)
1929.
[108] M.A.K. Al-So, A.M. Hassan, G.M. Mustafa, A.G.I. Dalvi, M.N.M. Kither,
Nanoltration as a means of achieving higher TBT of 120 C in MSF, Desalination
118 (1998) 123129.
[109] A.E. Al-Rawajfeh, Inuence of nanoltration pretreatment on scale deposition in
multi-stage ash thermal desalination plants, Therm. Sci. 15 (2011) 11.
[110] A.M. Hassan, M.A.K. Al-So, A.S. Al-Amoudi, A.T.M. Jamaluddin, A.M. Farooque, A.
Rowaili, A.G.I. Dalvi, N.M. Kither, G.M. Mustafa, I.A.R. Al-Tisan, A new approach to
membrane and thermal seawater desalination processes using nanoltration
membranes (part 1), Desalination 118 (1998) 3551.
[111] M.A.K. Al-So, A.M. Hassan, O.A. Hamed, A.G.I. Dalvi, M.N.M. Kither, G.M. Mustafa,
K. Bamardouf, Optimization of hybridized seawater desalination process, Desalination 131 (2000) 147156.
[112] A. Altaee, G. Zaragoza, A conceptual design of low fouling and high recovery FO
MSF desalination plant, Desalination 343 (2014) 27.
[113] C. Bellona, J.E. Drewes, P. Xu, G. Amy, Factors affecting the rejection of organic solutes during NF/RO treatmenta literature review, Water Res. 38 (2004)
27952809.
[114] V.L. Cunningham, S.P. Binks, M.J. Olson, Human health risk assessment from the
presence of human pharmaceuticals in the aquatic environment, Regul. Toxicol.
Pharmacol. 53 (2009) 3945.
[115] M. Schriks, M.B. Heringa, M.M.E. van der Kooi, P. de Voogt, A.P. van Wezel, Toxicological relevance of emerging contaminants for drinking water quality, Water Res.
44 (2010) 461476.
[116] M. Xie, L.D. Nghiem, W.E. Price, M. Elimelech, Comparison of the removal of hydrophobic trace organic contaminants by forward osmosis and reverse osmosis, Water
Res. 46 (2012) 26832692.
[117] B.D. Coday, B.G.M. Yaffe, P. Xu, T.Y. Cath, Rejection of trace organic compounds by
forward osmosis membranes: a literature review, Environ. Sci. Technol. 48 (2014)
36123624.
[118] C. Kim, S. Lee, H.K. Shon, M. Elimelech, S. Hong, Boron transport in forward osmosis: measurements, mechanisms, and comparison with reverse osmosis, J. Membr.
Sci. 419420 (2012) 4248.
[119] D.L. Shaffer, N.Y. Yip, J. Gilron, M. Elimelech, Seawater desalination for agriculture
by integrated forward and reverse osmosis: improved product water quality for
potentially less energy, J. Membr. Sci. 415416 (2012) 18.
[120] A. D'Haese, P. Le-Clech, S. Van Nevel, K. Verbeken, E.R. Cornelissen, S.J. Khan, A.R.D.
Verliefde, Trace organic solutes in closed-loop forward osmosis applications: inuence of membrane fouling and modeling of solute build-up, Water Res. 47 (2013)
52325244.
[121] N.T. Hancock, N.D. Black, T.Y. Cath, A comparative life cycle assessment of hybrid
osmotic dilution desalination and established seawater desalination and wastewater reclamation processes, Water Res. 46 (2012) 11451154.
j.desal.2014.10.031

Forward Osmosis: Where Are We Now?

Hochgeladen von

Dokumentinformationen

Originaltitel

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

Forward Osmosis: Where Are We Now?

Hochgeladen von

Copyright:

Verfügbare Formate

DES-12318; No of Pages 14

Desalination xxx (2014) xxxxxx

Contents lists available at ScienceDirect

Forward osmosis: Where are we now?

present a critical review of the current state of forward osmosis (FO).

D.L. Shaffer et al. / Desalination xxx (2014) xxxxxx

Is FO a low fouling process? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

a separation process than other membrane processes. In fact, FO is not

D.L. Shaffer et al. / Desalination xxx (2014) xxxxxx

D.L. Shaffer et al. / Desalination xxx (2014) xxxxxx

3. What are the key required membrane properties for FO?

Since its introduction, the structural parameter has provided a simple

permeability and selectivity than polyamide TFC-RO membranes [5,

D.L. Shaffer et al. / Desalination xxx (2014) xxxxxx

where Js is reverse mass ux of draw solute, n is the number of dissolved

including polyamide-based TFC membranes. Recent analyses led to

where L is the thickness of the active layer, is a tting parameter, and

D.L. Shaffer et al. / Desalination xxx (2014) xxxxxx

nanoparticles [63], hydrogels [61,62], and switchable polarity solvents

Comparison of Eqs. (9) and (2) results in:

The UF minimum operating hydraulic pressure in an FOUF process is

D.L. Shaffer et al. / Desalination xxx (2014) xxxxxx

5. Is FO a low fouling process?

4.3. Some opportunity remains for draw solute innovation

D.L. Shaffer et al. / Desalination xxx (2014) xxxxxx

semipermeable membrane, foulants in the feed solution accumulate on

5.2. Membrane surface properties are important for fouling control

D.L. Shaffer et al. / Desalination xxx (2014) xxxxxx

6. Where does FO outperform other desalination technologies?

D.L. Shaffer et al. / Desalination xxx (2014) xxxxxx

electrodialysis, and thermal desalination technologies like MED, MSF,

D.L. Shaffer et al. / Desalination xxx (2014) xxxxxx

The ability of FO pretreatment to remove dissolved constituents

D.L. Shaffer et al. / Desalination xxx (2014) xxxxxx

D.L. Shaffer et al. / Desalination xxx (2014) xxxxxx

D.L. Shaffer et al. / Desalination xxx (2014) xxxxxx

Das könnte Ihnen auch gefallen